Luật Chơi Cờ Cá Ngựa

Tổng hợp các bài viết thuộc chủ đề Luật Chơi Cờ Cá Ngựa xem nhiều nhất, được cập nhật mới nhất ngày 17/01/2021 trên website Athena4me.com. Hy vọng nội dung bài viết sẽ đáp ứng được nhu cầu của bạn, chúng tôi sẽ thường xuyên cập nhật mới nội dung Luật Chơi Cờ Cá Ngựa để bạn nhận được thông tin nhanh chóng và chính xác nhất. Cho đến thời điểm hiện tại, chủ đề này đã đạt được 3.069 lượt xem.

Có 12 tin bài trong chủ đề【Luật Chơi Cờ Cá Ngựa】

【#1】Hướng Dẫn Chi Tiết Luật Chơi Cờ Cá Ngựa Online

Những điều cần biết về cờ cá ngựa online.

Với trò chơi đua ngựa dân gian, bàn cờ sẽ được chia thành 4 ô vuông bằng nhau và mỗi ô vuông được phân biệt bằng các màu sắc như: màu xanh dương, màu vàng, màu đỏ, màu xanh lá. Trong bàn cờ sẽ có hai hạt xúc xắc có 6 mặt ứng với số thứ tự từ 1 tới 6. Trò chơi sẽ có 16 quân cờ ứng với 4 màu màu sắc của bàn cờ. Trò chơi cá ngựa online cũng có cấu trúc tương tự, bao gồm 16 quân cờ được chia đều 4 màu, mỗi màu sẽ có 4 quân cờ.

Nguyên tắc di chuyển trong trò chơi này là mỗi đội sẽ di chuyển theo số mặt xúc xắc mà người đó lắc được. Để được bắt đầu xuất phát thì bắt buộc bạn phải lắc được mặt 6 chấm. Người chiến thắng là người có đủ 4 quân cờ xếp vào các vị trí 6,5,4,3 nhanh nhất, 3 người còn lại sẽ tranh những vị trí 3,2 và cuối cùng.

Trong game cờ cá ngựa online thì mỗi trận sẽ kéo dài từ 30 tới 90 phút tùy vào tốc độ lắc xúc xắc của từng đội. Trò chơi này cũng không quy định hoặc giới hạn người chơi bên bạn có thể bắt đầu cuộc chơi khi có 2 người trở lên. Cách để chiến thắng trò chơi này là bạn cần lắc mạnh tay để đưa quân cờ về chuồng nhanh nhất.

Hướng dẫn chi tiết luật chơi cờ cá ngựa online.

Rất nhiều người thắc mắc không biết chơi cờ cá ngựa online có giống như chơi cờ cá ngựa online hay không, nó đơn giản hơn hay phức tạp hơn so với chơi trực tiếp? Với câu hỏi này thì bạn có thể hoàn toàn yên tâm vì chơi game này rất dễ, bạn chỉ cần một chút khéo léo và kiên nhẫn là có thể chiến thắng trò chơi này.

Trong luật chơi cờ cá ngựa, để có thể đá được ngựa của đối thủ bạn cần lắc được đúng số chấm trên mặt xúc xắc khi cần thiết. Lúc này bạn sẽ tước được quyền di chuyển của đối thủ và nhanh chóng giành chiến thắng.

Khi tham gia trò chơi này, bạn cần chú ý rằng nếu ngựa của bạn đã đứng ở cửa chuồng bạn cần nhanh chóng đưa quân của mình lên chuồng để đề phòng đối phương đá bạn. Cách chắc chắn nhất cho chiến thắng của mình là bạn lên thắng số 6 để đảm bảo an toàn.


【#2】Tìm Hiểu Về Luật Chơi Cờ Cá Ngựa Ăn Tiền Cho Người Mới Tham Gia

Nhiều bạn sẽ còn lạ lẫm với môn trò chơi được coi là kinh điển một thời của giới trẻ. Vì vậy, chúng tôi sẽ cùng các bạn điểm qua một vài nét về trò chơi được rất nhiều người yêu thích này. Trò chơi cờ cá ngựa còn có một cái tên gọi khác là cờ đua ngựa. Trò chơi này đã tồn tại từ rất lâu và được bắt nguồn từ Pachisi của Ấn Độ, sau một quá trình cải tiến ở bên Mỹ, trò chơi này đã được biết tới với phiên bản giản thể Ludo.

Người Pháp gọi bộ môn trò chơi này là ngựa nhỏ, còn khi về tới Việt Nam nó có tên là cờ cá ngựa. Trò chơi này được diễn ra trên một bàn cờ có hình vuông với sự tham gia của 2 đến 4 người chơi.

Tuy chỉ có sự tham gia của 2-4 người chơi nhưng trên thực tế, khi có bàn cờ cá ngựa ở đâu thì số lượng người tham gia xem cũng phải lên thêm từ 3-5 người nữa. Vì trò cá ngựa là một trò khá vui vì tính chất bất ngờ, gây cười của nó. Bạn cứ thử nghĩ xem khi bạn mới đổ xí ngầu và vừa mới đứng trước cửa chuồng thì tới lượt của đối phương, bạn lại bị đá bay lại vào trong chuồng. Tâm lý ức chế của bạn nhưng lại là niềm hào hứng của cả những người còn lại. Vì là trò khó dự đoán trước nên một số ít người ngoài tham gia xem thường cảm thấy nhanh chán và họ thường chuyển sang bộ môn khác dễ dự đoán hơn, đó là dự đoán kèo đêm nay của bộ môn bóng đá để mặc những người chơi và người xem khác vẫn đang hăng say trong sự vui thú của trò chơi cờ cá ngựa

Để có thể chơi tốt trò chơi này, ngoài việc hiểu rõ luật chơi các bạn cần phải chuẩn bị một bàn cờ tiêu chuẩn. Một bộ bàn cờ cá ngựa sẽ bao gồm:

– Một bàn cờ hình vuông được chia thành 4 ô vuông nhỏ với 4 màu khác biệt như đỏ, vàng, xanh lá cây và xanh dương.

– Bốn viên xúc xắc

– 16 quân cờ cá ngựa được chia đều với 4 màu của 4 ô cờ.

– Bốn chiếc cốc nhỏ để đổ xúc xắc.

Với bí quyết đánh cờ tướng giỏi này, tôi tin chắc bạn sẽ có thể đánh bại các thế cờ khó của đối thủ đấy.

Tìm hiểu về luật chơi cờ cá ngựa ăn tiền

Sau khi đã biết được cờ cá ngựa là gì và những vật dụng cần chuẩn bị để chơi cờ cá ngựa, chúng ta sẽ cùng nhau tìm hiểu về luật cờ cá ngựa để giúp các bạn hiểu rõ hơn khi chơi.

Nguyên tắc chơi cờ cá ngựa

Nguyên tắc để chơi trò chơi này khá đơn giản, các bạn sẽ phải di chuyển quân cờ sao cho nó đi quanh bàn cờ và về được đến đích. Khả năng di chuyển của các quân cờ sẽ phụ thuộc vào những lần đổ xúc xắc của bạn.

Người chơi chiến thắng là người đưa được 4 quân cờ vào các vị trí 6, 5, 4, 3 trong chuồng. Các người chơi tiếp theo sẽ tiếp tục thi đấu để tranh vị trí 2, 3 và 4.

Luật chơi cờ cá ngựa ăn tiền

Luật phân thắng bại: Người nào khi đưa được 4 quân cờ của mình xếp vào chuồng theo thứ tự 6, 5, 4, 3 sẽ là người chiến thắng. các người chơi còn lại sẽ tiếp tục chơi để tìm ra vị trí thứ 2 và 3.

Mỗi người chơi sẽ có 4 quân cờ, trò chơi này vẫn có thể áp dụng khi chơi với 2 hoặc 3 người. Từng người sẽ chọn quân cờ và phân lượt lắc xí ngầu trước khi chơi.

Cách di chuyển quân cờ trong mỗi lần đi: Điều kiện tiên quyết là bạn sẽ phải có ít nhất 1 quân cờ đã được xuất chuồng. Kết quả của việc lắc xí ngầu sẽ quyết định bạn được đi mấy nước. Trong khi di chuyển sẽ xảy ra một số tình huống sau:

– Bị cản: Khi bạn bị 1 quân cờ của đối phương cản ở phái trước, nếu muốn đi qua bạn sẽ phải lắc được số nước đi vượt qua khoảng cách tương ứng giữa 2 quân cờ.

– Đá: Đây là trường hợp rất đặc biệt và là một điểm khá thú vị trong cờ cá ngựa. Điều này chỉ xảy ra khi bạn đổ xúc xắc được đúng số nước đi vào vị trí của quân cờ khác. Khi đó bạn sẽ có 2 sự lựa chọn là đá quân cờ đó về chuồng hoặc di chuyển một quân cờ khác.

Để có thể xuất chuồng được thì lần lắc xí ngầu của bạn phải đạt được mặt 6 điểm, nếu không được sẽ bị mất lượt và chuyển qua lượt cho người khác. Khi quân cờ của bạn vào đến chuồng, các bạn sẽ không lo bị đá nữa.

Tuy nhiên, khi quân cờ mới ở cửa chuồng bạn sẽ vẫn còn nguy cơ bị đối phương đưa về vị trí xuất phát. Đặt cược khi chơi cờ cá ngựa: đây chính là một yếu tố sẽ tang sự kịch tính cho các ván đấu.

Khi bạn bị đối phương đá thì người chơi sẽ được cộng tiền theo mức cược định sẵn của chủ phòng. Cuối những cuộc chơi sẽ tổng tiền lại để trả cho người chơi theo quy định.


【#3】Tìm Hiểu Luật Chơi Cờ Cá Ngựa Chuẩn Để Có Những Cuộc Chơi Vui Vẻ

Một bàn cờ cá ngựa thường được chia làm 4 ô có kích thước và vùng đất được chia đều. Chúng được phân biệt bằng màu sắc của từng ô và quân, thông thường các màu bao gồm: đỏ, vàng, xanh dương, xanh lá. Tổng số quân cờ là 16 mỗi người 4 con và chia làm 4 đội. Mỗi ô là có 1 quân xúc xắc, trên bàn cờ online cũng như vậy, luật cờ cá ngựa cũng giống như truyền thống về cách chơi và số người chơi

Ngoài việc bạn tìm hiểu về những luật chơi cờ cá ngựa như nào để đánh bại đối thủ. Bạn muốn tìm kiếm một tựa để chơi giải. Chúng tôi chia sẻ những cách chơi game bắn cá ăn xu giành cho người mới để giúp các bạn mới bắt đầu chơi game này một cách dễ dàng hơn và giành được điểm cao nhất trong game.

Điều đặc biệt là trên bàn cờ sẽ có số thứ tự để đánh dấu sự di chuyển quân của mình. 4 người chơi lần lượt tung xúc xắc để di chuyển quân của mình về tới đích, người chơi phải tung ít nhất và về đích đầu tiên sẽ giành chiến thắng. Tuy nhiên có 1 một lưu ý đó là để quân xuất phát thì bắt buộc phải đổ được mặt 6 chấm mới xuất chuồng. Người nào xếp đủ về đích trước giành ngôi vị số 1, những người còn lại tiếp tục đua tranh vị trí 2,3,4.

Thời gian chơi của một ván cờ cá ngựa mất khoảng tầm 30 phút hoặc 90 phút hay thậm chí còn kéo dài đến 2 tiếng. Tuy dầu cá ngựa có 4 ô nhưng không nhất thiết phải đủ 4 người mới chơi được. Mà chỉ cần 2 hoặc 3 người là có thể chơi được. Cờ cá ngựa được xem là trò chơi mang lại tính giải trí khi tụ tập những lúc rảnh rỗi.

Hướng dẫn cách chơi cờ cá ngựa dễ nhất

Trước khi khai trận, chúng tôi muốn bạn hiểu qua về luật chơi cơ bản của môn này. Vậy luật chơi cờ cá ngựa chuẩn và đúng cách là gì?

Chơi cờ cá ngựa đúng cách

Đổ xúc xắc: Người chơi sẽ lắc xúc xắc và đổ khi đến lượt của mình di chuyển cá ngựa. Thí dụ, nếu bạn tung xúc xắc được 2 điểm đôi thì được thêm 1 lượt nữa và được xuất quân.

Xuất quân: Người chưa có quân nào hoặc mới bắt đầu chơi sẽ được xúc xắc 3 lần xuất quân. Để có thể xuất quân tốt thì kết quả tung xúc xắc phải lục hoặc kép. Có nghĩa là, tổng 2 lần xúc xắc của bạn cộng lại là 6 (3:3, 2:4: 5:1) thì mới được đưa 1 quân ra khỏi vị trí ban đầu.

Di chuyển: Trên bàn cờ cá ngựa có 1 quân cờ của mình, tổng điểm sẽ tương ứng với bước đi của quân cờ.

Trong khi di chuyển có một số tình huống xảy ra:

Bị cản: Cách chơi cờ cá ngựa khá đơn giản về di chuyển và chuồng. Mà luật chơi của cờ đua ngựa còn cho phép cản trong khi chơi. Nếu 1 quân bị cản là khi đó có quân cờ của người khác hoặc của mình chắn trước nó. Bước đi của 2 quân này sẽ dựa vào kết quả của việc tung xúc xắc. Có nghĩa là bạn sẽ không thể vượt qua được quân cờ trước đó. Nếu bạn không còn quân cờ nào để đi thì coi như bạn bị tịt (có nghĩa là mất lượt)

Đá: Trong luật chơi cờ cá ngựa đá là hành động khá thú vị là buộc đối phương mất quyền di chuyển quân trên bàn cờ. Trường hợp này chỉ xảy ra khi quân cờ của đối phương và mình nằm trên cùng đường đi và tung xúc xắc vào ô điểm đúng vị trí của quân đó.

Phân thắng bại: Cách chơi của cờ đua ngựa sẽ được quyết định vào 4 ô đầu là 6, 5, 4 và 3 (phải xếp đủ ngựa vào 4 ô này). Nếu đứng trong 4 ô này là người chiến thắng. Những người chơi còn lại sẽ lần lượt chơi tiếp để giành thứ hạng của giải. Trận đấu có 4 người mà 2 người về đích rồi thì 2 người kia sẽ dừng lại


【#4】Luật Chơi Cờ Cá Ngựa Ăn Tiền Chi Tiết Nhất Cho Người Mới

1. Trò chơi cờ cá ngựa là gì?

Được bắt nguồn từ Ấn Độ với tên gọi Cờ đua ngựa, cờ cá ngựa nhanh chóng du nhập vào Việt Nam và trở thành trò chơi được nhiều người yêu thích.

Đây là trò chơi có thể tham gia với số lượng đông từ 2-4 người. Là trò chơi thích hợp với nhiều độ tuổi khác nhau, từ trẻ nhỏ đến người lớn. Đặc biệt, về luật chơi cờ cá ngựa ăn tiền lại hoàn toàn có thể được nắm bắt một cách dễ dàng. Người chơi có thể truy cập trong khoảng thời gian 2 phút và thường kết thúc trò chơi trong khoảng 30 phút.

Một bàn cờ có thể chia thành 4 phần, cùng với các ô có màu sắc riêng biệt. Thông thường các màu được sử dụng nhiều hơn trong trò chơi này được sử dụng như màu đỏ, xanh dương, xanh lá cây, vàng. Trong 1 bộ cờ này thông thường có 16 quân cờ chia đều theo 4 màu của bàn cờ, 4 chiếc cốc nhỏ để đổ, 4 viên xúc xắc,…

2. Những ưu điểm của trò chơi cờ cá ngựa ăn tiền

Đây là trò chơi được rất nhiều người yêu thích, cùng với việc có thể phù hợp cho mọi lứa tuổi. Bạn có thể tham gia trò chơi này để thư giãn, giải tỏa và xua tan các căng thẳng mệt mỏi đang gặp phải.

Là trò chơi thú vị đem lại kích thích cho người chơi, trong những trò chơi chạy đua gay cấn, bạn cần sử dụng trí thông minh để có thể giành chiến thắng một cách tuyệt đối.

Trò chơi cờ cá ngựa ăn tiền còn có thể đem lại thu nhập khá lớn cho người chơi. Cùng với đó là bạn có thể chiến thắng và nhận về những phần quà hấp dẫn, có giá trị mà trò chơi mang lại.

3. Tìm hiểu về luật chơi cờ cá ngựa ăn tiền

Nguyên tắc chơi cờ cá ngựa

Để chơi được trò chơi cờ cá ngựa ăn tiền thì 4 người chơi phải nắm được các nguyên tắc cơ bản nhất của trò chơi. Tất cả điều bạn cần làm đó là di chuyển quân cờ sao cho di chuyển một cách đơn giản và dễ dàng. Đồng thời, với khoảng cách di chuyển của bạn có thể đổ vào lượt đổ xúc xắc, đổ được bao nhiêu thì có thể đổ vào và có được bước đi tương ứng.

Người chiến thắng trong game cờ cá ngựa muốn chiến thắng thì đồng nghĩa với việc, họ phải đưa được 4 quân cờ vào các vị trí trong chuồng một cách nhanh nhất. Như vậy, mới có thể giành được chiến thắng về tay của mình.

Thông thường, quy định về chiến thắng là người chơi phải di chuyển 4 quân cờ của mình xếp vào chuồng theo số thứ tự từ 6-5-4-3 nhanh nhất. Ai là người di chuyển nhanh thì người đó sẽ giành được chiến thắng. Người chơi có thể tiếp tục thi để giành được những vị tí thứ 2 và thứ 3.

Người chơi phải tiến hành lắc xúc xắc đạt được 6 điểm để xuất chuồng 1 quân cờ. Trong trường hợp nếu không thành công, bạn có thể bị mất lượt và chuyển qua lượt của người khác.

Trong luật chơi cờ cá ngựa ăn tiền, người chơi có thể thực hiện xuất chuồng được 1 con. Kết quả từ việc lắc xúc xắc bạn có thể phân biệt được mình di chuyển bao nhiêu bước. Tuy vậy, bạn sẽ có thể thấy sẽ có 1 số trường hợp cần lưu ý cụ thể như sau:

Nếu trong trường hợp 1 quân cờ của bạn bị di chuyển, trong khi đó lại bị 1 quân cờ khác cản ở phía trước, bạn sẽ tiến hành lắc xúc xắc vào số để vượt qua nó bằng cách vượt qua giữa các quân cờ.

Đây có thể là phương pháp làm cho cờ cá ngựa trở nên hấp dẫn hơn đối với người chơi. Nếu bạn nghĩ rằng trò chơi này nhàm chán thì luật chơi cờ cá ngựa ăn tiền lại ngày càng thú vị hơn với luật đá.

Nếu bạn lắc xúc xắc đúng số nước đi vào vị trí của quân mà đối phương đang sở hữu, tuy nhiên, bạn có thể có một sự lựa chọn khác nếu không muốn đá đối thủ đó là thực hiện di chuyển quân của mình. Vì thế, với những cách trên, chúng ta có thể hoàn toàn chơi trò chơi này đúng luật mà không sợ phạm quy.


【#5】Hướng Dẫn Cách Chơi, Luật Chơi Cờ Cá Ngựa Chi Tiết Nhất

Cờ cá ngựa là trò chơi bắt nguồn từ Mỹ. Là trò chơi chiến lược, các bạn có nhiệm vụ đưa quân của mình về chuồng trước các đối thủ. Các quân cờ có hình đầu ngựa, có thể đứng được trên mặt phẳng. Đây là trò chơi cho tối đa 4 người tham gia.

Bộ cờ cá ngựa được bán với giá khá rẻ. Ai cũng có thể mua và chơi được. Một bộ cờ gồm có:

  • Bàn cờ hình vuông chia làm 4 phần, mỗi phần một màu khác nhau. Thường là 4 màu xanh dương, xanh lá, vàng, đỏ. Bàn cờ có hai loại: loại bằng giấy và loại bằng nhựa.
  • Xúc xắc: 4 hoặc 8 xúc xắc.
  • Cốc lắc xúc xắc.
  • 16 quân cờ cá ngựa thuộc 4 màu khác nhau như màu bàn cờ.

Nếu bạn yêu thích cờ tướng có thể tìm hiểu thêm: Tổng hợp các thể cờ tướng hay và cách giải chi tiết.

Để chơi cờ cá ngựa, ta cần sắm đầy đủ bộ cờ và đọc hướng dẫn. Tốt nhất nên chọn những bộ cờ còn đủ quân, đủ xúc xắc, không bị nhòe các hình vẽ trên bàn cờ. Bạn có thể chơi từ 2-4 người đều được. Đầu tiên cần đặt các quân cờ vào ô có hình tròn màu tương tự. Chỗ trống đó gọi là “đồng cỏ” của ngựa. Các hình tròn nhỏ được sắp xếp vuông góc bao bên ngoài đồng cỏ là đường đi của ngựa. Chấm tròn được in đậm sẽ là điểm xuất phát cho ngựa của bạn. Để đưa ngựa về chuồng bạn cần cho ngựa đi một vòng quanh bàn cờ. Cách đi là đi theo số điểm được xúc xắc. Nên cẩn thận trong từng bước, nếu không sẽ bị đá bay khỏi bàn cờ.

Thống nhất thứ tự lượt đi xong, bạn sẽ tiến hành xúc xắc để xem có đủ điều kiện “xuất quân” hay không. Chỉ khi xúc xắc ra được mặt 6 chấm thì ngựa mới được đưa từ “đồng cỏ” ra điểm xuất phát. Bạn sẽ tiếp tục xúc xắc khi đến lượt của mình. Xúc được bao nhiêu điểm thì đi bấy nhiêu bước. Cứ như vậy cho đến khi về chuồng.

Xem hướng dẫn luật chơi cờ cá ngựa

Tuy nhiên trong cờ cá ngựa cũng có những luật chơi vô cùng thú vị. Bạn và bạn bè chắc chắn sẽ có những phen ngã ngửa khi rơi vào những tình huống như:

  • Nếu tung xúc xắc, xúc xắc bị rơi ra khỏi bàn cờ thì sẽ bị coi như mất lượt. Kết quả trên xúc xắc không được tính.
  • Nếu xúc xắc ra mặt 1 hoặc 6 thì người chơi phải lắc tiếp cho đến khi ra số khác.
  • Bị đá: Bạn sẽ công cốc cả quá trình dài nếu bị quân đối thủ đá bay ra khỏi bàn cờ đó. Đó là khi số xúc xắc đối thủ xóc ra bằng đúng số bước giữa bạn và đối thủ. Họ sẽ đá quân của bạn ra và thế quân của họ vào đó. Dĩ nhiên, bạn cũng có thể làm như thế. Bạn có thể lựa chọn “đá” đối thủ hoặc đi quân khác (nếu bạn có trên 1 quân trên bàn cờ). Ngay khi bạn về đến cửa chuồng, bạn vẫn có thể bị đá. Chỉ khi quân của bạn vào chuồng mới tính là an toàn.
  • Bị cản: bị cản là khi một quân cờ khác đứng trước quân cờ của bạn. Và số mà bạn xúc xắc ra lại không đủ để đá nó. Trong trường hợp này, bạn gần như không thể di chuyển.
  • Sập hầm: Nếu bạn ra số xúc xắc vào đúng chuồng số 1, bạn sẽ bị sập hầm và bị phạt. Ngoài ra, bạn phải lên lần lượt từng chuồng (2,3,4, rồi lên 5,6). Nếu xúc một lần ra được số 6, vào chuồng 6 thì sẽ được thưởng. Tuy nhiên luật chơi này vẫn tùy người giao kèo.

Đôi khi bạn sẽ phải tính toán trong lúc chơi cá ngựa. Cờ cá ngựa tuy dễ chơi nhưng vẫn là trò may rủi. Vì vậy, ai biết cách kết hợp với sự tính toán sẽ có khả năng thắng cao hơn. Tuy mạo hiểm nhưng lại thu về kết quả tốt. Ngại gì mà không thử những mẹo sau:

  • Khi bàn cờ có trên 1 quân cờ của bạn, bạn cần tính toán đi bước nào là hợp lý. Đôi khi bạn phải bỏ qua một số quân cờ để nhường quân có lợi thế đi trước.
  • Tính toán sao cho “đá” được ngựa đối thủ. Nếu bạn có 4 quân cờ trên bàn cờ, bạn cần tính toán đổ được xúc xắc bao nhiêu thì đi quân nào. Như thế bạn sẽ dễ dàng gài bẫy đối thủ, thậm chí khiến quân đối thủ bị chặn bởi quân mình.

Cách chơi cờ cá ngựa không khó như bạn tưởng phải không. Chỉ cần nắm rõ cách đi, bạn đã có thể nhập cuộc cùng bạn bè rồi. Mong những chia sẻ về cờ cá ngựa trên sẽ giúp bạn tất thắng.

Nguồn: nhacaiplus.com


【#6】Luật & Cách Chơi Game Cờ Cá Ngựa Online Chi Tiết Nhất

Cờ cá ngựa là gì?

Cá cá ngựa (Horse Race Chess) hay còn có tên gọi khác là cờ đua ngựa hoặc đơn giản là cờ ngựa. Đây là một trong các board game nổi tiếng với cả trẻ em và người lớn ở Việt Nam. Nói cách khác đây là trò chơi kinh điển hay trò chơi quốc dân với nhiều người chơi.

Game cá ngựa phổ biến ở nhiều quốc gia với các tên gọi khác nhau nhưng cách chơi thì không khác biệt. Ví dụ như ở Pháp, trò chơi này có tên gọi là Ngựa nhỏ (Petits Chevaux). Còn ở Việt Nam nó được gọi là cờ cá ngựa hay đổ cá ngựa. Bởi hình dáng các quân cờ khá giống con cá ngựa.

Trò chơi được đánh giá cao bởi cách chơi thú vị mang nặng tính chất giải trí. Nó đem đến các giây phút thực sự sảng khoái cho các người chơi. Đồng thời thích hợp để chơi trong gia đình, trường học hay văn phòng. Cá ngựa là một trong các thể loại cờ hay nhất mà bạn nên thử ít nhất một lần.

Hướng dẫn cách chơi cờ cá ngựa chi tiết cho người mới

1/ Giới thiệu về bộ cờ cá ngựa

Để có thể chơi trò chơi cá ngựa, các người chơi cần chuẩn bị một bộ cờ cá ngựa. Bộ cờ này có thể mua trên các shop chuyên bán board game hoặc mua online trên mạng. Do là game chơi cờ phổ biến nên không khó để bạn có được một bộ chơi cờ. Vậy một bộ cờ ngựa cho 4 người chơi có gì?

– Một bàn cờ chơi cá ngựa. Bàn cờ này thường có hình vuông được làm bằng giấy để gấp gọn cho tiện lợi. Trên bàn cờ có 4 khu vực chơi với 4 màu sắc khác nhau (Xanh dương, Xanh lá cây, Vàng, Đỏ). Ở từng khu vực chơi trên bàn cờ lại có các vùng riêng biệt dành cho một người chơi cụ thể.

  • Thảo nguyên là khu vực đặt các chú ngựa trước khi bắt đầu bàn chơi. Các người chơi khi bắt đầu một ván cờ cá ngựa có 4 con ngựa đặt tự do trong vùng Thảo nguyên.
  • Điểm xuất phát chính là nơi chú ngựa được đặt đầu tiên khi được đưa ra từ Thảo nguyên. Khi vào điểm xuất phát, chú ngựa bắt đầu hành trình di chuyển để về chuồng.
  • Chuồng ngựa là các ô được số từ 1 tới 6 tương ứng với màu sắc của quân cờ. Đây là nơi các chú ngựa di chuyển về (hay gọi cách khác chính là đích đến).

– Cùng với bàn cờ, một bộ cờ cá ngựa còn có 16 quân cờ. Các quân cờ này thường có hình đầu ngựa với 4 màu sắc khác nhau tương ứng với 4 khu vực và chuồng trên bàn cờ. Một người chơi khi bắt đầu trò chơi có 4 quân cờ hình đầu ngựa (giống với con cá ngựa).

– 4 viên xúc xắc cho 4 người chơi. Hoặc các người chơi dùng chung một con xúc xắc đều được. Các quân xúc xắc này giống với trong game cờ tỷ phú Monopoly.

– 4 cốc lắc xúc xắc dùng để đổ xí ngầu. Tuy nhiên nếu không có thì các người chơi có thể sử dụng các dụng cụ lắc xúc xắc khác như bát thay thế.

2/ Cách thiết lập một bàn cờ cá ngựa cơ bản

Tiếp theo các người chơi đặt các quân cờ của mình vào đúng khu vực vùng Thảo Nguyên trên bàn cờ. Sau đó, các người chơi có thể đổ xúc xắc để xác định người chơi đầu tiên. Người nào đổ được điểm cao nhất thì chơi đầu. Hoặc có thể Oẳn Tù Tì hay bốc thăm cho nhanh.

Thứ tự chơi ở các lượt tiếp theo là người chơi ngồi cạnh người đầu tiên cùng chiều hay ngược chiều kim đồng hồ do các người chơi quy định. Tuy nhiên thứ tự này không được thay đổi từ khi bắt đầu cho tới khi kết thúc ván cờ.

3/ Luật chơi cờ cá ngựa bạn cần biết

a/ Số lượng người chơi

Một game cờ cá ngựa cơ bản giới hạn số lượng người chơi từ 2 tới 4 người. Trong đó, chơi 4 người là phổ biến nhất và vui nhất trong trò chơi này. Bởi càng chơi đông thì một ván cờ càng thêm kịch tính. Đồng thời tính đối kháng, cạnh tranh cao hơn so với bàn cờ chơi 2 người.

b/ Mục tiêu của trò chơi

c/ Quy trình một ván cờ cá ngựa

+ Gieo xúc xắc

Gieo xúc xắc hay đổ xí ngầu là hành động các người chơi đều phải thực hiện trong các lượt chơi. Chính vì vậy ngoài tên gọi là cá ngựa thì trò chơi còn có tên là đổ cá ngựa. Xúc xắc được lắc trong một cốc nhựa hoặc khay, bát có độ nảy sau đó đổ ra ngoài.

Một quân xúc xắc có sáu mặt với các chấm nhỏ từ một (Nhất) tới sáu (Lục). Đây chính là điểm quyết định các hành động tiếp theo của người chơi. Tới lượt người chơi nào, người chơi đó được tung xúc xắc.

+ Xuất quân cờ

Người chơi chỉ thực sự bắt đầu ván chơi khi xuất được ngựa ra khỏi vùng Thảo nguyên vào điểm xuất phát. Để làm điều này, người chơi phải gieo xúc xắc được mặt Lục (6 chấm) hoặc Nhất (1 chấm). Khi đó người chơi được đưa một quân cờ từ Thảo nguyên ra điểm xuất phát. Đồng thời có thêm một lượt gieo sau khi xuất quân cờ. Khi chưa có quân cờ nào ra khỏi Thảo nguyên, người chơi phải gieo tới khi được mặt Nhất hoặc Lục.

+ Di chuyển các quân cờ

Sau khi xuất được ít nhất một quân cờ vào điểm xuất phát, người chơi có thể di chuyển quân cờ đó từ điểm xuất phát (Start) về chuồng theo hướng ngược chiều kim đồng hồ. Số nước được di chuyển tương ứng với số chấm hiển thị trên quân xúc xắc sau khi gieo. Ví dụ bạn gieo được mặt bốn chấm thì quân cờ được đi 4 nước trên bàn cờ.

Tuy nhiên trong trường hợp người chơi gieo được mặt xúc xắc 6 điểm thì có quyền xuất thêm quân cờ tiếp theo. Hoặc di chuyển 6 nước với một quân cờ đang ở bàn chơi.

Ngoài ra, trong quá trình di chuyển quân cờ có hai trường hợp xảy ra với người chơi mà bạn nên lưu ý.

– Quân cờ bị cản đường nếu có quân cờ khác của đối phương (hoặc của bạn) đứng ở phía trước trên đường di chuyển về chuồng. Số chấm mà bạn gieo xúc xắc lớn hơn khoảng cách số nước đi từ quân cờ của bạn tới quân cờ của đối phương. Lúc này, bạn bị quân cờ của đối phương cản lại không thể di chuyển hay lùi lại.

Ví dụ khoảng cách của quân cờ ngựa Đỏ đứng trước và ngựa Xanh dương đứng sau là 3 nước. Người chơi cầm ngựa Đỏ gieo xúc xắc được 5 chấm thì không thể di chuyển được ngựa Đỏ do phía trước là ngựa Xanh dương. Chỉ được đi quân cờ khác (nếu hợp lệ) hoặc chờ gieo được xúc xắc 1 hoặc 2 chấm.

Ví dụ khoảng cách của quân cờ ngựa Vàng (sau) và quân cờ ngựa Đỏ (trước) là 3 nước. Người chơi ngựa Vàng gieo được xúc xắc được 3 chấm và di chuyển 3 nước đúng vào vị trí ngựa Đỏ. Ngựa Đỏ lúc đó bị đá ra khỏi bàn cờ về lại Thảo nguyên.

+ Vào/Về chuồng

Muốn an toàn và không bị đối phương đá thì bạn phải vào chuồng tương ứng với màu quân cờ. Khi quân cờ của bạn đang đứng trước cửa chuồng hoặc rất gần cửa chuồng. Bạn gieo xúc xắc với số chấm đủ để vào chuồng thì được vào chuồng. Số chấm này tương ứng với số nước mà bạn di chuyển được.

Chuồng được đánh số từ 1 tới 6, bạn phải đưa 4 con ngựa vào chuồng theo đúng thứ tự chuồng 6, 5, 4, 3. Trường hợp này được gọi là về chuồng. Để đưa ngựa về chuồng đúng vị trí, bạn phải tung xúc xắc được số chấm đúng theo trình tự. Ví dụ ngựa đang ở ô chuồng số 3, muốn lên ô 4 thì bạn phải tung được xúc xắc 4 chấm chứ không được nhảy cóc. Trường hợp tung được số khác thì không di chuyển được ngựa về chuồng.

+ Kết thúc ván cờ cá ngựa

Một ván chơi cá ngựa kết thúc khi tìm được người chơi chiến thắng nếu ván chơi chỉ có hai người. Tức là người nào đưa được đủ 4 người về chuồng theo đúng thứ tự thì thắng. Trong ván chơi có 3 người hay 4 người thì phải tiếp tục chơi cho tới khi xác định được người về hạng Nhì hay Ba. Người cuối cùng không đưa được ngựa về chuồng là người thua cuộc hay về Bét.

d/ Điều kiện thắng khi chơi cờ cá ngựa

Điều kiện để thắng khi chơi cờ ngựa đó là người chơi phải đưa ngựa về chuồng càng nhanh càng tốt. Người nào đưa được hết các quân cờ về chuồng đầu tiên là người chiến thắng.

Tuy nhiên, để thắng thì phải đưa được cả 4 quân cờ của mình về chuồng cùng màu quân cờ. Đồng thời các quân cờ (ngựa) phải xếp trong chuồng theo thứ tự các ô liên tiếp 6, 5, 4 và 3 để giành chiến thắng. Chứ không phải cứ đưa được hết ngựa vào chuồng là thắng.

4/ Một số thuật ngữ cờ cá ngựa online phổ biến

  • Thảo nguyên – Khu vực vòng tròn đặt 4 quân cờ của một người chơi trước khi chơi cờ.
  • Xuất quân – Đưa quân cờ từ khu vực Thảo nguyên ra điểm xuất phát trên bàn cờ.
  • Cản đường – Quân cờ phía trước của đối phương (hoặc của chính bạn) không cho quân cờ phía sau của bạn di chuyển. Bởi số chấm trên xúc xắc vượt quá các nước hợp lệ so với khoảng cách của hai quân cờ.
  • Đá ngựa – Trường hợp một người chơi gieo xúc xắc và di chuyển đúng vào vị trí mà quân cờ của đối phương đang đứng. Lúc này quân cờ bị đá ra khỏi bàn cờ và về lại vùng Thảo nguyên.
  • Sát nút – Hai quân cờ trong trò chơi cờ cá ngựa đứng sát cạnh nhau. Khoảng cách chỉ là một nước cờ.
  • Vạch xuất phát – Điểm mà quân cờ đứng đầu tiên sau khi ra khỏi vùng Thảo nguyên.
  • Cửa chuồng – Vị trí chấm tròn trước chuồng ngựa. Ngựa đứng ở cửa chuồng có thể bị đá về lại Thảo nguyên nếu như chưa vào chuồng.
  • Chuồng – Đích đến của các quân cờ. Chuồng cùng màu với quân cờ được đánh số thứ tự từ 1 tới 6.
  • Về chuồng – Người chơi di chuyển quân cờ một vòng quanh bàn cờ về các vị trí cửa chuồng hoặc vào chuồng có màu tương ứng với màu quân cờ.
  • Thầu mạ – Đối với quân cờ đầu tiên đứng ở vị trí cửa chuồng. Người chơi gieo được mặt xúc xắc có 6 chấm. Lúc này, quân cờ di chuyển thẳng về vị trí số 6 trong chuồng ngựa. Trường hợp này có thể được thưởng x2 hoặc x3 theo giao kèo.
  • Sập hầm – Bất kỳ quân cờ ngựa nào di chuyển vào chuồng mà lọt vào vị trí chuồng đầu tiên (chuồng số 1) thì coi là sập hầm. Đồng thời bị phạt cho cả làng theo như giao kèo.
  • Mất lượt – Người chơi không được di chuyển trong lượt chơi của mình. Có thể do quân cờ của đối phương chặn. Hoặc do không có nước đi hay xuất quân đúng luật khi gieo xúc xắc.

Các biến thể của game cờ cá ngựa

Cách chơi cờ cá ngựa nhìn chung không có nhiều biến thể. Ngoài cách chơi truyền thống với một xúc xắc thì các người chơi có thể chơi gieo 2 xúc xắc để ván cờ thực hiện nhanh hơn. Tuy nhiên cách chơi này ít phổ biến hơn so với gieo một xúc xắc. Khi gieo hai xúc xắc nếu được hai mặt giống nhau thì người chơi được thêm một lượt gieo tiếp.

Một biến thể khác của cờ ngựa truyền thống đó là cờ cá ngựa online. Bạn có thể chơi ngay trên điện thoại, máy tính hay máy tính bảng nhanh chóng. Đồng thời trò chơi mô phỏng bàn cờ và các quân cờ sinh động hơn. Đi kèm âm thanh cuốn hút giúp tăng tính giải trí của trò chơi. Bạn có thể tải game cờ cá ngựa nhanh chóng trên các kho ứng dụng Android hay iOS trên mạng.

Kinh nghiệm chơi cờ cá ngựa online giỏi từ các cao thủ

1/ Phải nắm chắc luật chơi đổ cá ngựa cơ bản

Khi chơi cờ cá ngựa online hay kiểu truyền thống, bạn đều phải hiểu cụ thể về luật chơi và cách chơi trước khi thực hiện các ván cờ. Trong đó đặc biệt là luật xuất quân, di chuyển, đá ngựa hay về chuồng. Đây chính là các hành động cơ bản nhất trong trò chơi. Dù chơi cá ngựa khá đơn giản nhưng có hiểu luật thì chơi mới đúng và không sợ bị phạm luật khi chơi.

2/ Thường xuyên chơi game cờ cá ngựa online

Một cách để rèn luyện khả năng chơi cờ của bạn đó thường xuyên chơi cờ. Giờ đây với sự xuất hiện của các game cờ online bạn có thể chơi mọi lúc, mọi nơi với các người chơi khác hay chơi với máy. Giống như các bộ môn đánh cờ khác như cờ vây hay , muốn chơi giỏi thì nhất định bạn phải chơi nhiều. Chơi để rút kinh nghiệm và tìm ra các chiến thuật chơi cờ hợp lý cho bản thân.

3/ Ưu tiên đưa ngựa về chuồng trước

Một kinh nghiệm khi chơi cờ cá ngựa đó là bạn nên cố gắng đưa ngựa về chuồng và vào đúng vị trí của chuồng trước. Bởi nếu để ngựa ở cửa chuồng có thể cản nước đi của đối phương. Tuy nhiên đối mặt với nguy hiểm là bị đá và quân cờ phải trở lại Thảo nguyên.

Chắc hẳn bạn không muốn bao công sức di chuyển quân cờ của mình bị đổ xuống sông xuống biển đúng không? Ngoài ra nên ưu tiên đưa ngựa về đúng vị trí trong chuồng hơn là việc di chuyển các quân cờ. Bởi đưa ngựa về đúng chuồng thường khó khăn hơn. Do đó khi có cơ hội bạn phải thực hiện ngay.

4/ Xuất nhiều ngựa cùng một lúc

Càng xuất nhiều ngựa ra cùng một lúc thì cơ hội giành chiến thắng của bạn càng cao. Không chỉ có thể cản chân các quân cờ của đối phương mà còn gia tăng cơ hội đá được ngựa đối phương ra khỏi bàn cờ. Đồng thời giúp an toàn cho các quân cờ di chuyển phía trước. Vì vậy, khi đổ được xúc xắc mặt Nhất hoặc Lục thì nên xuất thêm quân. Trừ trường hợp phải đưa ngựa về chuồng trước thì nên ưu tiên hơn.

5/ Đừng quên đá và cản chân ngựa đối phương

Khi chơi cờ cá ngựa, bạn nên tính toán các nước đi sao có thể cản được đường di chuyển của ngựa đối phương. Hoặc đá ngựa đối phương ra khỏi bàn cờ trong trò chơi cá ngựa thì càng tốt. Tốt nhất là nên để một quân cờ di chuyển phía sau quân cờ của đối phương. Khi có cơ hội thì đá quân cờ của đối phương ra khỏi bàn cờ.

Ngược lại nên tạo khoảng cách với ngựa phía sau của đối phương càng xa càng tốt (trên 6 nước) để tránh bị đá. Đây là cách chơi cá ngựa mà bạn nên chú ý.

Kết luận

Nhà cái W88 tặng ngay 90.000 VND miễn phí cho người mới vừa đăng ký thành công tài khoản chơi cá cược thể thao, đánh bài, xổ số,…Thắng sẽ được rút tiền mặt về tài khoản ngân hàng

Link đăng ký nhận 90.000 VND miễn phí

Link 1 – Link 2 – Link 3


【#7】Hướng Dẫn Cách Chơi Cờ Cá Ngựa Và Luật Chơi Chi Tiết Nhất

Hướng dẫn cách chơi Cờ Cá Ngựa chi tiết

Không như những trò chơi may rủi khác, Cờ Cá Ngựa còn phụ thuộc vào tính toán cũng như khả năng đọc tình huống của bạn. Muốn chiến thắng được trong các ván cược, bạn cần phải hiểu rõ luật chơi cũng như cách chơi Cờ Cá Ngựa để có các kế hoạch chiến lược rõ ràng.

Đến lượt ai thì người đó tung. Xúc xắc được tung vào một chiếc cốc đi kèm. Tuy nhiên, không được làm rơi xúc xắc ra ngoài vật đựng đó. Nếu ra ngoài, bạn lập tức mất lượt và kết quả việc gieo xúc xắc không được công nhận. Ai tung được kết quả là “nhất” (một) hoặc “lục” (sáu) khi chơi 1 xúc xắc; “nhất” (một) và “lục” (sáu), cặp nhất (một), cặp nhị (hai), cặp tam (ba), cặp tứ (bốn), cặp ngũ (năm), cặp lục (sáu) khi chơi 2 xúc xắc thì được đi thêm lượt nữa cho đến khi ngoài kết quả trên.

Cách chơi game Cờ Cá Ngựa online rất đơn giản và dễ chơi. Nếu bạn là “fan” trung thành với trò chơi này thì đừng bỏ qua hướng dẫn cách chơi Cờ Cá Ngựa Online của chúng tôi sau đây.

Xuất quân là quyền đưa ra một quân cờ để tham gia di chuyển trên bàn cờ. Để có được quyền này thì kết quả của việc tung xúc xắc phải là một trong các kết quả có thể đi tiếp ở trên mới được ra một quân và quân này phải đứng ngay vị trí bắt đầu.

Một khi trên bàn cờ đã có ít nhất một quân cờ của mình được tham gia di chuyển thì ta có thể căn cứ vào kết quả của việc tung xúc xắc để di chuyển nó. Kết quả bao nhiêu thì đó là số bước phải di chuyển. Trong trường hợp đổ được 6 nước sẽ được thêm lượt đổ.

Sau khi quân cờ đầu tiên của bạn đã được “xuất chuồng” thì trò chơi của bạn mới thực sự bắt đầu. Sau khi lắc xúc xắc để di chuyển thì trên quãng đường di chuyển bạn có thể gặp một số vấn đề sau:

Tức là trên đường đi mà quân của bạn bị cản bởi một quân phía trước. Bạn muốn vượt qua được thì kết quả lắc xúc xắc phải lớn hơn bước đi giữa 2 quân.

Trường hợp này xảy ra khi khi xúc xắc bạn tung được trùng hợp với vị trí của quân cờ khác. Khi đó bạn có 2 sự lựa chọn là đá quân cờ đối phương, hoặc bỏ để đi một quân cờ khác. Khi người chơi đá quân của đối phương thì họ sẽ được cộng số điểm bằng mức cược định sẵn của chủ phòng. Khi quân mình đứng sát quân đối phương thì được gọi là sát nút.

Khi quân cờ của bạn đã đến được chuồng sau khi di chuyển 1 vòng quanh bàn cờ thì bạn sẽ không bị đá nữa. Còn khi đến được cửa chuồng thì bạn vẫn có nguy cơ bị đưa về vạch xuất phát.

Theo quy định của luật chơi Cờ Cá Ngựa, những người chơi nào đã xếp đủ ngựa vào đúng bốn ô đầu 6, 5, 4 và 3 sẽ là người chiến thắng. Những người chơi tiếp theo sẽ phải hoàn thành lượt chơi của mình đến khi phân định được người về thứ 2 và thứ 3.

Chơi trò gì cũng cần có mẹo chơi để tăng tỷ lệ giành chiến thắng cho mình. Và cách chơi cờ cá ngựa cũng vậy!

Đối với bộ môn này, đôi lúc bạn sẽ phải cân nhắc việc bỏ đi quân khác. Việc này đôi khi là mạo hiểm tuy nhiên nó lại là một chiến thuật. Nhất là khi bạn phải cân nhắc giữa việc đá một quân cờ khác của đối phương hay cho ngựa của mình về chuồng ở vị trí số 1 và rơi vào tình huống “sập hầm”. Trong những trường hợp đó chọn bỏ lượt đi đó lại là thượng sách. Đây là một mẹo chơi cờ cá ngựa hay mà các bạn có thể tham khảo.

Để có thể xuất chuồng được thì lần lắc xúc xắc của bạn phải đạt được mặt 6 điểm. Nếu không được sẽ bị mất lượt và chuyển qua lượt cho người khác. Khi quân cờ của bạn vào đến chuồng, các bạn sẽ không lo bị đá nữa.

Tuy nhiên, khi quân cờ mới ở cửa chuồng bạn sẽ vẫn còn nguy cơ bị đối phương đưa về vị trí xuất phát. Đặt cược khi chơi Cờ Cá Ngựa: đây chính là một yếu tố sẽ tăng sự kịch tính cho các ván đấu.

Khi bạn bị đối phương đá thì người chơi sẽ được cộng tiền theo mức cược định sẵn của chủ phòng. Cuối những cuộc chơi sẽ tổng tiền lại để trả cho người chơi theo quy định.

  • Ra chuồng đầu tiên thắng Gold các người chơi khác chơi cùng với mức bằng 1 lần tiền cược.
  • Lên chuồng số 1 thua Gold cho các người chơi khác chơi cùng với mức 1 lần tiền cược.
  • Bàn đang chơi có user vào tham gia chơi được không?
  • Để game được thú vị và hấp dẫn hơn BQT sẽ cập nhật thêm các loại cược như
  • Cược đá ngựa: 2 con ngựa trong phạm vi có thể đá nhau (12 nút), người chơi có thể cược đá ngựa về chuồng với nhau hoặc cược với hệ thống
  • Trường hợp Đổ cặp xí ngầu giống nhau mà đá được ngựa người chơi sẽ ăn gấp đôi
  • Cược Chẵn/lẽ: trong mỗi lượt đỗ, người chơi có thể cược với hệ thống Chẵn/Lẻ theo tổng điểm 2 xí ngầu.
  • Lên chuồng số 6 đầu tiên ăn Gold các người chơi chơi cùng đã góp cược ban đầu ( mỗi nhà sẽ bỏ ra 1 lần tiền cược thì người chơi lên chuồng số 6 đầu tiên sẽ ăn tất cả số tiền cược này)
  • Thắng nhất ăn hết (tính các quân chưa lên chuồng) hoặc tính về nhất, về nhì, thắng về 3, về 4.
  • Thêm tính năng cảnh báo số lần vi phạm luật có mà không đá, cố tình để qua lượt… để người chơi có thể kiểm soát số lần vi phạm của mình. Nếu quá 3 lần thì sẽ tính thua chót ván đấu.

Qua bài viết này, chắc hẳn bạn đã nắm được những thông tin hữu ích về cách chơi Cờ Cá Ngựa và một số luật chơi bộ môn này. Mong rằng với những chia sẻ trên sẽ giúp bạn có cái nhìn tổng quát và chiến lược chơi đúng đắn hơn.


【#8】Sensory Evaluation Of Food: Principles And Practices, Second Edition (Food Science Text Series)

Food Science Text Series

The Food Science Text Series provides faculty with the leading teaching tools. The Editorial Board has outlined the most appropriate and complete content for each food science course in a typical food science program and has identified textbooks of the highest quality, written by the leading food science educators.

Series Editor Dennis R. Heldman Editorial Board David A. Golden, Ph.D., Professor of Food Microbiology, Department of Food Science and Technology, University of Tennessee Richard W. Hartel, Professor of Food Engineering, Department of Food Science, University of Wisconsin Hildegarde Heymann, Professor of Food Sensory Science, Department of Food Science and Technology, University of California-Davis Joseph H. Hotchkiss, Professor, Institute of Food Science and Institute for Comparative and Environmental Toxicology, and Chair, Food Science Department, Cornell University Michael G. Johnson, Ph.D., Professor of Food Safety and Microbiology, Department of Food Science, University of Arkansas Joseph Montecalvo, Jr., Professor, Department of Food Science and Nutrition, California Polytechnic and State University-San Luis Obispo S. Suzanne Nielsen, Professor and Chair, Department of Food Science, Purdue University Juan L. Silva, Professor, Department of Food Science, Nutrition and Health Promotion, Mississippi State University

For further volumes: http://www.springer.com/series/5999

Harry T. Lawless · Hildegarde Heymann

Sensory Evaluation of Food Principles and Practices Second Edition

123

Harry T. Lawless Department of Food Science Cornell University Stocking Hall, Room 106 14853 Ithaca NY, USA

ISSN 1572-0330 ISBN 978-1-4419-6487-8 e-ISBN 978-1-4419-6488-5 DOI 10.1007/978-1-4419-6488-5 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2010932599 © Springer Science+Business Media, LLC 2010 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expssion of opinion as to whether or not they are subject to proprietary rights. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)

Preface

vi

Preface

Preface

vii

Peryam, David Stevens, Herb Meiselman, Elaine Skinner, Howard Schutz, Howard Moskowitz, Rose Marie Pangborn, Beverley Kroll, W. Frank Shipe, Lawrence E. Marks, Joseph C. Stevens, Arye Dethmers, Barbara Klein, Ann Noble, Harold Hedrick, William C Stringer, Roger Boulton, Kay McMath, Joel van Wyk, and Roger Mitchell. Ithaca, New York Davis, California

Harry T. Lawless Hildegarde Heymann

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Introduction and Overview . . . . . . . . . . . . . . . . . . . 1.1.1 Definition . . . . . . . . . . . . . . . . . . . . . . . 1.1.2 Measurement . . . . . . . . . . . . . . . . . . . . . 1.2 Historical Landmarks and the Three Classes of Test Methods . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.1 Difference Testing . . . . . . . . . . . . . . . . . . . 1.2.2 Descriptive Analyses . . . . . . . . . . . . . . . . . 1.2.3 Affective Testing . . . . . . . . . . . . . . . . . . . 1.2.4 The Central Dogma-Analytic Versus Hedonic Tests . . . . . . . . . . . . . . . . . . . . . 1.3 Applications: Why Collect Sensory Data? . . . . . . . . . . . 1.3.1 Differences from Marketing Research Methods . . . 1.3.2 Differences from Traditional Product Grading Systems . . . . . . . . . . . . . . . . . . . 1.4 Summary and Conclusions . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Physiological and Psychological Foundations of Sensory Function . . . . . . . . . . . . . . . . . . . . . . . 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . 2.2 Classical Sensory Testing and Psychophysical Methods 2.2.1 Early Psychophysics . . . . . . . . . . . . . 2.2.2 The Classical Psychophysical Methods . . . . 2.2.3 Scaling and Magnitude Estimation . . . . . . 2.2.4 Critiques of Stevens . . . . . . . . . . . . . . 2.2.5 Empirical Versus Theory-Driven Functions . 2.2.6 Parallels of Psychophysics and Sensory Evaluation . . . . . . . . . . . . . . . . . . . 2.3 Anatomy and Physiology and Functions of Taste . . . . 2.3.1 Anatomy and Physiology . . . . . . . . . . . 2.3.2 Taste Perception: Qualities . . . . . . . . . . 2.3.3 Taste Perception: Adaptation and Mixture Interactions . . . . . . . . . . . . . . . . . . 2.3.4 Inpidual Differences and Taste Genetics . . 2.4 Anatomy and Physiology and Functions of Smell . . . . 2.4.1 Anatomy and Cellular Function . . . . . . . .

1 1 1 3 4 5 6 7 8 10 13 15 16 17

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

19 19 20 20 21 23 25 25

. . . .

. . . .

. . . .

. . . .

26 27 27 30

. . . .

. . . .

. . . .

. . . .

30 33 34 34 ix

x

Contents

2.4.2 2.4.3 2.4.4 2.4.5

Retronasal Smell . . . . . . . . . . . . . . Olfactory Sensitivity and Specific Anosmia Odor Qualities: Practical Systems . . . . . Functional Properties: Adaptation, Mixture Suppssion, and Release . . . . . . . . . . 2.5 Chemesthesis . . . . . . . . . . . . . . . . . . . . . . 2.5.1 Qualities of Chemesthetic Experience . . . 2.5.2 Physiological Mechanisms of Chemesthesis 2.5.3 Chemical “Heat” . . . . . . . . . . . . . . 2.5.4 Other Irritative Sensations and Chemical Cooling . . . . . . . . . . . . . . . . . . . 2.5.5 Astringency . . . . . . . . . . . . . . . . . 2.5.6 Metallic Taste . . . . . . . . . . . . . . . . 2.6 Multi-modal Sensory Interactions . . . . . . . . . . . 2.6.1 Taste and Odor Interactions . . . . . . . . . 2.6.2 Irritation and Flavor . . . . . . . . . . . . . 2.6.3 Color-Flavor Interactions . . . . . . . . . . 2.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Principles of Good Practice . . . . . . . . . . . . . 3.1 Introduction . . . . . . . . . . . . . . . . . 3.2 The Sensory Testing Environment . . . . . . 3.2.1 Evaluation Area . . . . . . . . . . 3.2.2 Climate Control . . . . . . . . . . 3.3 Test Protocol Considerations . . . . . . . . 3.3.1 Sample Serving Procedures . . . . 3.3.2 Sample Size . . . . . . . . . . . . 3.3.3 Sample Serving Temperatures . . 3.3.4 Serving Containers . . . . . . . . 3.3.5 Carriers . . . . . . . . . . . . . . 3.3.6 Palate Cleansing . . . . . . . . . . 3.3.7 Swallowing and Expectoration . . 3.3.8 Instructions to Panelists . . . . . . 3.3.9 Randomization and Blind Labeling 3.4 Experimental Design . . . . . . . . . . . . . 3.4.1 Designing a Study . . . . . . . . . 3.4.2 Design and Treatment Structures . 3.5 Panelist Considerations . . . . . . . . . . . 3.5.1 Incentives . . . . . . . . . . . . . 3.5.2 Use of Human Subjects . . . . . . 3.5.3 Panelist Recruitment . . . . . . . 3.5.4 Panelist Selection and Screening . 3.5.5 Training of Panelists . . . . . . . 3.5.6 Panelist Performance Assessment . 3.6 Tabulation and Analysis . . . . . . . . . . . 3.6.1 Data Entry Systems . . . . . . . . 3.7 Conclusion . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

36 37 38

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

39 41 41 42 43

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

44 45 46 47 47 49 49 50 50

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

57 57 58 59 62 63 63 63 64 64 65 65 66 66 66 66 66 69 72 72 73 74 74 75 75 75 75 76 76

Contents

xi

4 Discrimination Testing . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Discrimination Testing . . . . . . . . . . . . . . . . . . . . . . 4.2 Types of Discrimination Tests . . . . . . . . . . . . . . . . . . 4.2.1 Paired Comparison Tests . . . . . . . . . . . . . . . 4.2.2 Triangle Tests . . . . . . . . . . . . . . . . . . . . . 4.2.3 Duo-Trio Tests . . . . . . . . . . . . . . . . . . . . 4.2.4 n-Alternative Forced Choice (n-AFC) Methods . . . 4.2.5 A-Not-A tests . . . . . . . . . . . . . . . . . . . . . 4.2.6 Sorting Methods . . . . . . . . . . . . . . . . . . . . 4.2.7 The ABX Discrimination Task . . . . . . . . . . . . 4.2.8 Dual-Standard Test . . . . . . . . . . . . . . . . . . 4.3 Reputed Strengths and Weaknesses of Discrimination Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Data Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.1 Binomial Distributions and Tables . . . . . . . . . . 4.4.2 The Adjusted Chi-Square (χ2 ) Test . . . . . . . . . . 4.4.3 The Normal Distribution and the Z-Test on Proportion . . . . . . . . . . . . . . . . . . . . . 4.5 Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5.1 The Power of the Statistical Test . . . . . . . . . . . 4.5.2 Replications . . . . . . . . . . . . . . . . . . . . . . 4.5.3 Warm-Up Effects . . . . . . . . . . . . . . . . . . . 4.5.4 Common Mistakes Made in the Interptation of Discrimination Tests . . . . . . . . . . . . . . . . Appendix: A Simple Approach to Handling the A, Not-A, and Same/Different Tests . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

79 79 80 80 83 84 85 85 87 88 88

5 Similarity, Equivalence Testing, and Discrimination Theory . . . . 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Common Sense Approaches to Equivalence . . . . . . . . . . 5.3 Estimation of Sample Size and Test Power . . . . . . . . . . . 5.4 How Big of a Difference Is Important? Discriminator Theory . . . . . . . . . . . . . . . . . . . . . . 5.5 Tests for Significant Similarity . . . . . . . . . . . . . . . . . 5.6 The Two One-Sided Test Approach (TOST) and Interval Testing . . . . . . . . . . . . . . . . . . . . . . . 5.7 Claim Substantiation . . . . . . . . . . . . . . . . . . . . . . . 5.8 Models for Discrimination: Signal Detection Theory . . . . . . 5.8.1 The Problem . . . . . . . . . . . . . . . . . . . . . . 5.8.2 Experimental Setup . . . . . . . . . . . . . . . . . . 5.8.3 Assumptions and Theory . . . . . . . . . . . . . . . 5.8.4 An Example . . . . . . . . . . . . . . . . . . . . . . 5.8.5 A Connection to Paired Comparisons Results Through the ROC Curve . . . . . . . . . . . . . . . 5.9 Thurstonian Scaling . . . . . . . . . . . . . . . . . . . . . . . 5.9.1 The Theory and Formulae . . . . . . . . . . . . . . . 5.9.2 Extending Thurstone’s Model to Other Choice Tests . . . . . . . . . . . . . . . . . . . . . .

101 101 103 104

88 89 89 90 90 92 92 94 97 97 98 99

105 108 110 111 111 112 112 113 114 116 116 116 118

xii

Contents

5.10

Extensions of the Thurstonian Methods, R-Index . . . 5.10.1 Short Cut Signal Detection Methods . . . . 5.10.2 An Example . . . . . . . . . . . . . . . . . 5.11 Conclusions . . . . . . . . . . . . . . . . . . . . . . Appendix: Non-Central t-Test for Equivalence of Scaled Data References . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

119 119 120 120 122 122

6 Measurement of Sensory Thresholds . . . . . . . . . . . . . . . . . 6.1 Introduction: The Threshold Concept . . . . . . . . . . . . . . 6.2 Types of Thresholds: Definitions . . . . . . . . . . . . . . . . 6.3 Practical Methods: Ascending Forced Choice . . . . . . . . . . 6.4 Suggested Method for Taste/Odor/Flavor Detection Thresholds . . . . . . . . . . . . . . . . . . . . . . 6.4.1 Ascending Forced-Choice Method of Limits . . . . . 6.4.2 Purpose of the Test . . . . . . . . . . . . . . . . . . 6.4.3 Preliminary Steps . . . . . . . . . . . . . . . . . . . 6.4.4 Procedure . . . . . . . . . . . . . . . . . . . . . . . 6.4.5 Data Analysis . . . . . . . . . . . . . . . . . . . . . 6.4.6 Alternative Graphical Solution . . . . . . . . . . . . 6.4.7 Procedural Choices . . . . . . . . . . . . . . . . . . 6.5 Case Study/Worked Example . . . . . . . . . . . . . . . . . . 6.6 Other Forced Choice Methods . . . . . . . . . . . . . . . . . . 6.7 Probit Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 6.8 Sensory Adaptation, Sequential Effects, and Variability . . . . 6.9 Alternative Methods: Rated Difference, Adaptive Procedures, Scaling . . . . . . . . . . . . . . . . . . . . . . . 6.9.1 Rated Difference from Control . . . . . . . . . . . . 6.9.2 Adaptive Procedures . . . . . . . . . . . . . . . . . 6.9.3 Scaling as an Alternative Measure of Sensitivity . . . 6.10 Dilution to Threshold Measures . . . . . . . . . . . . . . . . . 6.10.1 Odor Units and Gas-Chromatography Olfactometry (GCO) . . . . . . . . . . . . . . . . . 6.10.2 Scoville Units . . . . . . . . . . . . . . . . . . . . . 6.11 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix: MTBE Threshold Data for Worked Example . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

125 125 127 128

140 142 142 143 145

7 Scaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Some Theory . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 Common Methods of Scaling . . . . . . . . . . . . . . . . 7.3.1 Category Scales . . . . . . . . . . . . . . . . . . 7.3.2 Line Scaling . . . . . . . . . . . . . . . . . . . . 7.3.3 Magnitude Estimation . . . . . . . . . . . . . . . 7.4 Recommended Practice and Practical Guidelines . . . . . . 7.4.1 Rule 1: Provide Sufficient Alternatives . . . . . . 7.4.2 Rule 2: The Attribute Must Be Understood . . . . 7.4.3 Rule 3: The Anchor Words Should Make Sense . 7.4.4 To Calibrate or Not to Calibrate . . . . . . . . . . 7.4.5 A Warning: Grading and Scoring are Not Scaling

149 149 151 152 152 155 156 158 159 159 159 159 160

. . . . . . . . . . . . .

. . . . . . . . . . . . .

129 129 129 130 131 131 131 133 133 134 136 136 137 137 138 140 140

Contents

xiii

7.5

Variations-Other Scaling Techniques . . . . . . . . . . . . . 7.5.1 Cross-Modal Matches and Variations on Magnitude Estimation . . . . . . . . . . . . . . . . . 7.5.2 Category-Ratio (Labeled Magnitude) Scales . . . . . 7.5.3 Adjustable Rating Techniques: Relative Scaling . . . 7.5.4 Ranking . . . . . . . . . . . . . . . . . . . . . . . . 7.5.5 Indirect Scales . . . . . . . . . . . . . . . . . . . . . 7.6 Comparing Methods: What is a Good Scale? . . . . . . . . . . 7.7 Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.7.1 “Do People Make Relative Judgments” Should They See Their Previous Ratings? . . . . . . 7.7.2 Should Category Rating Scales Be Assigned Integer Numbers in Data Tabulation? Are They Interval Scales? . . . . . . . . . . . . . . . . . 7.7.3 Is Magnitude Estimation a Ratio Scale or Simply a Scale with Ratio Instructions? . . . . . . . 7.7.4 What is a “Valid” Scale? . . . . . . . . . . . . . . . 7.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix 1: Derivation of Thurstonian-Scale Values for the 9-Point Scale . . . . . . . . . . . . . . . . . . . . . . . Appendix 2: Construction of Labeled Magnitude Scales . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Time-Intensity Methods . . . . . . . . . . . . . . . . . . . . . . 8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 8.2 A Brief History . . . . . . . . . . . . . . . . . . . . . . . 8.3 Variations on the Method . . . . . . . . . . . . . . . . . . 8.3.1 Discrete or Discontinuous Sampling . . . . . . . 8.3.2 “Continuous” Tracking . . . . . . . . . . . . . . 8.3.3 Temporal Dominance Techniques . . . . . . . . . 8.4 Recommended Procedures . . . . . . . . . . . . . . . . . . 8.4.1 Steps in Conducting a Time-intensity Study . . . 8.4.2 Procedures . . . . . . . . . . . . . . . . . . . . . 8.4.3 Recommended Analysis . . . . . . . . . . . . . . 8.5 Data Analysis Options . . . . . . . . . . . . . . . . . . . . 8.5.1 General Approaches . . . . . . . . . . . . . . . . 8.5.2 Methods to Construct or Describe Average Curves 8.5.3 Case Study: Simple Geometric Description . . . 8.5.4 Analysis by Principal Components . . . . . . . . 8.6 Examples and Applications . . . . . . . . . . . . . . . . . 8.6.1 Taste and Flavor Sensation Tracking . . . . . . . 8.6.2 Trigeminal and Chemical/Tactile Sensations . . . 8.6.3 Taste and Odor Adaptation . . . . . . . . . . . . 8.6.4 Texture and Phase Change . . . . . . . . . . . . 8.6.5 Flavor Release . . . . . . . . . . . . . . . . . . . 8.6.6 Temporal Aspects of Hedonics . . . . . . . . . . 8.7 Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

160 160 162 164 165 166 167 168 168

169 169 169 170 171 172 174 179 179 180 182 182 183 184 185 185 186 186 187 187 188 189 192 193 193 194 194 195 195 196 197 198 198

xiv

Contents

9 Context Effects and Biases in Sensory Judgment . . . . . . . 9.1 Introduction: The Relative Nature of Human Judgment . 9.2 Simple Contrast Effects . . . . . . . . . . . . . . . . . 9.2.1 A Little Theory: Adaptation Level . . . . . . 9.2.2 Intensity Shifts . . . . . . . . . . . . . . . . 9.2.3 Quality Shifts . . . . . . . . . . . . . . . . . 9.2.4 Hedonic Shifts . . . . . . . . . . . . . . . . . 9.2.5 Explanations for Contrast . . . . . . . . . . . 9.3 Range and Frequency Effects . . . . . . . . . . . . . . 9.3.1 A Little More Theory: Parducci’s Range and Frequency Principles . . . . . . . . . . . 9.3.2 Range Effects . . . . . . . . . . . . . . . . . 9.3.3 Frequency Effects . . . . . . . . . . . . . . . 9.4 Biases . . . . . . . . . . . . . . . . . . . . . . . . . . 9.4.1 Idiosyncratic Scale Usage and Number Bias . 9.4.2 Poulton’s Classifications . . . . . . . . . . . 9.4.3 Response Range Effects . . . . . . . . . . . . 9.4.4 The Centering Bias . . . . . . . . . . . . . . 9.5 Response Correlation and Response Restriction . . . . 9.5.1 Response Correlation . . . . . . . . . . . . . 9.5.2 “Dumping” Effects: Inflation Due to Response Restriction in Profiling . . . . . 9.5.3 Over-Partitioning . . . . . . . . . . . . . . . 9.6 Classical Psychological Errors and Other Biases . . . . 9.6.1 Errors in Structured Sequences: Anticipation and Habituation . . . . . . . . . . . . . . . . 9.6.2 The Stimulus Error . . . . . . . . . . . . . . 9.6.3 Positional or Order Bias . . . . . . . . . . . . 9.7 Antidotes . . . . . . . . . . . . . . . . . . . . . . . . . 9.7.1 Avoid or Minimize . . . . . . . . . . . . . . 9.7.2 Randomization and Counterbalancing . . . . 9.7.3 Stabilization and Calibration . . . . . . . . . 9.7.4 Interptation . . . . . . . . . . . . . . . . . 9.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Descriptive Analysis . . . . . . . . . . . . . . . . . . . . . 10.1 Introduction . . . . . . . . . . . . . . . . . . . . . 10.2 Uses of Descriptive Analyses . . . . . . . . . . . . 10.3 Language and Descriptive Analysis . . . . . . . . . 10.4 Descriptive Analysis Techniques . . . . . . . . . . R 10.4.1 Flavor Profile . . . . . . . . . . . . . . R 10.4.2 Quantitative Descriptive Analysis . . . R . . . . . . . . . . . . . 10.4.3 Texture Profile R 10.4.4 Sensory Spectrum . . . . . . . . . . . . 10.5 Generic Descriptive Analysis . . . . . . . . . . . . 10.5.1 How to Do Descriptive Analysis in Three Easy Steps . . . . . . . . . . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

203 203 206 206 207 207 208 209 210

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

210 210 211 212 212 213 214 215 216 216

. . . . . . . . . . . .

217 218 218

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

218 219 219 219 219 220 221 222 222 223

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

227 227 228 228 231 231 234 237 238 240

. . . . . .

240

. . . . . . . . . .

. . . . . . . . . .

Contents

xv

10.5.2

Studies Comparing Different Conventional Descriptive Analysis Techniques . . . . . . 10.6 Variations on the Theme . . . . . . . . . . . . . . . . 10.6.1 Using Attribute Citation Frequencies Instead of Attribute Intensities . . . . . . . . . . . 10.6.2 Deviation from Reference Method . . . . . 10.6.3 Intensity Variation Descriptive Method . . . 10.6.4 Combination of Descriptive Analysis and Time-Related Intensity Methods . . . . 10.6.5 Free Choice Profiling . . . . . . . . . . . . 10.6.6 Flash Profiling . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . .

246 247

. . . . . . . . . . . . . . .

247 248 249

. . . .

. . . .

. . . .

. . . .

. . . .

249 249 252 253

Texture Evaluation . . . . . . . . . . . . . . . . . . . . . . . 11.1 Texture Defined . . . . . . . . . . . . . . . . . . . . . 11.2 Visual, Auditory, and Tactile Texture . . . . . . . . . . 11.2.1 Visual Texture . . . . . . . . . . . . . . . . . 11.2.2 Auditory Texture . . . . . . . . . . . . . . . 11.2.3 Tactile Texture . . . . . . . . . . . . . . . . . 11.2.4 Tactile Hand Feel . . . . . . . . . . . . . . . 11.3 Sensory Texture Measurements . . . . . . . . . . . . . 11.3.1 Texture Profile Method . . . . . . . . . . . . 11.3.2 Other Sensory Texture Evaluation Techniques 11.3.3 Instrumental Texture Measurements and Sensory Correlations . . . . . . . . . . . 11.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

259 259 262 262 262 264 268 270 270 272

. . . . . . . . . . . .

274 276 276

12

Color and Appearance . . . . . . . . . . . . . . . . . . 12.1 Color and Appearance . . . . . . . . . . . . . . . 12.2 What Is Color? . . . . . . . . . . . . . . . . . . . 12.3 Vision . . . . . . . . . . . . . . . . . . . . . . . 12.3.1 Normal Human Color Vision Variations 12.3.2 Human Color Blindness . . . . . . . . . 12.4 Measurement of Appearance and Color Attributes 12.4.1 Appearance . . . . . . . . . . . . . . . 12.4.2 Visual Color Measurement . . . . . . . 12.5 Instrumental Color Measurement . . . . . . . . . 12.5.1 Munsell Color Solid . . . . . . . . . . . 12.5.2 Mathematical Color Systems . . . . . . 12.6 Conclusions . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

283 283 284 285 286 286 286 286 289 293 293 294 299 299

13

Preference Testing . . . . . . . . . . . . . . . . . . 13.1 Introduction-Consumer Sensory Evaluation 13.2 Preference Tests: Overview . . . . . . . . . 13.2.1 The Basic Comparison . . . . . . 13.2.2 Variations . . . . . . . . . . . . . 13.2.3 Some Cautions . . . . . . . . . . 13.3 Simple Paired Preference Testing . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

303 303 305 305 305 306 306

11

. . . . . . .

. . . . . . .

. . . . . . .

xvi

Contents

13.3.1 Recommended Procedure . . . . . . . . . . . . . . . 13.3.2 Statistical Basis . . . . . . . . . . . . . . . . . . . . 13.3.3 Worked Example . . . . . . . . . . . . . . . . . . . 13.3.4 Useful Statistical Approximations . . . . . . . . . . 13.3.5 The Special Case of Equivalence Testing . . . . . . . 13.4 Non-forced Preference . . . . . . . . . . . . . . . . . . . . . . 13.5 Replicated Preference Tests . . . . . . . . . . . . . . . . . . . 13.6 Replicated Non-forced Preference . . . . . . . . . . . . . . . . 13.7 Other Related Methods . . . . . . . . . . . . . . . . . . . . . 13.7.1 Ranking . . . . . . . . . . . . . . . . . . . . . . . . 13.7.2 Analysis of Ranked Data . . . . . . . . . . . . . . . 13.7.3 Best-Worst Scaling . . . . . . . . . . . . . . . . . . 13.7.4 Rated Degree of Preference and Other Options . . . . 13.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix 1: Worked Example of the Ferris k-Visit Repeated Preference Test Including the No-Preference Option . . . . . . Appendix 2: The “Placebo” Preference Test . . . . . . . . . . . . . . . Appendix 3: Worked Example of Multinomial Approach to Analyzing Data with the No-Preference Option . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Acceptance Testing . . . . . . . . . . . . . . . . . . . . . . 14.1 Introduction: Scaled Liking Versus Choice . . . . . . 14.2 Hedonic Scaling: Quantification of Acceptability . . . 14.3 Recommended Procedure . . . . . . . . . . . . . . . 14.3.1 Steps . . . . . . . . . . . . . . . . . . . . . 14.3.2 Analysis . . . . . . . . . . . . . . . . . . . 14.3.3 Replication . . . . . . . . . . . . . . . . . 14.4 Other Acceptance Scales . . . . . . . . . . . . . . . . 14.4.1 Line Scales . . . . . . . . . . . . . . . . . 14.4.2 Magnitude Estimation . . . . . . . . . . . . 14.4.3 Labeled Magnitude Scales . . . . . . . . . 14.4.4 Pictorial Scales and Testing with Children . 14.4.5 Adjustable Scales . . . . . . . . . . . . . . 14.5 Just-About-Right Scales . . . . . . . . . . . . . . . . 14.5.1 Description . . . . . . . . . . . . . . . . . 14.5.2 Limitations . . . . . . . . . . . . . . . . . 14.5.3 Variations on Relative-to-Ideal Scaling . . . 14.5.4 Analysis of JAR Data . . . . . . . . . . . . 14.5.5 Penalty Analysis or “Mean Drop” . . . . . 14.5.6 Other Problems and Issues with JAR Scales 14.6 Behavioral and Context-Related Approaches . . . . . 14.6.1 Food Action Rating Scale (FACT) . . . . . 14.6.2 Appropriateness Scales . . . . . . . . . . . 14.6.3 Acceptor Set Size . . . . . . . . . . . . . . 14.6.4 Barter Scales . . . . . . . . . . . . . . . . 14.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . .

306 307 308 309 310 311 313 313 315 315 316 317 318 320 320 321 322 323 325 325 326 327 327 328 328 328 328 330 331 332 333 334 334 335 336 336 339 340 340 341 341 342 343 343 344

Contents

xvii

15

16

Consumer Field Tests and Questionnaire Design . . . . . . . . . . . 15.1 Sensory Testing Versus Concept Testing . . . . . . . . . . . . 15.2 Testing Scenarios: Central Location, Home Use . . . . . . . . 15.2.1 Purpose of the Tests . . . . . . . . . . . . . . . . . . 15.2.2 Consumer Models . . . . . . . . . . . . . . . . . . . 15.2.3 Central Location Tests . . . . . . . . . . . . . . . . 15.2.4 Home Use Tests (HUT) . . . . . . . . . . . . . . . . 15.3 Practical Matters in Conducting Consumer Field Tests . . . . . 15.3.1 Tasks and Test Design . . . . . . . . . . . . . . . . . 15.3.2 Sample Size and Stratification . . . . . . . . . . . . 15.3.3 Test Designs . . . . . . . . . . . . . . . . . . . . . . 15.4 Interacting with Field Services . . . . . . . . . . . . . . . . . 15.4.1 Choosing Agencies, Communication, and Test Specifications . . . . . . . . . . . . . . . . 15.4.2 Incidence, Cost, and Recruitment . . . . . . . . . . . 15.4.3 Some Tips: Do’s and Don’ts . . . . . . . . . . . . . 15.4.4 Steps in Testing with Research Suppliers . . . . . . . 15.5 Questionnaire Design . . . . . . . . . . . . . . . . . . . . . . 15.5.1 Types of Interviews . . . . . . . . . . . . . . . . . . 15.5.2 Questionnaire Flow: Order of Questions . . . . . . . 15.5.3 Interviewing . . . . . . . . . . . . . . . . . . . . . . 15.6 Rules of Thumb for Constructing Questions . . . . . . . . . . 15.6.1 General Principles . . . . . . . . . . . . . . . . . . . 15.6.2 Brevity . . . . . . . . . . . . . . . . . . . . . . . . . 15.6.3 Use Plain Language . . . . . . . . . . . . . . . . . . 15.6.4 Accessibility of the Information . . . . . . . . . . . 15.6.5 Avoid Vague Questions . . . . . . . . . . . . . . . . 15.6.6 Check for Overlap and Completeness . . . . . . . . . 15.6.7 Do Not Lead the Respondent . . . . . . . . . . . . . 15.6.8 Avoid Ambiguity and Double Questions . . . . . . . 15.6.9 Be Careful in Wording: Present Both Alternatives . . 15.6.10 Beware of Halos and Horns . . . . . . . . . . . . . . 15.6.11 Pre-test . . . . . . . . . . . . . . . . . . . . . . . . 15.7 Other Useful Questions: Satisfaction, Agreement, and Open-Ended Questions . . . . . . . . . . . . . . . . . . . 15.7.1 Satisfaction . . . . . . . . . . . . . . . . . . . . . . 15.7.2 Likert (Agree-Disagree) Scales . . . . . . . . . . . . 15.7.3 Open-Ended Questions . . . . . . . . . . . . . . . . 15.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix 1: Sample Test Specification Sheet . . . . . . . . . . . . . . Appendix 2: Sample Screening Questionnaire . . . . . . . . . . . . . . Appendix 3: Sample Product Questionnaire . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Qualitative Consumer Research Methods . . . . . . . 16.1 Introduction . . . . . . . . . . . . . . . . . . . 16.1.1 Resources, Definitions, and Objectives 16.1.2 Styles of Qualitative Research . . . . 16.1.3 Other Qualitative Techniques . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

349 349 351 351 352 353 354 355 355 355 356 358 358 359 360 360 362 362 362 363 364 364 364 364 365 365 365 365 366 366 366 366 367 367 367 367 368 370 371 374 378 379 380 380 380 382

xviii

Contents

16.2

Characteristics of Focus Groups . . . . . . . . . . . . . . . . . 383 16.2.1 Advantages . . . . . . . . . . . . . . . . . . . . . . 383 16.2.2 Key Requirements . . . . . . . . . . . . . . . . . . . 384 16.2.3 Reliability and Validity . . . . . . . . . . . . . . . . 384 16.3 Using Focus Groups in Sensory Evaluation . . . . . . . . . . . 385 16.4 Examples, Case Studies . . . . . . . . . . . . . . . . . . . . . 386 16.4.1 Case Study 1: Qualitative Research Before Conjoint Measurement in New Product Development 387 16.4.2 Case Study 2: Nutritional and Health Beliefs About Salt 387 16.5 Conducting Focus Group Studies . . . . . . . . . . . . . . . . 388 16.5.1 A Quick Overview . . . . . . . . . . . . . . . . . . 388 16.5.2 A Key Requirement: Developing Good Questions . . 389 16.5.3 The Discussion Guide and Phases of the Group Interview . . . . . . . . . . . . . . . . 390 16.5.4 Participant Requirements, Timing, Recording . . . . 391 16.6 Issues in Moderating . . . . . . . . . . . . . . . . . . . . . . . 392 16.6.1 Moderating Skills . . . . . . . . . . . . . . . . . . . 392 16.6.2 Basic Principles: Nondirection, Full Participation, and Coverage of Issues . . . . . . . . . 393 16.6.3 Assistant Moderators and Co-moderators . . . . . . . 394 16.6.4 Debriefing: Avoiding Selective Listening and Premature Conclusions . . . . . . . . . . . . . . 395 16.7 Analysis and Reporting . . . . . . . . . . . . . . . . . . . . . 395 16.7.1 General Principles . . . . . . . . . . . . . . . . . . . 395 16.7.2 Suggested Method (“Sorting/Clustering Approach”), also Called Classical Transcript Analysis . . . . . . . . . . . . . . . . . . . . . . . . 396 16.7.3 Report Format . . . . . . . . . . . . . . . . . . . . . 397 16.8 Alternative Procedures and Variations of the Group Interview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398 16.8.1 Groups of Children, Telephone Interviews, Internet-Based Groups . . . . . . . . . . . . . . . . 398 16.8.2 Alternatives to Traditional Questioning . . . . . . . . 399 16.9 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . 400 Appendix: Sample Report Group Report . . . . . . . . . . . . . . . . . 402 Boil-in-bag Pasta Project Followup Groups . . . . . . . . . . . . . . . 402 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404 17 Quality Control and Shelf-Life (Stability) Testing . 17.1 Introduction: Objectives and Challenges . . . 17.2 A Quick Look at Traditional Quality Control . 17.3 Methods for Sensory QC . . . . . . . . . . . 17.3.1 Cuttings: A Bad Example . . . . . . 17.3.2 In-Out (Pass/Fail) System . . . . . 17.3.3 Difference from Control Ratings . . 17.3.4 Quality Ratings with Diagnostics . . 17.3.5 Descriptive Analysis . . . . . . . . 17.3.6 A Hybrid Approach: Quality Ratings with Diagnostics . . . . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

407 408 409 409 409 410 411 412 413

. . . . . . . . .

414

Contents

xix

17.3.7 The Multiple Standards Difference Test . . . . . . . Recommended Procedure: Difference Scoring with Key Attribute Scales . . . . . . . . . . . . . . . . . . . . . . . 17.5 The Importance of Good Practice . . . . . . . . . . . . . . . . 17.6 Historical Footnote: Expert Judges and Quality Scoring . . . . 17.6.1 Standardized Commodities . . . . . . . . . . . . . . 17.6.2 Example 1: Dairy Product Judging . . . . . . . . . . 17.6.3 Example 2: Wine Scoring . . . . . . . . . . . . . . . 17.7 Program Requirements and Program Development . . . . . . . 17.7.1 Desired Features of a Sensory QC System . . . . . . 17.7.2 Program Development and Management Issues . . . 17.7.3 The Problem of Low Incidence . . . . . . . . . . . . 17.8 Shelf-Life Testing . . . . . . . . . . . . . . . . . . . . . . . . 17.8.1 Basic Considerations . . . . . . . . . . . . . . . . . 17.8.2 Cutoff Point . . . . . . . . . . . . . . . . . . . . . . 17.8.3 Test Designs . . . . . . . . . . . . . . . . . . . . . . 17.8.4 Survival Analysis and Hazard Functions . . . . . . . 17.8.5 Accelerated Storage . . . . . . . . . . . . . . . . . . 17.9 Summary and Conclusions . . . . . . . . . . . . . . . . . . . Appendix 1: Sample Screening Tests for Sensory Quality Judges . . . . Appendix 2: Survival/Failure Estimates from a Series of Batches with Known Failure Times . . . . . . . . . . . . . . Appendix 3: Arrhenius Equation and Q10 Modeling . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

414

17.4

18

19

415 417 419 419 419 420 422 422 423 424 424 424 426 426 427 428 428 429 429 430 431

Data Relationships and Multivariate Applications . . . . . . . . . . 18.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.2 Overview of Multivariate Statistical Techniques . . . . . . . . 18.2.1 Principal Component Analysis . . . . . . . . . . . . 18.2.2 Multivariate Analysis of Variance . . . . . . . . . . . 18.2.3 Discriminant Analysis (Also Known as Canonical Variate Analysis) . . . . . . . . . . . . 18.2.4 Generalized Procrustes Analysis . . . . . . . . . . . 18.3 Relating Consumer and Descriptive Data Through Preference Mapping . . . . . . . . . . . . . . . . . . . . . . . 18.3.1 Internal Preference Mapping . . . . . . . . . . . . . 18.3.2 External Preference Mapping . . . . . . . . . . . . . 18.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

440 442 442 445 446

Strategic Research . . . . . . . . . . . . . . . 19.1 Introduction . . . . . . . . . . . . . . . 19.1.1 Avenues for Strategic Research 19.1.2 Consumer Contact . . . . . . . 19.2 Competitive Surveillance . . . . . . . . 19.2.1 The Category Review . . . . . 19.2.2 Perceptual Mapping . . . . . . 19.2.3 Multivariate Methods: PCA . . 19.2.4 Multi-dimensional Scaling . .

451 451 451 453 453 453 455 456 458

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

433 433 434 434 437 438 439

xx

Contents

19.2.5

Cost-Efficient Methods for Data Collection: Sorting . . . . . . . . . . . . . . . . . . . . . . 19.2.6 Vector Projection . . . . . . . . . . . . . . . . 19.2.7 Cost-Efficient Methods for Data Collection: Projective Mapping, aka Napping . . . . . . . . 19.3 Attribute Identification and Classification . . . . . . . . . 19.3.1 Drivers of Liking . . . . . . . . . . . . . . . . 19.3.2 The Kano Model . . . . . . . . . . . . . . . . 19.4 Preference Mapping Revisited . . . . . . . . . . . . . . . 19.4.1 Types of Preference Maps . . . . . . . . . . . . 19.4.2 Preference Models: Vectors Versus Ideal Points 19.5 Consumer Segmentation . . . . . . . . . . . . . . . . . . 19.6 Claim Substantiation Revisited . . . . . . . . . . . . . . 19.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . 19.7.1 Blind Testing, New Coke, and the Vienna Philharmonic . . . . . . . . . . . . . . . . . . 19.7.2 The Sensory Contribution . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . .

459 460

. . . . . . . . . .

. . . . . . . . . .

461 462 462 463 464 464 464 465 467 468

. . . . . . . . .

468 469 469

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

473 473 474 475 476 478 478 478 479 480 482 482

. . . . .

. . . . .

. . . . .

484 485 485 486 487

. . . . . . . .

. . . . . . . .

. . . . . . . .

489 489 490 493 493 494 494 495

. . . . .

496

Appendix A Basic Statistical Concepts for Sensory Evaluation . . A.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . A.2 Basic Statistical Concepts . . . . . . . . . . . . . . . . . A.2.1 Data Description . . . . . . . . . . . . . . . . A.2.2 Population Statistics . . . . . . . . . . . . . . . A.3 Hypothesis Testing and Statistical Inference . . . . . . . A.3.1 The Confidence Interval . . . . . . . . . . . . . A.3.2 Hypothesis Testing . . . . . . . . . . . . . . . A.3.3 A Worked Example . . . . . . . . . . . . . . . A.3.4 A Few More Important Concepts . . . . . . . . A.3.5 Decision Errors . . . . . . . . . . . . . . . . . A.4 Variations of the t-Test . . . . . . . . . . . . . . . . . . . A.4.1 The Sensitivity of the Dependent t-Test for Sensory Data . . . . . . . . . . . . . . . . . . A.5 Summary: Statistical Hypothesis Testing . . . . . . . . . A.6 Postscript: What p-Values Signify and What They Do Not A.7 Statistical Glossary . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix B Nonparametric and Binomial-Based Statistical Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . B.1 Introduction to Nonparametric Tests . . . . . . . . . B.2 Binomial-Based Tests on Proportions . . . . . . . . . B.3 Chi-Square . . . . . . . . . . . . . . . . . . . . . . . B.3.1 A Measure of Relatedness of Two Variables B.3.2 Calculations . . . . . . . . . . . . . . . . . B.3.3 Related Samples: The McNemar Test . . . . B.3.4 The Stuart-Maxwell Test . . . . . . . . . . B.3.5 Beta-Binomial, Chance-Corrected Beta-Binomial, and Dirichlet Multinomial Analyses . . . . . . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . . . .

Contents

xxi

B.4

Useful Rank Order Tests . . . . . . . . . . . . . . . . B.4.1 The Sign Test . . . . . . . . . . . . . . . . B.4.2 The Mann-Whitney U-Test . . . . . . . . . B.4.3 Ranked Data with More Than Two Samples, Friedman and Kramer Tests . . . . . . . . . B.4.4 Rank Order Correlation . . . . . . . . . . . B.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . B.6 Postscript . . . . . . . . . . . . . . . . . . . . . . . . B.6.1 Proof showing equivalence of binomial approximation Z-test and χ 2 test for difference of proportions . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

499 499 500

. . . .

. . . .

501 502 503 503

. . . . . . . . . .

503 504

. . . .

Appendix C Analysis of Variance . . . . . . . . . . . . . . . . . . . C.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . C.1.1 Overview . . . . . . . . . . . . . . . . . . . . C.1.2 Basic Analysis of Variance . . . . . . . . . . . C.1.3 Rationale . . . . . . . . . . . . . . . . . . . . C.1.4 Calculations . . . . . . . . . . . . . . . . . . . C.1.5 A Worked Example . . . . . . . . . . . . . . . C.2 Analysis of Variance from Complete Block Designs . . . C.2.1 Concepts and Partitioning Panelist Variance from Error . . . . . . . . . . . . . . . . . . . . C.2.2 The Value of Using Panelists As Their Own Controls . . . . . . . . . . . . . C.3 Planned Comparisons Between Means Following ANOVA C.4 Multiple Factor Analysis of Variance . . . . . . . . . . . C.4.1 An Example . . . . . . . . . . . . . . . . . . . C.4.2 Concept: A Linear Model . . . . . . . . . . . . C.4.3 A Note About Interactions . . . . . . . . . . . C.5 Panelist by Product by Replicate Designs . . . . . . . . . C.6 Issues and Concerns . . . . . . . . . . . . . . . . . . . . C.6.1 Sensory Panelists: Fixed or Random Effects? . C.6.2 A Note on Blocking . . . . . . . . . . . . . . . C.6.3 Split-Plot or Between-Groups (Nested) Designs C.6.4 Statistical Assumptions and the Repeated Measures ANOVA . . . . . . . . . . . . . . . C.6.5 Other Options . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix D Correlation, Regression, and Measures of Association D.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . D.2 Correlation . . . . . . . . . . . . . . . . . . . . . . . . . D.2.1 Pearson’s Correlation Coefficient Example . . . D.2.2 Coefficient of Determination . . . . . . . . . . D.3 Linear Regression . . . . . . . . . . . . . . . . . . . . . D.3.1 Analysis of Variance . . . . . . . . . . . . . . D.3.2 Analysis of Variance for Linear Regression . . D.3.3 Prediction of the Regression Line . . . . . . . . D.3.4 Linear Regression Example . . . . . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . . . . . .

507 507 507 508 508 509 509 510

. . .

510

. . . . . . . . . . .

. . . . . . . . . . .

512 513 514 514 515 516 516 519 519 520 520

. . . . . . . . .

521 522 522

. . . . . . . . . .

525 525 527 528 529 529 530 530 530 531

. . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

xxii

Contents

D.4 D.5

Multiple Linear Regression . . . . . . . . . . . . . . Other Measures of Association . . . . . . . . . . . . D.5.1 Spearman Rank Correlation . . . . . . . . . D.5.2 Spearman Correlation Coefficient Example D.5.3 Cramér’s V Measure . . . . . . . . . . . . . D.5.4 Cramér Coefficient Example . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

531 531 531 532 532 533 533

Appendix E Statistical Power and Test Sensitivity . . . . . . . . E.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . E.2 Factors Affecting the Power of Statistical Tests . . . . . E.2.1 Sample Size and Alpha Level . . . . . . . . . E.2.2 Effect Size . . . . . . . . . . . . . . . . . . . E.2.3 How Alpha, Beta, Effect Size, and N Interact E.3 Worked Examples . . . . . . . . . . . . . . . . . . . . E.3.1 The t-Test . . . . . . . . . . . . . . . . . . . E.3.2 An Equivalence Issue with Scaled Data . . . E.3.3 Sample Size for a Difference Test . . . . . . . E.4 Power in Simple Difference and Preference Tests . . . . E.5 Summary and Conclusions . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

535 535 537 537 538 539 541 541 542 544 545 548 549

Appendix F Statistical Tables . . . . . . . . . . . . . . . . . . . . . . . . . Table F.A Cumulative probabilities of the standard normal distribution. Entry area 1-α under the standard normal curve from −∞ to z(1-α) . . . . . . . . . . . . . Table F.B Table of critical values for the t-distribution . . . . . . . . Table F.C Table of critical values of the chi-square (χ 2 ) distribution . . . . . . . . . . . . . . . . . . . . . . . . . Table F.D1 Critical values of the F-distribution at α = 0.05 . . . . . . Table F.D2 Critical values of the F-distribution at α = 0.01 . . . . . . Table F.E Critical values of U for a one-tailed alpha at 0.025 or a two-tailed alpha at 0.05 . . . . . . . . . . . . . . . . Table F.F1 Table of critical values of ρ (Spearman Rank correlation coefficient) . . . . . . . . . . . . . . . . . . . Table F.F2 Table of critical values of r (Pearson’s correlation coefficient) . . . . . . . . . . . . . . . . . . . . . . . . . Table F.G Critical values for Duncan’s multiple range test (p, df, α = 0.05) . . . . . . . . . . . . . . . . . . . . . . . Table F.H1 Critical values of the triangle test for similarity (maximum number correct as a function of the number of observations (N), beta, and proportion discriminating) . . . . . . . . . . . . . . . . . . . . . . . Table F.H2 Critical values of the duo-trio and paired comparison tests for similarity (maximum number correct as a function of the number of observations (N), beta, and proportion discriminating) . . . . . . . . . . Table F.I Table of probabilities for values as small as observed values of x associated with the binomial test (p=0.50) . . . . . . . . . . . . . . . . . . . . . . . .

551

552 553 554 555 556 556 557 558 559

560

561

562

Contents

xxiii

Table F.J

Critical values for the differences between rank sums (α = 0.05) . . . . . . . . . . . . . . . . . . . . . . . Critical values of the beta binomial distribution . . . . . . Minimum numbers of correct judgments to establish significance at probability levels of 5 and 1% for paired difference and duo-trio tests (one tailed, p = 1/2) and the triangle test (one tailed, p = 1/3) . . . . . . . . . . . . . . . . . . Minimum numbers of correct judgments to establish significance at probability levels of 5 and 1% for paired pference test (two tailed, p = 1/2) . . . . . . . . . . . . . . . . . . . . . . . . . . . Minimum number of responses (n) and correct responses (x) to obtain a level of Type I and Type II risks in the triangle test. Pd is the chance-adjusted percent correct or proportion of discriminators . . . . . . . . . . . . . . . . . . . . . . Minimum number of responses (n) and correct responses (x) to obtain a level of Type I and Type II risks in the duo-trio test. Pc is the chance-adjusted percent correct or proportion of discriminators . . . . . . . . . . . . . . . . . . . . . . d and B (variance factor) values for the duo-trio and 2-AFC (paired comparison) difference tests . . . . . . d and B (variance factor) values for the triangle and 3-AFC difference tests . . . . . . . . . . . . . . . . . Random permutations of nine . . . . . . . . . . . . . . . Random numbers . . . . . . . . . . . . . . . . . . . . . .

569 571 572

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

573

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

587

Table F.K Table F.L

Table F.M

Table F.N1

Table F.N2

Table F.O1 Table F.O2 Table F.P Table F.Q

563 564

565

566

567

567 568

Chapter 1

Introduction

Abstract In this chapter we carefully parse the definition for sensory evaluation, briefly discuss validity of the data collected before outlining the early history of the field. We then describe the three main methods used in sensory evaluation (discrimination tests, descriptive analysis, and hedonic testing) before discussing the differences between analytical and consumer testing. We then briefly discuss why one may want to collect sensory data. In the final sections we highlight the differences and similarities between sensory evaluation and marketing research and between sensory evaluation and commodity grading as used in, for example, the dairy industry.

Sensory evaluation is a child of industry. It was spawned in the late 40’s by the rapid growth of the consumer product companies, mainly food companies. . . . Future development in sensory evaluation will depend upon several factors, one of the most important being the people and their pparation and training. – Elaine Skinner (1989)

Contents Introduction and Overview . . . . . . . . . 1.1.1 Definition . . . . . . . . . . . . . . 1.1.2 Measurement . . . . . . . . . . . . 1.2 Historical Landmarks and the Three Classes of Test Methods . . . . . . . . . . . . . . 1.2.1 Difference Testing . . . . . . . . . . 1.2.2 Descriptive Analyses . . . . . . . . . 1.2.3 Affective Testing . . . . . . . . . . 1.2.4 The Central Dogma-Analytic Versus Hedonic Tests . . . . . . . . . . . . 1.3 Applications: Why Collect Sensory Data? . . 1.3.1 Differences from Marketing Research Methods . . . . . . . . . . . . . . 1.3.2 Differences from Traditional Product Grading Systems . . . . . . . . . . 1.4 Summary and Conclusions . . . . . . . . . References . . . . . . . . . . . . . . . . . . . 1.1

1.1 Introduction and Overview . . .

1 1 3

. . . .

4 5 6 7

. .

8 10

.

13

. . .

15 16 17

1.1.1 Definition The field of sensory evaluation grew rapidly in the second half of the twentieth century, along with the expansion of the processed food and consumer products industries. Sensory evaluation comprises a set of techniques for accurate measurement of human responses to foods and minimizes the potentially biasing effects of brand identity and other information influences on consumer perception. As such, it attempts to isolate the sensory properties of foods themselves and provides important and useful information to product developers, food scientists, and managers about the sensory characteristics of their products. The field was comphensively reviewed by Amerine, Pangborn, and Roessler in 1965, and more recent texts have been published by Moskowitz et al. (2006), Stone and Sidel (2004), and Meilgaard et al. (2006). These three later sources are practical works aimed at sensory specialists

H.T. Lawless, H. Heymann, Sensory Evaluation of Food, Food Science Text Series, DOI 10.1007/978-1-4419-6488-5_1, © Springer Science+Business Media, LLC 2010

1

2

in industry and reflect the philosophies of the consulting groups of the authors. Our goal in this book is to provide a comphensive overview of the field with a balanced view based on research findings and one that is suited to students and practitioners alike. Sensory evaluation has been defined as a scientific method used to evoke, measure, analyze, and interpt those responses to products as perceived through the senses of sight, smell, touch, taste, and hearing (Stone and Sidel, 2004). This definition has been accepted and endorsed by sensory evaluation committees within various professional organizations such as the Institute of Food Technologists and the American Society for Testing and Materials. The principles and practices of sensory evaluation involve each of the four activities mentioned in this definition. Consider the words “to evoke.” Sensory evaluation gives guidelines for the pparation and serving of samples under controlled conditions so that biasing factors are minimized. For example, people in a sensory test are often placed in inpidual test booths so that the judgments they give are their own and do not reflect the opinions of those around them. Samples are labeled with random numbers so that people do not form judgments based upon labels, but rather on their sensory experiences. Another example is in how products may be given in different orders to each participant to help measure and counterbalance for the sequential effects of seeing one product after another. Standard procedures may be established for sample temperature, volume, and spacing in time, as needed to control unwanted variation and improve test pcision. Next, consider the words, “to measure.” Sensory evaluation is a quantitative science in which numerical data are collected to establish lawful and specific relationships between product characteristics and human perception. Sensory methods draw heavily from the techniques of behavioral research in observing and quantifying human responses. For example, we can assess the proportion of times people are able to discriminate small product changes or the proportion of a group that expsses a pference for one product over another. Another example is having people generate numerical responses reflecting their perception of how strong a product may taste or smell. Techniques of behavioral research and experimental psychology offer guidelines as to how such measurement techniques should be employed and what their potential pitfalls and liabilities may be.

1 Introduction

The third process in sensory evaluation is analysis. Proper analysis of the data is a critical part of sensory testing. Data generated from human observers are often highly variable. There are many sources of variation in human responses that cannot be completely controlled in a sensory test. Examples include the mood and motivation of the participants, their innate physiological sensitivity to sensory stimulation, and their past history and familiarity with similar products. While some screening may occur for these factors, they may be only partially controlled, and panels of humans are by their nature heterogeneous instruments for the generation of data. In order to assess whether the relationships observed between product characteristics and sensory responses are likely to be real, and not merely the result of uncontrolled variation in responses, the methods of statistics are used to analyze evaluation data. Hand-in-hand with using appropriate statistical analyses is the concern of using good experimental design, so that the variables of interest are investigated in a way that allows sensible conclusions to be drawn. The fourth process in sensory evaluation is the interptation of results. A sensory evaluation exercise is necessarily an experiment. In experiments, data and statistical information are only useful when interpted in the context of hypotheses, background knowledge, and implications for decisions and actions to be taken. Conclusions must be drawn that are reasoned judgments based upon data, analyses, and results. Conclusions involve consideration of the method, the limitations of the experiment, and the background and contextual framework of the study. The sensory evaluation specialists become more than mere conduits for experimental results, but must contribute interptations and suggest reasonable courses of action in light of the numbers. They should be full partners with their clients, the end-users of the test results, in guiding further research. The sensory evaluation professional is in the best situation to realize the appropriate interptation of test results and the implications for the perception of products by the wider group of consumers to whom the results may be generalized. The sensory specialist best understands the limitations of the test procedure and what its risks and liabilities may be. A sensory scientist who is ppared for a career in research must be trained in all four of the phases mentioned in the definition. They must understand products, people as measuring instruments, statistical

1.1 Introduction and Overview

1.1.2 Measurement Sensory evaluation is a science of measurement. Like other analytical test procedures, sensory evaluation is concerned with pcision, accuracy, sensitivity, and avoiding false positive results (Meiselman, 1993). Precision is similar to the concept in the behavioral sciences of reliability. In any test procedure, we would like to be able to get the same result when a test is repeated. There is usually some error variance around an obtained value, so that upon repeat testing, the value will not always be exactly the same. This is especially true of sensory tests in which human perceptions are necessarily part of the generation of data. However, in many sensory test procedures, it is desirable to minimize this error variance as much as possible and to have tests that are low in error associated with repeated measurements. This is achieved by several means. As noted above, we isolate the sensory response to the factors of interest, minimizing extraneous influences, controlling sample pparation and psentation. Additionally, as necessary, sensory scientists screen and train panel participants. A second concern is the accuracy of a test. In the physical sciences, this is viewed as the ability of a test instrument to produce a value that is close to the “true” value, as defined by independent measurement from another instrument or set of instruments that have been appropriately calibrated. A related idea in the behavioral sciences, this principle is called the validity of a test. This concerns the ability of a test procedure to measure what it was designed and intended to measure. Validity is established in a number of ways. One useful criterion is pdictive validity, when a test result is of value in pdicting what would occur in another situation or another measurement. In sensory testing, for example, the test results should reflect the perceptions and opinions of consumers that might buy the product. In other words, the results of the sensory test should generalize to the larger population. The test results might correlate with instrumental measures, process or ingredient variables, storage factors, shelf life times,

3

or other conditions known to affect sensory properties. In considering validity, we have to look at the end use of the information provided by a test. A sensory test method might be valid for some purposes, but not others (Meiselman, 1993). A simple difference test can tell if a product has changed, but not whether people will like the new version. A good sensory test will minimize errors in measurement and errors in conclusions and decisions. There are different types of errors that may occur in any test procedure. Whether the test result reflects the true state of the world is an important question, especially when error and uncontrolled variability are inherent in the measurement process. Of primary concern in sensory tests is the sensitivity of the test to differences among products. Another way to phrase this is that a test should not often miss important differences that are psent. “Missing a difference” implies an insensitive test procedure. To keep sensitivity high, we must minimize error variance wherever possible by careful experimental controls and by selection and training of panelists where appropriate. The test must involve sufficient numbers of measurements to insure a tight and reliable statistical estimate of the values we obtain, such as means or proportions. In statistical language, detecting true differences is avoiding Type II error and the minimization of β-risk. Discussion of the power and sensitivity of tests from a statistical perspective occurs in Chapter 5 and in the Appendix. The other error that may occur in a test result is that of finding a positive result when none is actually psent in the larger population of people and products outside the sensory test. Once again, a positive result usually means detection of a statistically significant difference between test products. It is important to use a test method that avoids false positive results or Type I error in statistical language. Basic statistical training and common statistical tests applied to scientific findings are oriented toward avoiding this kind of error. The effects of random chance deviations must be taken into account in deciding if a test result reflects a real difference or whether our result is likely to be due to chance variation. The common procedures of inferential statistics provide assurance that we have limited our possibility of finding a difference where one does not really exist. Statistical procedures reduce this risk to some comfortable level, usually with a ceiling of 5% of all tests we conduct.

4

1 Introduction

(Lawless and Klein, 1989). An important function of sensory scientists in an academic setting is to provide consulting and resources to insure that quality tests are conducted by other researchers and students who seek to understand the sensory impact of the variables they are studying. In government services such as food inspection, sensory evaluation plays a key role (York, 1995). Sensory principles and appropriate training can be key in insuring that test methods reflect the current knowledge of sensory function and test design. See Lawless (1993) for an overview of the education and training of sensory scientists-much of this piece still rings true more than 15 years later.

1.2 Historical Landmarks and the Three Classes of Test Methods The human senses have been used for centuries to evaluate the quality of foods. We all form judgments about foods whenever we eat or drink (“Everyone carries his own inch-rule of taste, and amuses himself by applying it, triumphantly, wherever he travels.”-Henry Adams, 1918). This does not mean that all judgments are useful or that anyone is qualified to participate in a sensory test. In the past, production of good quality foods often depended upon the sensory acuity of a single expert who was in charge of production or made decisions about process changes in order to make sure the product would have desirable characteristics. This was the historical tradition of brewmasters, wine tasters, dairy judges, and other food inspectors who acted as the arbiters of quality. Modern sensory evaluation replaced these single authorities with panels of people participating in specific test methods that took the form of planned experiments. This change occurred for several reasons. First, it was recognized that the judgments of a panel would in general be more reliable than the judgments of single inpidual and it entailed less risk since the single expert could become ill, travel, retire, die, or be otherwise unavailable to make decisions. Replacement of such an inpidual was a nontrivial problem. Second, the expert might or might not reflect what consumers or segments of the consuming public might want in a product. Thus for issues of product quality and overall appeal, it was safer (although often more time consuming and expensive)

1.2 Historical Landmarks and the Three Classes of Test Methods

to go directly to the target population. Although the tradition of informal, qualitative inspections such as benchtop “cuttings” persists in some industries, they have been gradually replaced by more formal, quantitative, and controlled observations (Stone and Sidel, 2004). The current sensory evaluation methods comprise a set of measurement techniques with established track records of use in industry and academic research. Much of what we consider standard procedures comes from pitfalls and problems encountered in the practical experience of sensory specialists over the last 70 years of food and consumer product research, and this experience is considerable. The primary concern of any sensory evaluation specialist is to insure that the test method is appropriate to answer the questions being asked about the product in the test. For this reason, tests are usually classified according to their primary purpose and most valid use. Three types of sensory testing are commonly used, each with a different goal and each using participants selected using different criteria. A summary of the three main types of testing is given in Table 1.1.

1.2.1 Difference Testing The simplest sensory tests merely attempt to answer whether any perceptible difference exists between two types of products. These are the discrimination tests or simple difference testing procedures. Analysis is usually based on the statistics of frequencies and proportions (counting right and wrong answers). From the test results, we infer differences based on the proportions of persons who are able to choose a test product correctly from among a set of similar or control products. A classic example of this test was the triangle procedure, used in the Carlsberg breweries and in the Seagrams distilleries in the 1940s (Helm and Trolle,

5

1946; Peryam and Swartz, 1950). In this test, two products were from the same batch while a third product was different. Judges would be asked to pick the odd sample from among the three. Ability to discriminate differences would be inferred from consistent correct choices above the level expected by chance. In breweries, this test served primarily as a means to screen judges for beer evaluation, to insure that they possessed sufficient discrimination abilities. Another multiple-choice difference test was developed at about the same time in distilleries for purposes of quality control (Peryam and Swartz, 1950). In the duo-trio procedure, a reference sample was given and then two test samples. One of the test samples matched the reference while the other was from a different product, batch or process. The participant would try to match the correct sample to the reference, with a chance probability of one-half. As in the triangle test, a proportion of correct choices above that expected by chance is considered evidence for a perceivable difference between products. A third popular difference test was the paired comparison, in which participants would be asked to choose which of two products was stronger or more intense in a given attribute. Partly due to the fact that the panelist’s attention is directed to a specific attribute, this test is very sensitive to differences. These three common difference tests are shown in Fig. 1.1. Simple difference tests have proven very useful in application and are in widespad use today. Typically a discrimination test will be conducted with 25-40 participants who have been screened for their sensory acuity to common product differences and who are familiar with the test procedures. This generally provides an adequate sample size for documenting clear sensory differences. Often a replicate test is performed while the respondents are psent in the sensory test facility. In part, the popularity of these tests is due to the simplicity of data analysis. Statistical tables derived from the binomial distribution give the minimum number of correct responses needed to conclude statistical

Table 1.1 Classification of test methods in sensory evaluation Class Question of interest

Type of test

Panelist characteristics

Discrimination

Are products perceptibly different in any way

“Analytic”

Descriptive

How do products differ in specific sensory characteristics How well are products liked or which products are pferred

“Analytic”

Screened for sensory acuity, oriented to test method, sometimes trained Screened for sensory acuity and motivation, trained or highly trained Screened for products, untrained

Affective

“Hedonic”

6

1 Introduction

Fig. 1.1 Common methods for discrimination testing include the triangle, duo-trio, and paired comparison procedures.

significance as a function of the number of participants. Thus a sensory technician merely needs to count answers and refer to a table to give a simple statistical conclusion, and results can be easily and quickly reported.

the consensus of a panel was likely to be more reliable and accurate than the judgment of a single inpidual. Second, it provided a means to characterize the inpidual attributes of flavor and provide a comphensive analytical description of differences among a group of products under development. Several variations and refinements in descriptive analysis techniques were forthcoming. A group at the General Foods Technical Center in the early 1960s developed and refined a method to quantify food texture, much as the flavor profile had enabled the quantification of flavor properties (Brandt et al., 1963, Szczesniak et al., 1975). This technique, the Texture Profile method, used a fixed set of force-related and shape-related attributes to characterize the rheological and tactile properties of foods and how these changed over time with mastication. These characteristics have parallels in the physical evaluation of food breakdown or flow. For example, perceived hardness is related to the physical force required to penetrate a sample. Perceived thickness of a fluid or semisolid is related in part to physical viscosity. Texture profile panelists were also trained to recognize specific intensity points along each scale, using standard products or formulated pseudo-foods for calibration. Other approaches were developed for descriptive analysis problems. At Stanford Research Institute in the early 1970s, a group proposed a method for descriptive analysis that would remedy some of the R apparent shortcomings of the Flavor Profile method

1.2 Historical Landmarks and the Three Classes of Test Methods

Table 1.2 Descriptive evaluation of cookies-texture attributes Phase Attributes Word anchors Surface

First bite

First chew Chew down

Residual

Roughness Particles Dryness Fracturability Hardness Particle size Denseness Uniformity of chew Moisture absorption Cohesiveness of mass Toothpacking Grittiness Oiliness Particles Chalky

Smooth-rough None-many Oily-dry Crumbly-brittle Soft-hard Small-large Airy-dense Even-uneven None-much Loose-cohesive None-much None-much Dry-oily None-many Not chalky-very chalky

7

a uniform and controlled manner, typical of an analytical sensory test procedure. For example, the first bite may be defined as cutting with the incisors. The panel for such an analysis would consist of perhaps 10- 12 well-trained inpiduals, who were oriented to the meanings of the terms and given practice with examples. Intensity references to exemplify scale points are also given in some techniques. Note the amount of detailed information that can be provided in this example and bear in mind that this is only looking at the product’s texture-flavor might form an equally detailed sensory analysis, perhaps with a separate trained panel. The relatively small number of panelists (a dozen or so) is justified due to their level of calibration. Since they have been trained to use attribute scales in a similar manner, error variance is lowered and statistical power and test sensitivity are maintained in spite of fewer observations (fewer data points per product). Similar examples of texture, flavor, fragrance, and tactile analyses can be found in Meilgaard et al. (2006).

8

1 Introduction

the food product (the stimulus). However, since a sensory perception involves the necessary interaction of a person with a stimulus, it should be apparent that similar test methods are necessary to characterize this person-product interaction.

1.2.4 The Central Dogma-Analytic Versus Hedonic Tests

Fig. 1.2 The 9-point hedonic scale used to assess liking and disliking. This scale, originally developed at the U.S. Army Food and Container Institute (Quartermaster Corps), has achieved widespad use in consumer testing of foods.

The central principle for all sensory evaluation is that the test method should be matched to the objectives of the test. Figure 1.3 shows how the selection of the test procedure flows from questions about the objective of the investigation. To fulfill this goal, it is necessary to have clear communication between the sensory test manager and the client or end-user of the information. A dialogue is often needed. Is the important question whether or not there is any difference at all among the products? If so, a discrimination test is indicated. Is the question one of whether consumers like the new product better than the pvious version? A consumer acceptance test is needed. Do we need to know what attributes have changed in the sensory characteristics of the new product? Then a descriptive analysis procedure is called for. Sometimes there are multiple objectives and a sequence of different tests is required (Lawless and Claassen, 1993). This can psent problems if all the answers are required at once or under severe time pssure during competitive product development. One of the most important jobs of the sensory specialist in the food industry is to insure a clear understanding and specification of the type of information needed by the end-users. Test design may require a number of conversations, interviews with different people, or even written test requests that specify why the information is to be collected and how the results will be used in making specific decisions and subsequent actions to be taken. The sensory specialist is the best position to understand the uses and limitations of each procedure and what would be considered appropriate versus inappropriate conclusions from the data. There are two important corollaries to this principle. The sensory test design involves not only the selection of an appropriate method but also the selection of appropriate participants and statistical analyses. The three classes of sensory tests can be pided into two types, analytical sensory tests including discrimination

1.2 Historical Landmarks and the Three Classes of Test Methods

9

Fig. 1.3 A flowchart showing methods determination. Based on the major objectives and research questions, different sensory test methods are selected. Similar decision processes are made in panelist selection, setting up response scales, in choosing experimental designs, statistical analysis, and other tasks in designing a sensory test (reprinted with permission from Lawless, 1993).

and descriptive methods and affective or hedonic tests such as those involved in assessing consumer liking or pferences (Lawless and Claassen, 1993). For the analytical tests, panelists are selected based on having average to good sensory acuity for the critical characteristics (tastes, smells, textures, etc.) of products to be evaluated. They are familiarized with the test procedures and may undergo greater or lesser amounts of training, depending upon the method. In the case of descriptive analysis, they adopt an analytical frame of mind, focusing on specific aspects of the product as directed by the scales on their questionnaires. They are asked to put personal pferences and hedonic reactions aside, as their job is only to specify what attributes are psent in the product and at what levels of sensory intensity, extent, amount, or duration. In contrast to this analytical frame of mind, consumers in an affective test act in a much more integrative fashion. They perceive a product as a whole pattern. Although their attention is sometimes captured by a specific aspect of a product (especially if it is a bad, unexpected, or unpleasant one), their reactions to the product are often immediate and based on the integrated pattern of sensory stimulation from the product and expssed as liking or disliking. This occurs without a great deal of thought or dissection of the product’s specific profile. In other words, consumers are effective at rendering impssions based on the integrated pattern of perceptions. In such consumer

tests, participants must be chosen carefully to insure that the results will generalize to the population of interest. Participants should be frequent users of the product, since they are most likely to form the target market and will be familiar with similar products. They possess reasonable expectations and a frame of reference within which they can form an opinion relative to other similar products they have tried. The analytic/hedonic distinction gives rise to some highly important rules of thumb and some warnings about matching test methods and respondents. It is unwise to ask trained panelists about their pferences or whether they like or dislike a product. They have been asked to assume a different, more analytical frame of mind and to place personal pference aside. Furthermore, they have not necessarily been selected to be frequent users of the product, so they are not part of the target population to which one would like to generalize hedonic test results. A common analogy here is to an analytical instrument. You would not ask a gas chromatograph or a pH meter whether it liked the product, so why ask your analytical descriptive panel (O’Mahony, 1979). Conversely, problems arise when consumers are asked to furnish very specific information about product attributes. Consumers not only act in a non-analytic frame of mind but also often have very fuzzy concepts about specific attributes, confusing sour and bitter tastes, for example. Inpiduals often differ markedly

10

in their interptations of sensory attribute words on a questionnaire. While a trained texture profile panel has no trouble in agreeing how cohesive a product is after chewing, we cannot expect consumers to provide pcise information on such a specific and technical attribute. In summary, we avoid using trained panelists for affective information and we avoid asking consumers about specific analytical attributes. Related to the analytic-hedonic distinction is the question of whether experimental control and pcision are to be maximized or whether validity and generalizability to the real world are more important. Often there is a tradeoff between the two and it is difficult to maximize both simultaneously. Analytic tests in the lab with specially screened and trained judges are more reliable and lower in random error than consumer tests. However, we give up a certain amount of generalizability to real-world results by using artificial conditions and a special group of participants. Conversely, in the testing of products by consumers in their own homes we have not only a lot of real-life validity but also a lot of noise in the data. Brinberg and McGrath (1985) have termed this struggle between pcision and validity one of “conflicting desiderata.” O’Mahony (1988) has made a distinction between sensory evaluation Type I and Type II. In Type I sensory evaluation, reliability and sensitivity are key factors, and the participant is viewed much like an analytical instrument used to detect and measure changes in a food product. In Type II sensory evaluation, participants are chosen to be repsentative of the consuming population, and they may evaluate food under more naturalistic conditions. Their emphasis here is on pdiction of consumer response. Every sensory test falls somewhere along a continuum where reliability versus real-life extrapolation are in a potential tradeoff relationship. This factor must also be discussed with end-users of the data to see where their emphasis lies and what level of tradeoff they find comfortable. Statistical analyses must also be chosen with an eye to the nature of the data. Discrimination tests involve choices and counting numbers of correct responses. The statistics derived from the binomial distribution or those designed for proportions such as chi-square are appropriate. Conversely, for most scaled data, we can apply the familiar parametric statistics appropriate to normally distributed and continuous data, such as means, standard deviations, t-tests, analysis of variance. The choice of an appropriate statistical test is not

1 Introduction

always straightforward, so sensory specialists are wise to have thorough training in statistics and to involve statistical and design specialists in a complex project in its earliest stages of planning. Occasionally, these central principles are violated. They should not be put aside as a matter of mere expediency or cost savings and never without a logical analysis. One common example is the use of a discrimination test before consumer acceptance. Although our ultimate interest may lie in whether consumers will like or dislike a new product variation, we can conduct a simple difference test to see whether any change is perceivable at all. The logic in this sequence is the following: if a screened and experienced discrimination panel cannot tell the difference under carefully controlled conditions in the sensory laboratory, then a more heterogeneous group of consumers is unlikely to see a difference in their less controlled and more variable world. If no difference is perceived, there can logically be no systematic pference. So a more time consuming and costly consumer test can sometimes be avoided by conducting a simpler but more sensitive discrimination test first. The added reliability of the controlled discrimination test provides a “safety net” for conclusions about consumer perception. Of course, this logic is not without its pitfalls-some consumers may interact extensively with the product during a home use test period and may form stable and important opinions that are not captured in a short duration laboratory test, and there is also always the possibility of a false negative result (the error of missing a difference). MacRae and Geelhoed (1992) describe an interesting case of a missed difference in a triangle test where a significant pference was then observed between water samples in a paired comparison. The sensory professional must be aware that these anomalies in experimental results will sometimes arise, and must also be aware of some of the reasons why they occur.

1.3 Applications: Why Collect Sensory Data? Human perceptions of foods and consumer products are the results of complex sensory and interptation processes. At this stage in scientific history, perceptions of such multidimensional stimuli as conducted

1.3 Applications: Why Collect Sensory Data?

by the parallel processing of the human nervous system are difficult or impossible to pdict from instrumental measures. In many cases instruments lack the sensitivity of human sensory systems-smell is a good example. Instruments rarely mimic the mechanical manipulation of foods when tasted nor do they mimic the types of peri-receptor filtering that occur in biological fluids like saliva or mucus that can cause chemical partitioning of flavor materials. Most importantly, instrumental assessments give values that miss an important perceptual process: the interptation of sensory experience by the human brain prior to responding. The brain lies interposed between sensory input and the generation of responses that form our data. It is a massively parallel-distributed processor and computational engine, capable of rapid feats of pattern recognition. It comes to the sensory evaluation task complete with a personal history and experiential frame of reference. Sensory experience is interpted, given meaning within the frame of reference, evaluated relative to expectations and can involve integration of multiple simultaneous or sequential inputs. Finally judgments are rendered as our data. Thus there is a “chain of perception” rather than simply stimulus and response (Meilgaard et al., 2006). Only human sensory data provide the best models for how consumers are likely to perceive and react to food products in real life. We collect, analyze, and interpt sensory data to form pdictions about how products have changed during a product development program. In the food and consumer products industries, these changes arise from three important factors: ingredients, processes, and packaging. A fourth consideration is often the way a product ages, in other words its shelf life, but we may consider shelf stability to be one special case of processing, albeit usually a very passive one (but also consider products exposed to temperature fluctuation, light-catalyzed oxidation, microbial contamination, and other “abuses”). Ingredient changes arise for a number of reasons. They may be introduced to improve product quality, to reduce costs of production, or simply because a certain supply of raw materials has become unavailable. Processing changes likewise arise from the attempt to improve quality in terms of sensory, nutritional, microbiological stability factors, to reduce costs or to improve manufacturing productivity. Packaging changes arise from considerations of product stability or other quality factors, e.g., a certain

11

amount of oxygen permeability may insure that a fresh beef product remains red in color for improved visual appeal to consumers. Packages function as carriers of product information and brand image, so both sensory characteristics and expectations can change as a function of how this information can be carried and displayed by the packaging material and its print overlay. Packaging and print ink may cause changes in flavor or aroma due to flavor transfer out of the product and sometimes transfer of off-flavors into the product. The package also serves as an important barrier to oxidative changes, to the potentially deleterious effects of light-catalyzed reactions, and to microbial infestations and other nuisances. The sensory test is conducted to study how these product manipulations will create perceived changes to human observers. In this sense, sensory evaluation is in the best traditions of psychophysics, the oldest branch of scientific psychology, that attempts to specify the relationships between different energy levels impinging upon the sensory organs (the physical part of psychophysics) and the human response (the psychological part). Often, one cannot pdict exactly what the sensory change will be as a function of ingredients, processes, or packaging, or it is very difficult to do so since foods and consumer products are usually quite complex systems. Flavors and aromas depend upon complex mixtures of many volatile chemicals. Informal tasting in the lab may not bring a reliable or sufficient answer to sensory questions. The benchtop in the development laboratory is a poor place to judge potential sensory impact with distractions, competing odors, nonstandard lighting, and so on. Finally, the nose, eyes, and tongue of the product developer may not be repsentative of most other people who will buy the product. So there is some uncertainty about how consumers will view a product especially under more natural conditions. Uncertainty is the key here. If the outcome of a sensory test is perfectly known and pdictable, there is no need to conduct the formal evaluation. Unfortunately, useless tests are often requested of a sensory testing group in the industrial setting. The burden of useless routine tests arises from overly entrenched product development sequences, corporate traditions, or merely the desire to protect oneself from blame in the case of unexpected failures. However, the sensory test is only as useful as the amount of reduction in uncertainty that occurs. If there is no uncertainty, there

12

is no need for the sensory test. For example, doing a sensory test to see if there is a perceptible color difference between a commercial red wine and a commercial white wine is a waste of resources, since there is no uncertainty! In the industrial setting, sensory evaluation provides a conduit for information that is useful in management business decisions about directions for product development and product changes. These decisions are based on lower uncertainty and lower risk once the sensory information is provided. Sensory evaluation also functions for other purposes. It may be quite useful or even necessary to include sensory analyses in quality control (QC) or quality assurance. Modification of traditional sensory practices may be required to accommodate the small panels and rapid assessments often required in online QC in the manufacturing environment. Due to the time needed to assemble a panel, ppare samples for testing, analyze and report sensory data, it can be quite challenging to apply sensory techniques to quality control as an on-line assessment. Quality assurance involving sensory assessments of finished products are more readily amenable to sensory testing and may be integrated with routine programs for shelf life assessment or quality monitoring. Often it is desirable to establish correlations between sensory response and instrumental measures. If this is done well, the instrumental measure can sometimes be substituted for the sensory test. This is especially applicable under conditions in which rapid turnaround is needed. Substitution of instrumental measurements for sensory data may also be useful if the evaluations are likely to be fatiguing to the senses, repetitive, involve risk in repeated evaluations (e.g., insecticide fragrances), and are not high in business risk if unexpected sensory problems arise that were missed. In addition to these product-focused areas of testing, sensory research is valuable in a broader context. A sensory test may help to understand the attributes of a product that consumers view as critical to product acceptance and thus success. While we keep a wary eye on the fuzzy way that consumers use language, consumer sensory tests can provide diagnostic information about a product’s points of superiority or shortcomings. Consumer sensory evaluations may suggest hypotheses for further inquiry such as exploration of new product opportunities. There are recurrent themes and enduring problems in sensory science. In 1989, the ASTM Committee

1 Introduction

1.3 Applications: Why Collect Sensory Data?

(4)

(5)

(6)

(7)

control in multi-plant manufacturing situations, as well as international product development and the problem of multiple sensory testing sites and panels. Environmental and biochemical factors. Kamen recognized that pferences may change as a function of the situation (food often tastes better outdoors and when you are hungry). Meiselman (1993) questioned whether sufficient sensory research is being performed in realistic eating situations that may be more pdictive of consumer reactions, and recently sensory scientists have started to explore this area of research (for example, Giboreau and Fleury, 2009; Hein et al., 2009; Mielby and Frøst, 2009). Resolving discrepancies between laboratory and field studies. In the search for reliable, detailed, and pcise analytical methods in the sensory laboratory, some accuracy in pdicting field test results may be lost. Management must be aware of the potential of false positive or negative results if a full testing sequence is not carried out, i.e., if shortcuts are made in the testing sequence prior to marketing a new product. Sensory evaluation specialists in industry do not always have time to study the level of correlation between laboratory and field tests, but a prudent sensory program would include periodic checks on this issue. Inpidual differences. Since Kamen’s era, a growing literature has illuminated the fact that human panelists are not identical, interchangeable measuring instruments. Each comes with different physiological equipment, different frames of reference, different abilities to focus and maintain attention, and different motivational resources. As an example of differences in physiology, we have the growing literature on specific anosmias-smell “blindnesses” to specific chemical compounds among persons with otherwise normal senses of smell (Boyle et al., 2006; Plotto et al., 2006; Wysocki and Labows, 1984). It should not be surprising that some olfactory characteristics are difficult for even trained panelists to evaluate and to come to agreement (Bett and Johnson, 1996). Relating sensory differences to product variables. This is certainly the meat of sensory science in industrial practice. However, many product developers do not sufficiently involve their sensory specialists in the underlying research questions.

13

They also may fall into the trap of never ending sequences of paired tests, with little or no planned designs and no modeling of how underlying physical variables (ingredients, processes) create a dynamic range of sensory changes. The relation of graded physical changes to sensory response is the essence of psychophysical thinking. (8) Sensory interactions. Foods and consumer products are multidimensional. The more sensory scientists understand interactions among characteristics such as enhancement and masking effects, the better they can interpt the results of sensory tests and provide informed judgments and reasoned conclusions in addition to reporting just numbers and statistical significance. (9) Sensory education. End-users of sensory data and people who request sensory tests often expect one tool to answer all questions. Kamen cited the simple dichotomy between analytical and hedonic testing (e.g., discrimination versus pference) and how explaining this difference was a constant task. Due to the lack of widespad training in sensory science, the task of sensory education is still with us today, and a sensory professional must be able to explain the rationale behind test methods and communicate the importance and logic of sensory technology to non-sensory scientists and managers.

14 Table 1.3 Contrast of sensory evaluation consumer tests with market research tests

1 Introduction Sensory testing with consumers Participants screened to be users of the product category Blind-labeled samples-random codes with minimal conceptual information Determines if sensory properties and overall appeal met targets Expectations based on similar products used in the category Not intended to assess response/appeal of product concept Market research testing (concept and/or product test) Participants in product-testing phase selected for positive response to concept Conceptual claims, information, and frame of reference are explicit Expectations derived from concept/claims and similar product usage Unable to measure sensory appeal in isolation from concept and expectations

designed to make the product conceptually appealing (e.g., bringing attention to convenience factors in pparation). In a sensory test all these potentially biasing factors are stripped away in order to isolate the opinion based on sensory properties only. In the tradition of scientific inquiry, we need to isolate the variables of interest (ingredients, processing, packaging changes) and assess sensory properties as a function of these variables, and not as a function of conceptual influences. This is done to minimize the influence of a larger cognitive load of expectations generated from complex conceptual information. There are many potential response biases and task demands that are entailed in “selling” an idea as well as in selling a product. Participants often like to please the experimenter and give results consistent with what they think the person wants. There is a large literature on the effect of factors such as brand label on consumer response. Product information interacts in complex ways with consumer attitudes and expectancies (Aaron et al., 1994; Barrios and Costell, 2004; Cardello and Sawyer, 1992; Costell et al., 2009; Deliza and MacFie, 1996; Giménez et al., 2008; Kimura et al., 2008; Mielby and Frøst, 2009; Park and Lee, 2003; Shepherd et al., 1991/1992). Expectations can cause assimilation of sensory reactions toward what is expected under some conditions and under other conditions will show contrast effects, enhancing differences when expectations are not met (Siegrist and Cousin, 2009; Lee et al., 2006; Yeomans et al., 2008; Zellner et al., 2004). Packaging and brand information will also affect sensory judgments (Dantas et al., 2004; Deliza et al., 1999; Enneking et al., 2007). So the apparent resemblance of a blind sensory test and a fully concept-loaded market research test are quite illusory. Corporate management needs to be reminded of this important distinction. There continues to be

1.3 Applications: Why Collect Sensory Data?

1.3.2 Differences from Traditional Product Grading Systems A second arena of apparent similarity to sensory evaluation is with the traditional product quality grading systems that use sensory criteria. The grading of agricultural commodities is a historically important influence on the movement to assure consumers of quality standards in the foods they purchase. Such techniques were widely applicable to simple products such as fluid milk and butter (Bodyfelt et al., 1988, 2008), where an ideal product could be largely agreed upon and the defects that could arise in poor handling and processing gave rise to well-known sensory effects. Further impetus came from the fact that competitions could be held to examine whether novice judges-in-training could match the opinions of experts. This is much in the tradition of livestock grading-a young person could judge a cow and receive awards at a state fair for learning to use the same criteria and critical eye as the expert judges. There are noteworthy differences in the ways in which sensory testing and quality judging are performed. Some of these are outlined in Table 1.4. The commodity grading and the inspection tradition have severe limitations in the current era of highly processed foods and market segmentation. There are fewer and fewer “standard products” relative to the wide variation in flavors, nutrient levels (e.g., low fat), convenience pparations, and other choices that

15

line the supermarket shelves. Also, one person’s product defect may be another’s marketing bonanza, as in the glue that did not work so well that gave us the ubiquitous post-it notes. Quality judging methods are poorly suited to research support programs. The techniques have been widely criticized on a number of scientific grounds (Claassen and Lawless, 1992; Drake, 2007; O’Mahony, 1979; Pangborn and Dunkley, 1964; Sidel et al., 1981), although they still have their proponents in industry and agriculture (Bodyfelt et al., 1988, 2008). The defect identification in quality grading emphasizes root causes (e.g., oxidized flavor) whereas the descriptive approach uses more elemental singular terms to describe perceptions rather than to infer causes. In the case of oxidized flavors, the descriptive analysis panel might use a number of terms (oily, painty, and fishy) since oxidation causes a number of qualitatively different sensory effects. Another notable difference from mainstream sensory evaluation is that the quality judgments combine an overall quality scale (psumably reflecting consumer dislikes) with diagnostic information about defects, a kind of descriptive analysis looking only at the negative aspects of products. In mainstream sensory evaluation, the descriptive function and the consumer evaluation would be clearly separate in two distinct tests with different respondents. Whether the opinion of a single expert can effectively repsent consumer opinion is highly questionable at this time in history.

Table 1.4 Contrast of sensory evaluation tests with quality inspection Sensory testing Separates hedonic (like-dislike) and descriptive information into separate tests Uses repsentative consumers for assessment of product appeal (liking/disliking) Uses trained panelists to specify attributes, but not liking/disliking Oriented to research support Flexible for new, engineered, and innovative products Emphasizes statistical inference for decision making, suitable experimental designs, and sample sizes Quality inspection Used for pass-fail online decisions in manufacturing Provides quality score and diagnostic information concerning defects in one test Uses sensory expertise of highly trained inpiduals May use only one or very few trained experts Product knowledge, potential problems, and causes are stressed Traditional scales are multi-dimensional and poorly suited to statistical analyses Decision-making basis may be qualitative Oriented to standard commodities

16

1 Introduction

1.4 Summary and Conclusions Sensory evaluation comprises a set of test methods with guidelines and established techniques for product psentation, well-defined response tasks, statistical methods, and guidelines for interptation of results. Three primary kinds of sensory tests focus on the

existence of overall differences among products (discrimination tests), specification of attributes (descriptive analysis), and measuring consumer likes and dislikes (affective or hedonic testing). Correct application of sensory technique involves correct matching of method to the objective of the tests, and this requires good communication between sensory specialists and

Methods Selection

yes Consumer Acceptability Question? no Choose from: Preference/choice Ranking Rated Acceptability

no go to Panel Setup

Sensory Analytical Question? yes

yes

Simple Same/different Question?

re-open discussion of objectives

no

no

Choose from: Overall difference tests n-alternative forced choice Rated difference from control

go to Panel Setup Nature of Difference Question? yes

Choose from: descriptive analysis techniques or modifications

go to Panel Setup

Fig. 1.4 A sensory evaluation department may interact with many other departments in a food or consumer products company. Their primary interaction is in support of product research and development, much as marketing research supports the

References

end-users of the test results. Logical choices of test participants and appropriate statistical analyses form part of the methodological mix. Analytic tests such as the discrimination and descriptive procedures require good experimental control and maximization of test pcision. Affective tests on the other hand require use of repsentative consumers of the products and test conditions that enable generalization to how products are experienced by consumers in the real world. Sensory tests provide useful information about the human perception of product changes due to ingredients, processing, packaging, or shelf life. Sensory evaluation departments not only interact most heavily with new product development groups but may also provide information to quality control, marketing research, packaging, and, indirectly, to other groups throughout a company (Fig. 1.4). Sensory information reduces risk in decisions about product development and strategies for meeting consumer needs. A wellfunctioning sensory program will be useful to a company in meeting consumer expectations and insuring a greater chance of marketplace success. The utility of the information provided is directly related to the quality of the sensory measurement. . . . , sensory food science stands at the intersection of many disciplines and research traditions, and the stakeholders are many (Tuorila and Monteleone, 2009). Quantities derive from measurement, ps from quantities, comparisons from ps, and victory from comparisons (Sun Tzu – The Art of War (Ch. 4, v.18)).

References Aaron, J. I., Mela, D. J. and Evans, R. E. 1994. The influence of attitudes, beliefs and label information on perceptions of reduced-fat spad. Appetite, 22(1), 25-38. Adams, H. 1918. The Education of Henry Adams. The Modern Library, New York. Amerine, M. A., Pangborn, R. M. and Roessler, E. B. 1965. Principles of Sensory Evaluation of Food. Academic, New York. ASTM E1958. 2008. Standard guide for sensory claim substantiation. ASTM International, West Conshohocken, PA. ASTM. 1989. Sensory evaluation. In celebration of our beginnings. Committee E-18 on Sensory Evaluation of Materials and Products. ASTM, Philadelphia. Barrios, E. X. and Costell, E. 2004. Review: use of methods of research into consumers’ opinions and attitudes in food research. Food Science and Technology International, 10, 359-371.

18 Giboreau, A. and Fleury, H. 2009. A new research platform to contribute to the pleasure of eating and healthy food behaviors through academic and applied food and hospitality research. Food Quality and Preference, 20, 533-536 Giménez, A., Ares, G. and Gámbaro, A. 2008. Consumer attitude toward shelf-life labeling: does it influence acceptance? Journal of Sensory Studies, 23, 871-883. Hein, K. A., Hamid, N., Jaeger, S. R. and Delahunty, C. M. 2009. Application of a written scenario to evoke a consumption context in a laboratory setting: effects on hedonic ratings. Food Quality and Preference. doi:10.1016/j.foodqual.2009.10.003 Helm, E. and Trolle, B. 1946. Selection of a taste panel. Wallerstein Laboratory Communications, 9, 181-194. Jones, L. V., Peryam, D. R. and Thurstone, L. L. 1955. Development of a scale for measuring soldier’s food pferences. Food Research, 20, 512-520. Kamen, J. 1989. Observations, reminiscences and chatter. In: Sensory Evaluation. In celebration of our Beginnings. Committee E-18 on Sensory Evaluation of Materials and Products. ASTM, Philadelphia, pp. 118-122. Kimura, A., Wada, Y., Tsuzuki, D., Goto, S., Cai, D. and Dan, I. 2008. Consumer valuation of packaged foods. Interactive effects of amount and accessibility of information. Appetite, 51, 628-634. Lawless, H. T. 1993. The education and training of sensory scientists. Food Quality and Preference, 4, 51-63. Lawless, H. T. and Claassen, M. R. 1993. The central dogma in sensory evaluation. Food Technology, 47(6), 139-146. Lawless, H. T. and Klein, B. P. 1989. Academic vs. industrial perspectives on sensory evaluation. Journal of Sensory Studies, 3, 205-216. Lee, L., Frederick, S. and Ariely, D. 2006. Try it, you’ll like it. Psychological Science, 17, 1054-1058. MacRae, R. W. and Geelhoed, E. N. 1992. Preference can be more powerful than detection of oddity as a test of discriminability. Perception and Psychophysics, 51, 179-181. Meilgaard, M., Civille, G. V. and Carr, B. T. 2006. Sensory Evaluation Techniques. Fourth Second edition. CRC, Boca Raton. Meiselman, H. L. 1993. Critical evaluation of sensory techniques. Food Quality and Preference, 4, 33-40. Mielby, L. H. and Frøst, M. B. 2009. Expectations and surprise in a molecular gastronomic meal. Food Quality and Preference. doi:10.1016/j.foodqual.2009.09.005 Moskowitz, H. R., Beckley, J. H. and Resurreccion, A. V. A. 2006. Sensory and Consumer Research in Food Product Design and Development. Wiley-Blackwell, New York. Moskowitz, H. R. 1983. Product Testing and Sensory Evaluation of Foods. Food and Nutrition, Westport, CT. Oliver, T. 1986. The Real Coke, The Real Story. Random House, New York. O’Mahony, M. 1988. Sensory difference and pference testing: The use of signal detection measures. Chpater 8 In: H. R. Moskowitz (ed.), Applied Sensory Analysis of Foods. CRC, Boca Raton, FL, pp. 145-175.

1 Introduction O’Mahony, M. 1979. Psychophysical aspects of sensory analysis of dairy products: a critique. Journal of Dairy Science, 62, 1954-1962. Pangborn, R. M. and Dunkley, W. L. 1964. Laboratory procedures for evaluating the sensory properties of milk. Dairy Science Abstracts, 26, 55-121. Park, H. S. and Lee, S. Y. 2003. Genetically engineered food labels, information or warning to consumers? Journal of Food Products Marketing, 9, 49-61. Peryam, D. R. and Swartz, V. W. 1950. Measurement of sensory differences. Food Technology, 4, 390-395. Plotto, A., Barnes, K. W. and Goodner, K. L. 2006. Specific anosmia observed for β-ionone, but not for α-ionone: Significance for flavor research. Journal of Food Science, 71, S401-S406. Shepherd, R., Sparks, P., Belleir, S. and Raats, M. M. 1991/1992. The effects of information on sensory ratings and pferences: The importance of attitudes. Food Quality and Preference, 3, 1-9. Sidel, J. L., Stone, H. and Bloomquist, J. 1981. Use and misuse of sensory evaluation in research and quality control. Journal of Dairy Science, 61, 2296-2302. Siegrist, M. and Cousin, M-E. 2009. Expectations influence sensory experience in a wine tasting. Appetite, 52, 762-765. Skinner, E. Z. 1989. (Commentary). Sensory evaluation. In celebration of our beginnings. Committee E-18 on Sensory Evaluation of Materials and Products. ASTM, Philadelphia, pp. 58-65. Stone, H. and Sidel, J. L. 2004. Sensory Evaluation Practices, Third Edition. Academic, San Deigo. Stone, H., Sidel, J., Oliver, S., Woolsey, A. and Singleton, R. C. 1974. Sensory evaluation by quantitative descriptive analysis. Food Technology 28(1), 24, 26, 28, 29, 32, 34. Sun Tzu (Sun Wu) 1963 (trans.), orig. circa 350 B.C.E. The Art of War. S.B. Griffith, trans. Oxford University. Szczesniak, A. S., Loew, B. J. and Skinner, E. Z. 1975. Consumer texture profile technique. Journal of Food Science, 40, 1253-1257. Tuorila, H. and Monteleone, E. 2009. Sensory food science in the changing society: opportunities, needs and challenges. Trends in Food Science and Technology, 20, 54-62. Wysocki, C. J. and Labows, J. 1984. Inpidual differences in odor perception. Perfumer and Flavorist, 9, 21-24. Yeomans, M. R., Chambers, L., Blumenthal, H. and Blake, A. 2008. The role of expectation in sensory and hedonic evaluation: The case of salmon smoked ice-cream. Food Quality and Preference, 19, 565-573. York, R. K. 1995. Quality assessment in a regulatory environment. Food Quality and Preference, 6, 137-141. Zellner, D. A., Strickhouser, D. and Tornow, C. E. 2004. Disconfirmed hedonic expectations produce perceptual contrast, not assimilation. The American Journal of Psychology, 117, 363-387.

Chapter 2

Physiological and Psychological Foundations of Sensory Function

Abstract This chapter reviews background material underpinning sensory science and sensory evaluation methodologies. Basic and historical psychophysical methods are reviewed as well as the anatomy, physiology, and function of the chemical senses. The chapter concludes with a discussion of multi-modal sensory interactions. There is no conception in man’s mind which hath not at first, totally or in parts, been begotten upon by the organs of sense. -Thomas Hobbes, Leviathan (1651)

Contents 2.1 2.2

2.3

2.4

2.5

Introduction . . . . . . . . . . . . . . . . . Classical Sensory Testing and Psychophysical Methods . . . . . . . . . . . . . . . . . . . 2.2.1 Early Psychophysics . . . . . . . . . . 2.2.2 The Classical Psychophysical Methods . . . . . . . . . . . . . . . 2.2.3 Scaling and Magnitude Estimation . . . 2.2.4 Critiques of Stevens . . . . . . . . . . 2.2.5 Empirical Versus Theory-Driven Functions . . . . . . . . . . . . . . . 2.2.6 Parallels of Psychophysics and Sensory Evaluation . . . . . . . . . . . . . . Anatomy and Physiology and Functions of Taste . . . . . . . . . . . . . . . . . . . 2.3.1 Anatomy and Physiology . . . . . . . . 2.3.2 Taste Perception: Qualities . . . . . . . 2.3.3 Taste Perception: Adaptation and Mixture Interactions . . . . . . . . . . . . . . 2.3.4 Inpidual Differences and Taste Genetics . . . . . . . . . . . . . . . Anatomy and Physiology and Functions of Smell . . . . . . . . . . . . . . . . . . . 2.4.1 Anatomy and Cellular Function . . . . . 2.4.2 Retronasal Smell . . . . . . . . . . . 2.4.3 Olfactory Sensitivity and Specific Anosmia . . . . . . . . . . . . . . . 2.4.4 Odor Qualities: Practical Systems . . . . 2.4.5 Functional Properties: Adaptation, Mixture Suppssion, and Release . . . . . . . . Chemesthesis . . . . . . . . . . . . . . . . 2.5.1 Qualities of Chemesthetic Experience . . . . . . . . . . . . . .

2.5.2

19 20 20 21 23 25 25

Physiological Mechanisms of Chemesthesis . . . . . . . . . . . 2.5.3 Chemical “Heat” . . . . . . . . . . 2.5.4 Other Irritative Sensations and Chemical Cooling . . . . . . . . . . . . . . 2.5.5 Astringency . . . . . . . . . . . . . 2.5.6 Metallic Taste . . . . . . . . . . . . 2.6 Multi-modal Sensory Interactions . . . . . 2.6.1 Taste and Odor Interactions . . . . . . 2.6.2 Irritation and Flavor . . . . . . . . . 2.6.3 Color-Flavor Interactions . . . . . . 2.7 Conclusions . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . .

. .

42 43

. . . . . . . . .

44 45 46 47 47 49 49 50 50

26

2.1 Introduction 27 27 30 30 33 34 34 36 37 38 39 41 41

In order to design effective sensory tests and provide insightful interptation of the results, a sensory professional must understand the functional properties of the sensory systems that are responsible for the data. By a functional property, we mean a phenomenon like mixture interactions such as masking or suppssion. Another example is sensory adaptation, a commonly observed decrease in responsiveness to conditions of more or less constant stimulation. In addition, it is useful to understand the anatomy and physiology of the senses involved as well as their functional limitations. A good example of a functional limitation is the threshold or minimal amount of a stimulus needed for perception. Knowing about the anatomy of the senses

H.T. Lawless, H. Heymann, Sensory Evaluation of Food, Food Science Text Series, DOI 10.1007/978-1-4419-6488-5_2, © Springer Science+Business Media, LLC 2010

19

20

2

Physiological and Psychological Foundations of Sensory Function

can help us understand how consumers and panelists interact with the products to stimulate their senses and by what routes. Different routes of smelling, for example, are the orthonasal or sniffing route, when odor molecules enter the nose from the front (nostrils), versus retronasal smell, when odor molecules pass into the nose from the mouth or from breathing out, and thus have a reversed airflow pathway from that of external sniffing. Another basic area that the sensory professional should have as background knowledge involves the sensory testing methods and human measurement procedures that are the historical antecedents to the tests we do today. This is part of the science of psychophysics, the quantification and measurement of sensory experiences. Psychophysics is a very old discipline that formed the basis for the early studies in experimental psychology. Parallels exist between psychophysics and sensory evaluation. For example, the difference test using paired comparisons is a version of the method used for measuring difference thresholds called the method of constant stimuli. In descriptive analysis with trained panels, we work very hard to insure that panelists use singular uni-dimensional scales. These numerical systems usually refer to a single sensory continuum like sweetness or odor strength and are thus based on changes in perceived intensity. They do not consider multiple attributes and fold them into a single score like the old-quality grading methods. Thus there is a clear psychophysical basis for the attribute scales used in descriptive analysis. This chapter is designed to provide the reader some background in the sensory methods of psychophysics. A second objective is to give an overview of the structure and function of the chemical senses of taste, smell, and the chemesthetic sense. Chemesthesis refers to chemically induced sensations that seem to be at least partly tactile in nature, such as pepper heat, astringency, and chemical cooling. These three senses together comprise what we loosely call flavor and are the critical senses for appciating foods, along with the tactile, force, and motion-related experiences that are part of food texture and mouthfeel. Texture is dealt with in Chapter 11 and color and appearance evaluations in Chapter 12. The auditory sense is not a large part of food perception, although many sounds can be perceived when we eat or manipulate foods. These provide another sense modality to accompany and reinforce our texture perceptions, as in the case of crisp

or crunchy foods, or the audible hissing sound we get from carbonated beverages (Vickers, 1991). One growing area of interest in the senses concerns our human biopersity, differences among people in sensory function. These differences can be due to genetic, dietary/nutritional, physiological (e.g., aging), or environmental factors. The research into the genetics of the chemical senses, for example, has experienced a period of enormous expansion since the first edition of this book. The topic is too large and too rapidly changing to receive a comphensive treatment here. We will limit our discussion of inpidual differences and genetic factors to those areas that are well understood, such as bitter sensitivity, smell blindness, and color vision anomalies. The sensory practitioner should be mindful that people exist in somewhat different sensory worlds. These differences contribute to the persity of consumer pferences. They also limit the degree to which a trained panel can be “calibrated” into uniform ways of responding. Inpidual differences can impact sensory evaluations in many ways.

2.2 Classical Sensory Testing and Psychophysical Methods 2.2.1 Early Psychophysics The oldest branch of experimental psychology is that of psychophysics, the study of relationships between physical stimuli and sensory experience. The first true psychophysical theorist was the nineteenth century German physiologist, E. H. Weber. Building on earlier observations by Bernoulli and others, Weber noted that the amount that a physical stimulus needed to be increased to be just perceivably different was a constant ratio. Thus 14.5 and 15 ounces could be told apart, but with great difficulty, and the same could be said of 29 and 30 ounces or 14.5 and 15 drams (Boring, 1942). This led to the formulation of Weber’s law, generally written nowadays as I/I = k

(2.1)

where I is the increase in the physical stimulus that was required to be just discriminably different from some starting level, I. The fraction, I/I, is sometimes called the “Weber fraction” and is an index of

2.2 Classical Sensory Testing and Psychophysical Methods

how well the sensory system detects changes. This relationship proved generally useful and provided the first quantitative operating characteristic of a sensory system. Methods for determining the difference threshold or just-noticeable-difference (j.n.d.) values became the stock in trade of early psychological researchers. These methods were codified by G. T. Fechner in a book called Elemente der Psychophysik (Elements of Psychophysics) in 1860. Fechner was a philosopher as well as a scientist and developed an interest in Eastern religions, in the nature of the soul, and in the Cartesian mind-body dichotomy. Fechner’s broader philosophical interests have been largely overlooked, but his little book on sensory methods was to become a classic text for the psychology laboratory. Fechner also had a valuable insight. He realized that the j.n.d. might be used as a unit of measurement and that by adding up j.n.d.s one could construct a psychophysical relationship between physical stimulus intensity and sensory intensity. This relationship approximated a log function, since the integral of 1/x dx is proportional to the natural log of x. So a logarithmic relationship appeared useful as a general psychophysical “law:” S = k log I

(2.2)

where S is sensation intensity and I is once again the physical stimulus intensity. This relationship known as Fechner’s law was to prove a useful rule of thumb for nearly 75 years, until it was questioned by acoustical researchers who supplanted it with a power law (see Section 2.2.3).

21

with a particular type of measured response of sensory systems. The method of limits was well suited to determine absolute or detection thresholds. The method of constant stimuli could be used to determine difference thresholds and the method of adjustment to establish sensory equivalence. In the method of limits the physical stimulus is changed by successive discrete steps until a change in response is noted. For example, when the stimulus is increasing in intensity, the response will change from “no sensation” to “I detect something.” When the stimulus is decreasing in intensity, at some step the response will change back to “no sensation.” Over many trials, the average point of change can be taken as the person’s absolute threshold (see Fig. 2.1). This is the minimum intensity required for detection of the stimulus. Modern variations on this method often use only an ascending series and force the participants to choose a target sample among alternative “blank” samples at each step. Each concentration must be discriminated from a background level such as plain water in the case of taste thresholds. Forced-choice methods for determining thresholds are discussed in detail in Chapter 6. In the method of constant stimuli, the test stimulus is always compared against a constant reference level (a standard), usually the middle point on a series of physical intensity levels. The subject’s job is to respond to each test item as “greater than” or “less

2.2.2 The Classical Psychophysical Methods Fechner’s enduring contribution was to assemble and publish the details of sensory test methods and how several important operating characteristics of sensory systems could be measured. Three important methods were the method of limits, the method of constant stimuli (called the method of right and wrong cases in those days), and the method of adjustment or average error (Boring, 1942). The methods are still used today in some research situations and variations on these methods form part of the toolbox of applied sensory evaluation. Each of the three methods was associated

CONCENTRATION

16.0 8.0 4.0 2.0 1.0 0.5 .25 A1

D1

A2

D2

A3

D3

TRIAL

Fig. 2.1 An example of the method of limits. The circled reversal points would be averaged to obtain the person’s threshold. A: ascending series. D: descending series. In taste and smell, only ascending series are commonly used to pvent fatigue, adaptation or carry-over of persistent sensations.

22

2

Physiological and Psychological Foundations of Sensory Function

Frequency (percent) judged “sweeter” than the standard

80

60

40 UDL

20 standard

2

4

6

8

10

12

14

16

18

SUCROSE CONCENTRATION (%)

Fig. 2.2 A psychometric function derived from the Method of Constant Stimuli, a repeated series of paired comparisons against a constant (standard) stimulus, in this case 10% sucrose. Frequency of judgments in which the comparison stimulus is

judged sweeter than the standard are plotted against concentration. The difference threshold is determined by the concentration difference between the standard and the interpolated 75% (or 25%) point. UDL: Upper difference limen (threshold).

than” the standard. Many replications of each intensity level are psented. The percentage of times the response is “greater than” can be plotted as in Fig. 2.2. This S-shaped curve is called a psychometric function (Boring, 1942). The difference threshold was taken as the difference between the 50 and 75% points interpolated on the function. The method of constant stimuli bears a strong resemblance to current techniques of paired comparison, with two exceptions. One point of difference is that the method was geared toward interval estimation, rather than testing for statistically significant differences. That is, the technique estimated points on the psychometric function (25, 50, and 75%) and researchers were not concerned with statistical significance of difference tests. Also, a range of comparison stimuli were tested against the standard and not just a single paired comparison of products. The third major method in classical psychophysics was the method of adjustment or average error. The subject was given control over a variable stimulus like a light or a tone and asked to match a standard in brightness or loudness. The method could be used to determine difference thresholds based on the variability of the subject over many attempts at matching, for example, using the standard deviation as a measure

of difference threshold. A modern application is in measuring sensory tradeoff relationships. In this type of experiment the duration of a very brief tone could be balanced against a varying sound pssure level to yield a constant perception of loudness. Similarly, the duration of a flash of light could be traded off against its photometric intensity to create a constant perceived brightness. For very brief tones or brief flashes, there is summation of the intensity over time in the nervous system, so that increasing duration can be balanced against decreasing physical intensity to create a constant perception. These methods have proven useful in understanding the physiological response of different senses to the temporal properties of stimuli, for example, how the auditory and visual systems integrate energy over time. Adjustment methods have not proven so useful for assessing sensory equivalence in applied food testing, although adjustment is one way of trying to optimize an ingredient level (Hernandez and Lawless, 1999; Mattes and Lawless, 1985). Pangborn and co-workers employed an adjustment method to study inpidual pferences (Pangborn, 1988; Pangborn and Braddock, 1989). Adding flavors or ingredients “to taste” at the benchtop is a common way of initially formulating

2.2 Classical Sensory Testing and Psychophysical Methods

products. It is also fairly common to make formula changes to produce approximate sensory matches to some target, either a standard formula or perhaps some competitor’s successful product. However, the method as applied in the psychophysics laboratory is an unwieldy technique for the senses of taste and smell where elaborate equipment is needed to provide adjustable stimulus control. So methods of equivalency adjustment are somewhat rare with food testing.

23

they are a log scale of sound pssure relative to a reference (db = 20 log (P/P0 ) where P is the sound pssure and P0 is the reference sound pssure, usually a value for absolute threshold). However, discrepancies were observed between decibels and loudness proportions. Instead, Stevens found with the direct magnitude estimation procedure that loudness was a power function of stimulus intensity, with an exponent of about 0.6. Scaling of other sensory continua also gave power functions, each with its characteristic exponent (Stevens, 1957, 1962). Thus the following relationship held:

2.2.3 Scaling and Magnitude Estimation S = kI n or log S = n log I + log k A very useful technique for sensory measurement has been the direct application of rating scales to measure the intensity of sensations. Historically known as the “method of single stimuli,” the procedure is highly cost efficient since one stimulus psentation yields one data point. This is in contrast to a procedure like the method of constant stimuli, where the psentation of many pairs is necessary to give a frequency count of the number of times each level is judged stronger than a standard. Rating scales have many uses. One of the most common is to specify a psychophysical function, a quantitative relationship between the perceived intensity of a sensation and the physical intensity of the stimulus. This is another way of describing a dose-response curve or in other words, capturing the input-output function of a sensory system over its dynamic range. The technique of magnitude estimation grew out of earlier procedures in which subjects would be asked to fractionate an adjustable stimulus. For example, a subject would be asked to adjust a light or tone until it seemed half as bright as a comparison stimulus. The technique was modified so that the experimenter controlled the stimulus and the subject responded using (unrestricted) numbers to indicate the proportions or ratios of the perceived intensities. Thus if the test stimulus was twice as bright as the standard, it would be assigned a number twice as large as the rating for the standard and if one-third as bright, a number one-third as large. An important observation in S. S. Stevens’ laboratory at Harvard was that the loudness of sounds was not exactly proportional to the decibel scale. If Fechner’s log relationship was correct, rated loudness should grow in a linear fashion with decibels, since

(2.3)

where n was the characteristic exponent and k was a proportionality constant determined by the units of measurement. In other words, the function formed a straight line in a log-log plot with the exponent equal to the slope of the linear function. This was in contrast to the Fechnerian log function which was a straight line in a semilog plot (response versus log physical intensity). One of the more important characteristics of a power function is that it can accommodate relationships that are expanding or positively accelerated while the log function does not. The power function with an exponent less than one fits a law of diminishing returns, i.e., larger and larger physical increases are required to maintain a constant proportional increase in the sensation level. Other continua such as response to electric shocks and some tastes were found to have a power function exponent greater than one (Meiselman, 1971; Moskowitz, 1971; Stevens, 1957). A comparison of power functions with different exponents is shown in Fig. 2.3. Many sensory systems show an exponent less that one. This shows a compssive energy relationship that may have adaptive value for an organism responding to a wide range of energy in the environment. The range from the loudest sound one can tolerate to the faintest audible tone is over 100 dB. This repsents over 10 log units of sound energy, a ratio of 10 billion to one. The dynamic range for the visual response of the eyes to different levels of light energy is equally broad. Thus exponents less than one have ecological significance for sensory systems that are tuned to a broad range of physical energy levels.

24

2

Physiological and Psychological Foundations of Sensory Function

Psychological Magnitude, S

Electric Shock

Apparent Length n=1

Brightness n1 n=1 n

1 2

(4.3)

The verbal form of the alternative hypothesis is that the population would be correct (saying that AB and BA pairs are different and that AA and BB pairs are the same) more than half the time. The results of the paired difference test will only indicate whether the panelists could significantly discriminate between the samples. Unlike the paired directional test, no specification or direction of difference is indicated. In other words, the sensory scientist will only know that the samples are perceptibly different but not in which attribute(s) the samples differed. An alternative analysis is psented in the Appendix to this chapter, where each panelist sees an identical pair (AA or BB) and one test pair (AB or BA) in randomized sequence.

4.2.2 Triangle Tests In the triangle test, three samples are psented simultaneously to the panelists, two samples are from the same formulation and one is from the different formulation. Each panelist has to indicate either which sample is the odd sample or which two samples are most similar. The usual form of the score sheet asks the panelist to indicate the odd sample. However, some sensory specialists will ask the panelist to indicate the pair of similar samples. It probably does not matter which question is asked. However, the sensory specialist should not change the format when re-using panelists since they will get confused. See Fig. 4.3 for a sample score sheet. Similarly to the paired difference test the panelist must be trained to understand the task as described by the score sheet. The null hypothesis for the triangle test states that the long-run probability (Pt ) of making a correct selection when there is no perceptible difference between the samples is one in three (H0 :Pt = 1/3). The alternative hypothesis states that the probability that the underlying population will make the correct decision

Fig. 4.3 Example of a triangle score sheet.

84

4

1 3

(4.4)

This is a one-sided alternative hypothesis and the test is one tailed. In this case there are six possible serving orders (AAB, ABA, BAA, BBA, BAB, ABB) which should be counterbalanced across all panelists. As with the difference paired comparison, the triangle test allows the sensory specialist to determine if two samples are perceptibly different but the direction of the difference is not indicated by the triangle test. Again, the sensory scientist will only know that the samples are perceptibly different but not in which attribute(s) the samples differed.

Discrimination Testing

Again, the panelists should be trained to perform the task as described by the score sheet correctly. Duo-trio tests allow the sensory specialist to determine if two samples are perceptibly different but the direction of the difference is not indicated by the duo-trio test. In other words, the sensory scientist will only know that the samples are perceptibly different but not in which attribute(s) the samples differed. There are two formats to the duo-trio test, namely the constant reference duo-trio test and the balanced reference duo-trio test. From the point of view of the panelists the two formats of the duo-trio test are identical (see Figs. 4.4a and b), but to the sensory specialist the two formats differ in the sample(s) used as the reference. 4.2.3.1 Constant Reference Duo-Trio Test

1 2

(4.5)

In this case, all panelists receive the same sample formulation as the reference. The constant reference duo-trio test has two possible serving orders (RA BA, RA AB) which should be counterbalanced across all panelists. The constant reference duo-trio test seems to be more sensitive especially if the panelists have had prior experience with the product (Mitchell, 1956). For example, if product X is the current formulation (familiar to the panelists) and product Z is a new reformulation then a constant reference duo-trio test with product X as reference would be the method of choice.

Before starting please rinse your mouth with water and expectorate. There are three samples in each of the two duo−trio sets for you to evaluate. In each set, one of the coded pairs is the same as the reference. For each set taste the reference first. Then taste each of the coded samples in the sequence psented, from left to right. Take the entire sample in your mouth. NO RETASTING. Circle the number of the sample which is most similar to the reference. Do not swallow any of the sample or the water. Expectorate into the container provided. Rinse your mouth with water between sets 1 and 2.

Fig. 4.4a Example of a constant reference duo-trio score sheet.

Set 1

Reference

2

Reference

4.2 Types of Discrimination Tests

85

Fig. 4.4b Example of a balanced reference duo-trio score sheet.

Reference

2

Reference

reference and the other half of the panelists receive the other sample formulation as the reference. In this case, there are four possible serving orders (RA BA, RA AB, RB AB, RB BA) which should be counterbalanced across all panelists. This method is used when both products are prototypes (unfamiliar to the panelists) or when there is not a sufficient quantity of the more familiar product to perform a constant reference duo-trio test.

differ in the specified dimension and which sample is higher in perceived intensity of the specified attribute. The danger is that other sensory changes will occur in a food when one attribute is modified and these may obscure the attribute in question. Another version of the n-AFC asks panelists to pick out the weakest or strongest in overall intensity, rather than in a specific attribute. This is a very difficult task for panelists when they are confronted with a complex food system.

4.2.5 A-Not-A tests There are two types of A-not-A tests referenced in the literature. The first and the more commonly used version has a training phase with the two products followed by monadic evaluation phase (Bi and Ennis, 2001a, b), we will call this the standard A-not-A test. The second version is essentially a sequential paired difference test or simple difference test (Stone and Sidel, 2004), which we will call the alternate A-notA test. The alternate A-not-A test is not frequently used. In the next section we will discuss the alternate A-not-A test first since the statistical analysis for this version is similar to that of the paired comparison discrimination test. The statistical analyses for the various standard A-not-A tests are based on a different theory and somewhat more complex and will be discussed later.

4.2.5.1 Alternate A-Not-A test This is a sequential same/difference paired difference test where the panelist receives and evaluates the first

86

4

Fig. 4.5 Example of a three-alternative forced choice score sheet.

relevant to the objective of the study. However, the differences in color or shape or size have to be very subtle and only obvious when the samples are psented simultaneously. If the differences are not subtle the panelists are likely to remember these and they will make their decision based on these extraneous differences.

4.2.5.2 Standard A-Not-A Test Panelists inspect multiple examples of products that are labeled “A” and usually also products that are labeled “not-A.” Thus there is a learning period. Then once the training period has been completed the panelists receive samples one at a time and are asked whether each one is either A or not-A. As discussed by Bi and Ennis (2001a) the standard A-not-A test potentially has four different designs. For the monadic A-not-A test the panelist, after the training phase, is psented with a single sample (either A or not-A). In the paired A-not-A version the panelist, after completion of the training phase, is psented with a pair of samples, sequentially (one A and one not-A, counter balanced across panelists). In the replicated monadic A-not-A version the panelist, after completion of training, receives a series of samples of either A or not-A but not both. This version is rarely used in practice. Lastly, in the replicated mixed A-not-A version the panelist, after completion of training, receives a series of A and not-A samples. Each of these different formats requires different statistical models and using an inappropriate model could lead to a misleading conclusion. As described by Bi and Ennis (2001a) “The statistical models for the A-Not A method are different from that of other discrimination methods such as the m-AFC, the triangle, and the duo-trio methods.” “Pearson’s and McNemar’s chi-square statistics with one degree of freedom can be used for the

4.2 Types of Discrimination Tests

standard A-Not A method while binomial tests based on the proportion of correct responses can be used for the m-AFC, the triangle, and the duo-trio methods. The basic difference between the two types of difference tests is that the former involves a comparison of two proportions (i.e., the proportion of “A” responses for the A sample versus that for the Not A sample) or testing independence of two variables (sample and response) while the latter is a comparison of one proportion with a fixed value (i.e., the proportion of correct responses versus the guessing probability)”. Articles by Bi and Ennis (2001a, b) clearly describe data analysis methods for these tests.. Additionally, the article by Brockhoff and Christensen (2009) describes a R-package called SensR (http://www.cran.rproject.org/package=sensR/) that may be used for the data analyses of some Standard A-not-A tests. The data analyses associated with the standard A-not-A tests are beyond the scope of this textbook, but see the Appendix of this chapter which shows the application of the McNemar chi-square for a simple A-not-A test where each panelist received one standard product (a “true” example of A) and one test product. Each is psented separately and a judgment is collected for both products.

87

4.2.6.2 The Harris-Kalmus Test

4.2.6 Sorting Methods In sorting tests the panelists are given a series of samples and they are asked to sort them into two groups. The sorting tests can be extremely fatiguing and are not frequently used for taste and aroma sensory evaluation but they are used when sensory specialists want to determine if two samples are perceptibly different in tactile or visual dimensions. The sorting tests are statistically very efficient since the long-run probability of the null hypotheses of the sorting tests can be very small. For example, the null hypothesis of the two-out-of-five test is 1 in 10 (P2/5 = 0.1) and for the Harris-Kalmus test the null hypothesis is 1 in 70 (P4/8 = 0.0143). These tests are discussed below. 4.2.6.1 The Two-Out-of-Five Test The panelists receive five samples and are asked to sort the samples into two groups, one group should contain the two samples that are different from the other

88

4

Discrimination Testing

4.2.8 Dual-Standard Test The dual standard was first used by Peryam and Swartz (1950) with odor samples. It is essentially a duo-trio test with two reference standards-the control and the variant. The two standards allow the panelists to create a more stable criterion as to the potential difference between the samples. The potential serving orders for this test are R(A) R(B) , AB, R(A) R(B) BA, R(B) R(A) AB, R(B) R(A) BA. The probability of guessing the correct answer by chance is 0.5 and the data analyses for this test are identical to that of the duo-trio test. Peryam and Swartz felt quite strongly that the technique would work best with odor samples due to the relatively quick recovery and that the longer recovery associated with taste samples would pclude the use of the test. The test was used by Pangborn and Dunkley (1966) to detect additions of lactose, algin gum, milk salts, and proteins to milk. O’Mahony et al. (1986) working with lemonade found that the dual-standard test elicited superior performance over the duo-trio test. But O’Mahony (personal communication, 2009) feels that this result is in error, since the panelists were not instructed to evaluate the standards prior to each pair evaluation and therefore the panelists were probably reverting to a 2-AFC methodology. This would be in agreement with Huang and Lawless (1998) who studied sucrose additions to orange juice and they did not find superiority in performance between the dual standard and the duo-trio or the ABX tests.

4.3 Reputed Strengths and Weaknesses of Discrimination Tests If the batch-to-batch variation within a sample formulation is as large as the variation between formulations

4.4 Data Analyses

then the sensory specialist should not use triangle or duo-trio tests (Gacula and Singh, 1984). In this case the paired comparison difference test could be used but the first question that the sensory specialist should ask is whether the batch-to-batch variation should not be studied and improved prior to any study of new or different formulations. The major weakness of all discrimination tests is that they do not indicate the magnitude of the sensory difference(s) between the sample formulations. As the simple discrimination tests are aimed at a yes/no decision about the existence of a sensory difference, they are not designed to give information on the magnitude of a sensory difference, only whether one is likely to be perceived or not. The sensory specialist should not be tempted to conclude that a difference is large or small based on the significance level or the probability (pvalue) from the statistical analysis. The significance and p-value depend in part upon the number of panelists in the test as well as the inherent difficulty of the particular type of discrimination test method. So these are no acceptable indices of the size of the perceivable difference. However, it is sensible that a comparison in which 95% of the judges answered correctly has a larger sensory difference between control and test samples than a comparison in which performance was only at 50% correct. This kind of reasoning works only if a sufficient number of judges were tested, the methods were the same, and all test conditions were constant. Methods for interval level scaling of sensory differences based on proportions of correct discriminations in forced choice tests are discussed further in Chapter 5 as Thurstonian scaling methods. These methods are indirect measures of small differences. They are also methodologically and mathematically complex and require certain assumptions to be met in order to be used effectively. Therefore we feel that the sensory specialist is wiser to base conclusions about the degree of difference between samples on scaled (direct) comparisons, rather than indirect estimates from choice performance in discrimination tests. However, there are alternative opinions in the sensory community and we suggest that interested parties read Lee and O’Mahony (2007). With the exception of the 2-AFC and 3-AFC tests the other discrimination tests also do not indicate the nature of the sensory difference between the samples. The major strength of the discrimination tests is that the task that the panelists perform is quite simple and

89

intuitively grasped by the panelists. However, it is frequently the very simplicity of these tests that lead to the generation of garbage data. Sensory specialists must be very aware of the power, replication, and counterbalancing issues associated with discrimination tests. These issues are discussed later in this chapter.

4.4 Data Analyses The data from discrimination tests may be analyzed by any of the following statistical methods. The three data analyses are based on the binomial, chi-square, or normal distributions, respectively. All these analyses assume that the panelists were forced to make a choice. Thus they had to choose one sample or another and could not say that they did not know the answer. In other words, each panelist either made a correct or incorrect decision, but they all made a decision.

4.4.1 Binomial Distributions and Tables The binomial distribution allows the sensory specialist to determine whether the result of the study was due to chance alone or whether the panelists actually perceived a difference between the samples. The following formula allows the sensory scientists to calculate the probability of success (of making a correct decision; p) or the probability of failure (of making an incorrect decision; q) using the following formula. P(y) =

n! py pn−y y!(n − y)!

(4.6)

where n = total number of judgments y = total number of correct judgments p = probability of making the correct judgment by chance In this formula, n! describes the mathematical factorial function which is calculated as n×(n-1)× (n-2). . .×2×1. Before the widespad availability of calculators and computers, calculation of the binomial formula was quite complicated, and even now

90

4

it remains somewhat tedious. Roessler et al. (1978) published a series of tables that use the binomial formula to calculate the number of correct judgments and their probability of occurrence. These tables make it very easy to determine if a statistical difference were detected between two samples in discrimination tests. However, the sensory scientist may not have these tables easily available thus he/she should also know how to analyze discrimination data using statistical tables that are more readily available. We abridged the tables from Roessler et al. (1978) into Table 4.3. Using this table is very simple. For example, in a duo-trio test using 45 panelists, 21 panelists correctly matched the sample to the reference. In Table 4.3, in the section for duo-trio tests, we find that the table value for 45 panelists at 5% probability is 29. This value is larger than 21 and therefore the panelists could not detect a difference between the samples. In a different study, using a triangle test, 21 of 45 panelists correctly identified the odd sample. In Table 4.3, in the section for triangle tests, we find that the table value for 45 panelists at 5% probability is 21. This value is equal to 21 and therefore the panelists could detect a significant difference between the samples at the alpha probability of 5%.

Discrimination Testing

O2 = observed number incorrect choices E1 = expected number of correct choices E1 is equal to total number of observations (n) times probability (p) of a correct choice, by chance alone in a single judgment where p = 0.100 for the two-out-of-five test p = 0.500 for duo-trio, paired difference, paired directional, alternate A-not-A tests p = 0.333 for triangle tests E2 = expected number of incorrect choices E2 is equal to total number of observations (n) times probability (q) of an incorrect choice, by chance alone in a single judgment where q = 1−p q = 0.900 for the two-out-of-five test q = 0.500 for duo-trio, paired difference, paired directional, alternate A-not-A tests, ABX tests q = 0.667 for triangle tests The use of discrimination tests allows the sensory scientist to determine whether two products are statistically perceived to be different, therefore the degrees of freedom equal one (1). Therefore, a χ 2 table using df = 1 should be consulted, for alpha (α) at 5% the critical χ2 value is 3.84. For other alpha levels consult the chi-square table in the Appendix.

4.4.2 The Adjusted Chi-Square (χ2 ) Test

4.4.3 The Normal Distribution and the Z-Test on Proportion The chi-square distribution allows the sensory scientist to compare a set of observed frequencies with a matching set of expected (hypothesized) frequencies. The chi-square statistic can be calculated from the following formula (Amerine and Roessler, 1983), which includes the number -0.5 as a continuity correction. The continuity correction is needed because the χ 2 distribution is continuous and the observed frequencies from discrimination tests are integers. It is not possible for one-half of a person to get the right answer and so the statistical approximation can be off by as much as 1/2, maximally. χ = 2

The sensory specialist can also use the areas under the normal probability curve to estimate the probability of chance in the results of discrimination tests. The tables associated with the normal curve specify areas under the curve (probabilities) associated with specified values of the normal deviate (z). The following two formulae (Eqs. (4.8) and (4.9)) can be used to calculate the z-value associated with the results of a specific discrimination test (Stone and Sidel, 1978):

z=

+ 1.645(1.5) (X/N)(1−X/N) N

(5.7) Here is a worked example. Suppose we do a triangle test with 60 panelists, and 30 get the correct answer. We can ask the following questions: What is the best estimate of the number of discriminators? The proportion of discriminators? What is the upper 95% confidence interval on the number of discriminators? What is the confidence interval on the proportion of discriminators? Finally, could we conclude that there is significant similarity, based on a maximum allowable proportion of discriminators of 50%? The solution is as follows: Let X = number correct, D = number of discriminators, so X = 30 and N = 60. We have 1.5(30)-0.5(60) = D or 15 discriminators, or 25% of our judges detecting the difference, as our best estimate.

110

5 Similarity, Equivalence Testing, and Discrimination Theory

The standard error is given by

1.5

(30/60)

(6.4)

variation was used by Stevens et al. (1988) in a landmark paper on the inpidual variability in olfactory thresholds. In this case, five correct pairs were required to score the concentration as correctly detected, and this performance was confirmed at the next highest concentration level. The most striking finding of this study was that among the three inpiduals tested 20 times, their inpidual thresholds for butanol, pyridine, and phenylethylmethylethyl carbinol (a rose odorant) varied over 2,000- to 10,000-fold in concentration. Variation within an inpidual was as wide as the variation typically seen across a population of test subjects. This surprising result suggests that day-to-day variation in olfactory sensitivity is large and that thresholds for an inpidual are not very stable (for an example, see Lawless et al., 1995). More recent work using extensive testing of inpiduals at each concentration step suggests that these estimates of variability may be high. Walker et al. (2003) used a simple yes/no procedure (like the A, not-A test, or signal detection test) with 15 trials of targets and 15 trials of blanks at each concentration level. Using a model for statistical significant differences between blank and target trials, they were able to get sharp gradients for the inpidual threshold estimates. In summary, an ascending forced-choice method is a reasonably useful compromise between the need to pcisely define a threshold level and the problems encountered in sensory adaptation and observer fatigue when extensive measurements are made. However, the user of an ascending forced-choice procedure should be aware of the procedural choices that can affect the obtained threshold value. The following choices will affect the measured value: the number of alternatives (both targets and blanks), the stopping rule, or the number of correct steps in a row required to establish a threshold, the number of replicated correct trials required at any one step, and the rule to determine at what level of concentration steps the threshold value is assigned. For example, the inpidual threshold might be assigned at the lowest level correct, the geometric mean between the lowest level correct and highest level incorrect. Other specific factors include the chosen step size of concentration units (factors of two or three are common in taste and smell), the method of averaging or combining replicated ascending runs on the same inpidual and finally the method of averaging or combining group data. Geometric means are commonly used for the last two purposes.

136

6 Measurement of Sensory Thresholds

6.7 Probit Analysis

7.0

97.5 95 (90)

90 (80)

(85)

6.0

Cumulative Percentage

80 70

(57)

60 5.0

50

(40)

40

PROBIT

It is often useful to apply some kind of transformation or graphing method to the group data to linearize the curve used to find the 50% point in a group. Both the psychometric curve that repsents the behavior of an inpidual in multiple trials of a threshold test and the cumulative distribution of a group will resemble an S-shaped function similar to the cumulative normal distribution. A number of methods for graphing such data are shown in the ASTM standard E-1432 (ASTM, 2008b). One simple way to graph the data is simply to plot the cumulative percentages on “probability paper.” This p-printed solution provides a graph in which equal standard deviations are marked off along the ordinate, effectively stretching the percentile intervals at the ends and compssing them in the midrange to conform to the density of the normal distribution. Another way to achieve the straightening of the S-shaped response curve is to transform the data by taking z-scores. Statistical packages for data analysis often provide options for transformation of the data. A related method was once widely used in threshold measurement, called Probit analysis (ASTM, 2008b; Dravnieks and Prokop, 1975; Finney, 1971). In this approach, the inpidual points are transformed relative to the mean value, pided by the standard deviation and then a constant value of +5 is added to translate all the numbers to positive values for convenience. A linear fitted function can now be interpolated at the value of 5 to estimate the threshold as in Fig. 6.6. The conversion (to a z-score +5) tends to make an S-shaped curve more linear. An example of this can be found in the paper by Brown et al. (1978), using data from a multiple paired test. First the percent correct is adjusted for chance. Then the data are transformed from the percent correct (across the group) at each concentration level by conversion to z-scores and a constant of 5 is added. The mean value or Probit equal to 5 can be found by interpolation or curve fitting. An example of this technique for estimating threshold from a group of 20 panelists is shown in Meilgaard et al. (1991) and in ASTM (2008b). Probit plots can be used for any cumulative proportions, as well as ranked data and analysis of inpiduals who are more extensively tested than in the 3-AFC method example shown earlier.

30 (18)

20 10

4.0 (6)

5 2.5

3.0 1

2 4 8 16 32 64 Threshold Concentration (log scale)

Fig. 6.6 An example of Probit analysis. Numbers in parentheses are the cumulative percentages of panelists reaching threshold at each concentration step. Note the uneven scale on the left axis. Probits mark off equal standard deviations and are based on the z-score for any proportion plus the constant, 5. Interpolation at the 50% or Probit 5.0 gives the threshold value.

6.8 Sensory Adaptation, Sequential Effects, and Variability Inpidual variability, both among a group of people and within an inpidual over repeated measurements psents a challenge to the idea that the threshold is anything like a fixed value. For example, stable olfactory thresholds of an inpidual are difficult to measure. The test-retest correlation for inpidual’s olfactory threshold is often low (Punter, 1983). Even within an inpidual, threshold values will generally decrease with practice (Engen, 1960; Mojet et al., 2001; Rabin and Cain, 1986), and superimposed upon this practice effect is a high level of seemingly random variation (Stevens et al., 1988). Inpiduals may become sensitive to odorants to which they were formerly anosmic, apparently through simple exposure (Wysocki et al., 1989). Increased sensitivity as a function of exposure may be a common phenomenon among women of childbearing age (Dalton et al., 2002; Diamond et al., 2005). Sensory adaptation and momentary changes in sensitivity due to sequences may have occurred in

6.9 Alternative Methods: Rated Difference, Adaptive Procedures, Scaling

the experiments of Stevens et al. (1988) and could have contributed to some instability in the measurements. As pdicted by sequential sensitivity analysis (Masuoka et al., 1995; O’Mahony and Odbert, 1985) the specific stimulus sequence will render discrimination more or less difficult. After the stronger of two stimuli, an additional second strong stimulus psented next may be partially adapted and seem weaker than normal. Stevens et al. remarked that sometimes subjects would get all five pairs correct at one level with some certainty that they “got the scent” but lost the signal at the next level before getting it back. This report and the reversals of performance in threshold data are consistent with adaptation effects temporarily lessening sensitivity. The sensory impssion will sometimes “fade in and out” at levels near threshold. In attempts to avoid adaptation effects, other researchers have gone to fewer psentations of the target stimulus. For example, Lawless et al. (1995) used one target among three blank stimuli, a 4-AFC test that has appeared in pvious studies (e.g., Engen, 1960; Punter, 1983). This lowers the chance performance level and lessens the potential adaptation at any one concentration step. To guard against the effects of correct guessing, threshold was taken as the lowest concentration step with a correct choice when all higher concentrations were also correct. Thresholds were measured in duplicate ascending runs in a test session, and a duplicate session of two more ascending runs was run on a second day. Correlations across the four ascending series ranged from 0.75 to 0.92 for cineole and from 0.51 to 0.92 for carvone. For carvone, thresholds were better duplicated within a day (r = 0.91 and 0.88) than across days (r from 0.51 to 0.70). This latter result suggests some drift over time in odor thresholds, in keeping with the variability seen by Stevens et al. (1988). However, results with this ascending method may not be this reliable for all compounds. Using the ascending 4-AFC test and a sophisticated olfactometer, Punter (1983) found median retest correlations for 11 compounds to be only 0.40. The sense of taste may fare somewhat better. In a study of electrogustometric thresholds with ascending paired tests requiring five correct responses, retest correlations for an elderly population were 0.95 (Murphy et al., 1995). In many forced-choice studies, high variability in smell thresholds is also noted across the testing pool.

137

Brown et al. (1978) stated that for any test compound, a number of insensitive inpiduals would likely be seen in the data set, when 25 or more persons were tested to determine an average threshold. Among any given pool of participants, a few people with otherwise normal smell acuity will have high thresholds. This is potentially important for sensory professionals who need to screen panelists for detection of specific flavor or odor notes such as defects or taints. In an extensive survey of thresholds for branched-chain fatty acids, Brennand et al. (1989) remarked that “some judges were unable to identify the correct samples in the pairs even in the highest concentrations provided” and that ” panelists who were sensitive to most fatty acids found some acids difficulty to perceive” (p. 109). Wide variation in sensitivity was also observed to the common flavor compound, diacetyl, a buttery-smelling by-product of lactic bacteria fermentation (Lawless et al., 1994). Also, simple exposure to some chemicals can modify specific anosmia and increase sensitivity (Stevens and O’Connell, 1995).

6.9 Alternative Methods: Rated Difference, Adaptive Procedures, Scaling 6.9.1 Rated Difference from Control Another practical procedure for estimating threshold has involved the use of ratings on degree-of-difference scales, where a sample containing the to-be-recognized stimulus is compared to some control or blank stimulus (Brown et al., 1978; Lundahl et al., 1986). Rated difference may use a line scale or a category scale, ranging from no difference or “exact same” to a large difference, as discussed in Chapter 4. In these procedures ratings for the sensory difference from the control sample will increase as the intensity of the target gets stronger. A point on the plot of ratings versus concentration is assigned as threshold. In some variations on this method, a blind control sample is also rated. This provides the opportunity to estimate a baseline or false alarm rate based on the ratings (often nonzero) of the control against itself. Identical samples will often get nonzero difference estimates due to the moment-to-moment variability in sensations.

138

of the inpidual thresholds, due to the larger number of observations, another oddity of using statistical significance to determine the threshold. Instead of using statistical significance as a criterion, Marin et al. determined the point of maximum curvature on the dose-response curve as the threshold. Such an approach makes sense from consideration of the general form of the dose-response (psychophysical) curve for most tastes and odors. Figure 6.7 shows a semi-log plot for the Beidler taste equation, a widely applied dose-response relationship in studies of the chemical senses (see Chapter 2). This function has two sections of curvature (change in slope, i.e., acceleration) when plotted as a function of log concentration. There is a point at which the response appears to be slowly increasing out from the background noise and then rises steeply to enter the middle of the dynamic range of response. The point of maximum curvature can be estimated graphically or determined from curve fitting and finding the maximum rate of change (i.e., maximum of the second derivative) (Marin et al., 1991).

Psychophysical function (semi-hyperbolic) 100

Response (percent of maximum)

In one application of this technique for taste and smell thresholds, a substance was added in various levels to estimate the threshold in a food or beverage. In each inpidual trial, three samples would be compared to the control sample with no added flavor-two adjacent concentration steps of the target compound and one blind control sample (Lundahl et al., 1986). Samples were rated on a simple 9-point scale, from zero (no difference) to eight (extreme difference). This provided a comparison of the control to itself and a cost-effective way of making three comparisons in one set of samples. Since sample concentrations within the three rated test samples were randomized, the procedure was not a true ascending series and was dubbed the “semi-ascending paired difference method.” How is the threshold defined in these procedures? One approach is to compare the difference ratings for a given level with the difference ratings given to the control sample. Then the threshold can be based on some measure of when these difference ratings perge, such as when they become significantly different by a t-test (see Brown et al., 1978). Another approach is simply to subtract the difference score given to the blind control from the difference score give to each test sample and treat these adjusted scores as a new data set. In the original paper of Lundahl et al. (1986), this latter method was used. In the analysis, they performed a series of ttests. Two values were taken to bracket the range of the threshold. The upper level was the first level yielding a significant t-test versus zero, and the lower level was the nearest lower concentration yielding a significant ttest versus the first. This provided an interval in which the threshold (as defined by this method) lies between the two bracketing concentrations. One problem with this approach is that when the threshold is based on the statistical significance of t-statistics (or any such significance test), the value of threshold will depend upon the number of observations in the test. This creates a nonsensical situation where the threshold value will decrease as a function of the number of panelists used in the test. This is an irrelevant variable, a choice of the experimenter, and has nothing to do with the physiological sensitivity of the panelist or the biological potency of the substance being tested, a problem recognized by Brown et al. (1978) and later by Marin et al. (1991). Marin et al. also pointed out that a group threshold, based on a larger number of observations than an inpidual threshold, would be lower than the mean

6 Measurement of Sensory Thresholds

80

60

R = (Rmax C) / (K + C) K = 1000, Rmax = 100

40

20

zone of inflection = threshold range

0 0

2

4 log Concentration

6

8

Fig. 6.7 Beidler curve.

6.9.2 Adaptive Procedures Popular methods for threshold measurement for visual and auditory stimuli have been procedures in which the next stimulus intensity level to be tested depends

6.9 Alternative Methods: Rated Difference, Adaptive Procedures, Scaling

SAMPLE UDTR RECORD CONCENTRATION STEP

CONCENTRATION STEP

the part of the respondent. Psychophysical researchers have found ways to undo this sequential dependence to counteract observer expectancies. One example is the double random staircase procedure (Cornsweet, 1962) in which trials from two staircase sequences are randomly intermixed. One staircase starts above the threshold and descends, while the other starts below the threshold and ascends. On any given trial, the observer is unaware which of the two sequences the stimulus is chosen from. As in the simple staircase procedure, the level chosen for a trial depends upon detection or discrimination in the pvious trial, but of that particular sequence. Further refinements of the procedure involve the introduction of forced-choice (Jesteadt, 1980) to eliminate response bias factors involved in simple yes/no detection. Another modification to the adaptive methods has been to adjust the ascending and descending rules so that some number of correct or incorrect judgments is required before changing intensity levels, rather than the one trial as in the simple staircase (Jesteadt, 1980). An example is the “up down transformed response” rule or UDTR (Wetherill and Levitt, 1965). Wetherill and Levitt gave an example where two positive judgments were required before moving down, and only one negative judgment at a given level before moving up. An example is shown in Fig. 6.9. Rather than estimating the 50% point on a traditional psychometric function, this more stringent requirement now tends to converge on the 71% mark as an average of the peaks and valleys in the series. A forced choice can be added to an adaptive procedure. Sometimes

SAMPLE STAIRCASE RECORD

64 32 16 8 4

R 2 R R 2

3

4

R

R

1

1

5

139

6

7

R 8

9

10

TRIAL

Fig. 6.8 Staircase example.

TRIAL

Fig. 6.9 Staircase example.

140

6.9.3 Scaling as an Alternative Measure of Sensitivity Threshold measurements are not the only way to screen inpiduals for insensitivity to specific compounds like PTC or to screen for specific anosmia. Do thresholds bear any relation to suprathreshold responding? While it has been widely held that there is no necessary relationship between threshold sensitivity and suprathreshold responding (Frijters, 1978; Pangborn, 1981), this assertion somewhat overstates the case. Counter-examples of good correlations can be seen in tests involving compounds like PTC where there are insensitive groups. For example, there is a -0.8 correlation between simple category taste intensity ratings for PTC and the threshold, when the rated concentration is near the antimode or center between the modes of a bimodal threshold frequency distribution (Lawless, 1980). Thus ratings of a carefully chosen level can be used for a rapid screening method for PTC taster status (e.g., Mela, 1989). Similar results have been noted for smell. Berglund and Högman (1992) reported better reliability of suprathreshold ratings than threshold determinations in screening for olfactory sensitivity. Stevens and O’Connell (1991) used category ratings of perceived intensity as well as qualitative descriptors as a screening tool before threshold testing for specific anosmia. Threshold versus rating correlations were in the range of -0.6 for cineole, -0.3 for carvone, and -0.5 for diacetyl (Lawless et al., 1994, 1995). The correlations were obtained after subtraction of ratings to a blank stimulus, in order to correct for differences in scale usage. Thus there is a moderate negative correlation of sensitivity and rated intensity when one

6 Measurement of Sensory Thresholds

examines the data across a highly variable group as is the case with specific anosmia or tasting PTC bitterness. The correlation is negative since higher thresholds indicate lower sensitivity and thus lower rated intensity.

6.10 Dilution to Threshold Measures 6.10.1 Odor Units and Gas-Chromatography Olfactometry (GCO) In this section, several applied methods will be described that make use of the threshold concept in trying to determine the sensory impact of various flavors and odor materials. The first group of methods concerns the olfactory potency of volatile aroma compounds as they are found in foods or food components. The issue here becomes one of not just threshold determination, but determination of both the threshold and the actual concentration psent in a food sample. The ratio of these concentrations (actual concentration to threshold concentration) can help indicate whether or not a given flavor substance is likely to contribute to the overall sensory impssion in a food. These ratios are commonly called “odor units.” The second much older method is similar in logic and was developed to determine the point at which the irritative or heat sensations from pepper compounds would be first detectable when diluted to a given extent, the Scoville procedure. Both of these techniques then use dilution-to-threshold as a measure of sensory impact. When a complex natural product like a fruit extract is analyzed for its chemical components, hundreds or even thousands of chemicals may be identified, many of which have odor properties. The number of potential flavor compounds identified in any product seems only to be limited by the resolution and sensitivity of the current methods in analytical chemistry. These methods are always improving leading to longer and longer lists of possible contributing flavor materials (Piggott, 1990). Flavor scientists need to find a way to narrow the list or to separate those compounds which are most likely contributing to the overall flavor from those compounds that are psent in

6.10 Dilution to Threshold Measures

such low concentrations that they are probably not important. Obviously, a sensory-based method is needed in conjunction with the analytical chemistry to provide a bioassay for possible sensory impact (Acree, 1993). Thresholds can be useful in addressing this kind of problem. The reasoning goes that only those compounds that are psent in the product in concentrations above their threshold are likely to be contributors to the flavor of the product. There are a number of potential flaws in this thinking discussed below, but for now let us see how this can be put to use. Given a concentration C psent in a natural product, a dimensionless quantity can be derived by piding that concentration by the threshold concentration Ct , and the ratio C/Ct defines the number of odor units (or flavor units) for compounds assessed by smell. According to this logic, only those compounds with odor units greater than one will contribute to the aroma of the product. This reasoning is extended sometimes to include the idea that the greater the number of odor units, the greater the potential contribution. However, it is now widely recognized that the odor unit is a concentration multiple and not a measure of subjective magnitude. Only direct scaling methods can assess the actual magnitude of sensation above threshold and the psychophysical relationship between concentration and odor intensity (Frijters, 1978). Furthermore, this idea ignores the possibility of subthreshold additivity or synergy (Day et al., 1963). A closely related group of chemical compounds might all be psent below their inpidual thresholds, but together could stimulate common receptors so as to produce an above-threshold sensation. Such additivity is not pdicted by the odor unit approach and such a group of compounds could be missed in dilution analysis. Nonetheless, thresholds provide at least one isointense reference point on the dose response curve, so they have some utility as a measure of potency used to compare different odor compounds. In analyzing a food, one could look up literature values for all the identified compounds in the product in one of the published compendia of thresholds (e.g., ASTM, 1978; van Gemert, 2003). If the concentration in the product is determined, then the odor unit value can be calculated by simply piding by threshold. However, it is important to remember that the literature values for thresholds depend upon the method and the medium

141

of testing. Unless the same techniques are used and the same medium was used as the carrier (rarely the case) the values may not be necessarily comparable for different compounds. A second approach is to actually measure the dilutions necessary to reach threshold for each compound, starting with the product itself. This necessitates the use of a separatory procedure, so that each compound may be inpidually perceived. The combination of gas chromatography with odor port sniffing of a dilution series is a popular technique (Acree, 1993). Various catchy names have been applied to such techniques in the flavor literature, including Aroma Extract Dilution Analysis (for examples, see Guth and Grosch, 1994; Milo and Grosch, 1993; Schieberle and Grosch, 1988), CHARM analysis (Acree et al., 1984) or more generically, gas chromatography olfactometry or GCO. The basis of these techniques is to have subjects respond when an odor is perceived when sniffing the exit port during a GC run. In recent years, the effluent has been embedded in a cooled, humidified air stream to improve the comfort of the observer and to increase sensory resolution of the eluting compounds. Over several dilutions, the response will eventually drop out, and the index of smell potency is related to the reciprocal of the dilution factor. The sniffer’s responses occur on a time base that can be cross-referenced to a retention index and then the identity of the compound can be determined by a combination of retention index, mass spectrometry, and aroma character. In practice, these techniques considerably shorten the list of potential aroma compounds contributing to flavor in a natural product (e.g., Cunningham et al., 1986). The method has also been used as an assessment technique for measuring the sensitivity of human panelists, as opposed to being a tool to determine the sensory impact of flavor compounds (Marin et al., 1988). In this approach, the gas chromatograph once again serves as an olfactometer. Known compounds can be psented as a dilution series of mixtures. Variation in inpidual thresholds can readily be assessed for a variety of compounds since they can be combined in a GC run as long as they have different retention times, i.e., do not co-elute on the column of choice. The potential use of GCO for screening panelists, for assessing odor responses in general, and for assessing specific anosmias has also been attempted (Friedrich and Acree, 2000; Kittel and Acree, 2008).

142

6.10.2 Scoville Units Another example of a dilution method is the traditional Scoville procedure for scoring pepper heat in the spice trade. This procedure was named for W. Scoville who worked in the pharmaceutical industry in the early twentieth century. He was interested in the topical application of spice compounds like pepper extracts as counterirritants, and he needed to establish units that could be used to measure their potency. His procedure consisted of finding the number of dilutions necessary for sensations to disappear and then using this number of dilutions as an estimate of potency. In other words, potency was defined as a reciprocal threshold. A variation of this procedure was adopted by the Essential Oil Association, British Standards Institution, International Standards Organization, American Spice Trade Association (ASTA), and adopted as an Indian standard method (for a review, see Govindarajan, 1987). The procedure defined units of pungency as the highest dilution at which a definite “bite” would be perceived and thus contains instructions consistent with a recognition threshold. Scoville units were dilution factors, now commonly given as mL/g. ASTA Method 21 (ASTA, 1968) is widely used and contains some modifications in an attempt to overcome problems with the original Scoville procedure. In brief, the method proceeds as follows: Panelists are screened for acuity relative to experienced persons. Dilution schedules are provided which simplify calculations of the eventual potency. Solutions are tested in 5% sucrose and negligible amounts of alcohol. Five panelists participate and concentrations are given in ascending order around the estimated threshold. Threshold is defined as the concentration at which three out of five judges respond positively. This method is difficult in practice and a number of additional variations have been tried to improve on the accuracy and pcision of the method (Govindarajan, 1987). Examples include the following: (1) substitution of other rules for 3/5, e.g., mean + SD of 20-30 judgments, (2) use of a triangle test, rather than simple yes/no at each concentration, (3) requiring recognition of pungency (Todd et al., 1977), (4) reduction of sucrose concentration in the carrier solution to 3%, and (5) use of a rating scale from 1 (definitely not detectable) to 6 (definitely detectable). This latter

6 Measurement of Sensory Thresholds

modification defined a detection threshold at mean scale value of 3.5. Almost all these methods specify mandatory rest periods between samples due to the long lasting nature of these sensations. The measurements are still difficult. One problem is that capsaicin, the active heat principle in red pepper, is prone to desensitize observers within a session and also regular consumers of hot spices also become less sensitive, leading to wide inpidual differences in sensitivity among panelists (Green, 1989; Lawless et al., 1985). Alternative procedures have been developed based on rating scales with fixed physical references (Gillette et al., 1984) and these have been endorsed by ASTM (2008c) as standard test methods. These rating scale procedures show good correlations with instrumental measures of capsaicin content in pepper samples and can be cross-referenced to Scoville units for those who pfer to do business in the traditional units (Gillette et al., 1984).

6.11 Conclusions Threshold measurements find three common uses in sensory analysis and flavor research. First, they can be used to compare the sensitivities of different panelists. Second, they can be used as an index of the biological potency of a flavor compound. Third, they can provide useful information regarding the maximum tolerable levels of an off-flavor or taint. A variety of different techniques have been used to find thresholds or have employed the threshold concept in practical flavor work. Examples of different threshold methods are given in Table 6.4. In spite of their practical applications, the usefulness of threshold measures is often questioned in sensory evaluation. One criticism is that thresholds are only one point on an intensity function and thus they do not tell us anything about abovethreshold responding. There are some good examples in which thresholds do not pdict or do not correlate very well with suprathreshold responses. For example, patients irradiated for cancer may lose their sense of taste temporarily and thresholds return to normal long before suprathreshold responsiveness is recovered (Bartoshuk, 1987). However, as we have seen both in the case of PTC tasting and in specific anosmias, insensitive inpiduals (as determined by their threshold) will also tend to be less responsive above

Appendix

143

Table 6.4 Threshold procedures Method Citations/examples

Response

Threshold

Ascending forced choice Ascending forced choice

ASTM E-679-79 Stevens et al. (1988)

3-AFC 2-AFC, 5 replicates

Semi-ascending paired difference

Lundahl et al. (1986) Rated difference from control

Adaptive, up-down transformed response rule

Reed et al. (1995)

2-AFC

Double Random Staircase CHARM analysis

Cornsweet (1962) Acree et al. (1986)

Yes/No Yes/No

Geometric mean of inpidual threshold Lowest correct set with confirmation at next concentration t-test for significant difference from zero with all blank trials (blind control scores) subtracted Average of responses ascending after one incorrect, descending after two correct at each level Average of reversal points Nonresponse on descending concentration runs (implied)

threshold. These correlations are strong when comparing inpiduals of very different sensitivities, but the threshold-suprathreshold parallel may not extend to all flavor compounds. A more complete understanding of the whole dynamic range of dose-response, as found in scaling studies, would be more informative. Other shortcomings need to be kept in mind by sensory workers who would use threshold measurements for making product decisions. First, thresholds are statistical constructs only. Signal detection theory warns us that the signal and noise perge in a continuous fashion and discontinuities in perception may be an idealized construction that is comfortable but unrealistic. There is no sudden transition from nondetection to 100% detection. Any modern concept of threshold must take into account that a range of values, rather than a single point is involved in the specification. Thresholds depend upon the conditions of measurement. For example, as the purity of the diluent increases, the threshold for taste will go down. So a measure of the true absolute threshold for taste (if such a thing existed) would require water of infinite purity. The threshold exists not in this abstract sense, but only

as a potentially useful construct of our methods and informational requirements. Finally, because of the problems mentioned above, sensory professionals need to keep the following principles firmly in mind when working with threshold procedures: First, changes in method will change the obtained values. Literature values cannot be trusted to extend to a new product or a new medium or if changes are made in test procedure. Next, threshold distributions do not always follow a normal bell curve. There are often high outliers and possibly cases of insensitivity due to inherited deficits like specific anosmia (Amoore, 1971; Brown et al., 1978; Lawless et al., 1994). Threshold values for an inpidual are prone to high variability and low test-retest reliability. An inpidual’s threshold measure on a given day is not necessarily a stable characteristic of that person (Lawless et al., 1995; Stevens et al., 1988). Practice effects can be profound, and thresholds may stabilize over a period of time (Engen, 1960; Rabin and Cain, 1986). However, group averaged thresholds are reliable (Brown et al., 1978; Punter, 1983) and provide a useful index of the biological activity of a stimulus.

Appendix: MTBE Threshold Data for Worked Example

Panelist 1 2 3 4 5 6

Concentration (μg/L) 2 3.5 + o o o o + o o + o o o

6 + o o o + +

10 + + o o o +

18 + + o o + +

30 + + o o + +

60 + + + + o +

100 + + + + + +

BET 4.6 7.7 42 42 77 4.6

log(BET) 0.663 0.886 1.623 1.623 1.886 0.663

144

Panelist 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 Prop. Corr.

6 Measurement of Sensory Thresholds

Concentration (μg/L) 2 3.5 6 o + + + + o o o + o o o + o + o o o + + + + + + o + + o o + + o + o o + + + + + o + + + + + + + + + o + + + o + + o + + + o o o o + o + o o + + + o o + + + + + + o o o o o o o o o + + + + o + + + o o o + + + + o + o o o o o + + o + + + + + o + o o + o o o o o o + o o o + + + + + + + o o o o + o o o o o + 0.44 0.49 0.61

10 o + o o + o + + o o + + + + + + o + + + o + o o + + o o + + + + o o + + o o o + o + + + + + + o + o + 0.58

18 + + + o + + + + + o + + + o + + + + o o + + + + o o + o o o + + + + + + o + + + o o o + + + + o + o o 0.65

30 + + + o + o + + + o + + + + + + + + + + o + + + o o o o + + + + + + + + o o + + + + + + + + + + + + + 0.77

60 + + + o + + + + + + + + + o + + + + + + o + + + + o + + + + + + + + + + + + + o + + + + + + + + + + o 0.89

100 + + + + + o + + + o + + + o + + + + + + + o + + o + + o o + + + + + + + + + + + o + + + + + + + + + + 0.86

BET 13 7.7 13 77 4.6 132 1.4 1.4 13 132 4.6 4.6 1.4 132 1.4 1.4 13 1.4 23 23 77 132 13 13 132 77 42 132 132 23 1.4 1.4 13 13 1.4 7.7 42 42 13 77 132 23 23 4.6 7.7 1.4 1.4 23 4.6 23 77 Mean (log(BET)) 10ˆ1.154=

log(BET) 1.114 0.886 1.114 1.886 0.663 2.121 0.146 0.146 1.114 2.121 0.663 0.663 0.146 2.121 0.146 0.146 1.114 0.146 1.362 1.362 1.886 2.121 1.114 1.114 2.121 1.886 1.623 2.121 2.121 1.362 0.146 0.146 1.114 1.114 0.146 0.886 1.623 1.623 1.114 1.886 2.121 1.362 1.362 0.663 0.886 0.146 0.146 1.362 0.663 1.362 1.886 1.154 14.24

References

References Acree, T. E. 1993. Bioassays for flavor. In: T. E. Acree and R. Teranishi (eds.), Flavor Science, Sensible Principles and Techniques. American Chemical Society Books, Washington, pp. 1-20. Acree, T. E., Barnard, J. and Cunningham, D. G. 1984. A procedure for the sensory analysis for gas chromatographic effluents. Food Chemistry, 14, 273-286. American Spice Trade Association 1968. Pungency of capsicum spices and oleoresins. American Spice Trade Association Official Analytical Methods, 21.0, 43-47. Amoore, J. E. 1971. Olfactory genetics and anosmia. In: L. M. Beidler (ed.), Handbook of Sensory Physiology. Springer, Berlin, pp. 245-256. Amoore, J. E. 1979. Directions for pparing aqueous solutions of primary odorants to diagnose eight types of specific anosmia. Chemical Senses and Flavor, 4, 153-161. Amoore, J. E.,Venstrom, D. and Davis, A. R. 1968. Measurement of specific anosmia. Perceptual and Motor Skills, 26, 143-164. Antinone, M. A., Lawless, H. T., Ledford, R. A. and Johnston, M. 1994. The importance of diacetyl as a flavor component in full fat cottage cheese. Journal of Food Science, 59, 38-42. ASTM. 1978. Compilation of Odor and Taste Threshold Values Data. American Society for Testing and Materials, Philadelphia. ASTM. 2008a. Standard practice for determining odor and taste thresholds by a forced-choice ascending concentration series method of limits, E-679-04. Annual Book of Standards, Vol. 15.08. ASTM International, Conshocken, PA, pp. 36-42. ASTM. 2008b. Standard practice for defining and calculating inpidual and group sensory thresholds from forcedchoice data sets of intermediate size, E-1432-04. ASTM International Book of Standards, Vol. 15.08. ASTM International, Conshocken, PA, pp.82-89. ASTM. 2008c. Standard test method for sensory evaluation of red pepper heat, E-1083-00. ASTM International Book of Standards, Vol. 15.08. ASTM International, Conshocken, PA, pp. 49-53. Ayya, N. and Lawless, H. T. 1992. Qualitative and quantitative evaluation of high-intensity sweeteners and sweetener mixtures. Chemical Senses, 17, 245-259. Bartoshuk, L. M. 1987. Psychophysics of taste. American Journal of Clinical Nutrition, 31, 1068-1077. Bartoshuk, L. M., Murphy, C. and Cleveland, C. T. 1978. Sweet taste of dilute NaCl: Psychophysical evidence for a sweet stimulus. Physiology and Behavior, 21, 609-613. Berglund, B. and Högman, L. 1992. Reliability and validity of odor measurements near the detection threshold. Chemical Senses, 17, 823-824. Blakeslee, A. F. 1932. Genetics of sensory thresholds: Taste for phenylthiocarbamide. Proceedings of the National Academy of Sciences, 18, 120-130. Brennand, C. P., Ha, J. K. and Lindsay, R. C. 1989. Aroma properties and thresholds of some branched chain and other minor volatile fatty acids occurring in milkfat and meat lipids. Journal of Sensory Studies, 4, 105-120.

145 Boring, E. G. 1942. Sensation and Perception in the History of Experimental Psychology. Appleton-CenturyCrofts, New York, pp. 41-42. Brown, D. G. W., Clapperton, J. F., Meilgaard, M. C. and Moll, M. 1978. Flavor thresholds of added substances. American Society of Brewing Chemists Journal, 36, 73-80. Bufe, B., Breslin, P. A. S., Kuhn, C., Reed, D., Tharp, C. D., Slack, J. P., Kim, U.-K., Drayna, D. and Meyerhof, W. 2005. The molecular basis of inpidual differences in phenylthiocarbamide and propylthiouracil bitterness perception. Current Biology, 15, 322-327. Cain, W. S. 1976. Olfaction and the common chemical sense: Some psychophysical contrasts. Sensory Processes, 1, 57-67. Cain, W. S. and Murphy, C. L. 1980. Interaction between chemoreceptive modalities of odor and irritation. Nature, 284, 255-257. Cornsweet, T. M. 1962. The staircase method in psychophysics. American Journal of Psychology, 75, 485-491. Cunningham, D. G., Acree, T. E., Barnard, J., Butts, R. M. and Braell, P. A. 1986. CHARM analysis of apple volatiles. Food Chemistry, 19, 137-147. Dale, M. S., Moylan, M. S., Koch, B. and Davis, M. K. 1997. Taste and odor threshold determinations using the flavor profile method. Proceedings of the American Water Works Association Water-Quality Technology Conference, Denver, CO. Dalton, P., Doolittle, N. and Breslin, P. A. S. 2002. Genderspecific induction of enhanced sensitivity to odors. Nature Neuroscience, 5, 199-200. Day, E. A., Lillard, D. A. and Montgomery, M. W. 1963. Autooxidation of milk chúng tôi Effect on flavor of the additive interactions of carbonyl compounds at subthreshold concentrations. Journal of Dairy Science, 46, 291-294. Diamond, J., Dalton, P., Doolittle, N. and Breslin, P. A. S. 2005. Gender-specific olfactory sensitization: Hormonal and cognitive influences. Chemical Senses, 30(suppl 1), i224-i225. Dravnieks, A. and Prokop, W. H. 1975. Source emission odor measurement by a dynamic forced-choice triangle olfactometer. Journal of the Air Pollution Control Association, 25, 28-35. Engen, T. E. 1960. Effects of practice and instruction on olfactory thresholds. Perceptual and Motor Skills, 10, 195-198. Finney, D. J. 1971. Probit Analysis, Third Edition. Cambridge University, London. Friedrich, J. E. and Acree, T. E. 2000. Design of a standard set of odorants to test inpiduals for specific anosmia. Frontiers in Flavour Science [Proc. 9th Weurman Flavour Res. Symp.], 230-234. Frijters, J. E. R. 1978. A critical analysis of the odour unit number and its use. Chemical Senses and Flavour, 3, 227-233. Gillette, M., Appel, C. E. and Lego, M. C. 1984. A new method for sensory evaluation of red pepper heat. Journal of Food Science, 49, 1028-1033. Green, B. G. 1989. Capsaicin sensitization and desensitization on the tongue produced by brief exposures to a low concentration. Neuroscience Letters, 107, 173-178. Govindarajan, V. S. 1987. Capsicum – production, technology, chemistry and quality. Part III. Chemistry of the color, aroma and pungency stimuli. CRC Critical Reviews in Food Science and Nutrition, 24, 245-311.

146 Guth, H. and Grosch, W. 1994. Identification of the character impact odorants of stewed beef juice by instrumental analysis and sensory studies. Journal of Agricultural and Food Chemistry, 42, 2862-2866. Harris, H. and Kalmus, H. 1949. The measurement of taste sensitivity to phenylthiourea (P.T.C.). Annals of Eugenics, 15, 24-31. Harvey, L. O. 1986. Efficient estimation of sensory thresholds. Behavior Research Methods, Instruments and Computers, 18, 623-632. James, W. 1913. Psychology. Henry Holt and Co., New York. Jesteadt, W. 1980. An adaptive procedure for subjective judgments. Perception and Psychophysics, 28(1), 85-88. Kittel, K. M and Acree, T. E. 2008. Investigation of olfactory deficits using gas-chromatography olfactometry. Manuscript submitted, available from the authors. Lawless, H. T. 1980. A comparison of different methods used to assess sensitivity to the taste of phenylthiocarbamide PTC. Chemical Senses, 5, 247-256. Lawless, H. T. 2010. A simple alternative analysis for threshold data determined by ascending forced-choice method of limits. Journal of Sensory Studies, 25, 332-346. Lawless, H. T., Antinone, M. J., Ledford, R. A. and Johnston, M. 1994. Olfactory responsiveness to diacetyl. Journal of Sensory Studies, 9(1), 47-56. Lawless, H. T., Rozin, P. and Shenker, J. 1985. Effects of oral capsaicin on gustatory, olfactory and irritant sensations and flavor identification in humans who regularly or rarely consumer chili pepper. Chemical Senses, 10, 579-589. Lawless, H. T., Thomas, C. J. C. and Johnston, M. 1995. Variation in odor thresholds for l-carvone and cineole and correlations with suprathreshold intensity ratings. Chemical Senses, 20, 9-17. Linschoten, M. R., Harvey, L. O., Eller, P. A. and Jafek, B. W. 1996. Rapid and accurate measurement of taste and smell thresholds using an adaptive maximum-likelihood staircase procedure. Chemical Senses, 21, 633-634. Lundahl, D. S., Lukes, B. K., McDaniel, M. R. and Henderson, L. A. 1986. A semi-ascending paired difference method for determining the threshold of added substances to background media. Journal of Sensory Studies, 1, 291-306. Masuoka, S., Hatjopolous, D. and O’Mahony, M. 1995. Beer bitterness detection: Testing Thurstonian and sequential sensitivity analysis models for triad and tetrad methods. Journal of Sensory Studies, 10, 295-306. Marin, A. B., Acree, T. E. and Barnard, J. 1988. Variation in odor detection thresholds determined by charm analysis. Chemical Senses, 13, 435-444. Marin, A. B., Barnard, J., Darlington, R. B. and Acree, T. E. 1991. Sensory thresholds: Estimation from dose-response curves. Journal of Sensory Studies, 6(4), 205-225. McBurney, D. H. and Collings, V. B. 1977. Introduction to Sensation and Perception. Prentice-Hall, Englewood Cliffs, NJ. Meilgaard, M., Civille, G. V. and Carr, B. T. 1991. Sensory Evaluation Techniques, Second Edition. CRC Press, Boca Raton, FL. Mela, D. 1989. Bitter taste intensity: The effect of tastant and thiourea taster status. Chemical Senses, 14, 131-135. Milo, C. and Grosch, W. 1993. Changes in the odorants of boiled trout (Salmo fario) as affected by the storage of the raw

6 Measurement of Sensory Thresholds material. Journal of Agricultural and Food Chemistry, 41, 2076-2081. Mojet, J., Christ-Hazelhof, E. and Heidema, J. 2001. Taste perception with age: Generic or specific losses in threshold sensitivity to the five basic tastes. Chemical Senses 26, 854-860. Morrison, D. G. 1978. A probability model for forced binary choices. American Statistician, 32, 23-25. Murphy, C., Quiñonez, C. and Nordin, S. 1995. Reliability and validity of electrogustometry and its application to young and elderly persons. Chemical Senses, 20(5), 499-515. O’Mahony, M. and Ishii, R. 1986. Umami taste concept: Implications for the dogma of four basic tastes. In: Y. Kawamura and M. R. Kare (eds.), Umami: A Basic Taste. Marcel Dekker, New York, pp. 75-93.. O’Mahony, M. and Odbert, N. 1985. A comparison of sensory difference testing procedures: Sequential sensitivity analysis and aspects of taste adaptation. Journal of Food Science, 50, 1055-1060. Pangborn, R. M. 1981. A critical review of threshold, intensity and descriptive analyses in flavor research. In: Flavor ’81. Walter de Gruyter, Berlin, pp. 3-32. Piggott, J. R. 1990. Relating sensory and chemical data to understand flavor. Journal of Sensory Studies, 4, 261-272. Prescott, J., Norris, L., Kunst, M. and Kim, S. 2005. Estimating a “consumer rejection threshold” for cork taint in white wine. Food Quality and Preference, 18, 345-349. Punter, P. H. 1983. Measurement of human olfactory thresholds for several groups of structurally related compounds. Chemical Senses, 7, 215-235. Rabin, M. D. and Cain, W. S. 1986. Determinants of measured olfactory sensitivity. Perception and Psychophysics, 39(4), 281-286. Reed, D. R., Bartoshuk, L. M., Duffy, V., Marino, S. and Price, R. A. 1995. Propylthiouracil tasting: Determination of underlying threshold distributions using maximum likelihood. Chemical Senses, 20, 529-533. Saliba, A. J., Bullock, J. and Hardie, W. J. 2009. Consumer rejection threshold for 1,8 cineole (eucalyptol) in Australian red wine. Food Qualithy and Preference, 20, 500-504. Schieberle, P. and Grosch, W. 1988. Identification of potent flavor compounds formed in an aqueous lemon oil/citric acid emulsion. Journal of Agricultural and Food Chemistry, 36, 797-800. Stevens, D. A. and O’Connell, R. J. 1991. Inpidual differences in threshold and quality reports of subjects to various odors. Chemical Senses, 16, 57-67. Stevens, D. A. and O’Connell, R. J. 1995. Enhanced sensitivity to adrostenone following regular exposure to pemenone. Chemical Senses, 20, 413-419. Stevens, J. C., Cain, W. S. and Burke, R. J. 1988. Variability of olfactory thresholds. Chemical Senses, 13, 643-653. Stocking, A. J., Suffet, I. H., McGuire, M. J. and Kavanaugh, M. C. 2001. Implications of an MTBE odor study for setting drinking water standards. Journal AWWA, March 2001, 95-105. Todd, P. H., Bensinger, M. G. and Biftu, T. 1977. Determination of pungency due to capsicum by gas-liquid chromatography. Journal of Food Science, 42, 660-665. Tuorila, H., Kurkela, R., Suihko, M. and Suhonen. 1981. The influence of tests and panelists on odour detection

References thresholds. Lebensmittel Wissenschaft und Technologie, 15, 97-101. USEPA (U.S. Environmental Protection Agency). 2001. Statistical analysis of MTBE odor detection thresholds in drinking water. National Service Center for Environmental Publications (NSCEP) # 815R01024, available from http://nepis.epa.gov. Van Gemert, L. 2003. Odour Thresholds. Oliemans, Punter & Partners, Utrecht, The Netherlands. Viswanathan, S., Mathur, G. P., Gnyp, A. W. and St. Peirre, C. C. 1983. Application of probability models to threshold determination. Atmospheric Environment, 17, 139-143.

147 Walker, J. C., Hall, S. B., Walker, D. B., Kendall-Reed, M. S., Hood, A. F. and Nio, X.-F. 2003. Human odor detectability: New methodology used to determine threshold and variation. Chemical Senses, 28, 817-826. Wetherill, G. B. and Levitt, H. 1965. Sequential estimation of points on a psychometric function. British Journal of Mathematical and Statistical Psychology, 18(1), 1-10. Wysocki, C. J., Dorries, K. M. and Beauchamp, G. K. 1989. Ability to perceive androstenone can be acquired by ostensibly anosmic people. Proceedings of the National Academy of Science, USA, 86, 7976-7989.

Chapter 7

Scaling

Abstract Scaling describes the application of numbers, or judgments that are converted to numerical values, to describe the perceived intensity of a sensory experience or the degree of liking or disliking for some experience or product. Scaling forms the basis for the sensory method of descriptive analysis. A variety of methods have been used for this purpose and with some caution, all work well in differentiating products. This chapter discusses theoretical issues as well as practical considerations in scaling. The vital importance of knowing the properties and limitations of a measuring instrument can hardly be denied by most natural scientists. However, the use of many different scales for sensory measurement is common within food science; but very few of these have ever been validated. . . . -(Land and Shepard, 1984, pp. 144-145)

Contents 7.1 7.2 7.3

7.4

7.5

7.6

Introduction . . . . . . . . . . . . . . . . . Some Theory . . . . . . . . . . . . . . . . Common Methods of Scaling . . . . . . . . . 7.3.1 Category Scales . . . . . . . . . . . . 7.3.2 Line Scaling . . . . . . . . . . . . . 7.3.3 Magnitude Estimation . . . . . . . . . Recommended Practice and Practical Guidelines . . . . . . . . . . . . . . . . . . 7.4.1 Rule 1: Provide Sufficient Alternatives . . . . . . . . . . . . . . 7.4.2 Rule 2: The Attribute Must Be Understood . . . . . . . . . . . . 7.4.3 Rule 3: The Anchor Words Should Make Sense . . . . . . . . . . . . . . . . . 7.4.4 To Calibrate or Not to Calibrate . . . . . 7.4.5 A Warning: Grading and Scoring are Not Scaling . . . . . . . . . . . . . . . . Variations-Other Scaling Techniques . . . . 7.5.1 Cross-Modal Matches and Variations on Magnitude Estimation . . . . . . . . . 7.5.2 Category-Ratio (Labeled Magnitude) Scales . . . . . . . . . . . . . . . . 7.5.3 Adjustable Rating Techniques: Relative Scaling . . . . . . . . . . . . . . . . 7.5.4 Ranking . . . . . . . . . . . . . . . 7.5.5 Indirect Scales . . . . . . . . . . . . Comparing Methods: What is a Good Scale? . . . . . . . . . . . . . . . . . . . .

Issues . . . . . . . . . . . . . . . . . . . . 7.7.1 “Do People Make Relative Judgments” Should They See Their Previous Ratings? . . . . . . . . . . . . . . . 7.7.2 Should Category Rating Scales Be Assigned Integer Numbers in Data Tabulation? Are They Interval Scales? . . . . . . . . . 7.7.3 Is Magnitude Estimation a Ratio Scale or Simply a Scale with Ratio Instructions? . . . . . . . . . . . . . 7.7.4 What is a “Valid” Scale? . . . . . . . . 7.8 Conclusions . . . . . . . . . . . . . . . . . Appendix 1: Derivation of Thurstonian-Scale Values for the 9-Point Scale . . . . . . . . . . . . . Appendix 2: Construction of Labeled Magnitude Scales . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . 7.7

149 151 152 152 155 156 158 159 159 159 159 160 160 160 162 164 165 166 167

168

168

169

169 169 170 171 172 174

7.1 Introduction People make changes in their behavior all the time based on sensory experience and very often this involves a judgment of how strong or weak something feels. We add more sugar to our coffee if it is not sweet enough. We adjust the thermostat in our home if it is too cold or too hot. If a closet is too dark to find your shoes you turn the light on. We apply more force to

H.T. Lawless, H. Heymann, Sensory Evaluation of Food, Food Science Text Series, DOI 10.1007/978-1-4419-6488-5_7, © Springer Science+Business Media, LLC 2010

149

150

chew a tough piece of meat if it will not disintegrate to allow swallowing. These behavioral decisions seem automatic and do not require a numerical response. But the same kinds of experiences can be evaluated with a response that indicates the strength of the sensation. What was subjective and private becomes public data. The data are quantitative. This is the basis of scaling. The methods of scaling involve the application of numbers to quantify sensory experiences. It is through this process of numerification that sensory evaluation becomes a quantitative science subject to statistical analysis, modeling, pdiction, and hard theory. However, as noted in the quote above, in the practical application of sensory test methods, the nature of this process of assigning numbers to experiences is rarely questioned and deserves scrutiny. Clearly numbers can be assigned to sensations by a panelist in a variety of ways, some by mere categorization, or by ranking or in ways that attempt to reflect the intensity of sensory experience. This chapter will illustrate these techniques and discuss the arguments that have been raised to substantiate the use of different quantification procedures. Scaling involves sensing a product or stimulus and then generating a response that reflects how the person perceives the intensity or strength of one or more of the sensations generated by that product. This process is based on a psychophysical model (see Chapter 2). The psychophysical model states that as the physical strength of the stimulus increases (e.g., the energy of a light or sound or the concentration of a chemical stimulus) the sensation will increase in some orderly way. Furthermore, panelists are capable of generating different responses to indicate these changes in what they experience. Thus a systematic relationship can be modeled of how physical changes in the real world result in changing sensations. Scaling is a tool used for showing differences and degrees of difference among products. These differences are usually above the threshold level or justnoticeable difference. If the products are very similar and there is a question of whether there is any difference at all, the discrimination testing methods are more suitable (Chambers and Wolf, 1996). Scaling is usually done in one of the two scenarios. In the first, untrained observers are asked to give responses to reflect changes in intensity and it is psumed that (1) they understand the attribute they are asked to scale, e.g., sweetness and (2) there is no need to train or calibrate them to use the scale. This is the kind of scaling done to study a

7

Scaling

Fig. 7.1 The two processes involved in scaling. The first, physiological process is the psychophysical translation of energy in the outside world into sensation, i.e., conscious experience. The second is the translation of that experience into some response. The psychophysical process can be modified by physiological processes such as adaptation and masking. The judgment function can be modified by cognitive processes such as contextual effects, number usage, and other response biases.

7.2 Some Theory

7.2 Some Theory Measurement theory tells us that numbers can be assigned to items in different ways. This distinction was popularized by S. S. Stevens, the major proponent of magnitude estimation (1951). At least four ways of assigning numbers to events exist in common usage. These are referred to as nominal scaling, ordinal scaling, interval scaling, and ratio scaling. In nominal scaling, numbers are assigned to events merely as labels. Thus gender may be coded as a “dummy variable” in statistical analysis by assigning a zero to males and a one to females; no assumption is made that these numbers reflect any ordered property

151

of the sexes. They merely serve as convenient labels. The meals at which a food might be eaten could be coded with numbers as categories-one for breakfast, two for lunch, three for supper, and four for snacks. The assignment of a number for analysis is merely a label, a category or pigeonhole. The appropriate analysis of such data is to make frequency counts. The mode, the most frequent response, is used as a summary statistic for nominal data. Different frequencies of response for different products or circumstances can be compared by chi-square analysis or other nonparametric statistical methods (Siegel, 1956; see Appendix B). The only valid comparisons between inpidual items with this scale is to say whether they belong to the same category or to different ones (an equal versus not equal decision). In ordinal scaling, numbers are assigned to recognize the rank order of products with regard to some sensory property, attitude, or opinion (such as pference). In this case increasing numbers assigned to the products repsent increasing amounts or intensities of sensory experience. So a number of wines might be rank ordered for perceived sweetness or a number of fragrances rank ordered from most pferred to least pferred. In this case the numbers do not tell us anything about the relative differences among the products. We cannot draw conclusions about the degree of difference perceived nor the ratio or magnitude of difference. In an analogy to the order of runners finishing in a race, we know who placed first, second, third, etc. But this order does indicate neither the finishing distances between contestants nor the differences in their elapsed times. In general, analyses of ranked data can report medians as the summary statistic for central tendency or other percentiles to give added information. As with nominal data, nonparametric statistical analyses (see Appendix B) are appropriate when ranking is done (Siegel, 1956). The next level of scaling occurs when the subjective spacing of responses is equal, so the numbers repsent equal degrees of difference. This is called interval-level measurement. Examples in the physical sciences would be the centigrade and Fahrenheit scales of temperature. These scales have arbitrary zero points but equal pisions between values. The scales are inter-convertible through a linear transformation, for example, ◦ C = 5/9 (◦ F-32). Few scales used in sensory science have been subjected to tests that would help establish whether they achieved an interval level of measurement and yet this level is often assumed.

152

7

Scaling

instructed to assign numbers in relative proportions that reflect the strength of their sensations (Stevens, 1956). However, ratio instructions are easy to give, but whether the scale has ratio properties in reflecting a person’s actual subjective experiences is difficult to determine, if not impossible. Because of these different measurement types with different properties, the sensory professional must be careful about two things. First, statements about differences or ratios in comparing the scores for two products should not be made when the measurement level is only nominal or ordinal. Second, it is risky to use parametric statistics for measurements that reflect only frequency counts or rankings (Gaito, 1980; Townsend and Ashby, 1980). Nonparametric methods are available for statistical analyses of such data.

7.3 Common Methods of Scaling Several different scaling methods have been used to apply numbers to sensory experience. Some, like magnitude estimation, are adapted from psychophysical research, and others, like category scaling have become popular through practical application and dissemination in a wide variety of situations. This section illustrates the common techniques of category scales, line marking, and magnitude estimation. The next section discusses the less frequently used techniques of hybrid category-ratio scales, indirect scales, and ranking as alternatives. Two other methods are illustrated. Intensity matching across sensory modalities, called cross-modality matching, was an important psychophysical technique and a pcedent to some of the category-ratio scales. Finally, adjustable rating techniques in which panelists make relative placements and are able to alter their ratings are also discussed.

7.3.1 Category Scales Perhaps the oldest method of scaling involves the choice of discrete response alternatives to signify increasing sensation intensity or degrees of liking and/or pference. The alternatives may be psented in a horizontal or vertical line and may offer choices of integer numbers, simple check boxes, or word phrases. Examples of simple category scales are shown in

7.3 Common Methods of Scaling

153

A) INTENSITY 1

2

3

4

5

6

7

8

9 Strong

Weak

B) Oxidized

not noticeable trace, not sure faint slight mild moderate definite strong very strong

D) Sweetness

Extremely sweet

Not sweet at all

E) Sweetness

R

Weaker

Stronger

F) Hedonic scale for children

Super Good

Really Good

Good

Maybe Good or Maybe Bad

Bad

Really Bad

Super Bad

Fig. 7.2 Examples of category scales. (a) a simple integer scale for sensation strength (after Lawless and Malone, 1986b); (b) a verbal scale for degree of oxidized flavor (after Mecredy et al., 1974; (c) a verbal scale for degree of difference from some reference or control sample (after Aust et al., 1985), (d) a simple

check-box scale for perceived intensity; (e) a simple check-box scale for difference in intensity from some reference sample, marked R (after Stoer and Lawless, 1993); (f) a facial scale suitable for use with children, after Chen et al. (1996).

Fig. 7.2. The job of the consumer or panelist is to choose the alternative that best repsents their reaction or sensation. In a category scale the number of alternative responses is limited. Seven to 15 categories are commonly used for intensity scaling depending upon the application and the number of gradations that the panelists are able to distinguish in the products. As panel training progresses, perceptual discrimination

of intensity levels will often improve and more scale points may be added to allow the panel to make finer distinctions. A key idea is to psent an easily understandable word like “sweetness” and ask the participant to evaluate the perceived intensity of that attribute. A second important factor concerns the verbal labels that appear along the alternatives. At the very least, the low and high ends of the scale must be labeled

154

with words that make sense, e.g., “not sweet at all” to “extremely sweet.” A wide variety of these scales have been used. A common version is to allow integer numerical responses of approximately nine points (e.g., Lawless and Malone, 1986a, b). Further gradations may be allowed. For example, Winakor et al. (1980) allowed options from 1 to 99 in rating attributes of fabric handfeel. In the Spectrum method (Meilgaard et al., 2006) a 15-point category scale is used, but allows intermediate points in tenths, rendering it (at least in theory) a 150point scale. In hedonic or affective testing, a bipolar scale is common, with a zero or neutral point of opinion at the center (Peryam and Girardot, 1952). These are often shorter than the intensity scales. For example, in the “smiling face” scale used with children, only three options may be used for very young respondents (Birch et al., 1980, 1982), although with older children as many as nine points may be used (Chen et al., 1996; Kroll, 1990). Lately there has been a move away from using labels or integers, in case these may be biasing to subjects. People seem to have favorite numbers or tendencies to use some numbers more than others (e.g., Giovanni and Pangborn, 1983). A solution to this problem is to use an unlabeled check-box scale as shown in Fig. 7.2. In early applications of category scaling, the procedure specifically instructed subjects to use the categories to repsent equal spacing. They might also be instructed to distribute their judgments over the available scale range, so the strongest stimulus was rated at the highest category and the weakest stimulus at the lowest category. This use of such explicit instructions surfaces from time to time. An example is in Anderson’s (1974) recommendation to show the subject specific examples of bracketing stimuli that are above and below the anticipated range of items in the set to be judged. A related method is the relative scaling procedure of Gay and Mead (1992) in which subjects place the highest and lowest stimuli at the scale endpoints (discussed below). The fact that there is an upper boundary to the allowable numbers in a category scaling task may facilitate the achievement of a linear interval scale (Banks and Coleman, 1981). A related issue concerns what kind of experience the high end anchor refers to. Muñoz and Civille (1998) pointed out that for descriptive analysis panels, the high end-anchor phrase could refer to different situations. For example, is the term “extremely sweet”

7

Scaling

7.3 Common Methods of Scaling

7.3.2 Line Scaling

155

a) very poor

excellent

AROMA

b) moderate

weak

strong

Pepper Heat

c)

O threshold

slight

moderate

strong

Sweetness

d) weaker

reference

stronger

overall opinion

e) dislike moderately

f)

X

least liked

neither

like moderately

X most liked

Fig. 7.3 Examples of line-marking scales: (a) with endpoints labeled (after Baten, 1946); (b) with indented “goal posts” (after Mecredy et al., 1974); (c) with additional points labeled as in ASTM procedure E-1083 (ASTM, 2008b); (d) a line for ratings relative to a reference as in Stoer and Lawless (1993); (e) hedonic scaling using a line; (f) the adjustable scale of Gay and Mead (1992) as pictured by Villanueva and D Silva (2009).

An important historical trend to use lines in descriptive analysis was instrumental in the popularization of line marking. Stone et al. (1974) recommended the use of line marking for Quantitative Descriptive Analysis (QDA), then a relatively new approach to specifying the intensities of all important sensory attributes. It was important to have a scaling method in QDA that approximated an interval scale, as analysis of variance was to become the standard statistical technique for comparing products in descriptive analysis. The justification for the application in QDA appears to rest on the pvious findings of Baten regarding the sensitivity of the method and the writings of Norman Anderson on functional measurement theory (Anderson, 1974). In his approach, Anderson used indented end anchors (see Weiss, 1972, for another example). Anderson also showed his subjects examples of the high and

156

7

Scaling

with a cursor that moves up and down via the computer mouse. Time-intensity methods are reviewed more fully in Chapter 8.

7.3.3 Magnitude Estimation 7.3.3.1 The Basic Techniques A popular technique for scaling in psychophysical studies has been the method of magnitude estimation. In this procedure, the respondent is instructed to assign numbers to sensations in proportion to how strong the sensation seems. Specifically, the ratios between the numbers are supposed to reflect the ratios of sensation magnitudes that have been experienced. For example, if product A is given the value of 20 for sweetness intensity and product B seems twice as sweet, B is given a magnitude estimate of 40. The two critical parts of the technique are the instructions given to the participant and the techniques for data analysis. Two primary variations of magnitude estimation have been used. In one method, a standard stimulus is given to the subject as a reference and that standard is assigned a fixed value such as 10. All subsequent stimuli are rated relative to this standard, sometimes called a “modulus.” It is often easier for panelists if the reference (i.e., the item used as the modulus) is chosen from somewhere near the middle of the intensity range. In the other variation of magnitude estimation, no standard stimulus is given and the participant is free to choose any number he or she wishes for the first sample. All samples are then rated relative to this first intensity, although in practice people probably “chain” their ratings to the most recent items in the series. Because people can choose different ranges of numbers in this “non-modulus” magnitude estimation, the data have to be treated to bring all judgments into the same range, an extra step in the analysis. Variations on magnitude estimation and guidelines for data analysis are found in ASTM Standard Test Method E 1697-05 (ASTM, 2008a). In the psychophysical laboratory, where magnitude estimation has found its primary usage, generally only one attribute is rated at a time. However, rating multiple attributes or profiling has been used in taste studies (McBurney and Bartoshuk, 1973; McBurney and Shick, 1971; McBurney et al., 1972) and this can

7.3 Common Methods of Scaling

naturally be extended to foods with multiple taste and aromatic attributes. Magnitude estimation has not been used very often for descriptive analysis, but in principle there is no reason why it could not be used for that purpose. Participants should be cautioned to avoid falling into pvious habits of using bounded category scales, e.g., limited ranges of numbers from zero to ten. This may be a difficult problem with pviously trained panels that have used a different scaling method, as people like to stick with a method they know and feel comfortable with. Panelists who show such behavior may not understand the ratio nature of the instructions. It is sometimes useful with a new panelist to have the participant perform a warm-up task to make sure they understand the scaling instructions. The warm-up task can involve estimation of the size or area of different geometric ps (Meilgaard et al., 2006) or the length of lines (McBurney et al., 1972). Sometimes it is desired to have panelists rate multiple attributes at the same time or to break down overall intensity into specific qualities. If this kind of “profiling” is needed, the geometric ps can include areas with different shading or the lines can be differently colored. A practice task is highly recommended so that the sensory scientist can check on whether the participant understands the task. Values of zero are allowed in this method as some of the products may in fact have or no sensation for a given attribute (like no sweetness in our example). Of course, the rating of zero should not be used for the reference material. While the value of zero is consistent with common sense for products with no sensation of some attributes, it does complicate the data analysis as discussed below. 7.3.3.2 Instructions The visual appearance of the ballot in magnitude estimation is not critical; it is the instructions and the participant’s comphension of the ratio nature of the judgments that are important. Some ballots even allow the subject/participant to view all pvious ratings. Here are sample instructions for the use of magnitude estimation with a reference sample or modulus with a fixed number assigned to it: Please taste the first sample and note its sweetness. This sample is given the value of “10” for its sweetness

157 intensity. Please rate all other samples in proportion to this reference. For example, if the next sample is twice as sweet, assign it a value of “20”, if half as sweet, assign it a value of “5” and if 3.5 times as sweet, assign in a value of 35. In other words, rate the sweetness intensity so that your numbers repsent the ratios among the intensities of sweetness. You may use any positive numbers including fractions and decimals.

The other major variation on this method uses no reference. In this case the instructions may read as follows: Please taste the first sample and note its sweetness. Please rate all other samples relative to this reference, applying numbers to the samples to repsent the ratios of sweetness intensity among the samples. For example, if the next sample was twice as sweet, you would give it a number twice as big as the rating assigned to the first sample, if half as sweet, assign it a number half as big and if 3.5 times as sweet, assign it a number 3.5 times as big. You may use any positive numbers including fractions and decimals.

7.3.3.3 Data Treatment In non-modulus methods, participants will generally choose some range of numbers they feel comfortable with. The ASTM procedure suggests having them pick a value between 30 and 100 for the first sample, and avoiding any number that seems small. If participants are allowed to choose their own number range, it becomes necessary to re-scale each inpidual’s data to bring them into a common range before statistical analysis (Lane et al., 1961). This will pvent subjects who choose very large numbers from having undue influence on measures of central tendency (means) and in statistical tests. This rescaling process has been referred to as “normalizing” (ASTM, 2008a) although it has nothing to do with the normal distribution or Z-scores. A common method for rescaling proceeds as follows: (1) Calculate the geometric mean of each inpidual’s ratings across their data set. (2) Calculate the geometric mean of the entire data set (of all subjects combined). (3) For each subject, construct a ratio of the grand geometric mean of the entire data set to each person’s geometric mean. The value of this ratio provides a post hoc inpidual rescaling factor for each subject. In place of the grand geometric mean,

158

any positive numerator may also be chosen in constructing this factor, e.g., a value of 100. (4) Multiply each data point for a given person by their inpidual rescaling factor. Do this for all participants using their own inpidual re-scaling factors. These re-scaled data are then analyzed. Note that due to the extra data treatment step in this method, it is simpler to use the modulus-based variation with a standard reference item. Magnitude estimation data are often transformed to logs before data analysis (Butler et al., 1987; Lawless, 1989). This is done primarily because the data tend to be positively skewed or log-normally distributed. There tends to be some high outlying values for any given sample. Perhaps this is not surprising because the scale is open-ended at the top, and bounded by zero at the bottom. Transformation into log data and/or taking geometric means psents some problems, however, when the data contain zeros. The log of zero is undefined. Any attempt to take a geometric mean by calculating the product of N items will yield a zero on multiplying. Several approaches have been taken to this problem. One is to assign a small positive value to any zeros in the data, perhaps one-half of the smallest rating given by a subject (ASTM, 2008a). The resulting analysis, however, will be influenced by this choice. Another approach is to use the median judgments in constructing the normalization factor for non-modulus methods. The median is less influenced by the high outliers in the data than the arithmetic mean.

7.3.3.4 Applications For practical purposes, the method of magnitude estimation may be used with trained panels, consumers, and even children (Collins and Gescheider, 1989). However, the data do tend to be a bit more variable than other bounded scaling methods, especially in the hands of untrained consumers (Lawless and Malone, 1986b). The unbounded nature of the scale may make it especially well suited to sensory attributes where an upper boundary might impose restrictions on the panelists’ ability to differentiate very intense sensory experiences in their ratings. For example, irritative or painful sensations such as chili pepper intensity might all be rated near the upper bound of a category scale

7

Scaling

of intensity, but an open-ended magnitude estimation procedure would allow panelists more freedom to differentiate and report variations among very intense sensations. With hedonic scaling of likes and dislikes, there is an additional decision in using magnitude estimation scaling. Two options have been adopted in using this technique, one employing a single continuum or unipolar scale for amount of liking and the other applying a bipolar scale with positive and negative numbers plus a neutral point (Pearce et al., 1986). In bipolar magnitude scaling of likes and dislikes, positive and negative numbers are allowed in order to signify ratios or proportions of both liking and disliking (e.g., Vickers, 1983). An alternative to positives and negatives is to have the respondent merely indicate whether the number repsents liking or disliking (Pearce et al., 1986). In unipolar magnitude estimation only positive numbers (and sometimes zeros) are allowed, with the lower end of the scale repsenting no liking and higher numbers given to repsent increasing proportions of liking (Giovanni and Pangborn, 1983; Moskowitz and Sidel, 1971). It is questionable whether a unipolar scale is a sensible response task for the participant, as it does not recognize the fact that a neutral hedonic response may occur, and that there are clearly two modes of reaction, one for liking and one for disliking. If one assumes that all items are on one side of the hedonic continuum-either all liked to varying degrees or all disliked to varying degrees then the one-sided scale makes sense. However, it is a rare situation with foods or consumer product testing in which at least some indifference or change of opinion was not visible in at least some respondents. So a bipolar scale fits common sense.

7.4 Recommended Practice and Practical Guidelines Both line scales and category scales may be used effectively in sensory testing and consumer work. So we will not expend much effort in recommending one of these two common techniques over another. Some practical concerns are given below to help the student or practitioner avoid some potential problems. The category-ratio or labeled magnitude scales may facilitate comparisons among different groups, and this issue is discussed below in Section 7.5.2.

7.4 Recommended Practice and Practical Guidelines

7.4.1 Rule 1: Provide Sufficient Alternatives One major concern is to provide sufficient alternatives to repsent the distinctions that are possible by panelists (Cox, 1980). In other words, a simple 3-point scale may not suffice if the panel is highly trained and capable of distinguishing among many levels of the stimuli. This is illustrated in the case of the flavor profile scale, which began with five points to repsent no sensation, threshold sensation, weak, moderate, and strong (Caul, 1957). It was soon discovered that additional intermediate points were desirable for panelists, especially in the middle range of the scale where many products would be found. However, there is a law of diminishing returns in allowing too many scale points-further elaboration allows better differentiation of products up to a point and then the gains diminish as the additional response choices merely capture random error variation (Bendig and Hughes, 1953). A related concern is the tendency, especially in consumer work, to simplify the scale by eliminating options or truncating endpoints. This brings in the danger caused by end-use avoidance. Some respondents may be reluctant to use the end categories, just in case a stronger or weaker item may be psented later in the test. So there is some natural human tendency to avoid the end categories. Truncating a 9-point scale to a 7-point scale may leave the evaluator with what is functionally only a 5-point scale for all practical purposes. So it is best to avoid this tendency to truncate scales in experimental planning.

7.4.2 Rule 2: The Attribute Must Be Understood Intensity ratings must be collected on an attribute which the participants understand and about which they have a general consensus and agreement as to its meaning. Terms like sweetness are almost universal but a term like “green aroma” might be interpted in different ways. In the case of a descriptive panel, a good deal of effort may be directed at using reference standards to illustrate what is meant by a specific term. In the case of consumer work, such training is not done, so if any intensity ratings are collected, they must be about simple terms about which people generally

159

agree. Bear in mind that most early psychophysics was done on simple attributes like the loudness of a sound or the heaviness of a weight. In the chemical senses, with their perse types of sensory qualities and fuzzy consumer vocabulary, this is not so straightforward. Other problems to avoid include mixing sensation intensity (strength) and hedonics (liking), except in the just-right scale where this is explicit. An example of a hedonically loaded sensory term is the adjective “fresh.” Whatever this means to consumers, it is a poor choice for a descriptive scale because it is both vague and connotes some degree of liking or goodness to most people. Vague terms are simply not actionable when it comes to giving feedback to product developers about what needs to be fixed. Another such vague term that is popular in consumer studies is “natural.” Even though consumers might be able to score products on some unknown basis using this word, the information is not useful as it does not tell formulators what to change if a product scores low. A similar problem arises with attempting to scale “overall quality.” Unless quality has been very carefully defined, it cannot be scaled.

7.4.3 Rule 3: The Anchor Words Should Make Sense In setting up the scales for descriptive analysis or for a consumer test, the panel leader should carefully consider the nature of the verbal end anchors for each scale as well as any intermediate anchors that may be needed. Should the scale be anchored from “very weak” to “very strong” or will there be cases in which the sensory attribute is simply not psent? If so, it makes sense to verbally anchor the bottom of the scale with “not at all” or “none.” For example, a sweetness scale could be anchored with “not at all sweet” and a scale for “degree of oral irritation” could be anchored using “none.”

160

7.4.5 A Warning: Grading and Scoring are Not Scaling In some cases pseudo-numerical scales have been set up to resemble category scales, but the examples cut

7

Scaling

across different sensory experiences, mixing qualities. An example is the pseudo-scale used for baked products, where the number 10 is assigned for perfect texture, 8 for slight dryness, 6 for gumminess, and 4 if very dry (AACC, 1986). Gumminess and dryness are two separate attributes and should be scaled as such. This is also an example of quality grading, which is not a true scaling procedure. When the numbers shift among different sensory qualities, this violates the psychophysical model for scaling the intensity of a single attribute. Although numbers may be applied to the grades, they cannot be treated statistically, as the average of “very dry ” (4) and slightly dry (8) is not gummy (6) (see Pangborn and Dunkley, 1964, for a critique of this in the dairy grading arena). The numbers in a quality-grading scheme do not repsent any kind of unitary psychophysical continuum.

7.5 Variations-Other Scaling Techniques An important idea in scaling theory is the notion that people may have a general idea of how weak or strong sensations are, and that they can compare different attributes of a product for their relative strength, even across different sensory modalities. So, for example, someone could legitimately say that this product tastes much more salty than it is sweet. Or that the trumpets are twice as loud as the flutes in a certain passage in a symphony. Given that this notion is correct, people would seem to have a general internal scale for the strength of sensations. This idea forms the basis for several scaling methods. It permits the comparison of different sensations cross-referenced by their numerical ratings, and even can be used to compare word responses. Methods derived from this idea are discussed next.

7.5.1 Cross-Modal Matches and Variations on Magnitude Estimation The method of magnitude estimation has a basis in earlier work such as fractionation methods and the method of sense ratios in the older literature (Boring,

7.5 Variations-Other Scaling Techniques

161

1942) where people would be asked to set one stimulus in a given sensation ratio to another. The notion of allowing any numbers to be generated in response to stimuli, rather than adjusting stimuli to repsent fixed numbers appeared somewhat later (Moskowitz, 1971; Richardson and Ross, 1930; Stevens, 1956). An important outcome of these studies was the finding that the resulting psychophysical function generally conformed to a power law of the following form:

brightness of the lights about in the same proportions as the loudness of the sounds. Stevens (1969) proposed that these experiments could validate the power law, since the exponents of the cross-modality matching function could be pdicted from the exponents derived from separate scaling experiments. Consider the following example: For one sensory attribute (using the log transform, Eq. (7.2)),

R = kI n

log R1 = n1 log I1 + log k1

(7.1)

and for a second sensory attribute

or after log transformation: log(R) = n log(I) + log(k)

(7.3)

(7.2)

where R was the response, e.g., perceived loudness (mean or geometric mean of the data) and I was the physical stimulus intensity, e.g., sound pssure, and k was a constant of proportionality that depends upon the units of measurement. The important characteristic value of any sensory system was the value for n, the exponent of the power function or slope of the straight line in a log-log plot (Stevens, 1957). The validity of magnitude estimation then came to hang on the validity of the power law-the methods and resulting functions formed an internally consistent theoretical system. Stevens also viewed the method as providing a direct window into sensation magnitude and did not question the idea that these numbers generated by subjects might be biased in some way. However, the generation of responses is properly viewed as combining at least two processes, the psychophysical transformation of energy into conscious sensation and the application of numbers to those sensations. This response process was not given due consideration in the early magnitude estimation work. Responses as numbers can exhibit nonlinear transformations of sensation (Banks and Coleman, 1981; Curtis et al., 1968) so the notion of a direct translation from sensation to ratings is certainly a dangerous oversimplification. Ratio-type instructions have been applied to other techniques as well as to magnitude estimation. A historically important psychophysical technique was that of cross-modality matching, in which the sensation levels or ratios would be matched in two sensory continua such as loudness and brightness. One continuum would be adjusted by the experimenter and the other by the subject. For example, one would try to make the

log R2 = n2 log I2 + log k2

(7.4)

Setting R1 = R2 in cross-modality matching gives n1 log I1 + log k1 = n2 log I2 + log k2

(7.5)

and rearranging, log I1 = (n2 /n1 ) log I2 + a constant

(7.6)

If one plots log I1 against log I2 from a crossmodality matching task, the slope of the function can be pdicted from the ratio of the slopes of the inpidual exponents (i.e., n2 /n1 , which you can derive from two separate magnitude estimation tasks). This pdiction holds rather well for a large number of compared sensory continua (Steven, 1969). However, whether it actually provides a validation for the power law or for magnitude estimation has been questioned (e.g., Ekman, 1964). For practical purposes, it is instructive that people can actually take disparate sensory continua and compare them using some generalized notion of sensory intensity. This is one of the underpinnings of the use of a universal scale in the Spectrum descriptive procedure (Meilgaard et al., 2006). In that method, different attributes are rated on 15-point scales that can (in theory) be meaningfully compared. In other words, a 12 in sweetness is twice as intense a sensation as a 6 in saltiness. Such comparisons seem to makes sense for tastes and flavors but may not cut across all other modalities. For example, it might seem less sensible to compare the rating given for the amount of chocolate chips in a cookie to the rating given for the cookie’s hardness-these seem like quite different experiences to quantify.

162

7

Scaling

or line scale? The idea of scaling word phrases takes shape in the labeled magnitude scales discussed next.

7.5.2 Category-Ratio (Labeled Magnitude) Scales A group of hybrid techniques for scaling has recently enjoyed some popularity in the study of taste and smell, for hedonic measurement and other applications. One of the problems with magnitude estimation data is that it does not tell in any absolute sense whether sensations are weak or strong, only giving the ratios among them. This group of scales attempts to provide ratio information, but combines it with common verbal descriptors along a line scale to provide a simple frame of reference. They are referred to as category-ratio scales, or more recently, labeled magnitude scales. They all involve a horizontal or vertical line with deliberately spaced labels and the panelists’ task is to make a mark somewhere along the line to indicate the strength of their perception or strength of their likes or dislikes. In general, these labeled line scales give data that are consistent with those from magnitude estimation (Green et al., 1993). An unusual characteristic of these scales is the verbal high endanchor phrase, which often refers to the “strongest imaginable.” The technique is based on early work by Borg and colleagues, primarily in the realm of perceived physical exertion (Borg, 1982, 1990; see Green et al., 1993). In developing this scale, Borg assumed that the semantic descriptors could be placed on a ratio scale and that they defined the level of perceptual intensity and that all inpiduals experienced the same perceptual range. Borg suggested that for perceived exertion, the maximal sensation is roughly equivalent across people for this sensory “modality” (Marks et al., 1983). For example, it is conceivable that riding a bicycle to the point of physical exhaustion produces a similar sensory experience for most people. So the scale came to have the highest label referring to the strongest sensation imaginable. This led to the development of the labeled magnitude scale (LMS) shown in Fig. 7.4. It is truly a hybrid method since the response is a vertical linemarking task but verbal anchors are spaced according to calibration using ratio-scaling instructions (Green

7.5 Variations-Other Scaling Techniques

Strongest imaginable

Very strong

(95.5)

(50.12)

Strong

(33.1)

Moderate

(16.2)

Weak Barely detectable

163

LMS

(5.8) (1.4)

Fig. 7.4 The labeled magnitude scale (LMS) of Green et al. (1993).

2007; Lawless et al., 2010a, b, c). A growing number of similar scales have been developed for various applications including oral pleasantness/unpleasantness (the “OPUS” scale, Guest et al., 2007), perceived satiety (the “SLIM” scale, Cardello et al., 2005), clothing fabric comfort (the “CALM” scale, Cardello et al., 2003), and odor dissimilarity (Kurtz et al., 2000). All of these scales depend upon a ratio scaling task to determine the spacing of the verbal descriptors and almost all use a Borg-type high end-anchor phrase. Others will surely be developed. Instructions to participants have differed in the use of these scales. In the first application of the LMS, Green et al. (1993) instructed subjects to first choose the most appropriate verbal descriptor, and then to “fine tune” their judgment by placing a mark on the line between that descriptor and the next most appropriate one. In current practice less emphasis may be placed on the consideration of the verbal labels and instructions may be given to simply make a mark “anywhere” on the line. A common observation with the hedonic versions of the scale is that some panelists will mark at or very near a verbal descriptor, seeming to use it as a category scale (Cardello et al., 2008; Lawless et al., 2010a). The proportion of people displaying this behavior may depend upon the physical length of the line (and not the instructions or examples that may be shown) (Lawless et al., 2010b). Results may depend in part on the nature of the high end-anchor example and the frame of reference of the subject in terms of the sensory modality they are thinking about. Green et al. (1996) studied the application of the LMS to taste and odor, using descriptors for the upper bound as “strongest imaginable” taste, smell, sweetness, etc. Steeper functions (a smaller response range) were obtained when mentioning inpidual taste qualities. This appears to be due to the omission of painful experiences (e.g., the “burn of hot peppers”) from the frame of reference when sensations were scaled relative to only taste. The steepening of the functions for the truncated frame of reference is consistent with the idea that subjects expanded their range of numbers as seen in other scaling experiments (e.g., Lawless and Malone, 1986b). The fact that subjects appear to adjust their perceptual range depending on instructions or frame of reference suggests that the scales have relative and not absolute properties, like most other scaling methods. Cardello et al. (2008) showed that the hedonic version of the scale (the LAM

164 Greatest imaginable like

Most imaginable

(+100)

Scaling (+100)

Like Extremely

(+74.2)

extremely

(+75.2)

Like Very Much

(+56.1)

very

(+56.1)

Like Moderately

(+36.2)

moderately

(+33.9)

Like Slightly

(+11.2)

slightly weakly

(+19.7) (+15.0)

PLEASANT

Fig. 7.5 Affective labeled magnitude scales, including the LAM scale (Cardello and Schutz, 2004) and the OPUS scale (Guest et al., 2007).

7

barely detectable Neither like nor dislike

(+6.5) (0)

(0)

(+6.5)

(-10.6)

barely detectable weakly slightly

(+15.0) (+19.7)

Dislike moderately

(-31.9)

moderately

(+33.9)

Dislike Very much

(-55.5)

very

(+56.1)

Dislike extremely

(-75.5)

extremely

(+75.2)

Greatest imaginable dislike

(-100)

LAM scale) will also show such range effects. A compssed range of responses is obtained when the frame of reference is greatest imaginable like (dislike) for an “experience of any kind” rather than something more delimited like “foods and beverages.” Apparently the compssion is not very detrimental to the ability of the LAM scale to differentiate products (Cardello et al., 2008, but see also Lawless et al., 2010a). Is this a suitable method for cross-subject comparisons? To the extent that Borg’s assumptions of common perceptual range and the similarity of the high end-anchor experience among people are true, the method might provide one approach to valid comparisons of the ratings among different respondents. This would facilitate comparisons of clinical groups or patients with sensory disorders or genetically different inpiduals such as anosmics, PTC/PROP taster groups. Bartoshuk and colleagues (1999, 2003,

UNPLEASANT

Dislike slightly

Most imaginable

(-100)

OPUS 2004a, b, 2006) have argued that the labeled magnitude scales should anchor their endpoints to “sensations of any kind” as such a reference experience would allow different inpiduals to use the scale in similar ways/and thus facilitate inter-inpidual comparisons. Scales with this kind of high end anchor have been termed “generalized” labeled magnitude scales (or gLMS). However, the sensory evaluation practitioner should be aware of the compssion effects that can occur with this kind of scale, which could potentially lead to lessened differentiation among products.

7.5.3 Adjustable Rating Techniques: Relative Scaling A few methods have been tried that allow consumers or panelists to change their ratings. An example is

7.5 Variations-Other Scaling Techniques

165

7.5.4 Ranking Another alternative to traditional scaling is the use of ranking procedures. Ranking is simply ordering the products from weakest to strongest on the stated

166

7.5.5 Indirect Scales A conservative approach to scaling is to use the variance in the data as units of measurement, rather than the numbers taken at face value. For example, we could

7

Scaling

ask how many standard deviations apart are the mean values for two products. This is a different approach to measurement than simply asking how many scale units separate the means on the response scale. On a 9-point scale, one product may receive mean rating of seven, and another nine, making the difference two scale units. If the pooled standard deviation is two units, however, they would only be one unit apart on a variability-based scale. As one example, Conner and Booth (1988) used both the slope and the variance of functions from just-right scales to derive a “tolerance discrimination ratio.” This ratio repsents a measure of the degree of difference along a physical continuum (such as concentration of sugar in lime drink) that observers find to make a meaningful change in their ratings of difference-from-ideal (or just-right). This is analogous to finding the size of a just-noticeable difference, but translated into the realm of hedonic scaling. Their insight was that it is not only the slope of the line that is important in determining this tolerance or liking-discrimination function, but also the variance around that function. Variability-based scales are the basis for scaling in Thurstone’s models for comparative judgment (Thurstone, 1927) and its extension into determining distances between category boundaries (Edwards, 1952). Since the scale values can be found from choice experiments as well as rating experiments, the technique is quite flexible. How this type of scaling can be applied to rating data is discussed below in the derivation of the 9-point hedonic scale words. When the scale values are derived from a choice method like the triangle test or paired comparison method, this is sometimes called “indirect scaling” (Baird and Noma, 1978; Jones, 1974). The basic operation in Thurstonian scaling of choice data is to convert the proportion correct in a choice experiment (or simply the proportions in a two-tailed test like paired pference) to Z-scores. The exact derivation depends upon the type of test (e.g., triangle versus 3-AFC) and the cognitive strategy used by the subject. Tables for Thurstonian scale values from various tests such as the triangle test were given by Frijters et al. (1980) and Bi (2006) and some tables are given in the Appendix. Mathematical details of Thurstonian scaling are discussed in Chapter 5. Deriving measures of sensory differences in such indirect ways psents several problems in applied sensory evaluation so the method has not been widely used. The first problem is one of economy in data

7.6 Comparing Methods: What is a Good Scale?

collection. Each difference score is derived from a separate discrimination experiment such as a paired comparison test. Thus many subjects must be tested to get a good estimate of the proportion of choice, and this yields just one scale value. In direct scaling, each participant gives at least one data point for each item tasted. Direct scaling allows for easy comparisons among multiple products, while the discrimination test must be done on one pair at a time. Thus the methods of indirect scaling are not cost-efficient. A second problem can occur if the products are too clearly different on the attribute in question, because then the proportion correct will approach 100%. At that point the scale value is undefined as they are some unknown number of standard deviations apart. So the method only works when there are small differences and some confusability of the items. In a study of many products, however, it is sometimes possible to compare only adjacent or similar items, e.g., products that differ in small degrees of some ingredient or process variable. This approach was taken by Yamaguchi (1967) in examining the synergistic taste combination of monosodium glutamate and disodium 5 inosinate. Many different levels of the two ingredients were tasted, but because the differences between some levels were quite apparent, an incomplete design was used in which only three adjacent levels were compared. Other applications have also used this notion of variability as a yardstick for sensory difference or sensation intensity. The original approach of Fechner in constructing a psychophysical function was to accumulate the difference thresholds or just-noticeable differences (JNDs) in order to construct the log function of psychophysical sensory intensity (Boring, 1942; Jones, 1974). McBride (1983a, b) examined whether JND-based scales might give similar results to category scales for taste intensity. Both types of scaling yielded similar results, perhaps not surprising since both tend to conform to log functions. In a study of children’s pferences for different odors Engen (1974) used a paired pference paradigm, which was well suited to the abilities of young children to respond in a judgment task. He then converted the paired pference proportions to Thurstonian scale values via Z-scores and was able to show that the hedonic range of children was smaller than that of adults. Another example of choice data that can be converted to scale values is best-worst scaling, in which

167

a consumer is asked to choose the best liked and least liked samples from a set of three or more items (Jaeger et al., 2008). With three products, it can be considered a form of a ranking task. When applied to sensory intensity, this is sometimes known as maximum-difference or “max-diff.” Best-worst scaling is also discussed in Section 13.7. Simple difference scores may be calculated based on the number of times an item is called best versus worst and these scores are supposed to have interval properties. If a multinomial logistic regression is performed on the data, they are theorized to have true ratio properties (Finn and Louviere, 1992). A practical problem with the method, however, is that so many products must be tasted and compared, rendering it difficult to perform with foods (Jaeger and Cardello, 2009). The sensory professional should bear in mind that in spite of their theoretical sophistication, the indirect methods are based on variability as the main determinant of degree of difference. Thus any influence, which increases variability, will tend to decrease the measured differences among products. In the wellcontrolled psychophysical experiment under constant standard conditions across sessions and days, this may not be important-the primary variability lies in the resolving power of the participant (and secondarily in the sample products). But in drawing conclusions across different days, batches, panels, factories, and such, one has a less pure situation to consider. Whether one considers the Thurstonian-type indirect measures comparable across different conditions depends upon the control of extraneous variation.

7.6 Comparing Methods: What is a Good Scale? A large number of empirical studies have been conducted comparing the results using different scaling methods (e.g., Birnbaum, 1982; Giovanni and Pangborn, 1983; Hein et al., 2008; Jaeger and Cardello, 2009; Lawless and Malone, 1986a, b; Lawless et al., 2010a; Marks et al., 1983; Moskowitz and Sidel, 1971; Pearce et al., 1986; Piggot and Harper, 1975; Shand et al., 1985; Vickers, 1983; Villanueva and Da Silva, 2009; Villanueva et al., 2005). Because scaling data are often used to identify differences between products, the ability to detect differences is one important

168

practical criterion for how useful a scaling method can be (Moskowitz and Sidel, 1971). A related criterion is the degree of error variance or similar measures such as size of standard deviations or coefficients of variation. Obviously, a scaling method with low interinpidual variability will result in more sensitive tests, more significant differences, and lower risk of Type II error (missing a true difference). A related issue is the reliability of the procedure. Similar results should be obtained upon repeated experimentation. Other practical considerations are important as well. The task should be user friendly and easy to understand for all participants. Ideally, the method should be applicable to a wide range of products and questions, so that the respondent is not confused by changes in response type over a long ballot or questionnaire. If panelists are familiar with one scale type and are using it effectively, there may be some liability in trying to introduce a new or unfamiliar method. Some methods, like category scales, line scales, and magnitude estimation, can be applied to both intensity and hedonic (like-dislike) responses. The amount of time required to code, tabulate and process the information may be a concern, depending upon computer-assisted data collection and other resources available to the experimenters. As in any method, validity or accuracy are also issues. Validity can only be judged by reference to some external criterion. For hedonic scaling, one might want the method to correspond to other behaviors such as choice or consumption (Lawless et al., 2010a). A related criterion is the ability of the scale to identify or uncover consumer segments with different pferences (Villanueva and Da Silva, 2009). Given these practical considerations, we may then ask how the different scaling methods fare. Most published studies have found about equal sensitivity for the different scaling methods, provided that the methods are applied in a reasonable manner. For example, Lawless and Malone (1986a, b) performed an extensive series of studies (over 20,000 judgments) with consumers in central location tests using different sensory continua including olfaction, tactile, and visual modalities. They compared line scales, magnitude estimation, and category scales. Using the degree of statistical differentiation among products as the criterion for utility of the methods, the scales performed about equally well. A similar conclusion was reached by Shand et al. (1985) for trained panelists. There

7

Scaling

was some small tendency for magnitude estimation to be marginally more variable in the hands of consumers as opposed to college students (Lawless and Malone, 1986b). Statistical differentiation increased over replicates, as would be expected as people came to understand the range of items to be judged (see Hein et al., 2008, for another example of improvement over replication in hedonic scaling). Similar findings for magnitude estimation and category scales in terms of product differentiation were found by Moskowitz and Sidel (1971), Pearce et al. (1986), Shand et al. (1985), and Vickers (1983) although the forms of the mathematical relations to underlying physical variables was often different (Piggot and Harper, 1975). In other words, as found by Stevens and Galanter (1957) there is often a curvilinear relationship between the data from the two methods. However, this has not been universally observed and sometimes simple linear relationships have been found (Vickers, 1983). Similar results for category scales and line scales were found by Mattes and Lawless (1985). Taken together, these empirical studies paint a picture of much more parity among methods than one might suppose given the number of arguments over the validity of scaling methods in the psychological literature. With reasonable spacing of the products and some familiarization with the range to be expected, respondents will distribute their judgments across the available scale range and use the scale appropriately to differentiate the products. A reasonable summary of the literature comparing scale types is that they work about equally well to differentiate products, given a few sensible pcautions.

7.7 Issues

responses (i.e., judgments that might fall in-between categories). The line scale offers a continuously graded choice of alternatives, limited only by the measurement abilities in data tabulation. Baten also noted that the line scale seemed to facilitate a relative comparison among the products. This was probably due to his placement of the scales one above the other on the ballot, so judges could see both markings at the same time. In order to minimize such contextual effects it is now more common to remove the prior ratings for products to achieve a more independent judgment of the products. However, whether that is ever achieved in practice is open to question-humans are naturally comparative when asked to evaluate items, as discussed in Chapter 9. Furthermore, there may be potential for increased discrimination in methods like the relative positioning technique. The naturally comparative nature of human judgment may be something we could benefit from rather than trying to fight this tendency by over-calibration.

7.7.2 Should Category Rating Scales Be Assigned Integer Numbers in Data Tabulation? Are They Interval Scales? There is also a strong suspicion that many numerical scaling methods may produce only ordinal data, because the spacing between alternatives is not subjectively equal. A good example is the common marketing research scale of “excellent-very good-good- fair-poor.” The subjective spacing between these adjectives is quite uneven. The difference between two products rated good and very good is a much smaller difference than that between products rated fair and poor. However, in analysis we are often tempted to assigned numbers one through five to these categories and take means and perform statistics as if the assigned numbers reflected equal spacing. This is a ptense at best. A reasonable analysis of the 5-point excellent to poor scale is simply to count the number of respondents in each category and to compare frequencies. Sensory scientists should not assume that any scale has interval properties in spite of how easy it is to tabulate data as an integer series.

169

7.7.4 What is a “Valid” Scale? An ongoing issue in psychophysics is what kind of scale is a true reflection of the subject’s actual sensations? From this perspective, a scale is valid when the numbers generated reflect a linear translation of subjective intensity (the private experience). It is well established that category scales and magnitude estimates, when given to the same stimuli, will form a curve when plotted against one another (Stevens and Galanter, 1957). Because this is not a linear relationship, one method or the other must result from a non-linear translation of the subjective intensities of the stimuli. Therefore, by this criterion, at least one scale must be “invalid.”

170

7

Anderson (1974, 1977) proposed a functional measurement theory to address this issue. In a typical experiment, he would ask subjects to do some kind of combination task, like judging the total combined intensity of two separately psented items (or the average lightness of two gray swatches). He would set up a factorial design in which every level of one stimulus was combined with every level of the other (i.e., a complete block). When plotting the response, a family of lines would be seen when the first stimulus continuum formed the X-axis and the second formed a family of lines. Anderson argued that only when the response combination rule was additive, and the response output function was linear, would a parallel plot be obtained (i.e., there would be no significant interaction term in ANOVA). This argument is illustrated in Fig. 7.6. In his studies using simple line and category scales, parallelism was obtained in a number of studies, and thus he reasoned that magnitude estimation was invalid by this criterion. If magnitude estimation is invalid, then its derivatives such as the LMS and LAM scales are similarly suspect.

S1

p1

S2

p2

P Psychophysical process

Response output process

Integration process

Response for Combination Judgment

R

(High)

(Med.)

Levels of Stimulus or Variable 2

(Low)

(Low)

(Med.)

(High)

Levels of Stimulus or Variable 1

Fig. 7.6 The functional measurement scheme of Anderson (1974).

Scaling

Others have found support for the validity of magnitude estimation in studies of binaural loudness summation (Marks, 1978). This argument continues and is difficult to resolve. A review of the matter was published by Gescheider (1988). For the purposes of sensory evaluation, the issue is not terribly important for two reasons. First, any scale that produces statistically significant differentiation of products is a useful scale. Second, the physical ranges over which category scaling and magnitude estimation produce different results is usually quite large in any psychophysical study. In most product tests, the differences are much more subtle and generally do not span such a wide dynamic range. The issue dissolves from any practical perspective.

7.8 Conclusions Much sound and fury has been generated over the years in the psychophysical literature concerning what methods yield valid scales. For the sensory practitioner, these issues are less relevant because the scale values do not generally have any absolute meaning. They are only convenient indices of the relative intensities or appeal of different products. The degree of difference may be a useful piece of information, but often we are simply interested in which product is stronger or weaker in some attribute, and whether the difference is both statistically significant and practically meaningful. Scaling provides a quick and useful way to get intensity or liking information. In the case of descriptive analysis, scaling allows collection of quantitative data on multiple attributes. The degree of variability or noise in the system is, to a large part, determined by whether the panelists have a common frame of reference. Thus reference standards for both the attribute terms and for the intensity anchors are useful. Of course, with consumer evaluations or a psychophysical study such calibration is not possible and usually not desired. The variability of consumer responses should offer a note of caution in the interptation of consumer scaling data. Students and sensory practitioners should examine their scaling methods with a critical eye. Not every task that assigns numbers will have useful scale properties like equal intervals. Bad examples abound

Appendix 1

in the commodity grading (quality scoring) literature (Pangborn and Dunkley, 1964). For example, different numbers may be assigned to different levels of oxidation, but that is scoring a physical condition based on inferences from sensory experience. It is not a report of the intensity of some experience itself. It is not tracking changes along a single perceptual continuum in the psychophysical sense. Scoring is not scaling. All hedonic scales seem to measure what they are intended to measure rather effectively, as long as no gross mistakes are made (Peryam, 1989, p. 23).

Appendix 1: Derivation of Thurstonian-Scale Values for the 9-Point Scale The choice of adjective words for the 9-point hedonic scale is a good example of how carefully a scale can be constructed. The long-standing track record of this tool demonstrates its utility and wide applicability in consumer testing. However, few sensory practitioners actually know how the adjectives were found and what criteria were brought to bear in selecting these descriptors (slightly, moderately, very much, and extremely like/dislike) from a larger pool of possible words. The goal of this section is to provide a shorthand description of the criteria and mathematical method used to select the words for this scale. One concern was the degree to which the term had consensual meaning in the population. The most serious concern was when a candidate word had an ambiguous or double meaning across the population. For example, the word “average” suggests an intermediate response to some people, but in the original study by Jones and Thurstone (1955) there were a group of people who equated it with “like moderately” perhaps since an average product in those days was one that people would like. These days, one can think of negative connotations to the word “average” as in “he was only an average student.” Other ambiguous or bimodal terms were “like not so much” and “like not so well.” Ideally, a term should have low variability in meaning, i.e., a low standard deviation, no bimodality, and little skew. Part of this concern with the normality of the distribution of psychological reactions to a word was the fact that the developers used Thurstone’s model

171

172

7

that marks the ordinate in equal standard deviations units according to the cumulative normal distribution. Either of these methods will tend to make the cumulative S-shaped curve for the item into a straight line. The X-axis value for each point is the “category z-value” for that bucket. 6. Fit a line to the data and interpolate the 50% point on the X-axis (the re-scaled category boundary estimates). These interpolated values for the median for each item now form the new scale values for the items. An example of this interpolation is shown in Fig. 7.7. Three of the phrases used in the original scaling study of Jones and Thurstone (1955) are pictured, three that were not actually chosen but for which we have approximate proportions and z-scores from their ps. The small vertical arrows on the X-axis show the scale values for the original categories of −4 to +3 (+4 has cumulative proportion of 100% and thus the z-score is infinite). Table 7.1 gives the values and

Scaling

proportions for each phrase and the original categories. The dashed vertical lines dropped from the intersection at the zero z-score (50% point) show the approximate mean values interpolated on the X-axis (i.e., about -1.1 for “do not care for it” and about +2.1 for “pferred.”). Note that “pferred” and “don’t care for it” have a linear fit and steep slope, suggesting a normal distribution and low standard deviation. In contrast, “highly unfavorable” has a lower slope and some curvilinearity, indicative of higher variability, skew, and/or pockets of disagreement about the connotation of this term. The actual scale values for the original adjectives are shown in Table 7.2, as found with a soldier population circa 1950 (Jones et al., 1955). You may note that the words are not equally spaced, and that the “slightly” values are closer to the neutral point than some of the other intervals, and the extreme points are a little farther out. This bears a good deal of similarity to the intervals found with the LAM scale as shown in the column where the LAM values are re-scaled to the same range as the 9-point Thurstonian Values.

Appendix 2: Construction of Labeled Magnitude Scales “

Fig. 7.7 An illustration of the method used to establish spacings and scale values for the 9-point hedonic scale using Thurstonian theory. Arrows on the X-axis show the scale points for the z-scores based on the complete distribution of the original -4 to +4 ratings. The Y-axis shows the actual z-scores based on the proportion of respondents using that category for each specific term. Re-plotted from data provided in Jones et al. (1955).

There are two primary methods for constructing labeled magnitude scales and they are very similar. Both require magnitude estimates from the participants to scale the word phrases used on the lines. In one case, just the word phrases are scaled, and in the second method, the word phrases are scaled among a list of common everyday experiences or sensations that most people are familiar with. The values obtained by the simple scaling of just the words will depend upon the words that are chosen, and extremely high examples (e.g., greatest imaginable liking for any experience) will tend to compss the values of the interior phrases (Cardello et al., 2008). Whether this kind of context effect will occur for the more general method of scaling amongst common experiences is not known. But the use of a broad frame of reference could be a stabilizing factor. Here is an example of the instructions given to subjects in construction of a labeled affective magnitude scale. Note that for hedonics, which are a bipolar continuum with a neutral point, it is necessary to collect a tone or valence (plus or minus) value as well as the overall “intensity” rating.

Appendix 2

173

Table 7.1 Examples of scaled phrases used in Fig. 7.7 Original “Preferred” category Proportion Z-score Proportion 4 3 2 1 0 −1 −2 −3 −4

1.000 0.999 0.983 0.882 0.616 0.383 0.185 0.068 0.008

Table 7.2 Actual 9-point scale phrase values and comparison to the LAM values

(undef.) 3.0 2.1 1.2 0.3 −0.3 −0.9 −1.5 −2.4

0.80 0.50 0.20 0.07 0.03

Z-score 0.84 0.00 −0.84 −1.48 −1.88

Descriptor

Scale value (9-point)

Like extremely Like very much Like moderately Like slightly Neither like nor dislike Dislike slightly Dislike moderately Dislike very much Dislike extremely

4.16 2.91 1.12 0.69 0.00 −0.59 −1.20 −2.49 −4.32

Next to each word label a response area appeared similar to this: Phrase: Like extremely

Words or phrases are psented in random order. After reading a word they must decide whether the word is positive, negative or neutral and place the corresponding symbol on the first line. If the hedonic tone was not a neutral one (zero value), they are instructed to give a numerical estimate using modulus-free magnitude estimation. The following is a sample of the instructions taken from Cardello et al. (2008): After having determined whether the phrase is positive or negative or neutral and writing the appropriate symbol (+, −, 0) on the first line, you will then assess the strength or magnitude of the liking or disliking reflected by the phrase. You will do this by placing a number on the second blank line (under “How Much”). For the first phrase that you rate, you can write any number you want on the line. We suggest you do not use a small number for this word/phrase. The reason for this is that subsequent words/phrases may reflect much lower levels of liking or disliking. Aside from this restriction you can use any numbers you want. For each subsequent

“Do not care for it” Proportion Z-score

“Highly unfavorable” Proportion Z-score

0.96 0.83 0.55 0.30 0.14

1.75 0.95 0.13 −0.52 −1.08

0.96 0.93 0.92 0.90 0.86 0.84 0.82 0.46

Interval

LAM value

LAM rescaled

Interval

1.26 1.79 0.43 0.69 0.59 0.61 1.29 1.83

74.2 56.1 36.2 11.2 0.0 −10.6 −31.9 −55.5 −75.5

4.20 3.18 2.05 0.63 0.00 −0.60 −1.81 −3.14 −4.28

1.02 1.13 1.52 0.63 0.60 1.21 1.33 1.14

1.75 1.48 1.41 1.28 1.08 0.99 0.92 −0.10

word/phrase your numerical judgment should be made proportionally and in comparison to the first number. That is, if you assigned the number 800 to index the strength of the liking/disliking denoted by the first word/phrase and the strength of liking/disliking denoted by the second word/phrase were twice as great, you would assign the number 1,600. If it were three times as great you would assign the number 2,400, etc. Similarly, if the second word/phrase denoted only 1/10 the magnitude of liking as the first, you would assign it the number 80 and so forth. If any word/phrase is judged to be “neutral” (zero (0) on the first line) it should also be given a zero for its magnitude rating.

In the cased of Cardello et al. (2008), positive and negative word labels were analyzed separately. Raw magnitude estimates were equalized for scale range using the procedure of Lane et al. (1961). All positive and negative magnitude estimates for a given subject were multiplied by an inpidual scaling factor. This factor was equal to the ratio of the grand geometric mean (of the absolute value of all nonzero ratings) across all subjects pided by the geometric mean for that subject. The geometric mean magnitude estimates for each phrase were then calculated based on this range-equated data. These means became the distance

174

from the zero point for placement of the phrases along the scale, usually accompanied by a short cross-hatch mark at that point.

7

Scaling

Basker, D. 1988. Critical values of differences among rank sums for multiple comparisons. Food Technology, 42(2), 79, 80-84. Baten, W. D. 1946. Organoleptic tests pertaining to apples and pears. Food Research, 11, 84-94. Bendig, A. W. and Hughes, J. B. 1953. Effect of number of verbal anchoring and number of rating scale categories upon transmitted information. Journal of Experimental Psychology, 46(2), 87-90. Bi, J. 2006. Sensory Discrimination Tests and Measurement. Blackwell, Ames, IA. Birch, L. L., Zimmerman, S. I. and Hind, H. 1980. The influence of social-affective context on the formation of children’s food pferences. Child Development, 51, 865-861. Birch, L. L., Birch, D., Marlin, D. W. and Kramer, L. 1982. Effects of instrumental consumption on children’s food pferences. Appetite, 3, 125-143. Birnbaum, M. H. 1982. Problems with so-called “direct” scaling. In: J. T. Kuznicki, R. A. Johnson and A. F. Rutkiewic (eds.), Selected Sensory Methods: Problems and Approaches to Hedonics. American Society for Testing and Materials, Philadelphia, pp. 34-48. Borg, G. 1982. A category scale with ratio properties for intermodal and interinpidual comparisons. In: H.-G. Geissler and P. Pextod (Eds.), Psychophysical Judgment and the Process of Perception. VEB Deutscher Verlag der Wissenschaften, Berlin, pp. 25-34. Borg, G. 1990. Psychophysical scaling with applications in physical work and the perception of exertion. Scandinavian Journal of Work and Environmental Health, 16, 55-58. Boring, E. G. 1942. Sensation and Perception in the History of Experimental Psychology. Appleton-Century-Crofts, New York. Brandt, M. A., Skinner, E. Z. and Coleman, J. A. 1963. The texture profile method. Journal of Food Science, 28, 404-409. Butler, G., Poste, L. M., Wolynetz, M. S., Agar, V. E. and Larmond, E. 1987. Alternative analyses of magnitude estimation data. Journal of Sensory Studies, 2, 243-257. Cardello, A. V. and Schutz, H. G. 2004. Research note. Numerical scale-point locations for constructing the LAM (Labeled affective magnitude) scale. Journal of Sensory Studies, 19, 341-346. Cardello, A. V., Lawless, H. T. and Schutz, H. G. 2008. Effects of extreme anchors and interior label spacing on labeled magnitude scales. Food Quality and Preference, 21, 323-334. Cardello, A. V., Winterhaler, C. and Schutz, H. G. 2003. Predicting the handle and comfort of military clothing fabrics from sensory and instrumental data: Development and application of new psychophysical methods. Textile Research Journal, 73, 221-237. Cardello, A. V., Schutz, H. G., Lesher, L. L. and Merrill, E. 2005. Development and testing of a labeled magnitude scale of perceived satiety. Appetite, 44, 1-13. Caul, J. F. 1957. The profile method of flavor analysis. Advances in Food Research, 7, 1-40. Chambers, E. C. and Wolf, M. B. 1996. Sensory Testing Methods. ASTM Manual Series, MNL 26. ASTM International, West Conshohocken, PA. Chen, A. W., Resurreccion, A. V. A. and Paguio, L. P. 1996. Age appropriate hedonic scales to measure the food pferences of young children. Journal of Sensory Studies, 11, 141-163.

References Chung, S.-J. and Vickers, 2007a. Long-term acceptability and choice of teas differing in sweetness. Food Quality and Preference 18, 963-974. Chung, S.-J. and Vickers, 2007b. Influence of sweetness on the sensory-specific satiety and long-term acceptability of tea. Food Quality and Preference, 18, 256-267. Coetzee, H. and Taylor, J. R. N. 1996. The use and adaptation of the paired comparison method in the sensory evaluation of hamburger-type patties by illiterate/semi-literate consumers. Food Quality and Preference, 7, 81-85. Collins, A. A. and Gescheider, G. A. 1989. The measurement of loudness in inpidual children and adults by absolute magnitude estimation and cross modality matching. Journal of the Acoustical Society of America, 85, 2012-2021. Conner, M. T. and Booth, D. A. 1988. Preferred sweetness of a lime drink and pference for sweet over non-sweet foods. Related to sex and reported age and body weight. Appetite, 10, 25-35. Cordinnier, S. M. and Delwiche, J. F. 2008. An alternative method for assessing liking: Positional relative rating versus the 9-point hedonic scale. Journal of Sensory Studies, 23, 284-292. Cox, E. P. 1980. The optimal number of response alternatives for a scale: A review. Journal of Marketing Research, 18, 407-422. Curtis, D. W., Attneave, F. and Harrington, T. L. 1968. A test of a two-stage model of magnitude estimation. Perception and Psychophysics, 3, 25-31. Edwards, A. L. 1952. The scaling of stimuli by the method of successive intervals. Journal of Applied Psychology, 36, 118-122. Ekman, G. 1964. Is the power law a special case of Fechner’s law? Perceptual and Motor Skills, 19, 730. Einstein, M. A. 1976. Use of linear rating scales for the evaluation of beer flavor by consumers. Journal of Food Science, 41, 383-385. El Dine, A. N. and Olabi, A. 2009. Effect of reference foods in repeated acceptability tests: Testing familiar and novel foods using 2 acceptability scales. Journal of Food Science, 74, S97-S105. Engen, T. 1974. Method and theory in the study of odor pferences. In: A. Turk, J. W. Johnson and D. G. Moulton (Eds.), Human Responses to Environmental Odors. Academic, New York. Finn, A. and Louviere, J. J. 1992. Determining the appropriate response to evidence of public concern: The case of food safety. Journal of Public Policy and Marketing, 11, 12-25. Forde, C. G. and Delahunty, C. M. 2004. Understanding the role cross-modal sensory interactions play in food acceptability in younger and older consumers. Food Quality and Preference, 15, 715-727. Frijters, J. E. R., Kooistra, A. and Vereijken, P. F. G. 1980. Tables of d’ for the triangular method and the 3-AFC signal detection procedure. Perception and Psychophysics, 27, 176-178. Gaito, J. 1980. Measurement scales and statistics: Resurgence of an old misconception. Psychological Bulletin, 87, 564-587. Gay, C., and Mead, R. 1992 A statistical appraisal of the problem of sensory measurement. Journal of Sensory Studies, 7, 205-228.

175 Gent, J. F. and Bartoshuk, L. M. 1983. Sweetness of sucrose, neohesperidin dihydrochalcone and sacchar in is related to genetic ability to taste the bitter substance 6-npropylthiouracil. Chemical Senses, 7, 265-272. Gescheider, G. A. 1988. Psychophysical scaling. Annual Review of Psychology, 39, 169-200. Giovanni, M. E. and Pangborn, R. M. 1983. Measurement of taste intensity and degree of liking of beverages by graphic scaling and magnitude estimation. Journal of Food Science, 48, 1175-1182. Gracely, R. H., McGrath, P. and Dubner, R. 1978a. Ratio scales of sensory and affective verbal-pain descriptors. Pain, 5, 5-18. Gracely, R. H., McGrath, P. and Dubner, R. 1978b. Validity and sensitivity of ratio scales of sensory and affective verbal-pain descriptors: Manipulation of affect by Diazepam. Pain, 5, 19-29. Green, B. G., Shaffer, G. S. and Gilmore, M. M. 1993. Derivation and evaluation of a semantic scale of oral sensation magnitude with apparent ratio properties. Chemical Senses, 18, 683-702. Green, B. G., Dalton, P., Cowart, B., Shaffer, G., Rankin, K. and Higgins, J. 1996. Evaluating the “Labeled Magnitude Scale” for measuring sensations of taste and smell. Chemical Senses, 21, 323-334. Greene, J. L., Bratka, K. J., Drake, M. A. and Sanders, T. H. 2006. Effective of category and line scales to characterize consumer perception of fruity fermented flavors in peanuts. Journal of Sensory Studies, 21, 146-154. Guest, S., Essick, G., Patel, A., Prajpati, R. and McGlone, F. 2007. Labeled magnitude scales for oral sensations of wetness, dryness, pleasantness and unpleasantness. Food Quality and Preference, 18, 342-352. Hein, K. A., Jaeger, S. R., Carr, B. T. and Delahunty, C. M. 2008. Comparison of five common acceptance and pference methods. Food Quality and Preference, 19, 651-661. Huskisson, E. C. 1983. Visual analogue scales. In: R. Melzack (Ed.), Pain Measurement and Assessment. Raven, New York, pp. 34-37. Jaeger, S. R.; Jørgensen, A. S., AAslying, M. D. and Bredie, W. L. P. 2008. Best-worst scaling: An introduction and initial comparison with monadic rating for pference elicitation with food products. Food Quality and Preference, 19, 579-588. Jaeger, S. R. and Cardello, A. V. 2009. Direct and indirect hedonic scaling methods: A comparison of the labeled affective magnitude (LAM) scale and best-worst scaling. Food Quality and Preference, 20, 249-258. Jones, F. N. 1974. History of psychophysics and judgment. In: E. C. Carterette and M. P. Friedman (Eds.), Handbook of Perception. Psychophysical Judgment and Measurement, Vol. 2. Academic, New York, pp. 11-22. Jones, L. V. and Thurstone, L. L. 1955. The psychophysics of semantics: An experimental investigation. Journal of Applied Psychology, 39, 31-36. Jones, L. V., Peryam, D. R. and Thurstone, L. L. 1955. Development of a scale for measuring soldier’s food pferences. Food Research, 20, 512-520. Keskitalo, K. Knaapila, A., Kallela, M., Palotie, A., Wessman, M., Sammalisto, S., Peltonen, L., Tuorila, H. and Perola, M. 2007. Sweet taste pference are partly genetically

176 determined: Identification of a trait locus on Chromosome 161-3 . American Journal of Clinical Nutrition, 86, 55-63. Kim, K.-O. and O’Mahony, M. 1998. A new approach to category scales of intensity I: Traditional versus rank-rating. Journal of Sensory Studies, 13, 241-249. King, B. M. 1986. Odor intensity measured by an audio method. Journal of Food Science, 51, 1340-1344. Koo, T.-Y., Kim, K.-O., and O’Mahony, M. 2002. Effects of forgetting on performance on various intensity scaling protocols: Magnitude estimation and labeled magnitude scale (Green scale). Journal of Sensory Studies, 17, 177-192. Kroll, B. J. 1990. Evaluating rating scales for sensory testing with children. Food Technology, 44(11), 78-80, 82, 84, 86. Kurtz, D. B., White, T. L. and Hayes, M. 2000. The labeled dissimilarity scale: A metric of perceptual dissimilarity. Perception and Psychophysics, 62, 152-161. Land, D. G. and Shepard, R. 1984. Scaling and ranking methods. In: J. R. Piggott (ed.), Sensory Analysis of Foods. Elsevier Applied Science, London, pp. 141-177. Lane, H. L., Catania, A. C. and Stevens, S. S. 1961. Voice level: Autophonic scale, perceived loudness and effect of side tone. Journal of the Acoustical Society of America, 33, 160-167. Larson-Powers, N. and Pangborn, R. M. 1978. Descriptive analysis of the sensory properties of beverages and gelatins containing sucrose or synthetic sweeteners. Journal of Food Science, 43, 47-51. Lawless, H. T. 1977. The pleasantness of mixtures in taste and olfaction. Sensory Processes, 1, 227-237. Lawless, H. T. 1989. Logarithmic transformation of magnitude estimation data and comparisons of scaling methods. Journal of Sensory Studies, 4, 75-86. Lawless, H. T. and Clark, C. C. 1992. Psychological biases in time intensity scaling. Food Technology, 46, 81, 84-86, 90. Lawless, H. T. and Malone, J. G. 1986a. The discriminative efficiency of common scaling methods. Journal of Sensory Studies, 1, 85-96. Lawless, H. T. and Malone, G. J. 1986b. A comparison of scaling methods: Sensitivity, replicates and relative measurement. Journal of Sensory Studies, 1, 155-174. Lawless, H. T. and Skinner, E. Z. 1979. The duration and perceived intensity of sucrose taste. Perception and Psychophysics, 25, 249-258. Lawless, H. T., Popper, R. and Kroll, B. J. 2010a. Comparison of the labeled affective magnitude (LAM) scale, an 11-point category scale and the traditional nine-point hedonic scale. Food Quality and Preference, 21, 4-12. Lawless, H. T., Sinopoli, D. and Chapman, K. W. 2010b. A comparison of the labeled affective magnitude scale and the nine point hedonic scale and examination of categorical behavior. Journal of Sensory Studies, 25, S1, 54-66. Lawless, H. T., Cardello, A. V., Chapman, K. W., Lesher, L. L., Given, Z. and Schutz, H. G. 2010c. A comparison of the effectiveness of hedonic scales and end-anchor compssion effects. Journal of Sensory Studies, 28, S1, 18-34. Lee, H.-J., Kim, K.-O., and O’Mahony, M. 2001. Effects of forgetting on various protocols for category and line scales of intensity. Journal of Sensory Studies, 327-342. Likert, R. 1932. Technique for the measurement of attitudes. Archives of Psychology, 140, 1-55.

7

Scaling

Lindvall, T. and Svensson, L. T. 1974. Equal unpleasantness matching of malodourous substances in the community. Journal of Applied Psychology, 59, 264-269. Mahoney, C. H., Stier, H. L. and Crosby, E. A. 1957. Evaluating flavor differences in canned foods. II. Fundamentals of the simplified procedure. Food Technology 11, Supplemental Symposium Proceedings, 37-42. Marks, L. E. 1978. Binaural summation of the loudness of pure tones. Journal of the Acoustical Society of America, 64, 107-113. Marks, L. E., Borg, G. and Ljunggren, G. 1983. Inpidual differences in perceived exertion assessed by two new methods. Perception and Psychophysic, 34, 280-288. Marks, L. E., Borg, G. and Westerlund, J. 1992. Differences in taste perception assessed by magnitude matching and by category-ratio scaling. Chemical Senses, 17, 493-506. Mattes, R. D. and Lawless, H. T. 1985. An adjustment error in optimization of taste intensity. Appetite, 6, 103-114. McBride, R. L. 1983a. A JND-scale/category scale convergence in taste. Perception and Psychophysics, 34, 77-83. McBride, R. L. 1983b. Taste intensity and the case of exponents greater than 1. Australian Journal of Psychology, 35, 175-184. McBurney, D. H. and Shick, T. R. 1971. Taste and water taste for 26 compounds in man. Perception and Psychophysics, 10, 249-252. McBurney, D. H. and Bartoshuk, L. M. 1973. Interactions between stimuli with different taste qualities. Physiology and Behavior, 10, 1101-1106. McBurney, D. H., Smith, D. V. and Shick, T. R. 1972. Gustatory cross-adaptation: Sourness and bitterness. Perception and Psychophysics, 11, 228-232. Mead, R. and Gay, C. 1995. Sequential design of sensory trials. Food Quality and Preference, 6, 271-280. Mecredy, J. M. Sonnemann, J. C. and Lehmann, S. J. 1974. Sensory profiling of beer by a modified QDA method. Food Technology, 28, 36-41. Meilgaard, M., Civille, G. V. and Carr, B. T. 2006. Sensory Evaluation Techniques, Fourth Edition. CRC, Boca Raton, FL. Moore, L. J. and Shoemaker, C. F. 1981. Sensory textural properties of stabilized ice cream. Journal of Food Science, 46, 399-402. Moskowitz, H. R. 1971. The sweetness and pleasantness of sugars. American Journal of Psychology, 84, 387-405. Moskowitz, H. R. and Sidel, J. L. 1971. Magnitude and hedonic scales of food acceptability. Journal of Food Science, 36, 677-680. Muñoz, A. M. and Civille, G. V. 1998. Universal, product and attribute specific scaling and the development of common lexicons in descriptive analysis. Journal of Sensory Studies, 13, 57-75. Newell, G. J. and MacFarlane, J. D. 1987. Expanded tables for multiple comparison procedures in the analysis of ranked data. Journal of Food Science, 52, 1721-1725. Olabi, A. and Lawless, H. T. 2008. Persistence of context effects with training and reference standards. Journal of Food Science, 73, S185-S189. O’Mahony, M., Park, H., Park, J. Y. and Kim, K.-O. 2004. Comparison of the statistical analysis of hedonic data using

References analysis of variance and multiple comparisons versus and R-index analysis of the ranked data. Journal of Sensory Studies, 19, 519-529. Pangborn, R. M. and Dunkley, W. L. 1964. Laboratory procedures for evaluating the sensory properties of milk. Dairy Science Abstracts, 26-55-62. Parducci, A. 1965. Category judgment: A range-frequency model. Psychological Review, 72, 407-418. Park, J.-Y., Jeon, S.-Y., O’Mahony, M. and Kim, K.-O. 2004. Induction of scaling errors. Journal of Sensory Studies, 19, 261-271. Pearce, J. H., Korth, B. and Warren, C. B. 1986. Evaluation of three scaling methods for hedonics. Journal of Sensory Studies, 1, 27-46. Peryam. D. 1989. Reflections. In: Sensory Evaluation. In Celebration of our Beginnings. American Society for Testing and Materials, Philadelphia, pp. 21-30. Peryam, D. R. and Girardot, N. F. 1952. Advanced taste-test method. Food Engineering, 24, 58-61, 194. Piggot, J. R. and Harper, R. 1975. Ratio scales and category scales for odour intensity. Chemical Senses and Flavour, 1, 307-316. Pokor´ny, J., Davídek, J., Prnka, V. and Davídková, E. 1986. Nonparametric evaluation of graphical sensory profiles for the analysis of carbonated beverages. Die Nahrung, 30, 131-139. Poulton, E. C. 1989. Bias in Quantifying Judgments. Lawrence Erlbaum, Hillsdale, NJ. Richardson, L. F. and Ross, J. S. 1930. Loudness and telephone current. Journal of General Psychology, 3, 288-306. Rosenthal, R. 1987. Judgment Studies: Design, Analysis and Meta-Analysis. University Press, Cambridge. Shand, P. J., Hawrysh, Z. J., Hardin, R. T. and Jeremiah, L. E. 1985. Descriptive sensory analysis of beef steaks by category scaling, line scaling and magnitude estimation. Journal of Food Science, 50, 495-499. Schutz, H. G. and Cardello, A. V. 2001.. A labeled affective magnitude (LAM) scale for assessing food liking/disliking. Journal of Sensory Studies, 16, 117-159. Siegel, S. 1956. Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill, New York. Sriwatanakul, K., Kelvie, W., Lasagna, L., Calimlim, J. F., Wels, O. F. and Mehta, G. 1983. Studies with different types of visual analog scales for measurement of pain. Clinical Pharmacology and Therapeutics, 34, 234-239. Stevens, J. C. and Marks, L. M. 1980. Cross-modality matching functions generated by magnitude estimation. Perception and Psychophysics, 27, 379-389. Stevens, S. S. 1951. Mathematics, measurement and psychophysics. In: S. S. Stevens (ed.), Handbook of Experimental Psychology. Wiley, New York, pp. 1-49.

177 Stevens, S. S. 1956. The direct estimation of sensory magnitudes-loudness. American Journal of Psychology, 69, 1-25. Stevens, S. S. 1957. On the psychophysical law. Psychological Review, 64, 153-181. Stevens, S. S. 1969. On pdicting exponents for cross-modality matches. Perception and Psychophysics, 6, 251-256. Stevens, S. S. and Galanter, E. H. 1957. Ratio scales and category scales for a dozen perceptual continua. Journal of Experimental Psychology, 54, 377-411. Stoer, N. L. and Lawless, H. T. 1993. Comparison of single product scaling and relative-to-reference scaling in sensory evaluation of dairy products. Journal of Sensory Studies, 8, 257-270. Stone, H., Sidel, J., Oliver, S., Woolsey, A. and Singleton, R. C. 1974. Sensory Evaluation by quantitative descriptive analysis. Food Technology, 28, 24-29, 32, 34. Teghtsoonian, M. 1980. Children’s scales of length and loudness: A developmental application of cross-modal matching. Journal of Experimental Child Psychology, 30, 290-307. Thurstone, L. L. 1927. A law of comparative judgment. Psychological Review, 34, 273-286. Townsend, J. T. and Ashby, F. G. 1980. Measurement scales and statistics: The misconception misconceived. Psychological Bulletin, 96, 394-401. Vickers, Z. M. 1983. Magnitude estimation vs. category scaling of the hedonic quality of food sounds. Journal of Food Science, 48, 1183-1186. Villanueva, N. D. M. and Da Silva, M. A. A. P. 2009. Performance of the nine-point hedonic, hybrid and selfadjusting scales in the generation of internal pference maps. Food Quality and Preference, 20, 1-12. Villanueva, N. D. M., Petenate, A. J., and Da Silva, M. A. A. P. 2005. Comparative performance of the hybrid hedonic scale as compared to the traditional hedonic, self-adjusting and ranking scales. Food Quality and Preference, 16, 691-703. Ward, L. M. 1986. Mixed-modality psychophysical scaling: Double cross-modality matching for “difficult” continua. Perception and Psychophysics, 39, 407-417. Weiss, D. J. 1972. Averaging: an empirical validity criterion for magnitude estimation. Perception and Psychophysics, 12, 385-388. Winakor, G., Kim, C. J. and Wolins, L. 1980. Fabric hand: Tactile sensory assessment. Textile Research Journal, 50, 601-610. Yamaguchi, S. 1967. The synergistic effect of monosodium glutamate and disodium 5 inosinate. Journal of Food Science 32, 473-477.

Chapter 8

Time-Intensity Methods

Contents 8.1 8.2 8.3

8.4

8.5

8.6

Introduction . . . . . . . . . . . . . . . . . A Brief History . . . . . . . . . . . . . . . Variations on the Method . . . . . . . . . . 8.3.1 Discrete or Discontinuous Sampling . . . 8.3.2 “Continuous” Tracking . . . . . . . . . 8.3.3 Temporal Dominance Techniques . . . . Recommended Procedures . . . . . . . . . . 8.4.1 Steps in Conducting a Time-intensity Study . . . . . . . . . . . . . . . . . 8.4.2 Procedures . . . . . . . . . . . . . . 8.4.3 Recommended Analysis . . . . . . . . Data Analysis Options . . . . . . . . . . . . 8.5.1 General Approaches . . . . . . . . . . 8.5.2 Methods to Construct or Describe Average Curves . . . . . . . . . . . . . . . . 8.5.3 Case Study: Simple Geometric Description . . . . . . . . . . . . . . 8.5.4 Analysis by Principal Components . . . Examples and Applications . . . . . . . . . . 8.6.1 Taste and Flavor Sensation Tracking . . . 8.6.2 Trigeminal and Chemical/Tactile Sensations . . . . . . . . . . . . . . 8.6.3 Taste and Odor Adaptation . . . . . . .

179 180 182 182 183 184 185 185 186 186 187 187 188 189 192 193 193 194 194

8.6.4 Texture and Phase Change . . 8.6.5 Flavor Release . . . . . . . 8.6.6 Temporal Aspects of Hedonics 8.7 Issues . . . . . . . . . . . . . . . 8.8 Conclusions . . . . . . . . . . . . References . . . . . . . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

195 195 196 197 198 198

8.1 Introduction Perception of aroma, taste, flavor, and texture in foods is a dynamic not a static phenomenon. In other words, the perceived intensity of the sensory attributes change from moment to moment. The dynamic nature of food sensations arises from processes of chewing, breathing, salivation, tongue movements, and swallowing (Dijksterhuis, 1996). In the texture profile method for instance, different phases of food breakdown were recognized early on as evidenced by the separation of characteristics into first bite, mastication, and residual phases (Brandt et al., 1963). Wine tasters often discuss

H.T. Lawless, H. Heymann, Sensory Evaluation of Food, Food Science Text Series, DOI 10.1007/978-1-4419-6488-5_8, © Springer Science+Business Media, LLC 2010

179

180

how a wine “opens in the glass,” recognizing that the flavor will vary as a function of time after opening the bottle and exposing the wine to air. It is widely believed that the consumer acceptability of different intensive sweeteners depends on the similarity of their time profile to that of sucrose. Intensive sweeteners with too long a duration in the mouth may be less pleasant to consumers. Conversely, a chewing gum with longlasting flavor or a wine with a “long finish” may be desirable. These examples demonstrate how the time profile of a food or beverage can be an important aspect of its sensory appeal. The common methods of sensory scaling ask the panelists to rate the perceived intensity of the sensation by giving a single (uni-point) rating. This task requires that the panelists must “time-average” or integrate any changing sensations or to estimate only the peak intensity in order to provide the single intensity value that is required. Such a single value may miss some important information. It is possible, for example, for two products to have the same or similar time-averaged profiles or descriptive specifications, but differ in the order in which different flavors occur or when they reach their peak intensities. Time-intensity (TI) methods provide panelists with the opportunity to scale their perceived sensations over time. When multiple attributes are tracked, the profile of a complex food flavor or texture may show differences between products that change across time after a product is first tasted, smelled, or felt. For most sensations the perceived intensity increases and eventually decreases but for some, like perceived toughness of meat, the sensations may only decrease as a function of time. For perceived melting, the sensation may only increase until a completely melted state is reached. When performing a TI study the sensory specialist can obtain a wealth of detailed information such as the following for each sample: the maximum intensity perceived, the time to maximum intensity, the rate and shape of the increase in intensity to the maximum point, the rate and shape of the decrease in intensity to half-maximal intensity and to the extinction point, and the total duration of the sensation. Some of the common time-intensity parameters are illustrated in Fig. 8.1. The additional information derived from time-intensity methods is especially useful when studying sweeteners or products like chewing gums and hand lotions that have a distinctive time profile.

8

Time-Intensity Methods

INTENSITY Other Parameters: DUR, Total duration = Tend – Tstart AUC, Area under the curve

TIME

Fig. 8.1 Example of a time-intensity curve and common curve parameters extracted from the record.

The remainder of this chapter will be devoted to an overview of the history and applications of this method, as well as recommended procedures and analyses. For the student who wants only the basic information, the following sections are key: variations on the method (Section 8.3), steps and recommended procedures (Section 8.4), data analysis options (Section 8.5), and conclusions (8.8).

8.2 A Brief History Holway and Hurvich (1937) published an early report of tracking taste intensity over time. They had their subjects trace a curve to repsent the sensations from a drop of either 0.5 or 1.0 M NaCl placed on the anterior tongue surface over 10 s. They noted several general effects that were later confirmed as common trends in other studies. The higher concentration led to a higher peak intensity, but the peak occurred later, in spite of a steeper rising slope. Most importantly they noted that taste intensity was not strictly a function of concentration: “while the concentration originally placed on the tongue is ‘fixed,’ the intensity varies in a definite manner from moment to moment. Saline intensity depends on time as well as concentration.” A review of studies of temporal factors in taste is found in Halpern (1991). A review of TI studies of the 1980s and early 1990s can be found in Cliff and Heymann (1993b). Sjostrom (1954) and Jellinek (1964) also made early attempts to quantify the temporal response to perceived sensory intensities. These authors asked their panelists to indicate their perceived bitterness of beer at 1 s intervals on a ballot, using a clock to indicate time.

8.2 A Brief History

They then constructed TI curves by plotting the x-y coordinates (time on the x-axis and perceived intensity on the y-axis) on graph paper. Once panelists had some experience with the method it was possible to ask them simultaneously to rate the perceived intensities of two different attributes at 1 s intervals. Neilson (1957) in an attempt to make the production of the TI curves easier, asked panelists to indicate perceived bitterness directly on graph paper at 2 s timed intervals. The clock could be distracting to the panelists and thus Meiselman (1968), studying taste adaptation and McNulty and Moskowitz (1974), evaluating oilin-water emulsions, improved the TI methodology by eliminating the clock. These authors used audible cues to tell the panelists when to enter perceived intensities on a ballot, placing the timekeeping demands on the experimenter rather than the participant. Larson-Powers and Pangborn (1978), in another attempt to eliminate the distractions of the clock or audible cues, employed a moving strip-chart recorder equipped with a foot pedal to start and stop the movement of the chart. Panelists recorded their responses to the perceived sweetness in beverages and gelatins sweetened with sucrose or synthetic sweeteners, by moving a pen along the cutter bar of the strip-chart recorder. The cutter bar was labeled with an unstructured line scale, from none to extreme. A cardboard cover was placed over the moving chart paper to pvent the panelists from watching the evolving curves and thus pventing them from using any visual cues to bias their responses. A similar setup was independently developed and used at the General Foods Technical Center in 1977 to track sweetness intensity (Lawless and Skinner, 1979). In this apparatus, the actual pen carriage of the chart recorder was grasped by the subject, eliminating the need for them to position a pen; also the moving chart was obscured by a line scale with a pointer attached to the pen carriage. In yet another laboratory at about the same time, Birch and Munton (1981) developed the “SMURF” version of TI scaling (short for “Sensory Measuring Unit for Recording Flux”). In the SMURF apparatus, the subject turned a knob graded from 1 to 10, and this potentiometer fed a variable signal to a strip-chart recorder out of sight of the panelist. The use of strip-chart recorders provided the first continuous TI data-collection methods and freed the panelists from any distractions caused by a clock or auditory signal. However, the methods required a fair degree of mental and physical

181

coordination by the participants. For example, in the Larson-Powers setup, the strip-chart recorder required the panelists to use a foot pedal to run the chart recorder, to place the sample in the mouth, and to move the pen to indicate the perceived intensity. Not all panelists were suitably coordinated and some could not do the evaluation. Although the strip-chart recorder made continuous evaluation of perceived intensities possible, the TI curves had to be digitized manually, which was extremely time consuming. The opportunity to use computers to time sample an analog voltage signal quite naturally led to online data collection to escape the problem of manual measurement of TI curves. To the best of our knowledge, the first computerized system was developed at the US Army Natick Food Laboratories in 1979 to measure bitter taste adaptation. It employed an electric sensor in the spout just above the subject’s tongue in order to determine the actual onset of stimulus arrival. Subthreshold amounts of NaCl were added to the stimulus and thus created a conductance change as the flow changed from the pliminary water rinse to the stimulus interface in the tube. A special circuit was designed to detect the conductance change and to connect the response knob to the visual line scale. Like the SMURF apparatus developed by Birch and Munton, the subject turned a knob controlling a variable resistor. The output of this potentiometer moved a pointer on a line scale for visual feedback while a parallel analog signal was sent to an analog-to-digital converter and then to the computer. The programming was done using FORTRAN subroutines on a popular lab computer of that era. The entire system is shown in Fig. 8.2. The appearance of desktop computers led to an explosion in the use of TI methodology in the 1980s and 1990s. A number of thesis research projects from U.C. Davis served as useful demonstrations of the method (Cliff, 1987; Dacanay, 1990; Rine, 1987) and the method was championed by Pangborn and coworkers (e.g., Lee and Pangborn, 1986). Several scientists (Barylko-Pikielna et al., 1990; Cliff, 1987; Guinard et al., 1985; Janusz et al., 1991; Lawless, 1980; Lee, 1985; Rine, 1987; Yoshida, 1986) developed computerized TI systems using a variety of hardware and software products. Computerized TI systems are now commercially available as part of data-collection software ensembles, greatly enhancing the ease and availability of TI data collection and data processing.

182

8

Time-Intensity Methods

Fig. 8.2 An early computerized system for time-intensity scaling used for tracking bitterness adaptation in a flow system for stimulating the anterior tongue. Heavy lines indicate the flow of stimulus solution, solid lines the flow of information, and dashed lines the experimenter-driven process control. Stimulus arrival at the tongue was monitored by conductivity sensors fitted into the glass tube just above the subject’s tongue. The subjects’

responses change on a line scale while the experimenter could view the conductivity and response voltage outputs on the computer’s display screen, which simultaneously were output to a data file. The system was programmed in FORTRAN subroutines controlling the clock sampling rate and analog-to-digital conversion routine. From Lawless and Clark (1992), reprinted with permission.

However, despite the availability of computerized systems, some research was still conducted using the simple time cueing at discrete intervals (e.g., Lee and Lawless, 1991; Pionnier et al., 2004), and the semi-manual strip-chart recorder method (Ott et al., 1991; Robichaud and Noble, 1990). A discussion of some common applications of TI methods is given in Section 8.6.

oldest approach is simply to ask the panelists to rate the intensity of sensation during different phases of consuming a food. This is particularly applicable to texture which may be evaluated in phases such as initial bite, first chew, mastication, and residual. An example of time pision during the texture evaluation of a baked product is shown in Table 8.1. When using a descriptive panel, it may be useful to have residual flavor or mouthfeel sensations rated at a few small intervals, e.g., every 30 s for 2 min, or immediately after tasting and then again after expectoration. For an example of this approach used with hot pepper “burn” see Stevens and Lawless (1986). Each measurement is then treated like a separate descriptive attribute and analyzed as a separate variable, with little or no attempt to reconstruct a time-connected record like the TI curve shown in Fig. 8.1. For researchers

8.3 Variations on the Method 8.3.1 Discrete or Discontinuous Sampling The sensory scientist has several options for collecting time-dependent sensory data. The methods for timerelated scaling can be pided into four groups. The

8.3 Variations on the Method

183

Table 8.1 Texture attributes at different phases of descriptive analysis Phase Attributes Surface

First bite

First chew Chew down

Residual

Roughness Particles Dryness Fracturability Hardness Particle size Denseness Uniformity of chew Moisture absorption Cohesiveness of mass Toothpacking Grittiness Oiliness Particles Chalky

interested in some simple aspect like “strength of bitter aftertaste” this method may suffice. Another related approach is to ask for repeated ratings of a single or just a few attributes at repeated smaller time intervals, usually cued by the panel leader or experimenter. These ratings are then connected and graphed on time axis. This is a simple procedure that can be used to track changes in the intensity of a flavor or texture attribute and requires no special equipment other than a stopwatch or other timing device. The panel must be trained to rate their sensations upon the time cue and to move rapidly through the list of attributes. The cue may be given verbally or on a computer screen. It is not known how many attributes can be rated in this way, but with faster time cueing and shorter intervals, obviously fewer attributes may be rated. This method also requires some faith in the assumption that the attributes are being rated close to the actual time when the cue is given. The accuracy with which panelists can do this is unknown, but given that there is a reaction time delay in any perceptual judgment, there must be some inherent error or delayrelated variance built into the procedure. An example of the repeated, discrete time interval method with verbal cueing and multiple attributes can be found in studies of sweetener mixtures (e.g., Ayya and Lawless, 1992) and astringency (Lee and Lawless, 1991). The time record is treated as a connected series and time is analyzed as one factor (i.e., one independent variable) in the statistical analysis.

Word anchors Smooth-rough None-many Oily-dry Crumbly-brittle Soft-hard Small-large Airy-dense Even-uneven None-much Loose-cohesive None-much None-much Dry-oily None-much Not chalky-very chalky

184

trained to move a mouse diagonally and a visual scale included pointers on both horizontal and vertical scales to repsent the intensity of the inpidual attributes. With a slowly changing product like chewing gum, with a sampling time that is not too frequent (every 9-15 s in this case), the technique would seem to be within the capabilities of human observers to either rapidly shift their attention or to respond to the overall pattern of the combined flavors. However, as currently used, most TI tracking methods must repeat the evaluation in order to track additional attributes. Ideally, this could lead to a composite profile of all the dynamic flavor and texture attributes in a product and how they changed across time. Such an approach was proposed by DeRovira (1996), who showed how the descriptive analysis spider-web plot of multiple attributes could be extended into the time dimension to produce a set of TI curves and thus to characterize an entire profile.

8

Time-Intensity Methods

8.4 Recommended Procedures

attributes are rated in each trial, (2) that it is simple and easy to do, requiring little or no training, and (3) that it provides a picture of enhanced differences relative to the TI records. Because panelists are forced to respond to only one attribute at a time, differences in the temporal profile may be accentuated. However, at the time of this publication, no standard procedure seemed to be agreed upon. The technique requires specialized software to collect the information, but at least one of the major sensory software systems has implemented a TDS option. Attributes are assumed to have a score of zero before they become dominant, and some attributes may never be rated. This would seem to necessarily lead to an incomplete record. Different panelists are contributing at different times to different attributes, so statistical methods for comparison of differences between products using the raw data set are difficult. However, products can be statistically compared using the simplified summary scores (summed intensity by duration measures, but losing time information) or by comparison of the proportion of responders (losing intensity information) as in Labbe et al. (2009). Qualitative comparisons can be made from inspecting the curves, such as “this product is initially sweet, then becoming more astringent, compared to product X which is initially sour, then fruity.” Is information provided by TDS and by traditional TI tracking methods different? One study found a strong resemblance of the constructed time curves for both methods, providing virtually redundant information for some attributes (Le Reverend et al., 2008). In another study, correlations with TI parameters were high for intensity maxima versus dominance proportion maxima, as might be expected since a higher number of persons finding an attribute dominant should be related to the mean intensity in TI. Correlations with other time-dependent parameters such as time to Imax (Tmax ) and duration measures were low, due to the different information collected in TDS and the limited attention to one attribute at a time (Pineau et al., 2009).

8.4 Recommended Procedures 8.4.1 Steps in Conducting a Time-intensity Study The steps in conducting a time-intensity study are similar to those in setting up a descriptive analysis procedure. They are listed in Table 8.2. The first

185 Table 8.2 Steps in conducting a time-intensity study 1. Determine project objectives: Is TI the right method? 2. Determine critical attributes to be rated. 3. Establish products to be used with clients/researchers. 4. Choose system and/or TI method for data collection. a. What is the response task? b. What visual feedback is provided to the panelists? 5. Establish statistical analysis and experimental design. a. What parameters are to be compared? b. Are multivariate comparisons needed? 6. Recruit panelists. 7. Conduct training sessions. 8. Check panelist performance. 9. Conduct study. 10. Analyze data and report.

important question is to establish whether TI methods are appropriate to the experimental research objective. Is this a product with just one or a few critical attributes that are likely to vary in some important way in their time course? Is this difference likely to impact consumer acceptance of the product? What are these critical attributes? Next, the product test set should be determined with the research team, and this will influence the experimental design. The TI method itself must be chosen, e.g., discrete point, continuous tracking, or TDS. The sensory specialist should by this time have an idea of what the data set will look like and what parameters can be extracted from the TI records for statistical comparisons such as intensity maxima, time to maxima, areas under the curve, and total duration. Many of the TI curve parameters are often correlated, so there is little need to analyze more than about ten parameters. Practice is almost always essential. You cannot assume that a person sitting down in a test booth will know what to do with the TI system and feel comfortable with the mouse or other response device. A protocol for training TI panelists was outlined by Peyvieux and Dijksterhuis (2001) and this protocol or similar versions have been widely adopted. It is also wise that some kind of panel checking be done to make sure the panelists are giving reliable data (see Bloom et al., 1995; Peyvieux and Dijksterhuis, 2001) and to examine the reasonableness of their data records. At this time the researchers and statistical staff should also decide how to handle missing data or records that may have artifacts or be incomplete. As in any sensory study, extensive planning may save a lot of headaches and problems, and this is especially true for TI methods.

186

8.4.2 Procedures If only a few attributes are going to be evaluated, the continuous tracking methods are appropriate and will provide a lot of information. This usually requires the use of computer-assisted data collection. Many, if not all, of the commercial software packages for sensory evaluation data collection have TI modules. The start and stop commands, sampling rate, and inter-trial intervals can usually be specified. The mouse movement will generally produce some visual feedback such as the motion of a cursor or line indicator along a simple line scale. The display often looks like a vertical or horizontal thermometer with the cursor position clearly indicated by bar or line that rises and falls. The computer record can be treated as raw data for averaging across panelists. However, statistical analysis by simple averaging raises a number of issues (discussed in Section 8.5). A simple approach is to pull characteristic curve parameters off each record for purposes of statistical comparisons such as intensity maximum (Imax ), time to maximum (Tmax ), and area under the curve (AUC). These are sometimes referred to as “scaffolding parameters” as they repsent the fundamental structure of the time records. Statistical comparison of these parameters can provide a clear understanding of how different products are perceived with regard to the onset of sensations, time course of rising and falling sensations, total duration, and total sensory impact of that flavor or textural aspect of the products. If a computer-assisted software package is not available or cannot be programmed, the research can always choose to use the cued/discontinuous method (e.g., with a stopwatch and verbal commands). This may be suitable for products in which multiple attributes must be rated in order to get the full picture. However, given the widespad availability of commercial sensory data-collection systems in major food and consumer product companies, it is likely that a sensory professional will have access to a continuous tracking option. The starting position of the cursor on the visible scale or computer screen should be considered carefully. For most intensity ratings it makes sense to start at the lower end, but for hedonics (like/dislike) the cursor should begin at the neutral point. For meat tenderness or product melting, the track is usually unidirectional, so the cursor should start at “not tender” or

8

Time-Intensity Methods

“tough” for meat and “not melted” for a product that melts. If the cursor is started at the wrong end of a unidirectional tracking situation, a falsely bi-directional record may be obtained due to the initial movement. An outline of panel training for TI studies was illustrated in a case study by Peyvieux and Dijksterhuis (2001) and the sensory specialist should consider using this or a similar method to insure good performance of the panelists. These authors had panelists evaluate flavor and texture components of a complex meat product. The prospective panelists were first introduced to the TI method, and then given practice with basic taste solutions over several sessions. The basic tastes were considered simpler than the complex product and more suitable for initial practice. Panelist consistency was checked and a panelist was considered reliable if they could produce two out of three TI records on the same taste stimulus that did not differ more than 40% of the time. A vertical line scale was used and an important specification was when to move the cursor back to zero (when there was no flavor or when the sample was swallowed for the texture attribute of juiciness). Problems were noted with (1) nontraditional curve shapes such as having no return to zero, (2) poor replication by some panelists, (3) unusable curves due to lack of a landmark such as no Imax . The authors also conducted a traditional profiling (i.e., descriptive analysis) study before the TI evaluations to make sure the attributes were correctly chosen and understood by the panelists. If TI panelists are already chosen from an existing descriptive panel, this step may not be needed. The authors conducted several statistical analyses to check for consistent use of the attributes, to look for oddities in curve shapes by some panelists, and to examine inpidual replicates. Improvements in consistency and evidence of learning and practice were noted.

8.4.3 Recommended Analysis For purposes of comparing products, the simplest approach is to extract the curve parameters such as Imax , Tmax , AUC, and total duration from each record. Some sensory software systems will generate these measures automatically. Then these curve parameters can be treated like any data points in any sensory evaluation and compared statistically. For three or

8.5 Data Analysis Options

more products, analysis would be by ANOVA and then planned comparisons of means (see Appendix C). Means and significance of differences can be reported in graphs or tables for each curve attribute and product. If a time by intensity curve is desired, the curves can be averaged in the time direction by choosing points at specific time intervals. This averaging method is not without its pitfalls; however, a number of alternative methods are given in the next section. An example of how to produce a simplified averaged curve is given below in the case study of the trapezoidal method (Lallemand et al., 1999).

8.5 Data Analysis Options 8.5.1 General Approaches Two common statistical approaches have been taken to perform hypothesis testing on TI data. Perhaps the most obvious test is simply to treat the raw data at whatever time intervals were sampled as the input data to analyses of variance (ANOVA) (e.g., Lee and Lawless, 1991). This approach results in a very large ANOVA with at least three factors-time, panelists, and the treatments of interest. Time and panelists effects may not be of primary interest but will always show large F-ratios due to the fact that people differ and the sensations change over time. This is not news. Another common pattern is a time-by-treatment interaction since all curves will tend to converge near baseline at later time intervals. This is also to be

187

Table 8.3 Parameters extracted from time-intensity curves Parameter Other names Definition Peak intensity Imax , Ipeak Height of highest point on TI record Total duration DUR, Dtotal Time from onset to return to baseline Self-explanatory Area under the curve AUC, Atotal Plateau Dpeak Time difference between reaching maximum and beginning descent Area under plateau Apeak Self-explanatory Area bounded by onset of decline and reaching baseline Area under descending phase Ptotal Rising slope Ri Rate of increase (linear fit) or slope of line from onset to peak intensity Rate of decrease (linear fit) or slope of line from initial declining point to baseline. Declining slope Rf Extinction Time at which curve terminates at baseline Time to reach peak intensity Time to peak Tmax , Tpeak Time to half peak Half-life Time to reach half maximum in decay portion Modified from Lundahl (1992) Other shape parameters are given in Lundahl (1992), based on a half circle of equivalent area under the curve and piding the half circle into rising and falling phase segments

188

8

2006), an effect that is sometimes described as an inpidual “signature.” Examples of inpidual signatures are shown in Fig. 8.3. The causes of these inpidual patterns are unknown but could be attributed to differences in anatomy, differences in physiology such as salivary factors (Fischer et al., 1994), different types of oral manipulation or chewing efficiency (Brown et al., 1994; Zimoch and Gullet, 1997), and inpidual habits of scaling. Some of this information may be lost when analyzing only the extracted parameters.

INTENSITY

Judge 1 Judge 2 Judge 3

0

20

40

60 80 TIME (sec)

100

120

140

Fig. 8.3 Examples of time-intensity records showing characteristic signatures or shapes. Judge 1 shows a record with multiple plateaus, a common occurrence. Judge 2 shows a smooth and continuous curve. Judge 3 shows a steep rise and fall.

A third approach is to fit some mathematical model or set of equations to each inpidual record and then use the constants from the model as the data for comparisons of different products (Eilers and Dijksterhuis, 2004; Garrido et al., 2001; Ledauphin et al., 2005, 2006; Wendin et al., 2003). Given the increasing activity and ingenuity in the area of sensometrics, it is likely that such models will continue to be developed. The sensory scientist needs to ask how useful they are in the product testing arena and whether the model fitting is useful in differentiating products. Various approaches to modeling and mathematical description of TI curves are discussed in the next section.

8.5.2 Methods to Construct or Describe Average Curves The analysis of TI records has produced a sustained interest and response from sensometricians, who have

Time-Intensity Methods

8.5 Data Analysis Options

189

INTENSITY

JUDGE 1 JUDGE 2 AVERAGE

TIME

Fig. 8.4 Two curves with different peak times, if averaged, can lead to a double-peaked curve that resembles neither original data record.

style” either by simple visual inspection of the curves or by a clustering analysis or other statistical methods (van Buuren, 1992; Zimoch and Gullet, 1997). Then these subgroups can be analyzed separately. The analysis may proceed using the simple fixed-time averaging of curve heights or one of the other methods described next. An alternative approach is to average in both the intensity and time directions, by setting each inpidual’s maximum of the mean time to maximum across all curves, and then finding mean times to fixed percentages of maximum in the rising and falling phases of each curve. This procedure was originally published by Overbosch et al. (1986) and subsequent modifications were proposed by Liu and MacFie (1990). The steps in the procedure are shown in Fig. 8.5. Briefly, the method proceeds as follows: In the first step, the geometric mean value for the intensity maximum is found. Inpidual curves are multiplicatively scaled to have this Imax value. In the second step, the geometric mean time to Imax is calculated. In the next steps, geometric mean times are calculated for fixed percentage “slices” of each curve, i.e., at fixed percentages of Imax . For example, the rising and falling phases are “sliced” at 95% of Imax and 90% of Imax and the geometric mean time values to reach these heights are found. This procedure avoids the kind of double-peaked curve that can arise from simple averaging of two distinctly different curve shapes as shown in Fig. 8.4. The method results in several desirable properties that do not necessarily occur with simple averaging at fixed times. First, the Imax value from the mean curve is the geometric mean of the Imax of the inpidual curves. Second, the Tmax value from the mean curve is the geometric mean of the Tmax of the inpidual

curves. Third, the endpoint is the geometric mean of all endpoint times. Fourth, all judges contribute to all segments of the curve. With simple averaging at fixed times, the tail of the curve may have many judges returned to zero and thus the mean is some small value that is a poor repsentation of the data at those points. In statistical terms, the distribution of responses at these later time intervals is positively skewed and left-censored (bound by zero). One approach to this problem is to use the simple median as the measure of central tendency (e.g., Lawless and Skinner, 1979). In this case the summary curve goes to zero when over half the judges go to zero. A second approach is to use statistical techniques designed for estimating measures of central tendency and standard deviations from leftcensored positively skewed data (Owen and DeRouen, 1980). Overbosch’s method works well if all inpidual curves are smoothly rising and falling with no plateaus or multiple peaks and valleys, and all data begin and end at zero. In practice, the data are not so regular. Some judges may begin to fall after the first maximum and then rise again to a second peak. Due to various common errors, the data may not start and end at zero, for example, the record may be truncated within the allowable time of sampling. To accommodate these problems, Liu and MacFie (1990) developed a modification of the above procedure. In their procedure, Imax and four “time landmarks” were averaged, namely starting time, time to maximum, time at which the curve starts to descend from Imax and ending time. The ascending and descending phases of each curve were then pided into about 20 time interval slices. At each time interval, the mean I value is calculated. This method then allows for curves with multiple rising and falling phases and a plateau of maximum intensity that is commonly seen in some judges’ records.

8.5.3 Case Study: Simple Geometric Description A simple and elegant method for comparing curves and extracting parameters by a geometric approximation was described by Lallemand et al. (1999). The authors used the method with a trained texture panel to evaluate different ice cream formulations. The labor-intensive nature of TI studies was illustrated in the fact that

190

8

JUDGE 1 JUDGE 2

Time-Intensity Methods

JUDGE 1 JUDGE 2

A Imax 1

B Geo. mean Imax

Imax 2

JUDGE 1 JUDGE 2

Tmax1

C

JUDGE 1 JUDGE 2

D

Geo. mean Tmax

Tmax2

JUDGE 1 JUDGE 2

Geo. mean 95% Tmax

T1 @ 95% Imax

E

F Geo. mean T @95% Imax

Geo. mean Tmax,Imax

T2 @ 95% Imax

Fig. 8.5 Steps in the data analysis procedure recommended by Overbosch et al. (1986). (a) Two hypothetical time-intensity records from two panelists showing different intensity maxima at different times. (b) The geometric mean value for the intensity maximum is found. Inpidual curves may then by multiplicatively scaled to have this Imax value. (c) The two Tmax values. (d) The geometric mean time to maximum (Tmax ) is calculated.

(e) Geometric mean times are calculated for fixed percentage “slices” of each curve, i.e., at fixed percentages of Imax . The rising phase is “sliced” at 95% of Imax and the time values determined. A similar value will be determined at 95% of maximum for the falling phase. (f) The geometric mean times at each percent of maximum are plotted to generate the composite curve.

12 products were evaluated on 8 different attributes in triplicate sessions, requiring about 300 TI curves from each panelist! Texture panelists were given over 20 sessions of training, although only a few of the final sessions were specifically devoted to practice

with the TI procedure. Obviously, this kind of extensive research program requires a significant time and resource commitment. Data were collected using a computer-assisted rating program, where mouse movement was linked to

8.5 Data Analysis Options

191

the position of a cursor on a 10 cm 10-point scale. The authors noted a number of issues with the data records due to “mouse artifacts” or other problems. These included sudden unintended movement of the mouse leading to false peaks or ratings after the end of the sensation, mouse blockage leading to unusable records, and occasional inaccurate positioning by panelists causing data that did not reflect their actual perceptions. Such ergonomic difficulties are not uncommon in TI studies although they are rarely reported or discussed. Even with this highly trained panel, from 1 to 3% of the records needed to be discarded or manually corrected due to artifacts or inaccuracies. Sensory professionals should not assume that just because they have a computer-assisted TI system, the human factors in mouse and machine interactions will always work smoothly and as planned. Examples of response artifacts are shown in Fig. 8.6. T1 T2

T3

False Plateau

INTENSITY

“Mouse artifacts”

TIME (sec)

Fig. 8.6 Response artifacts in TI records. The solid line shows some perhaps unintended mouse movement (muscle spasm?) near the peak intensity. The dashed line shows a bump in the mouse after sensation ceased and returned to zero. The dotted line illustrates an issue in determining at what point the intensity plateau has ended. The short segment between T1 and T2 may have simply been an adjustment of the mouse after the sudden rise, when the panelist felt they overshot the mark. The actual end of the plateau might more reasonably be considered to occur at T3. (see Lallemand et al. 1999).

Lallemand and coworkers noted that TI curves often took a shape in which the sensation rose to a plateau near peak intensity for a period during which intensity ratings changed very little and then fell to the baseline. They reasoned that a simple geometric approximation by a trapezoid shape might suffice for extracting curve parameters and finding the area under the curve (not unlike the trapezoidal approximation method used for integration in calculus). In principle, four points could

be defined that would describe the curve: the onset time, the time at the intensity maximum or beginning of the plateau, the time at which the plateau ended and the decreasing phase began, and the time at which sensation stopped. These landmarks are those originally proposed by Lui and MacFie (1990). In practice, these points turned out to be more difficult to estimate than expected, so some compromises were made. For example, some records would show a gradually decreasing record during the “plateau” and before the segment with a more rapidly falling slope was evident. How much of a decrease would justify the falling phase or conversely, how little of a decrease would be considered still part of the plateau (see Fig. 8.6)? Also, what should one do if the panelist did not return to zero sensation or unintentionally bumped the mouse after reaching zero? In order to solve these issues, the four points were chosen at somewhat interior sections of the curve, namely the times at 5% of the intensity maximum for the onset and endpoint of the trapezoid, and the times at 90% of the intensity maximum for the beginning and end of the plateau. This approximation worked reasonably well, and its application to a hypothetical record is shown in Fig. 8.7. Given the almost 3,000 TI curves in this one study, the trapezoid points were not mapped by hand or by eye, but a special program written to extract the points. However, for smaller experiments it should be quite feasible to do this kind of analysis “by hand” on any collection of graphed records. The establishment of the four trapezoid vertices now allows extraction of the six basic TI curve parameters for statistical analysis (5 and 90% of maximum intensity points, the four times at those points), as well as the intensity maximum from the original record, and derived (secondary) parameters such as rising and falling slopes and the total area under the curve. Note that the total area becomes simply the sum of the two triangles and the rectangle described by the plateau. These are shown in the lower section of Fig. 8.7. A composite trapezoid can be drawn from the averaged points. The utility and validity of the method was illustrated in one sample composite record, showing the fruity flavor intensity from two ice creams differing in fat content. Consistent with what might be expected from the principles of flavor release, the higher fat sample had a slower and more delayed rise to the peak (plateau) but a longer duration. This would be pdicted if the higher fat level was better able to sequester

192

8

Time-Intensity Methods

with their single-point intensity estimates. Consistent with the latter notion, the profiling scores could be better modeled by a combination of several of the TI parameters. The simplicity and validity of this analysis method suggests that it should find wider application in industrial settings.

Intensity trapezoidal approxiamtion to T-I curve

A,D = points at 5% of maximum B,C = points at 90% of maximum

8.5.4 Analysis by Principal Components Time DERIVED PARAMETERS;

B

C

Rd Ri Ai

Am

Ad

A

D Time Di

Dm

Dd

Fig. 8.7 The trapezoidal method of Lallemand et al. (1999) for assessing curve parameters on TI records. The upper panel shows the basic scheme in which four points are found when the initial 5% of the intensity maximum (Imax ) occurs, when 90% of Imax is first reached on the ascending segment, when the plateau is finished at 90% of Imax on the descending phase and the endpoint approximation at 5% of Imax on the descending phase. The lower panel shows the derived parameters, namely Ri , Ai , and Di for the rate (slope), area, and duration of the initial rising phase; Am and Dm for the area and duration of the middle plateau section; and Rd , Ad , and Dd for the rate (slope), area, and duration of the falling phase. A total duration can be found from the sum of Di , Dm , and Dd . The total area is given by the sum of the A parameters or by the formula for the area of a trapezoid: Total area = (I90 -I5 ) (2Dm + Di + Dd )/2. (Height times the sum of the two parallel segments, then pided by 2).

a lipophilic or nonpolar flavor compound and thus delay the flavor release. They also examined the correlation with a traditional texture descriptive analysis and found very low correlations of inpidual TI parameters with texture profiling mean scores. This would be expected if the TI parameters were contributing unique information or if the texture profilers were integrating a number of time-dependent events in coming up

Another analysis uses principal components analysis (PCA, discussed in Chapter 18) (van Buuren, 1992). Briefly, PCA is a statistical method that “bundles” groups of correlated measurements and substitutes a new variable (a factor or principal component) in place of the original variables, thus simplifying the picture. In studying the time-intensity curves for bitterness or different brands of lager beer, van Buuren noticed that inpiduals once again produced their own characteristic “style” of curve shape. Most people showed a classic TI curve shape, but some subjects were classified as “slow starters,” with a delayed peak and some showed a tendency to persist and not come back down to baseline within the test period. Submission of the data to PCA allowed the extraction of a “principal curve” which captured the majority trend. This showed a peaked TI curve and a gradual return to baseline. The second principal curve captured the shape of the minority trends, with slow onset, a broad peak and slow decline without reaching baseline. The principal curves were thus able to extract judge trends and provide a cleaned-up view of the primary shape of the combined data (Zimoch and Gullet, 1997). Although a PCA program may extract a number of principal components, not all may be practically meaningful (for an example, see Reinbach et al., 2009), and the user should examine each one for the story it tells. Reasonable questions are whether the component reflects something important relative to the simple TI curve parameters, and whether it shows any patterns related to inpidual differences among panelists. Dijksterhuis explored the PCA approach in greater detail (Dijksterhuis, 1993; Dijksterhuis and van den Broek, 1995; Dijksterhuis et al., 1994). Dijksterhuis (1993) noted that the PCA method as applied by van Buuren was not discriminating of different bitter stimuli. An alternative approach was “non-centered PCA” in which curve height information was retained during

8.6 Examples and Applications

data processing, rather than normalizing curves to a common scale. The non-centered approach works on the raw data matrix. Stimuli or treatments were better distinguished. The first principal curve tends to look like the simple average, while the second principal curve contains rate information such as accelerations or inflection points (Dijksterhuis et al., 1994). This could be potentially useful information in differentiating the subtle patterns of TI curves for different flavors. The PCA approach also involves the possibility of generating weights for different assessors that indicate the degree to which they contribute to the different principal curves. This could be an important tool for differentiating outliers in the data or panelists with highly unusual TI signatures (Peyvieux and Dijksterhuis, 2001).

8.6 Examples and Applications A growing number of studies have used TI methods for the evaluations of flavor, texture, flavor release, hedonics, and basic studies of the chemical senses. A short review of these studies follows in this section, although the reader is cautioned that the list is not exhaustive. We have cited a few of the older studies to give credit to the pioneers of this field as well as some of the newer applications. The examples are meant to show the range of sensory studies for which TI methods are suitable.

8.6.1 Taste and Flavor Sensation Tracking A common application of continuous time-intensity scaling is tracking the sensation rise and decay from important flavor ingredients, such as sweeteners (Swartz, 1980). An early study of Jellinek and Pangborn reported that addition of salt to sucrose extended the time-intensity curve and made the taste “more rounded” in their words (Jellinek, 1964). One of the salient characteristics of many intensive or non-carbohydrate sweeteners is their lingering taste that is different from that of sucrose. Time-intensity tracking of sweet tastes was an active area of study (Dubois and Lee, 1983; Larson-Powers and Pangborn, 1978; Lawless and Skinner, 1979; Yoshida, 1986) and

193

194

8

Cliff and Noble, 1990; Matysiak and Noble, 1991). Using these techniques, enhancement of sweetness by fruity volatiles has been observed, and differences in the interactions were seen for different sweetening agents.

8.6.2 Trigeminal and Chemical/Tactile Sensations Reactions to other chemical stimuli affecting irritation or tactile effects in the mouth have been a fertile area for time-related sensory measurements. Compounds such as menthol produce extended flavor sensations and the time course is concentration dependent (Dacanay, 1990; Gwartney and Heymann, 1995). A large number of studies of the burning sensation from hot pepper compounds have used time-intensity scaling, using repeated ratings at discrete time intervals as well as continuous tracking (Cliff and Heymann, 1993a; Green, 1989; Green and Lawless, 1991; Lawless and Stevens, 1988; Stevens and Lawless, 1986). Given the slow onset and extended time course of the sensations induced by even a single sample of a food containing hot pepper, this is a highly appropriate application for time-related judgments. The repeated ingestion paradigm has also been used to study the short- and long-term desensitization to irritants such as capsaicin and zingerone (Prescott and Stevenson, 1996). The temporal profile of different irritative spice compounds is an important point of qualitative differentiation (Cliff and Heymann, 1992). Reinbach and colleagues used TI tracking to study the oral heat from capsaicin in various meat products (Reinbach et al., 2007, 2009) as well as the interactions of oral burn with temperature (see also Baron and Penfield, 1996). When examining the decay curves following different pepper compounds, different time courses can be fit by different decay constants in a simple exponential curve of the form R = Ro e−kt

Time-Intensity Methods

from peak during the decay portion of the time curve (Lawless, 1984). Another chemically induced tactile or feeling factor in the mouth that has been studied by TI methods is astringency. Continuous tracking has been applied to the astringency sensation over repeated ingestions (Guinard et al., 1986). Continuous tracking can provide a clear record of how sensations change during multiple ingestions, and how flavors may build as subsequent sips or tastings add greater sensations on an existing background. An example of this is shown in Fig. 8.8 where the saw-tooth curve shows an increase and builds astringency over repeated ingestions (Guinard et al., 1986). Astringency has also been studied using repeated discrete-point scaling. For example, Lawless and coworkers were able to show differences in the time profiles of some sensory sub-qualities related to astringency, namely dryness, roughness in the mouth, and puckery tightening sensations (Lawless et al., 1994; Lee and Lawless, 1991) depending on the astringent materials being evaluated.

Fig. 8.8 Continuous tracking record with multiple ingestions producing a “sawtooth” curve record. The abscissa shows the time axis in seconds and the ordinate the mean astringent intensity from 24 judges. The dashed curve is from a 15 ml sample of a base wine with 500 gm/l added tannin and the solid curve from the base wine with no added tannin. Sample intake and expectoration are indicated by stars and arrows, respectively. From Guinard et al. (1986), reprinted with permission of the American Society of Enology and Viticulture.

(8.1)

or

8.6.3 Taste and Odor Adaptation ln R = ln Ro − kt

(8.2)

where Ro is the peak intensity and t is time, k is the value determining how rapid the sensation falls off

The measurement of a flavor sensation by tracking over time has a close parallel in studies of taste and odor adaptation (Cain, 1974; Lawless and Skinner,

8.6 Examples and Applications

1979; O’Mahony and Wong, 1989). Adaptation may be defined as the decrease in responsiveness of a sensory system to conditions of constant stimulation providing a changing “zero point” (O’Mahony, 1986). Adaptation studies have sometimes used discrete, single bursts of stimulation (e.g., Meiselman and Halpern, 1973). An early study of adaptation used a flowing system through the subject’s entire mouth using pipes inserted through a dental impssion material that was held in the teeth (Abrahams et al., 1937). Disappearance of salt taste was achieved in under 30 s for a 5% salt solution, although higher concentrations induced painful sensations over time that were confusing to the subjects and sometimes masked the taste sensation. By flowing a continuous stream over a section of the subject’s tongue or stabilizing the stimulus with wet filter paper, taste often disappears in under a few minutes (Gent, 1979; Gent and McBurney, 1978; Kroeze, 1979; McBurney, 1966). Concentrations above the adapting level are perceived as having the characteristic taste of that substance, e.g., salty for NaCl. Concentrations below that level, to which pure water is the limiting case, take on other tastes so that water after NaCl, for example, is sour-bitter (McBurney, and Shick, 1971). Under other conditions adaptation may be incomplete (Dubose et al., 1977; Lawless and Skinner, 1979; Meiselman and Dubose, 1976). Pulsed flow or intermittent stimulation causes adaptation to be much less complete or absent entirely (Meiselman and Halpern, 1973).

8.6.4 Texture and Phase Change Tactile features of foods and consumer products have been evaluated using time-related measurements. Phase change is an important feature of many products that undergo melting when eaten. These include both frozen products like ice cream and other dairy desserts and fatty products with melting points near body temperature such as chocolate. Using the chartrecording method for time-intensity tracking, Moore and Shoemaker (1981) evaluated the degree of coldness, iciness, and sensory viscosity of ice cream with different degrees of added carboxymethyl cellulose. The added carbohydrate shifted the time for peak intensity of iciness and extended the sensations of coldness on the tongue. Moore and Shoemaker also

195

8.6.5 Flavor Release A potentially fertile area for the application of timerelated sensory measurements is in flavor release from foods during eating. Not only are texture changes

196

8

Time-Intensity Methods

cheese systems (Pionnier et al., 2004) using a discrete point TI method.

8.6.6 Temporal Aspects of Hedonics Since the pleasantness or appeal of a sensory characteristic is largely dependent on its intensity level, it is not surprising that one’s hedonic reaction to a product might shift over time as the strength of a flavor waxes and wanes. Time-related shifts in food likes and dislikes are well known. In the phenomenon known as alliesthesia, our liking for a food depends a lot on whether we are hungry or replete (Cabanac, 1971). The delightful lobster dinner we enjoyed last night may not look quite so appealing as leftovers at lunchtime the next day. Wine tasters may speak of a wine that is “closed in the glass, open on the palate, and having a long finish.” Accompanying such a description is the implicit message that this particular wine got better over the course of the sensory experience. Given the shorter time span of flavor and texture sensations in the mouth, we can ask whether there are shifts in liking and disliking. This has been measured in several studies. Taylor and Pangborn (1990) examined liking for chocolate milk with varying degrees of milk fat. Different inpidual trends were observed in liking for different concentrations, and this affected the degree of liking expssed over time. Another example of hedonic TI scaling was in a study of the liking/disliking for the burning oral sensation from chili (hot ) pepper (Rozin et al., 1982). They found different patterns of temporal shifting in liking as the burn rose and declined. Some subjects liked the burn at all time intervals, some disliked the burn at all time intervals, and some shifted across neutrality as strong burns became more tolerable. This method was revisited by Veldhuizen et al. (2006) who used a simple bipolar line scale for pleasantness and had subjects evaluated both intensity and hedonic reactions to a citrus beverage flowed over the tongue from a computer-controlled delivery system (see Fig. 8.2 for an early example of this kind of device). Note that with a bipolar hedonic scale, the mouse and cursor positions must begin at the center of the scale and not the lower end as with intensity scaling. The authors found a delayed pleasantness response compared to the intensity tracking, a similar time to maximum, but an unexpectedly quicker offset of response for pleasantness tracking. Some panelists

8.7 Issues

produced a double-peaked pleasantness response, as the sensation could rise in pleasantness, but then become too intense, but become more pleasant again as adaptation set in and the perceived strength decreased.

8.7 Issues Sensory scientists who wish to use time-intensity methods for any particular study need to weigh the potential for obtaining actionable information against the cost and time involved in collecting these data. Some orientation or training is required (Peyvieux and Dijksterhuis, 2001) and in some published studies, the training and practice is quite extensive. For example, Zimoch and Gullet (1997) trained their meat texture panel for 12 h. Panelists must be trained to use the response device and sufficiently practiced to feel comfortable with the requirements of the task in terms of maintaining focused attention to momentary sensation changes. With the use of online data collection the tabulation and processing of information is generally not very labor intensive; but without computer-assisted collection the time involved can be enormous. Even with computerized systems, the data collection is not foolproof. Responses may be truncated or fail to start at zero in some records (Liu and MacFie, 1990; McGowan and Lee, 2006) making automatic averaging of records infeasible. In one study of melting behavior (Lawless et al., 1996), some subjects mistakenly returned the indicating cursor to zero as soon as the product was completely melted, instead of leaving the cursor on maximum, producing truncated records. Such unexpected events remind us to never assume that panelists are doing what you think they should be doing. A fundamental issue is information gain. In the case where changes in duration are observed at equal maximum intensities, it can be argued that the traditional scaling might have missed important sensory differences. For example, TI can capture information such as when the TI curves cross over, e.g., the interesting case when a product with a lower peak intensity has a longer duration (e.g., Lallemand et al., 1999; Lawless et al., 1996). However, this pattern is not often seen. Usually products with stronger peak height have longer durations. In general there is a lot of redundant information in TI parameters. Lundahl (1992) studied the correlation of 15 TI parameters associated with TI

197

curves’ shapes, sizes, and rates of change. Curve size parameters were highly correlated and usually loaded on the first principal component of a PCA, capturing most of the variance (see also Cliff and Noble, 1990). Curve size parameters, including peak height, were correlated with simple category ratings of the same beverages. So an open question for the sensory scientist is whether there is any unique information in the TI parameters extracted from the records and whether there is information gain over what would be provided by more simple direct scaling using a single intensity rating. A potential problem in time-intensity methods is that factors affecting response behavior are not well understood. In TI measurements, there are dynamic physical processes (chewing, salivary dilution) leading to changes in the stimulus and resulting sensations (Fischer et al., 1994). A second group of processes concerns how the participant translates the conscious experience into an overt response, including a decision mechanism and motoric activation (Dijksterhuis, 1996). The notion that TI methods provide a direct link from the tongue of the subject to the hand moving the mouse is a fantasy. Even in continuous tracking, there must be some decision process involved. There is no information as to how often a panelist in the continuous procedure reflects upon the sensation and decides to change the position of the response device. Decisions are probably not continuous even though some records from some subjects may look like smooth curves. An indication that response tendencies are important is when the conditions of stimulation are held constant, but the response task changes. For example, using the graphic chart-recorder method, Lawless and Skinner (1979) found median durations for sucrose intensity that were 15-35% shorter than the same stimuli rated using repeated category ratings. Why would the different rating methods produce apparently different durations? Very different patterns may be observed when taste quality and intensity are tracked. Halpern (1991) found that tracked taste quality of 2 mM sodium saccharin had a delayed onset (by 400 ms) compared with tracked intensity. This might be understandable from the point of requiring a more complex decision process in the case of tracking intensity. However, it still alerts us to the fact that the behavior probably trails the actual experience by some unknown amount. What is more surprising in Halpern’s data is that tracked quality also stopped well before tracked intensity (by

198

600 ms). Can it be possible that subjects are still experiencing a taste of some definable intensity and yet the quality has disappeared? Or is there another response generation process at work in this task? A third area where potential response biases can operate is in contextual effects. Clark and Lawless (1994) showed that common contextual affects like successive contrast also operated with TI methods, as they do with other scaling tasks. Also, some ratings could be enhanced when a limited number of scales were used by subjects. As observed in singlepoint scaling, enhancement of sweetness by fruity flavors tends to occur when only sweetness is rated. When the fruity flavor is also rated, the sweetness enhancement often disappears (Frank et al., 1989), an effect sometimes referred to as halo dumping or simply “dumping.” Using the discrete-point version of TI scaling, so that multiple attributes could be rated, Clark and Lawless showed a similar effect. This is potentially troublesome for the continuous tracking methods, since they often limit subjects to responding to only one attribute at a time. This may explain in part why sweetness enhancement by flavors can occur so readily in TI studies (e.g., Matysiak and Noble, 1991). A final concern is the question of whether the bounded response scales often used in TI measurement produces any compssion of the differences among products. In analog tracking tasks, there is a limit as to how far the joystick, mouse, lever, dial, or other response device can be moved. With some practice, judges learn not to bump into the top. Yet the very nature of the tracking response encourages judges to sweep a wide range of the response scale. If this were done on every trial, it would tend to attenuate the differences in maximum tracked intensity between products. As an example, Overbosch et al. (1986, see Fig. 2) showed curves for pentanone where doubling the concentration changed peak heights by only about 8%. A similar sort of compssion is visible in Lawless and Skinner’s (1979) data for sucrose, compared to the psychophysical data in the literature.

8.8 Conclusions In most cases TI parameters show similar statistical differentiation as compared to traditional scales, but this is not universally the case (e.g., Moore and Shoemaker, 1981).

8

Time-Intensity Methods

Many sensory evaluation researchers have supported increased application of time-intensity measurements for characterization of flavor and texture sensations. In particular, the method was championed by Lee and Pangborn, who argued that the methods provide detailed information not available from single estimates of sensation intensity (Lee, 1989; Lee and Pangborn, 1986). TI methods can provide rate-related, duration, and intensity information not available from traditional scaling. However, the utility of the methods must be weighed against the enhanced cost and complexity in data collection and analysis. In deciding whether to apply TI methods over conventional scaling, the sensory scientist should consider the following criteria: (1) Is the attribute or system being studied known to change over time? Simply eating the food can often settle this issue; in many cases it is obvious. (2) Will the products differ in sensory time course as a function of ingredients, processing, packaging, or other variables of interest? (3) Will the time variation occur in such a way that it will probably not be captured by direct single ratings? (4) Is some aspect of the temporal profile likely to be related to consumer acceptability? (5) Does the added information provided by the technique outweigh any additional costs or time delays in panel training, data acquisition, and data analysis? Obviously, when more answers are positive on these criteria, a stronger case can be made for choosing a TI method from the available set of sensory evaluation tools.

References Abrahams, H., Krakauer, D. and Dallenbach, K. M. 1937. Gustatory adaptation to salt. American Journal of Psychology, 49, 462-469. Ayya, N. and Lawless, H. T. 1992. Qualitative and quantitative evaluation of high-intensity sweeteners and sweetener mixtures. Chemical Senses, 17, 245-259. Baron, R. F. and Penfield, M. P. 1996. Capsaicin heat intensity – concentration, carrier, fat level and serving temperature effects. Journal of Sensory Studies, 11, 295-316. Barylko-Pikielna, N., Mateszewska, I. and Helleman, U. 1990. Effect of salt on time-intensity characteristics of bread. Lebensmittel Wissenschaft und Technologie, 23, 422-426.

References Birch, G. G. and Munton, S. L. 1981. Use of the “SMURF” in taste analysis. Chemical Senses, 6, 45-52. Bloom, K., Duizer, L. M. and Findlay, C. J. 1995. An objective numerical method of assessing the reliability of time- intensity panelists. Journal of Sensory Studies, 10, 285-294. Bonnans, S. and Noble, A. C. 1993. Effect of sweetener type and of sweetener and acid levels on temporal perception of sweetness, sourness and fruitiness. Chemical Senses, 18, 273-283. Brandt, M. A., Skinner, E. Z. and Coleman, J. A. 1963. Texture profile method. Journal of Food Science, 28, 404-409. Brown, W. E., Landgley, K. R., Martin, A. and MacFie, H. J. 1994. Characterisation of patterns of chewing behavior in human subjects and their influence on texture perception. Journal of Texture Studies, 15, 33-48. Butler, G., Poste, L. M., Mackie, D. A., and Jones, A. 1996. Time-intensity as a tool for the measurement of meat tenderness. Food Quality and Preference, 7, 193-204. Cabanac, M. 1971. Physiological role of pleasure. Science, 173, 1103-1107. Cain, W. S. 1974. Perception of odor intensity and time-course of olfactory adaptation. ASHRAE transactions, 80, 53-75. Clark, C. C. and Lawless, H. T. 1994. Limiting response alternatives in time-intensity scaling: An examination of the Halo-Dumping effect. Chemical Senses, 19, 583-594. Cliff, M. 1987. Temporal perception of sweetness and fruitiness and their interaction in a model system. MS Thesis, University of California, Davis, USA. Cliff, M. and Heymann, H. 1992. Descriptive analysis of oral pungency. Journal of Sensory Studies, 7, 279-290. Cliff, M. and Heymann, H. 1993a. Time-intensity evaluation of oral burn. Journal of Sensory Studies, 8, 201-211. Cliff, M. and Heymann, 1993b. Development and use of time- intensity methodology for sensory evaluation: A review. Food Research International, 26, 375-385. Cliff, M. and Noble, A. C. 1990. Time-intensity evaluation of sweetness and fruitiness in a model solution. Journal of Food Science, 55, 450-454. Dacanay, L. 1990. Thermal and concentration effects on temporal sensory attributes of L – menthol. M.S. Thesis, University of California, Davis, USA. de Roos, K. B. 1990. Flavor release from chewing gums. In: Y. Bessiere and A. F. Thomas (eds.), Flavour Science and Technology. Wiley, Chichester, pp. 355-362. DeRovira, D. 1996. The dynamic flavor profile method. Food Technology, 50, 55-60. Dijksterhuis, G. 1993. Principal component analysis of time- intensity bitterness curves. Journal of Sensory Studies, 8, 317-328. Dijksterhuis, G. 1996. Time-intensity methodology: Review and pview. Proceedings, COST96 Meeting: Interaction of Food Matrix with Small Ligands Influencing Flavour and Texture, Dijon, France, November 20, 1995. Dijksterhuis, G. and van den Broek, E. 1995. Matching the shape of time-intensity curves. Journal of Sensory Studies, 10, 149-161. Dijksterhuis, G., Flipsen, M. and Punter, P. H. 1994. Principal component analysis of time-intensity data. Food Quality and Preference, 5, 121-127. DuBois, G. E. and Lee, J. F. 1983. A simple technique for the evaluation of temporal taste properties. Chemical Senses, 7, 237-247.

200 DuBois (eds.), Sweeteners: Discovery, Molecular Design and Chemoreception. ACS Symposium Series #450. American Chemical Society, Washington, DC, pp. 277-289. Jellinek, G. 1964. Introduction to and critical review of modern methods of sensory analysis (odor taste and flavor evaluation) with special emphasis on descriptive analysis. Journal of Nutrition and Dietetics, 1, 219-260. Jellinek, G. 1985. Sensory Evaluation of Food, Theory and Practice. Ellis Horwood, Chichester, England. Kroeze, J. H. A. 1979. Masking and adaptation of sugar sweetness intensity. Physiology and Behavior, 22, 347-351. Kuo, Y.-L., Pangborn, R. M. and Noble, A. C. 1993. Temporal patterns of nasal, oral and retronasal perception of citral and vanillin and interactions of these odorants with selected tastants. International Journal of Food Science and Technology, 28, 127-137. Labbe, D., Schlich, P., Pineau, N., Gilbert, F. and Martin, N. 2009. Temporal dominance of sensations and sensory profiling: A comparative study. Food Quality and Preference, 20, 216-221. Lallemand, M., Giboreau, A., Rytz, A. and Colas, B. 1999. Extracting parameters from time-intensity curves using a trapezoid model: The example of some sensory attributes of ice cream. Journal of Sensory Studies, 14, 387-399. Larson-Powers, N. and Pangborn, R. M. 1978. Paired comparison and time-intensity measurements of the sensory properties of beverages and gelatins containing sucrose or synthetic sweeteners. Journal of Food Science, 43, 41-46. Lawless, H. T. 1980. A computerized system for assessing taste intensity over time. Paper psented at the Chemical Senses and Intake Society, Hartford, CT, April 9, 1980. Lawless, H. T. 1984. Oral chemical irritation: Psychophysical properties. Chemical Senses, 9, 143-155. Lawless, H. T. and Clark, C. C. 1992. Psychological biases in time-intensity scaling. Food Technology, 46(11), 81, 84-86, 90. Lawless, H. T. and Skinner, E. Z. 1979. The duration and perceived intensity of sucrose taste. Perception and Psychophysics, 25, 249-258. Lawless, H. T. and Stevens, D. A. 1988. Responses by humans to oral chemical irritants as a function of locus of stimulation. Perception and Psychophysics, 43, 72-78. Lawless, H. T., Corrigan, C. L. and Lee, C. L. 1994. Interactions of astringent substances. Chemical Senses, 19, 141-154. Lawless, H. T., Tuorila, H., Jouppila, K., Virtanen, P. and Horne, J. 1996. Effects of guar gum and microcrystalline cellulose on sensory and thermal properties of a high fat model food system. Journal of Texture Studies 27, 493-516. Le Reverend, F. M., Hidrio, C., Fernandes, A. and Aubry, V. 2008. Comparison between temporal dominance of sensation and time intensity results. Food Quality and Preference, 19, 174-178. Leach, E. J. and Noble, A. C. 1986. Comparison of bitterness of caffeine and quinine by a time-intensity procedure. Chemical Senses, 11, 339-345. Ledauphin, S., Vigneau, E. and Causeur, D. 2005. Functional approach for the analysis of time intensity curves using B-splines. Journal of Sensory Studies, 20, 285-300. Ledauphin, S., Vigneau, E. and Qannari, E. M. 2006. A procedure for analysis of time intensity curves. Food Quality and Preference, 17, 290-295.

8

Time-Intensity Methods

Lee, C. B. and Lawless, H. T. 1991. Time-course of astringent materials. Chemical Senses, 16, 225-238. Lee, W. E. 1985. Evaluation of time-intensity sensory responses using a personal computer. Journal of Food Science, 50, 1750-1751. Lee, W. E. 1986. A suggested instrumental technique for studying dynamic flavor release from food products. Journal of Food Science, 51, 249-250. Lee, W. E. 1989. Single-point vs. time-intensity sensory measurements: An informational entropy analysis. Journal of Sensory Studies, 4, 19-30. Lee, W. E. and Pangborn, R. M. 1986. Time-intensity: The temporal aspects of sensory perception. Food Technology, 40, 71-78, 82. Liu, Y. H. and MacFie, H. J. H. 1990. Methods for averaging time-intensity curves. Chemical Senses, 15, 471-484. Lundahl, D. S. 1992. Comparing time-intensity to category scales in sensory evaluation. Food Technology, 46(11), 98-103. Lynch, J., Liu, Y.-H., Mela, D. J. and MacFie, H. J. H. 1993. A time-intensity study of the effect of oil mouthcoatings on taste perception. Chemical Senses, 18, 121-129. Matysiak, N. L. and Noble, A. C. 1991. Comparison of temporal perception of fruitiness in model systems sweetened with aspartame, aspartame + acesulfame K blend or sucrose. Journal of Food Science, 65, 823-826. McBurney, D. H. 1966. Magnitude estimation of the taste of sodium chloride after adaptation to sodium chloride. Journal of Experimental Psychology, 72, 869-873. McBurney, D. H. and Shick, T. R. 1971. Taste and water taste of 26 compounds for man. Perception and Psychophysics, 11, 228-232. McGowan, B. A. and Lee, S.-Y. 2006. Comparison of methods to analyze time-intensity curves in a corn zein chewing gum study. Food Quality and Preference 17, 296-306. McNulty, P. B. 1987. Flavour release-elusive and dynamic. In: J. M. V. Blanshard and P. Lillford (eds.), Food Structure and Behavior. Academic, London, pp. 245-258. McNulty, P. B. and Moskowitz, H. R. 1974. Intensity -time curves for flavored oil-in-water emulsions. Journal of Food Science, 39, 55-57. Meiselman, H. L. 1968. Magnitude estimation of the time course of gustatory adaptation. Perception and Psychophysics, 4, 193-196. Meiselman, H. L. and Dubose, C. N. 1976. Failure of instructional set to affect completeness of taste adaptation. Perception and Psychophysics, 19, 226-230. Meiselman, H. L. and Halpern, B. P. 1973. Enhancement of taste intensity through pulsatile stimulation. Physiology and Behavior, 11, 713-716. Moore, L. J. and Shoemaker, C. F. 1981. Sensory textural properties of stabilized ice cream. Journal of Food Science, 46, 399-402, 409. Neilson, A. J. 1957. Time-intensity studies. Drug and Cosmetic Industry, 80, 452-453, 534. O’Keefe, S. F., Resurreccion, A. P., Wilson, L. A. and Murphy, P. A. 1991. Temperature effect on binding of volatile flavor compounds to soy protein in aqueous model systems. Journal of Food Science, 56, 802-806. O’Mahony, M. 1986. Sensory adaptation. Journal of Sensory Studies, 1, 237-257.

References O’Mahony, M. and Wong, S.-Y. 1989. Time-intensity scaling with judges trained to use a calibrated scale: Adaptation, salty and umami tastes. Journal of Sensory Studies, 3, 217-236. Ott, D. B., Edwards, C. L. and Palmer, S. J. 1991. Perceived taste intensity and duration of nutritive and non-nutritive sweeteners in water using time-intensity (T-I) evaluations. Journal of Food Science, 56, 535-542. Overbosch, P. 1987. Flavour release and perception. In: M. Martens, G. A. Dalen and H. Russwurm (eds.), Flavour Science and Technology. Wiley, New York, pp. 291-300. Overbosch, P., Van den Enden, J. C., and Keur, B. M. 1986. An improved method for measuring perceived intensity/time relationships in human taste and smell. Chemical Senses, 11, 315-338. Owen, W. J. and DeRouen, T. A. 1980. Estimation of the mean for lognormal data containing zeroes and left-censored values, with application to the measurement of worker exposure to air contaminants. Biometrics, 36, 707-719. Pangborn, R. M. and Koyasako, A. 1981. Time-course of viscosity, sweetness and flavor in chocolate desserts. Journal of Texture Studies, 12, 141-150. Pangborn, R. M., Lewis, M. J. and Yamashita, J. F. 1983. Comparison of time-intensity with category scaling of bitterness of iso-alpha-acids in model systems and in beer. Journal of the Institute of Brewing, 89, 349-355. Peyvieux, C. and Dijksterhuis, G. 2001. Training a sensory panel for TI: A case study. Food Quality and Preference, 12, 19-28. Pineau, N., Schlich, P., Cordelle, S., Mathonniere, C., Issanchou, S., Imbert, A., Rogeaux, M., Eteviant, P. and Köster, E. 2009. Temporal dominance of sensations: Construction of the TDS curves and comparison with time- intensity. Food Quality and Preference, 20, 450-455. Pionnier, E., Nicklaus, S., Chabanet, C., Mioche, L., Taylor, A. J., LeQuere, J. L. and Salles, C. 2004. Flavor perception of a model cheese: relationships with oral and physico-chemical parameters. Food Quality and Preference, 15, 843-852. Prescott, J. and Stevenson, R. J. 1996. Psychophysical responses to single and multiple psentations of the oral irritant zingerone: Relationship to frequency of chili consumption. Physiology and Behavior, 60-617-624. Reinbach, H. C., Toft, M. and Møller, P. 2009. Relationship between oral burn and temperature in chili spiced pork patties evaluated by time-intensity. Food Quality and Preference, 20, 42-49. Reinbach, H. C., Meinert, L., Ballabio, D., Aayslyng, M. D., Bredie, W. L. P., Olsen, K. and Møller, P. 2007. Interactions between oral burn, meat flavor and texture in chili spiced pork patties evaluated by time-intensity. Food Quality and Preference, 18, 909-919. Rine, S. D. 1987. Computerized analysis of the sensory properties of peanut butter. M. S. Thesis, University of California, Davis, USA.

Chapter 9

Context Effects and Biases in Sensory Judgment

Abstract Human judgments about a sensation or a product are strongly influenced by items that surround the item of interest, either in space or in time. This chapter shows how judgments can change as a function of the context within which a product is evaluated. Various contextual effects and biases are described and categorized. Some solutions and courses of action to minimize these biases are psented. By such general principles of action as these everything looked at, felt, smelt or heard comes to be located in a more or less definite position relatively to other collateral things either actually psented or only imagined as possibly there. – James (1913, p. 342)

Contents

9.6.3 Positional or Order Bias . . . . . . Antidotes . . . . . . . . . . . . . . . . 9.7.1 Avoid or Minimize . . . . . . . . . 9.7.2 Randomization and Counterbalancing 9.7.3 Stabilization and Calibration . . . . 9.7.4 Interptation . . . . . . . . . . . 9.8 Conclusions . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . 9.7

9.1 9.2

9.3

9.4

9.5

9.6

Introduction: The Relative Nature of Human Judgment . . . . . . . . . . . . . . . . . . Simple Contrast Effects . . . . . . . . . . . 9.2.1 A Little Theory: Adaptation Level . . . . 9.2.2 Intensity Shifts . . . . . . . . . . . . 9.2.3 Quality Shifts . . . . . . . . . . . . . 9.2.4 Hedonic Shifts . . . . . . . . . . . . 9.2.5 Explanations for Contrast . . . . . . . . Range and Frequency Effects . . . . . . . . . 9.3.1 A Little More Theory: Parducci’s Range and Frequency Principles . . . . . . . . 9.3.2 Range Effects . . . . . . . . . . . . . 9.3.3 Frequency Effects . . . . . . . . . . . Biases . . . . . . . . . . . . . . . . . . . . 9.4.1 Idiosyncratic Scale Usage and Number Bias . . . . . . . . . . . 9.4.2 Poulton’s Classifications . . . . . . . . 9.4.3 Response Range Effects . . . . . . . . 9.4.4 The Centering Bias . . . . . . . . . . Response Correlation and Response Restriction . . . . . . . . . . . . . . . . . 9.5.1 Response Correlation . . . . . . . . . 9.5.2 “Dumping” Effects: Inflation Due to Response Restriction in Profiling . . . 9.5.3 Over-Partitioning . . . . . . . . . . . Classical Psychological Errors and Other Biases . . . . . . . . . . . . . . . . . . . . 9.6.1 Errors in Structured Sequences: Anticipation and Habituation . . . . . . . . . . . . 9.6.2 The Stimulus Error . . . . . . . . . .

203 206 206 207 207 208 209 210 210 210 211 212 212 213 214 215 216 216 217 218 218 218 219

. . . . . . . .

. . . . . . . .

219 219 219 220 221 222 222 223

9.1 Introduction: The Relative Nature of Human Judgment This chapter will discuss context effects and common biases that can affect sensory judgments. Context effects are conditions in which the judgment about a product, usually a scaled rating, will shift depending upon factors such as the other products that are evaluated in the same tasting session. A mediocre product evaluated in the context of some poor-quality items may seem very good in comparison. Biases refer to tendencies in judgment in which the response is influenced in some way to be an inaccurate reflection of the actual sensory experience. In magnitude estimation ratings, for example, people have a tendency to use numbers that are multiples of 2, 5, and 10, even though they can use any number or fraction they wish. At the

H.T. Lawless, H. Heymann, Sensory Evaluation of Food, Food Science Text Series, DOI 10.1007/978-1-4419-6488-5_9, © Springer Science+Business Media, LLC 2010

203

204

end of the chapter, some solutions to these problems are offered, although a sensory scientist should realize that we can never totally eliminate these factors. In fact, they are of interest and deserve study on their own for what they can tell us about human sensory and cognitive processes. An axiom of perceptual psychology has it that humans are very poor absolute measuring instruments but are very good at comparing things. For example, we may have difficulty estimating the exact sweetness level of our coffee, but we have little trouble in telling whether more sugar has been added to make it sweeter. The question arises, if people are prone to making comparisons, how can they give ratings when no comparison is requested or specified? For example, when asked to rate the perceived firmness of a food sample, how do they judge what is firm versus what is soft? Obviously, they must either choose a frame of reference for the range of firmness to be judged or be trained with explicit reference standards to understand what is high and low on the response scale. In other words, they must relate this sensory judgment to other products they have tried. For many items encountered in everyday life, we have established frames of reference based on our experiences. We have no trouble forming an image of a “large mouse running up the trunk of a small elephant” because we have established frames of reference for what constitutes the average mouse and the average elephant. In this case the judgment of large and small is context dependent. Some people would argue that all judgments are relative. This dependence upon a frame of reference in making sensory judgments demonstrates the influence of contextual factors in biasing or changing how products are evaluated. We are always prone to see things against a background or pvious experience and evaluate them accordingly. A 40◦ (Fahrenheit) day in Ithaca, New York, in January seems quite mild against the background of the northeastern American winter. However, the same 40◦ C temperature will feel quite cool on an evening in August in the same location. This principle of frame of reference is the source of many visual illusions, where the same physical stimulus causes very different perceptual impssions, due to the context within which it is embedded. Examples are shown in Fig. 9.1. A simple demonstration of context is the visual afterimage effect that gave rise to Emmert’s law (Boring, 1942). In 1881, Emmert formalized a

9

Context Effects and Biases in Sensory Judgment

A

B

C Fig. 9.1 Examples of contextual effects from simple visual illusions. (a) The dumbell version of the Muller-Lyer illusion. (b) The Ebbinghaus illusion. (c) Illusory contours. In this case the contexts induce the perceptions of shapes.

principle of size constancy based on the following effect: Stare for about 30 s at a brightly illuminated colored paper rectangle (it helps to have a small dot to aid in fixation in the center) about a meter away. Then shift your gaze to a white sheet on the table in front of you. You should see the rectangle afterimage in a complementary color and somewhat smaller in size as compared to the original colored rectangle. Next, shift your gaze to a white wall some distance off. The afterimage will now appear much larger, as the brain finds a fixed visual angle at greater distance to repsent larger physical objects. Since the mind does not immediately recognize that the afterimage is just a creation of the visual sensory system, it projects it at the distance of the surface upon which it is “seen.” The more distant frame of reference, then, demands a larger size perception. The close link between sensory judgments and context psents problems for anyone who wants to view ratings as absolute or comparable across different times, sessions, or settings. Even when the actual sensory impssion of two items is the same, we can shift the frame of reference and change the overt behavior of the person to produce a different response. This problem (or principle of sensory function) was glossed

9.1 Introduction: The Relative Nature of Human Judgment

over by early psychophysical scientists. In psychological terms, they used a simple stimulus-response (S-R) model, in which response was considered a direct and unbiased repsentation of sensory experiences. Certain biases were observed, but it was felt that suitable experimental controls could minimize or eliminate them (Birnbaum, 1982; Poulton, 1989). A more modern view is that there are two or three distinct processes contributing to ratings. The first is a psychophysical process by which stimulus energy is translated into physiological events that result in a subjective experience of some sensory intensity. The second, equally important process is the function by which the subjective experience is translated into the observed response, i.e., how the percept is translated onto the rating scale (see Fig. 9.2). Many psychophysical researchers now consider a “judgment function” to

A)

Response

Stimulus

B) Stimulus

Response

Sensation Judgment Process

Psychophysical Process C) Stimulus

Sensation

Psychophysical Process

ENCODED SENSATION, “PERCEPT”

Perceptual Process

Response

Judgment Process

Fig. 9.2 Models for sensory-response processes. (a) The simple stimulus-response model of twentieth-century behavioral psychology. (b) Two processes are involved in sensation and response, a psychophysical process and then a response output or a judgment process in which the participant decides what response to give for that sensation. (c) A more complex model in which the sensation may be transformed before the response is generated. It may exist in short-term memory as an encoded percept, different from the sensation. Contextual effects of simultaneous or sequential stimuli can influence the stimulus-response sequence in several ways. Peripheral physiological effects such as adaptation or mixture inhibition may change the transduction process or other early stages of neural processing. Other stimuli may give rise to separate percepts that are integrated into the final response. Contextual factors may also influence the frame of reference that determines how the response output function will be applied. In some models, an additional step allows transformation of the percept into covert responses that are then translated as a separate step into the overt response R.

205

be an important part of the sequence from stimulus to response (Anderson, 1974, Birnbaum, 1982, McBride and Anderson, 1990; Schifferstein and Frijters, 1992; Ward, 1987). This process is also sometimes referred to as a response output function. A third intermediate step is the conversion of the raw sensory experience into some kind of encoded percept, one that is available to memory for a short time, before the judgment is made (Fig. 9.2c). Given this framework, there are several points at which stimulus context may influence the sensory process. First, of course, the actual sensation itself may change. Many sensory processes involve interaction effects of simultaneous or sequential influences of multiple items. An item may be perceived differently due to the direct influence of one stimulus upon another that is nearby in time or space. Simultaneous color contrast is an example in color vision and some types of inhibitory mixture interactions and masking in taste and smell are similarly hard wired. Quinine with added salt is less bitter than quinine tasted alone, due to the ways that sodium ions inhibit bitterness transduction. Sensory adaptation weakens the perception of a stimulus because of what has pceded it. So the psychophysical process itself is altered by the milieu in which the stimulus is observed, sometimes because of physical effects (e.g., simple buffering of an acid) or physiological effects (e.g., neural inhibition causing mixture suppssion) in the peripheral sensory mechanisms. A second point of influence is when the context shifts the frame of reference for the response output function. That is, two sensations may have the same subjective intensity under two conditions, but because of the way the observer places them along the response continuum (due to different contexts), they are rated differently. A number of studies have shown that contextual factors such as the distribution of stimuli along the physical continuum affect primarily (although not exclusively) the response output function (Mellers and Birnbaum, 1982, 1983). A third process is sometimes added in which the sensation itself is translated into an implicit response or encoded image that may also be affected by context (Fig. 9.2c). This would provide another opportunity to influence the process if contextual factors affect this encoding step. Contextual change can be viewed as a form of bias. Bias, in this sense, is a process that causes a shift or a change in response to a constant sensation. If one situation is viewed as producing a true

206

and accurate response, then the contextual conditions that cause shifts away from this accurate response are “biased.” However, bias need only have a negative connotation if there is a reason to psume that one condition of judgment is more accurate than all others. A broader view is to accept the idea that all judgments are a function of observing conditions and therefore all judgments are biased from one another in different ways. Fortunately, many of these biases and the conditions that cause them are pdictable and well understood, and so they can be eliminated or minimized. At the very least, the sensory practitioner needs to understand how these influences operate so as to know when to expect changes in judgments and ratings. An important endpoint is the realization that few, if any, ratings have any absolute meaning. You cannot say that because a product received a hedonic rating of 7.0 today, it is better than the product that received a rating of 6.5 last week. The context may have changed.

9.2 Simple Contrast Effects By far the most common effect of sensory context is simple contrast. Any stimulus will be judged as more intense in the psence of a weaker stimulus and as less intense in the psence of a stronger stimulus, all other conditions being equal. This effect is much easier to find and to demonstrate than its opposite, convergence or assimilation. For example, an early sensory worker at the Quartermaster Corp., Kamenetsky (1957), noticed that the acceptability ratings for foods seemed to depend upon what other foods were psented during an evaluation session. Poor foods seemed even worse when pceded by a good sample. Convergence is more difficult to demonstrate, although under some conditions a group of items may seem more similar to each other when they are in the psence of an item that is very different from that group (Zellner et al., 2006).

9.2.1 A Little Theory: Adaptation Level As we noted above, a 40◦ day in January (in New York) seems a lot warmer than the same temperature in August. These kinds of effects are pdicted Helson’s

9

Context Effects and Biases in Sensory Judgment

theory of adaptation level. Helson (1964) proposed that we take as a frame of reference the average level of stimulation that has pceded the item to be evaluated. The mild temperature in the middle of a hot and humid summer seems a lot more cool and refreshing than is the mild temperature after a period of cold and icy weather. So we refer to our most recent experiences in evaluating the sensory properties of an item. Helson went on to elaborate the theory to include both immediate and distant pdecessors. That is, he appciated the fact that more recent items tend to have a stronger effect on the adaptation level. Of course, mere reference to the mean value of experience is not always sufficient to induce a contrast effect-it is more influential if the mean value comes to be centered near the middle of the response scale, an example of a centering bias, discussed below (Poulton, 1989). The notion of adaptation, a decrease in responsiveness under conditions of constant stimulation, is a major theme in the literature on sensory processes. Physiological adaptation or an adjustment to the ambient level of stimulation is obvious in light/dark adaptation in vision. The thermal and tactile senses also show profound adaptation effects-we become easily adjusted to the ambient room temperature (as long as it is not too extreme) and we become unaware of tactile stimulation from our clothing. So this mean reference level often passes from consciousness or becomes a new baseline from which deviations in the environment become noticeable. Some workers have even suggested that this improves discrimination-that the difference threshold is smallest right around the adaptation level or physiological zero, in keeping with Weber’s law (McBurney, 1966). Examples of adaptation effects are discussed in Chapter 2 for the senses of taste and smell. In the chemical, thermal, and tactile senses, adaptation is quite profound. However, we need not invoke the concept of neural adaptation to a pceding item or a physiological effect to explain all contrast effects. It may be simply that more or less extreme stimuli change our frame of reference or the way in which the stimulus range and response scales are to be mapped onto one another. The general principle of context is that human observers act like measuring instruments that constantly recalibrate themselves to the experienced frame of reference. What we think of as a small horse may depend upon whether the frame of reference includes Clydesdales, Shetland ponies, or tiny phistoric equine species. The

9.2 Simple Contrast Effects

207

following examples show simple effects of context on intensity, sensory quality, and hedonics or acceptability. Most of these examples are cases of perceptual contrast or a shift in judgment away from other stimuli psented in the same session.

9.2.2 Intensity Shifts Figure 9.3 shows a simple contrast effect of soups with varying salt levels psented in different contexts (Lawless, 1983). The central stimulus in the series was psented either with two lower or two higher concentrations of salt added to a low sodium soup. Ratings of saltiness intensity were made on a simple nine-point category scale. In the lower context, the central soup received a higher rating, and in the higher context, it received a lower rating, analogous to our perception of a mild day in winter (seemingly warmer) versus a mild day in summer (seemingly cooler). Note that the shift is quite dramatic, about two points on the nine-point scale or close to 25% of scale range. A simple classroom demonstration can show a similar shift for the tactile roughness of sandpapers varying in grit size. In the context of a rougher sample, a medium sample will be rated lower than it is in the context of a smoother sample. The effects of simple contrast are not limited to taste and smell.

Mean Category Rating

7 6 5

Low Context High Context

4

3

2

.12

.18

.25

.35

.50

NaCl Concentration (log scale)

Fig. 9.3 Saltiness ratings of soups with added NaCl. The sample at 0.25 M was evaluated in two contexts, one with higher concentrations and one with lower concentrations. The shift is typical of a simple contrast effect of contrast. Replotted from Lawless (1983). Copyright ASTM, used with permission.

Contrast effects are not always observed. In some psychophysical work with long series of stimuli, some item-to-item correlations have been observed. The effects of immediately pceding versus remotely pceding stimuli have been measured and a positive correlation among adjacent responses in the series was found. This can be taken as evidence for a type of assimilation, or underestimation of differences (Ward, 1979, 1987; but see also Schifferstein and Frijters, 1992).

9.2.3 Quality Shifts Visual examples such as color contrast were well known to early psychologists like William James: “Meanwhile it is an undoubted general fact that the psychical effect of incoming currents does depend on what other currents may be simultaneously pouring in. Not only the perceptibility of the object which the current brings before the mind, but the quality of it is changed by the other currents.” (1913, p. 25). A gray line against a yellow background may appear somewhat bluish, and the same line against a blue background may seem more yellowish. Paintings of the renowned artist Josef Albers made excellent use of color contrast. Similar effects can be observed for the chemical senses. During a descriptive panel training period for fragrance evaluation, the terpene aroma compound dihydromyrcenol was psented among a set of woody or pine-like reference materials. The panelists complained that the aroma was too citrus-like to be included among the woody reference materials. However, when the same odor was placed in the context of citrus reference materials, the same panelists claimed that it was far too woody and pine-like to be included among the citrus examples. This contextual shift is shown in Fig. 9.4. In a citrus context, the item is rated as more woody in character than when placed in a woody context. Conversely, ratings for citrus intensity decrease in the citrus context and increase in the woody context. The effect is quite robust and is seen whether or not a rest period is included to undo the potential effects of sensory adaptation. It even occurs when the contextual odor follows the target item and judgments are made after both are experienced (Lawless et al., 1991)! This striking effect is discussed further in Section 9.2.5.

208

9

Context Effects and Biases in Sensory Judgment

CITRUS CONTEXT

MEAN INTENSITY RATING, DIHYDROMYRCENOL

WOODY CONTEXT Citrus/woody scales

Intensity/pleasantness

used during exposure

scales during exposure

9 8 7 6 5 4 3 2 1 CITRUS RATING WOODY RATING

CITRUS RATING WOODY RATING

ODOR CHARACTERISTIC Fig. 9.4 Odor quality contrast noted for the ambiguous terpene aroma compound dihydromyrcenol. In a citrus context, woody ratings increase and citrus character decreases. In a woody context, the woody ratings decrease. The group using different

scales did not rate citrus and woody character during the contextual exposure phase, only overall intensity and pleasantness were rated. From Lawless et al. (1991) by permission of Oxford University Press.

Context effects can also alter how items are identified and characterized. When people categorize speech sounds, repeated exposure to one type of simple phoneme changes the category boundary for other speech sounds. Repeated exposure to the sound of the phoneme “bah,” which has an early voice onset time, can shift the phoneme boundary so that speech sounds near the boundary are more likely classified as “pah” sounds (a later voice onset) (Eimas and Corbit, 1973). Boundary-level examples are shifted across the boundary and into the next category. This shift resembles a kind of contrast effect.

an item of poor quality and less appealing if it followed something of better quality. The effect was known to Beebe-Center (1932), who also attributed it to Fechner in 1898. This kind of contrast has been observed for tastes (Riskey et al., 1979; Schifferstein, 1995), odors (Sandusky and Parducci, 1965), and art (Dolese et al., 2005). Another effect observed in these kinds of experiments is that a contrasting item causes other, generally lower rated stimuli to become more similar or less discriminable, an effect termed condensation (Parker et al., 2002; Zellner et al., 2006;). In the study by Zellner et al. (2006), p-exposure to a good-tasting juice reduced the magnitude of pference ratings among less appealing juices. Mediocre items were both worse and more similar. An example of hedonic shifting was found in a study on the optimization of the saltiness of tomato juice and also the sweetness of a fruit beverage using the method of adjustment (Mattes and Lawless, 1985). When trying to optimize the level of sweetness or saltiness in this study, subjects worked in two directions. In an ascending series, they would concentrate a dilute solution by mixing the beverage with a more concentrated

9.2.4 Hedonic Shifts Changes in the pference or acceptance of foods can be seen as a function of context. Hedonic contrast was a well-known effect to early workers in food acceptance testing (Hanson et al., 1955; Kamenetzky, 1959). An item seems more appealing if it followed

9.2 Simple Contrast Effects

version having the same color, aroma, and other flavor materials (i.e., only sweetness or saltiness was different). In a second descending series, they would be given a very intense sample as the starting point and then dilute down to their pferred level. This effect is shown in Fig. 9.5. The adjustment stops too soon and the discrepancy is remarkable, nearly a concentration range of 2:1. The effect was also robust-it could not be attributed to sensory adaptation or lack of discrimination and persisted even when subjects were financially motivated to try to achieve the same endpoints in both trials. This is a case of affective contrast. When compared to a very sweet or salty starting point, a somewhat lower item seems just about right, but when starting with a relatively sour fruit beverage

Fig. 9.5 Optimized concentrations of salt in tomato juice and sucrose in a fruit beverage. In the trials labeled D, the concentration was diluted from a concentrated version of the test sample. In the trials marked A, the concentration was increased from a dilute version of the test sample. Concentrations of other ingredients were held constant. The contextual shift is consistent with reaching the apparent optimum too soon as if the apparent optimum was shifted in contrast to the starting point. From Mattes and Lawless (1985) with permission.

209

or bland tomato juice, just a little bit of sugar or salt helps quite a bit. The stopping point contrasts with the starting material and seems to be better than it would be perceived in isolation. In an ascending or descending sequence of products, a change in responses that happens too soon is called an “error of anticipation.”

9.2.5 Explanations for Contrast At first glance, one is tempted to seek a physiological explanation for contrast effects, rather than a psychological or a judgmental one. Certainly sensory adaptation to a series of intense stimuli would cause any subsequent test item to be rated much lower. The importance of sensory adaptation in the chemical senses of taste and smell lends some credence to this explanation. However, a number of studies have shown that pcautions against sensory adaptation may be taken, such as sufficient rinsing or time delays between stimuli and yet the context effects persist (Lawless et al., 1991; Mattes and Lawless, 1985; Riskey, 1982). Furthermore, it is difficult to see how sensory adaptation to low-intensity stimuli would cause an increase in the ratings for a stronger item, as adaptation necessarily causes a decrement in physiological responsiveness compared to a no-stimulation baseline. Perhaps the best evidence against a simple adaptation explanation for contrast effects is from the reversed-pair experiments in which the contextual item follows the to-be-rated target item and therefore can have no physiologically adapting effect on it. This paradigm calls for a judgment of the target item from memory after the psentation of the contextual item, in what has been termed a reversed-pair procedure (Diehl et al., 1978). Due to the reversed order, the context effects cannot be blamed on physiological adaptation of receptors, since the contextual item follows rather than pcedes the item to be rated. Reversed-pair effects are seen for shifts in odor quality of aroma compounds like dihydromyrcenol and are only slightly smaller in magnitude than the contextual shift caused when the contextual item comes first (Lawless et al., 1991). The reversed-pair situation is also quite capable of causing simple contrast effects in sensory intensity. A sweetness shift was observed when a higher or a lower sweetness item was interpolated between the tasting and rating (from memory)

210

of a normal-strength fruit beverage (Lawless, 1994). Looking back at Fig. 9.2, it seems more likely that the effect changes the response function. However, not all workers in the field agree. In particular, Marks (1994) has argued that the contextual shifts are much like an adaptation process and that for auditory stimuli this is a peripheral event. It is possible that what changes is not the sensation/experience, but to some encoded version of the sensation, or to some kind of implicit response, not yet verbalized. If a person, when rating, is evaluating some memory trace of the experience, it is possible that this memory, for example, could be altered.

9.3 Range and Frequency Effects Two of the most common factors that can affect ratings are the sensory range of the products to be evaluated and the frequency with which people use the available response options. These factors were nicely integrated into a theory that helped to explain shifts in category ratings. They are also general tendencies that can affect just about any ratings or responses.

9.3.1 A Little More Theory: Parducci’s Range and Frequency Principles Parducci (1965, 1974) sought to go beyond Helson’s (1964) simple idea that people respond to the mean or the average of their sensory experiences in determining the frame of reference for judgment. Instead, they asserted that the entire distribution of items in a psychophysical experiment would influence the judgments of a particular stimulus. If this distribution was denser (bunched up) at the low ends and a lot of weak items were psented, product ratings would shift up. Parducci (1965, 1974) proposed that behavior in a rating task was a compromise between two principles. The first was the range principle. Subjects use the categories to sub-pide the available scale range and will tend to pide the scale into equal perceptual segments. The second was the frequency principle. Over many judgments, people like to use the categories an equal number of times (Parducci, 1974). Thus it is not only

9

Context Effects and Biases in Sensory Judgment

the average level that is important but also how stimuli may be grouped or spaced along the continuum that would determine how the response scale was used. Category scaling behavior could be pdicted as a compromise between the effects of the range and frequency principles (Parducci and Perrett, 1971).

9.3.2 Range Effects The range effect has been known for some time, both in category ratings and other judgments including ratio scaling (Engen and Levy, 1958; Teghtsoonian and Teghtsoonian, 1978). When expanding or shrinking the overall range of products, subjects will map their experiences onto the available categories (Poulton, 1989). Thus short ranges produce steep psychophysical functions and wide ranges produce flatter functions. An example of this can be seen in two published experiments on rating scales (Lawless and Malone, 1986a, b). In these studies, four types of response scales and a number of visual, tactile, and olfactory continua were used to compare the abilities of consumers to use the different scales to differentiate products. In the first study, the consumers had no trouble in differentiating the products and so in the second study, the stimuli were spaced more closely on the physical continua so that the task would be more challenging. However, when the experimenters closed the stimuli in, the range principle took over, and participants used more of the rating scale than expected. In Fig. 9.6, ratings for perceived thickness of (stirred) silicone samples are shown in the wide and narrow stimulus ranges. Note the steepening of the response function. For the same one log unit change in physical viscosity, the range of responses actually doubled from the wide range to the narrower range. Another kind of stimulus range effect occurs with anchor stimuli. Sarris (1967) pviously showed a strong effect of anchor stimuli on the use of rating scales, unless the anchors were very extreme, at which point their influence would tend to diminish, as if they had become irrelevant to the judgmental frame of reference. Sarris and Parducci (1978) found similar effects of both single and multiple end anchors that generally take the form of a contrast effect. For example, a low anchor stimulus, whether rated or unrated, will cause stronger stimuli to receive higher ratings than

9.3 Range and Frequency Effects

211

Fig. 9.6 A simple range effect. When products are psented over a wide range, a shallow psychophysical function is found. Over a narrow range, a steeper psychophysical function will be observed. This is in part due to the tendency of subjects to map the products (once known) onto the available scale range. From Lawless and Malone (1986b), with permission.

they would if no anchor were psented, unless the “anchor” is so extreme as to seem irrelevant. Sarris and Parducci (1978) provided the following analogy: A salesman will consider his commissions to be larger if coworkers receive less than he does. However, he is unlikely to extend or shift his scale of judgment by hearing about others who are in an entirely different bracket of income (1978, p. 39). Whether an outlying product is similar enough to have an influence (or whether it is a “horse of a different color”) should be of concern in product comparisons of perse items.

9.3.3 Frequency Effects The frequency effect is the tendency of people to try to use the available response options about the same number of times across a series of products or stimuli to be rated. The frequency effect can cause shifts that look like simple contrast and also a local steepening of the psychophysical function around points where stimuli were closely spaced or very numerous (compared with less “dense” portions of the stimulus range). The frequency principle dictates that when judging many samples, products that are numerous or bunched at the low or high ends of the distributions tend to be spad out into neighboring categories. This is illustrated in the two panels of Fig. 9.7. The upper panel shows four hypothetical experiments and how products might be bunched in different parts of the range. In the upper left panel, we see how a normal replicated psychophysical experiment would be conducted with

equal psentations of each stimulus level. The common outcome of such a study using category ratings would be a simple linear function of the log of stimulus intensity. However, if the stimulus psentations were more frequent at the high end of the distribution, i.e., negative skew, the upper categories would be overused, and subjects would begin to distribute their judgments into lower categories. If the samples were bunched at the lower end, the lower response categories would be overused and subjects would begin to move into higher categories. If the stimuli were bunched in the midrange, the adjacent categories would be used to take on some of the middle stimuli, pushing extreme stimuli into the ends of the response range, as shown in the panel for a quasi-normal distribution. Such behavior is relevant to applied testing situations. For example, in rating the psence of off-flavors or taints, there may be very few examples of items with high values on the scale and lots of weak (or zero) sensations. The frequency effect may explain why the low end of the scale is used less often than anticipated, and higher mean values are obtained than one would deem appropriate. Another example is screening a number of flavor or fragrance candidates for a new product. A large number of good candidates are sent for testing by suppliers or a flavor development group. Presumably these have been p-tested or at least have received a favorable opinion from a flavorist or a perfumer. Why do they then get only mediocre ratings from the test panel? The high end of the distribution is overrepsented (justifiably so and perhaps on purpose), so the tendency for the panel is to drop into lower categories. This may partly explain why in-house testing

212

9

Number of Stimuli Presented

Rectangular Distribution

Quasinormal

Positive Skew

Negative Skew

Predicted Rating

number of zones or categories. The effect of grouping or spacing products also intensifies as the exposure to the distributions increases. Lawless (1983) showed that the shift that occurred with a negative skew (bunching at the upper end) into lower response categories would intensify as the exposure to the skewed distribution went from none to a single exposure to three exposures. Thus the contextual effects do not suddenly appear but will take hold of the subjects’ behavior as they gain experience with the sample set.

9.4 Biases

Rectangular Distribution

Positive Skew

Quasi-normal Distribution

Context Effects and Biases in Sensory Judgment

Negative Skew

Stimulus Intensity (log scale)

Fig. 9.7 Predictions from the Parducci range-frequency theory. Distributions of stimuli that are concentrated at one part of the perceptual range (upper quartet) will show local steepening of the psychophysical functions (lower quartet). This is due to subjects’ tendencies to use categories with equal frequency, the resulting shifting into adjacent categories from those that are overused.

panels are sometimes more critical or negative than consumers when evaluating the same items. Although the great majority of experiments on the range and frequency effects have been performed with simple visual stimuli, there are also examples from taste evaluation (Lee et al., 2001; Riskey et al., 1979; Riskey, 1982; Schifferstein and Frijters, 1992; Vollmecke, 1987). Schifferstein and Frijters found similar effects of skewed distributions with line-marking responses as seen in pvious studies with category ratings. Perhaps line marking is not a response scale with infinite pisions, but panelists sub-pide the line into discrete sub-ranges as if they were using a limited

9.4.1 Idiosyncratic Scale Usage and Number Bias People appear to have pferred ranges or numbers on the response scale that they feel comfortable using. Giovanni and Pangborn (1983) noted that people using magnitude estimation very often used numbers that were multiples of 2 and 5 (or obviously 10), an effect that is well known in the psychophysical literature (Baird and Noma, 1978). With magnitude estimation, the idiosyncratic usage of a favorite range of numbers causes a correlation of the power function exponents across different sensory continua for an inpidual (Jones and Marcus, 1961; Jones and Woskow, 1966). This correlation can be explained if people are more or less expansive (versus restrictive) in their response output functions, i.e., in how they apply numbers to their sensations in magnitude estimation studies. Another version of such personal idiosyncrasy is the common observation in time-intensity scaling that people show a kind of personal “signature” or a characteristic curve shape (Dijksterhuis, 1993; McGowan and Lee, 2006). Another version of self-induced response restriction can be seen when people use only selected portions of the scale in a line-marking rating task. On a line scale with verbal labels, people may choose to make markings only near the verbal labels, rather than distributing them across the response scale. This was first observed by Eng (1948) with a simple hedonic line scale labeled Like Very Highly at one end, Dislike Very Highly at the other, and Neither Like nor Dislike at the center. In a group of 40 consumers, 24 used only the three labeled parts of the scale, and Eng deleted them

9.4 Biases

from the data analysis! This kind of behavior was also noted with the labeled affective magnitude scale (LAM scale) by Cardello et al. (2008) with both Army laboratory and student groups. Lawless et al. (2010a) found a very high frequency (sometimes above 80%) of people making marks within ±2 mm of a phrase mark on the LAM scale in a multi-city consumer central location test. Lawless et al. (2010b) found that instructions did not seem to change this behavior much but that expanding the physical size of the scale on the ballot (from about 120 to 200 mm) decreased the “categorical” behavior somewhat. Categorical rating behavior can also be seen as a step function in time-intensity records (rather than a smooth continuous curve). Finding product differences against the background of these inpidual response tendencies can be facilitated by within-subject experimental designs. Each participant is used as his or her own baseline in comparisons of products, as in dependent t-tests or repeated measure analysis of variance in complete block designs. Another approach is to compute a difference score for a comparison of products in each inpidual’s data, rather than merely averaging across people and looking at differences between mean values.

9.4.2 Poulton’s Classifications Poulton (1989) published extensively on biases in ratings and classified them. Biases in Poulton’s system go beyond Parducci’s theory but are documented in the psychophysical literature. These include centering biases, contraction biases, logarithmic response bias with numerical ratings, and a general transfer bias that is seen when subjects carry the context from a pvious session or study into a new experiment. The centering bias is especially relevant to just-right scales and is discussed in a later section. The response range bias is also a special case and follows this section. The contraction biases are all forms of assimilation, the opposite of contrast. According to Poulton, people may rate a stimulus relative to a reference or a mean value that they hold in memory for similar types of sensory events. They tend to judge new items as being close (perhaps too close) to this reference value, causing underestimation of high values and overestimation of low values. There may also be

213

overestimation of an item when it follows a stronger standard stimulus or underestimation when it follows a weaker standard stimulus, a sort of local contraction effect. Poulton also classifies the tendency to gravitate toward the middle of the response range as a type of contraction effect, called a response contraction bias. While all of these effects undoubtedly occur, the question arises as to whether contrast or assimilation is a more common and potent process in human sensory judgment. While some evidence for response assimilation has been found in psychophysical experiments through sequential analysis of response correlations (Ward, 1979), contrast seems much more to be the rule with taste stimuli (Schifferstein and Frijters, 1992) and foods (Kamenetzky, 1959). In our experience, assimilation effects are not as pvalent as contrast effects, although assimilation has certainly been observed in experiments on consumer expectation (e.g., Cardello and Sawyer, 1992). In that case, the assimilation is not toward other actual stimuli but toward expected levels. The logarithmic response bias can be observed with open-ended response scales that use numbers, such as magnitude estimation. There are several ways to view this type of bias. Suppose that a series of stimuli have been arranged in increasing magnitude and they are spaced in subjectively equal steps. As the intensity increases, subjects change their strategy as they cross into ranges of numerical responses where there are more digits. For example, they might be rating the series using numbers like 2, 4, 6, and 8, but then when they get to 10, they will continue by larger steps, perhaps 20, 40, 60, 80. In Poulton’s view they proceed through the larger numerical responses “too rapidly.” A converse way of looking at this problem is that the perceived magnitude of the higher numbers is in smaller arithmetic steps as numbers get larger. For example, the difference between one and two seems much larger compared to the difference between 91 and 92. Poulton also points out that in addition to contraction of stimulus magnitude at very high levels, the converse is also operating and that people seem to illogically expand their subjective number range when using responses smaller than the number 3. One obvious way to avoid the problems in number bias is to avoid numbers altogether or to substitute line scaling or cross-modality matching to line length as a response instead of numerical rating techniques like magnitude estimation (Poulton, 1989).

214

Transfer bias refers to the general tendency to use pvious experimental situations and remembered judgments to calibrate oneself for later tasks. It may involve any of the biases in Poulton’s or Parducci’s theories. The situation is common when subjects are used in multiple experiments or when sensory panelists are used repeatedly in evaluations (Ward, 1987). People have memories and a desire to be internally consistent. Thus the ratings given to a product on one occasion may be influenced by ratings given to similar products on pvious occasions. There are two ways to view this tendency. One is that the judgments may not shift appropriately when the panelists’ sensory experience, perception, or opinion of the product has in fact changed. On the other hand, one of the primary functions of panelist training and calibration in descriptive analysis is to build up exactly those sorts of memory references that may stabilize sensory judgments. So there is a positive light to this tendency as well. An open question for sensory evaluation is whether exposure to one continuum of sensory intensities or one type of product will transfer contextual effects to another sensory attribute or a related set of products (Murphy, 1982; Parducci et al., 1976; Rankin and Marks, 1991). And if so, how far does the transfer extend?

9.4.3 Response Range Effects One of Poulton’s biases was called the “response range equalizing bias” in which the stimulus range is held constant but the response range changes and so do the ratings. Ratings expand or contract so that the entire range is used (minus any end-category avoidance). This is consistent with the “mapping” idea mentioned for stimulus range effects (stimuli are mapped onto the available response range). Range stabilization is implicit in the way some scaling studies have been set up and in the instructions given to subjects. This is similar to the use of physical reference standards in some descriptive analysis training (Muñoz and Civille, 1998) and is related to Sarris’s work on anchor stimuli (Sarris and Parducci, 1978). In Anderson’s work with 20-point category scales and line marking, high and low examples or end anchors are given to subjects to show them the likely range of the stimuli to be encountered. The range of responses is known since

9

Context Effects and Biases in Sensory Judgment

it is visible upon the page of the response sheet or has been p-familiarized in a practice session (Anderson, 1974). Thus it is not surprising that subjects distribute their responses across the range in a nicely graded fashion, giving the appearance that there is a reasonably linear use of the scale. Anderson noted that there are end effects that work against the use of the entire range (i.e., people tend to avoid using the endpoints) but that these can be avoided by indenting the response marks for the stimulus end anchors, for example, at points 4 and 16 on the 20-point category scale. This will provide psychological insulation against the end effects by providing a comfort zone for unexpected or extreme stimuli at the ends of the scale while leaving sufficient gradations and room to move within the interior points. The “comfort zone” idea is one reason why early workers in descriptive analysis used line scales with indented vertical marks under the anchor phrases. An exception to the response range mapping rule is seen when anchor phrases or words on a scale are noted and taken seriously by participants. An example is in Green’s work on the labeled magnitude scale, which showed a smaller response range when it was anchored to “greatest imaginable sensation” that included all oral sensations including pain, as opposed to a wider range when the greatest imaginable referred only to taste (Green et al., 1996). This also looks like an example of contrast in which the high-end anchor can evoke a kind of stimulus context, at least in the participant’s mind. If the image evoked by the high-end phrase is very extreme, it acts like a kind of stimulus that compsses ratings into a smaller range of the scale. A similar kind of response compssion was seen with the LAM scale when it was anchored to greatest imaginable liking for “sensations of any kind” as opposed to a more delimited frame such as “foods and beverages” (Cardello et al., 2008). A sensory scientist should consider how the high anchor phrase is interpted, especially if he or she wants to avoid any compssion of ratings along the response range. As Muñoz and Civille (1998) pointed out, the use of a descriptive analysis scale also depends a lot on the conceptualization of the high extreme. Does “extremely strong” refer to the strongest possible taste among all sensations and products, the strongest sensation in this product type, or just how strong this particular attribute can become in this particular product? The strongest sweetness in this product might be more intense than the strongest saltiness. The definition needs to be a

9.4 Biases

215

deliberate choice of the panel leader and an explicit instruction to the panelist to give them a uniform frame of reference.

9.4.4 The Centering Bias The centering bias arises when subjects become aware of the general level of stimulus intensity they are likely to encounter in an experiment and tend to match the center or midpoint of the stimulus range with the midpoint of the response scale. Poulton (1989) distinguished a stimulus centering bias from a response centering bias, but this distinction is primarily a function of how experiments are set up. In both cases, people tend to map the middle of the stimulus range onto the middle of the response range and otherwise ignore the anchoring implications of the verbal labels on the response scale. Note that the centering bias works against the notion that respondents can use unbalanced scales with any effectiveness. For example, the “Excellent-very good-good-fair-poor” scale commonly used in marketing research with consumers is unbalanced. The problem with unbalanced scales is that over many trials, the respondents will come to center their responses on the middle category, regardless of its verbal label. The centering bias is an important problem when there is a need to interpolate some value on a psychophysical function or to find an optimal product in just-right scaling. Poulton gives the example of McBride’s method for considering bias in the justabout-right (JAR) scale (McBride, 1982; see also Johnson and Vickers, 1987). In any series of products to be tested, say for just-right level of sweetness, there is a tendency to center the series so that the middle product will come out closest to the just-right point. The function shifts depending upon the range that is tested. One way to find the true just-right point would be to actually have the experimental series centered on that value, but then of course you would not need to do the experiment. McBride gives a method for interpolation across several experiments with different ranges. The point at which the just-right function and the median of the stimulus series will cross shows the unbiased or true just-right level. This method of interpolation is shown in Fig. 9.8. In this method, you

Fig. 9.8 Adjusting for the centering bias in just-right ratings. Three series of sucrose concentrations in lemonade were tested, a low series (2-8%), a middle series (6-14%), and a high range (10-22%). In the upper panel, the method of Poulton is used to interpolate the unbiased just-right point from the series where the midpoint concentration would correspond to the just-right item. In the lower panel, the method of McBride is used to interpolate the just-right point from a series in which the average response would correspond to the just-right point. When the average response would be just right (zero on this scale), the hypothetical stimulus range would have been centered on the just-right level. Replotted from Johnson and Vickers (1987), with permission.

psent several ranges of the products in separate sessions and plot how the judgments of the JAR point shift up and down. You can then interpolate to find the range in which the just-right point would have been from the center product in the series. This obviously takes more work to do the test a couple of times, but it could avoid a mistaken estimate of the JAR level.

216

9.5 Response Correlation and Response Restriction

9

9.5.1 Response Correlation A simple example of a halo effect is shown in Fig. 9.9. In this case, a small amount of vanilla extract was added to low-fat milk, near the threshold of perceptibility. Ratings were then collected from 19 milk consumers for sweetness, thickness, creaminess and liking for the spiked sample, and for a control milk. In spite of the lack of relationship between vanilla aroma and sweet taste and between vanilla and texture characteristics, the introduction of this one positive aspect was sufficient to cause apparent enhancement in sweetness, creaminess, and thickness ratings. Apparent enhancement of sweetness is an effect long known for ethyl maltol, a caramelization product that has an odor similar to heated sugar (Bingham

ADDED VANILLA CONTROL MILK

13

MEAN RATING + 1 S.E.M.

11

Early experimental psychologists like Thorndike (1920) noted that one very positive attribute of a person could influence judgments on other, seemingly unrelated characteristics of that inpidual. In personnel evaluations of military officers, Thorndike noted a moderate positive correlation among the inpidual rated factors. People evaluate others like this in real life. If achievement in sports is influential in our assessment of a person, we might suppose a gifted athlete to also be kind to children, generous to charities, etc., even though there is no logical relationship between these characteristics. People like to have cognitive structures that form consistent wholes and are without conflicts or contradictions (called cognitive dissonance) that can make us uncomfortable. The halo effect has also been described as a carry-over from one positive product to another (Amerine et al., 1965), but its common usage is in reference to a positive correlation of unrelated attributes (Clark and Lawless, 1994). Of course, there can also be negative or horns effects, in which one salient negative attribute causes other, unrelated attributes to be viewed or rated negatively. If a product makes a mess in the microwave, it might be rated negatively for flavor, appearance, and texture as well.

Context Effects and Biases in Sensory Judgment

*Different from

* *

9

control, paired t-test N = 19.

*

7

5

3

1

SWEET

CREAMY THICK SCALED ATTRIBUTE

LIKE

Fig. 9.9 Adding a just perceivable level of vanilla extract to low-fat milk causes increases in rated sweetness, thickness, creaminess, and liking, an example of the Halo effect. From Lawless and Clark (1994), with permission.

et al., 1990). When maltol is added to various products, sweetness ratings may rise compared to products lacking this flavor. However, the effect seems to be a case of the misattribution of olfactory stimulation to the taste sense. Murphy and Cain (1980) showed that citral (a lemon odor) could enhance taste ratings, but only when the nostrils were open, which allows diffusion of the odor into the nose and stimulation of the olfactory receptors (i.e., retronasal smell). When the nostrils are pinched shut, the diffusion is effectively eliminated and the enhancement disappears. Murphy and Cain interpted this as convincing evidence that there was no true enhancement of taste intensity by citral, but only olfactory referral, a kind of confusion between taste and smell. Studies with other odors have also shown that the taste enhancement effect from volatile flavors can be eliminated by nose pinching (Frank and Byram, 1988) even for maltol (Bingham et al., 1990). The maltol effect is also minimized by training subjects who then learn to more effectively separate or localize their odor experiences from taste (Bingham et al., 1990). The sweetness enhancement may arise as a function of conditioning or experience with the pairing of sweet tastes with some odors in foods (Stevenson et al., 1995). Several lessons can be learned from the vanilla halo effect shown in Fig. 9.9. First, untrained consumers

9.5 Response Correlation and Response Restriction

9.5.2 “Dumping” Effects: Inflation Due to Response Restriction in Profiling It is part of the folklore of consumer testing that if there is one very negative and salient attribute of a product, it will influence other attributes in a negative direction, an example of a horns effect. The effect is even worse when the salient negative attribute is omitted from the questionnaire. Omission could be due to some oversight or failure to anticipate the outcome in a consumer test or simply that it was not observed in the laboratory conditions of pliminary phases of testing. In this case, consumers will find a way to dump their frustration from not being able to report their dissatisfaction by giving negative ratings on other scales or reporting negative opinions of other even unrelated attributes. In other words, restricting responses or failure to ask a relevant question may change ratings on a number of other scales. A common version of this restriction effect can be seen in sweetness enhancement. Frank et al. (1993) found that the enhancement of sweet ratings in the psence of a fruity odor was stronger when ratings were restricted to sweetness only. When

217

both sweetness and fruitiness ratings were allowed, no enhancement of sweetness was observed. Exactly the same effect was seen for sweetness and fruitiness ratings and for sweetness and vanilla ratings (Clark and Lawless, 1994). So allowing the appropriate number of attributes can address the problem of illusory enhancement. Schifferstein (1996) gave the example of hexenol, a fresh green aroma, which when added to a strawberry flavor mixture caused mean ratings in several other scales to increase. The enhancement of the other ratings occurred only when the “green” attribute was omitted from the ballot. When the “green” attribute was included in the ballot, the response was correctly assigned to that scale, and there was no apparent enhancement in the other attributes in the aroma profile. There is good news and bad news in these observations. From a marketing perspective, ratings can be easily obtained from consumers that will show apparent sweetness enhancements if the questionnaires cleverly omit the opportunity to report on sensations other than sweetness. However, the nose pinch conditions and the use of complete sets of attributes show us that these volatile odorants such as maltol are not sweet taste enhancers but they are sweet rating enhancers. That is, they are not affecting the actual perception of sweet taste intensity but are changing the response output function or perhaps broadening the concept of sweetness to go beyond taste and include pleasant aromas as well. It would not be wise to try to use maltol to sweeten your coffee. Are there other situations in sensory testing where the dumping effect can show up? One area in which responses are usually restricted to one attribute at a time is in time-intensity scaling (Chapter 8). In a common version of this technique, the subject moves a pointer, a mouse, or some other response device to provide a continuous record of sensory intensity for a specified attribute. Usually, just one attribute at a time is rated since it is very difficult to attend continuously or even by rapid shifting of attention to more than one attribute. This would seem to be a perfect opportunity for the dumping tendency to produce illusory enhancements (e.g., Bonnans and Noble, 1993). This idea was tested in experiments with repeated category ratings across time, a time-intensity procedure that allows for ratings of multiple attributes. These studies showed sweetness enhancement in sweet-fruity mixtures when sweetness alone was rated, but little or

218

no enhancement when both sweetness and fruit intensity were rated over time (Clark and Lawless, 1994). This is exactly parallel to the sweetness enhancement results seen by Frank and colleagues. Workers using single-attribute time-intensity ratings should be wary of apparent enhancements due to response restriction.

9.5.3 Over-Partitioning In the data of van der Klaauw and Frank (1996), one can also see cases in which having too many attributes causes a deflation in ratings. As in the dumping examples, their usual paradigm was to compare the sweetness ratings of a simple sucrose solution to the same concentration with a fruity odor added. When rating sweetness only, the rating is higher than when rating sweetness and fruitiness, the common dumping effect. But when total intensity and six additional attributes were rated, the sweetness rating was significantly lower than either of the other two conditions. In another example, including a bitterness rating (in addition to the sweetness and fruitiness ratings) lowered the sweetness rating compared to rating sweetness (the highest condition) and also compared to rating sweetness and fruitiness (an intermediate sweetness rating was obtained). This effect appears to be a deflation due to people over-partitioning their sensations into too many categories. The specific choices may be important in this effect. Adding only a bitter or a bitter and a floral rating had little or no effect and dumping inflation was still observed probably because there was no fruity rating. In the study by Clark and Lawless (1994), the control condition (sweetener only) showed some evidence of a decrement when the attributes for volatiles were also available. Even more dramatic was the complete reversal of sweet enhancement to inhibition when a large number of response categories were provided for simple mixtures (Frank et al., 1993). Although this effect has not been thoroughly studied, it serves to warn the sensory scientist that the number of choices given to untrained consumers may affect the outcome and that too many choices may be as dangerous as too few. Whether this effect might be seen with trained panels remains an open question. It is sometimes difficult to pdetermine the correct number of attributes to rate in order to guard against

9

Context Effects and Biases in Sensory Judgment

the dumping effect. Careful p-testing and discussion phases in descriptive training may help. It is obviously important to be inclusive and exhaustive, but also not to waste the panelists’ time with irrelevant attributes.

9.6 Classical Psychological Errors and Other Biases A number of psychological errors in judgment have been described in the literature and are commonly listed in references in sensory evaluation (e.g., Amerine et al., 1965; Meilgaard et al., 2006). They are only briefly listed here, as they serve primarily as empirical descriptions of behavior, without much reference to cause or any theoretical bases. It is important to distinguish between description and explanation, and not to confuse naming something with trying to explain why it occurred in terms of mechanism or larger theory. The sensory evaluation practitioner needs to be aware of these errors and the conditions under which they may occur.

9.6.1 Errors in Structured Sequences: Anticipation and Habituation Two errors may be seen when a non-random sequence of products is psented for evaluation, and the observer is aware that a sequence or a particular order of items is going to be psented. The error of anticipation is said to occur when the subject shifts responses in the sequence before the sensory information would indicate that it is appropriate to do so (Mattes and Lawless, 1985). An example is found in the method of limits for thresholds, where an ascending sequence is psented and the observer expects a sensation to occur at some point and “jumps the gun.” The opposite effect is said to be the error of habituation, in which the panelist stays put too long with one pvious response, when the sensory information would indicate that a change is overdue. Obviously, the psentation of samples in random order will help to undo the expectations involved in causing the error of anticipation. Perseveration is a little bit harder to understand but may have to do with lack of attention or motivation

9.7 Antidotes

on the part of the observer or having an unusually strict criterion for changing responses. Attention and motivation can be addressed by sufficient incentives and keeping the test session from being too long.

219

The stimulus error is another classical problem in sensory measurement. This occurs when the observer knows or psumes to know the identity of the stimulus and thus draws some inference about what it should taste, smell, or look like. The judgment is biased due to expectations about stimulus identity. In the old parlor game of trying to identify the origin and vintage of a wine, stimulus error is actually a big help. It is much easier to guess the wine if you know what the host is prone to drink or you have taken a peek at the bottles in the kitchen beforehand. In sensory evaluation, the principle of blind testing and the practice using random three-digit coding of samples mitigate against the stimulus error. However, panelists are not always completely in the dark about the origin or the identity of samples. Employee panels may have a fair amount of knowledge about what is being tested and they may make inferences, correctly or incorrectly. For example, in small-plant quality assurance, workers may be aware of what types of products are being manufactured that day and these same workers may serve as sensory panelists. In the worst possible scenario, the persons drawing the samples from production are actually doing the tasting. As much as possible, these situations should be avoided. In quality control panels, insertion of blind control samples (both positive controls and flawed samples) will tend to minimize the guesswork by panelists.

the sequential effects may be counterbalanced or averaged out in the group data. The second approach is to consider the order effects of interest. In this case, different orders are analyzed as a purpose of the experiment and, if order effects are observed, they are duly noted and discussed. Whether order effects are of interest will depend upon the circumstances of the product evaluation and its goals. If counterbalanced orders or randomization cannot be worked into the experimental design, the experimenter must consider whether product differences are true sensory differences or are artifacts of stimulus order. Purposeful experimentation and analysis of order effects in at least some of the routine evaluations may give one an appciation for where and when these effects occur in the product category of interest. Another well-known order effect in acceptance testing is the reception of a higher score for the first sample in a series (Kofes et al., 2009). Counterbalancing orders is of course appropriate, but one can also give a “dummy” product first to absorb the first product’s s


【#9】Luật Chơi Ma Sói One Night

Published on

Link chi tiết: http://boardgame.vn/ma-soi/ma-soi-one-night-262

Ma sói nay được tái hiện lại trong phiên bản “tình một đêm” với chỉ một đêm duy nhất! Ma sói One Night đặc biệt phù hợp với những fan Ma Sói nhưng có ít người chơi và không muốn bị chờ đợi do loại trừ. Trong Ma Sói One Night, cách chơi vẫn giữ tính chất suy luận, hack não, phán đoán và đánh lừa tâm lý, nhưng ở mức độ cao và căng thẳng hơn. Chỉ một ngày và một đêm, bạn phải chọn để treo cổ ai, nếu giết trúng Sói phe người thắng, ngược lại, nếu treo phải dân vô tội thì Sói thắng.

Nhiều nhân vật đặt biệt với chức năng khác nhau thêm vào thú vị của cách chơi. Bạn có thể thử combo những nhân vật này để tạo ra các ván chơi khác biệt. Không một ván nào giống ván nào. Nhanh, không loại trừ người chơi, kịch tính và cao trào, đó chính là tính chất của One Night – một đêm, một mạng!

  1. 1. T’ 41% I , _ 7 3A’I’zoGAr: sv§. :
  2. 3. K ca that gtan thuc hten hanh dong vao ban B u N dem. Qay la k_tch ban ma Nguat ‘Ihong Quac chudn b_t thay d6t my vao luang gag sg dang xuyen sugt dem; nguat chat (3-10). @6t vat Ian dazu, khz chat t6t nhdt hay chat ta 3-5 nguat de’ htéu ro Mot nguat at nga dt! cach chat ( J’/ ‘éu ca nhtéu nguat, cang chdng “fat Cd deu nhdm mdt lat (kg Cd sao daul. Trb chat dtén ra rdt nhanh va at Jx/ ‘gum Thong gdo) Egg co the’ thodt mdt ngbt nhtn ban chat Sm at day am ban dbng hank dt! Sat ma mdt day va ttm xem nguat 4 Neu-Q1 cu nao dang ma mat chung vat mlnh glflaual CH0! (co the’ se khong at ca) 1 Jcé gm R6, +1 gm, (Mg +2 90,, Lang Sat at, nhdm mdt lat! Tzen Trt at, day dt. Qan duac xem 1 la bat caa nguat 1 mn “tang kh h he’ 1 b ¢’hdt dam bdo rang s6 bat luon nhtéu han 81:: ode C6 t xem 2 mmg 3 d at 6 s6 nguat chat dung 3 la. J{do nhang la bat up va chta chung cho mat nguat. ©at 3 la ‘Mgum co Id bat Tie” T” ma mdt m va lam theo hteu lenh can lat va tdt ca thé tuang ling vat s6 nhan Chen T” at, nhdm mdt lat! gag ‘R16 01, vat duac chan vao chtnh gtaa. Mot nguat bt mat xem bat caa chtnh mtnh, ‘M9 d1’ 94″ co thé hodn ch”-ye, ” bat minh vat nguat khdc, va xem Id bat va dat chlzng 11p xuang canh 3 la a gttza 3 Nolrbl cm): 2.901 1 ‘Ben Tn/1 Qao Tao sao cho tdt ca déu dé dang cham dén tat ca ban mat dd’ duo” bat tren ban, nhu Vida danh cho 5 nguat a Nguot ea In bat @410 Tao mo mdt duat day; ra va lam theo hteu lenh Qao ‘Tac at, nhdm mdt lat. JCé Qhd Rat at, day (it. flan co thé trdo dd: bat caa 2 nguat bdt kt ma khong duac nhln chung J’/ ‘gab: co la bat JCé Qha R61 ma mdt ra va lam theo hteu lenh JCé Qha R6: at, ng12 dz Nguat ‘Ihong §Bdo , vdn nhdm mdt, dat cac ld bat xung quanh dé chdc chdn khong ca bat kt at co the’ khdng d_tnh rdng la bat dd duac chuyén dt / b_t xem bat vt no dd khong can ndm 6 v_t tri ban dau mza Mot nguat at, day dt! §Bu6t sang bdt ddu, mat nguat co quyén bten luan C6 rat nhtéu nhan vat thlzc gtdc dé thuc hten chtlc ndng vao ban dem. Tug nhten vdn ca nhang nguat chtm trong gtdc ng11,la ©an Lang, 1719 San va JCé Chan 901 JV at ra, cbn ca 1 nguat duac chi dtnh la .1’/ gm: ‘Ihong Qao va keu ten mat nhan vat theo thL’t tu va dém dén 10 dé mot nhan vat Trong dem, nguat khong co phan su khong duac phép ma mat hay lam bat kt hanh dong gt de’ ra hteu cho nguat dang ca luat
  3. 5. .l. It :5 Lm? [email protected] Co the ba tr! 1 gtat aau nho bang cach cap cho nhang nguot thang cuoc 1 la bat chtp hay la 1 man qua nho xtnh. JCht co at do nhan auoc so luong qua nhat atnh, chang han nhu la 5, nguat co so qua nhteu nhat se chten thang Qat vat so nguat chot dong hoac nhteu Ian chat, hay ap dang luat tht aau flat mang, me la duoc phép chot cho den kht ho bt thua 2 Ian %fl? HHW’TWTTNfi Qan nen sa dung vat nhan vat b6 sung trong tra chot ngoat nhang luat co ban. Chi vtec lay la bat ra va thay no bang la khac. flan hau nhu co the phat hop mot nhan vat vat nhau trong 1 van, tra kht dang gtat thteu cho nguat mat va khong muan chat cang lac qua nhteu nhan vat Du sao at naa, thi chenh lech gtaa so bat va so nguat chat buoc phat M 3 ‘Truac mot vang, at 1 the tuong ang vat mat la bat nam gan JV/ ‘guat ‘Thong Qao. J/ hu vay, mot nguat se btet nhan vat nao co trong tra chat, va Nguat ‘Thong Baa se btet tha tu ae got trong gtat doan Qan Qem. flay sang tao cach chon nhan vat caa ban: Tha chot vat 1 Sat, hoac tron aeu tat ca nhang la bat va chta ngau nhten! flan co the tao ra ngot lang khong co bat ki Dan (ang- hoac Ma Sat. . Va at at cang deu co the co chac nang ca! twfiHwfrr, ,. FLHFHL JVhtIng nhan vat nay se thac day theo tha ta nhat dtnh (Cuat trang 1 ca Danh sach tom tat, de nhin de doc hon! ). Cung vat mot nhan vat duoc mo ta can co cau mau danh cho J/ ‘guat ‘Ihong Qao: J, *1″‘II”! *t’ ‘A ‘I .1′ . .’ r. ..’t; “.o rm: -_L~”t: -: eICeMao¢)anhlalabatcucki 7 phac tap, bot vi Q’he caa co nang pha thuoc vao la bat auoc xem. J/ ‘e’u ban chua quen vat nhang nhan vat co ban kta, hay tam quen co nang rac rat nay at! J/ hung sau kht aa nam bat duoc tra chot thi day se la nhan vat cac ki tha vt JCe Mao SDanh luon thuc day truac tat ca nhang nhan vat khac. Vao ban aem, JCe Mao Qanh nhin (nhung khong auoc trao aat) 1 la bat caa nguat khac va lam theo nhung ateu sau, neu la bat em tha la: ‘ n N LANG. RE cl-IAN oat. mg sAN ém gto day la nhan vat do va khong lam bat ki ateu gl nua vao ban dem. sol. MASON Sm se thac day cang nhang Sat hoac Mason kht auoc got, va se theo phe Sat neu la co ta xem @ Sat, neu la Mason thi, nguoc lat NHA TIEN TRI. DAD T§C. RE PHA ROI. n’aN sou NI-Itlu Em se thac hten chac nang caa nhan vat do ngay lap tac ( co ta khong auoc thac day Ian ntza kht flang chtnh hteu auoc got). RE PHAN ac}: Sau kht ket thac gtat doan caa JCe Mao Danh, Nguat ‘Thong Qao se keu co nang nhdm mat lat tra kht em la JCe Qhan Got, lac nay Ma Sat se gto ngon my cat len. JCe Mao Danh lac na theo phe Sat cl’: 9 M Sau kht Ca Gem nham mat lat, Mao Danh Ca Qem se thac day va xem la bat caa minh ca con la JCe Mao Danh khong. JVeu 1 nguat nhan la JCe Mao Danh trong suat aem, co ta se mang vat tra caa nhan vat gac ma co ta thay. JCtch ban danh cho J/ han ban vao ban dem co dot cho khac btet, co kht se auoc keu tim ktem Sot neu em la JCe ? han Qot, va se thac day tre nhat vao ban aem neu Ca Qem da duoc got. Sa dang nhang cau thoat mau sau neu khong co JCe Q’hdn Got trong tra chat:
  4. 6. JCé Mao Danh at, thac day dt, va chon xem 1 la bat nhan vat caa nguot khac. flay gta ban la nhan vat do. J/ ‘an nhan vat do ca chac nang vao ban dem thi thuc hten ngay lap ttzc JCé Mao Qanh ma mat va xem bat 1 nguat khac. J/ ‘éu la Tten Trt, Qao Tac, JCé Hm Rat, hay ‘Ten Qom J/ hau, co nang thac hten chac nang do ltén. .J/ ‘éu la Sat hay Mason, se thac day cang vat nhang nhan vat do kht ho duoc got. ‘Irong truang hap JCé Chan 961 hoac ‘Tho San thi khong can lam gt ca eICéMao Oanh at, nga at Jt/ ‘éu ca JCé €Phdn Bot tht dung ktch ban sau: JCé Mao Qanh at, thuc day dt, va chon xem 1 la bat nhan vat caa nguat khac. Gay gto ban la nhan vat do. J/ Eu nhan vat do ca chuc nang vao ban dem tht thuc hten ngay lap tllc JCéMao ©anh ma mat va xem bat 1 nguoz khac; Néu la ‘I ten T rt, Qao ‘Tac, JCé §’ha Rat, hay ‘Ten §Bam . Nhau, co nang thac hten chac nang do ltén. Jréu la Sat hay Mason, se thac day cung vat nhang nhan vat do kht ho duoc got. ‘Irong truang hop JCé Chan Qat hoac ‘Iho San tht khong can lam gt ca JV/ ‘éu Say gto ban la JCé Qhan Qot, ca ma mdt. éu khong thi nhdm lat. Sat at, gto ngon cat caa ban len dé Mao Qanh – ? hdn @301 btét Sat va JCé Mao Danh – Q’han Q01 cang lam theo hteu lenh Sat at, dat ngon cat xuang dt. Jt/ han flan at, nhdm mdt lat J/ bu ca 1 Cu Qem trong tra chat, sau kht C11 99m duoc got, slz dung ktch ban mdu sau: JCé Mao Danh néu dd xem la Ca Qem, hay thac day va nhin lat bat caa minh at JCéMao Danh se ma mdt ra va nhln lat bat caa minh xem na ca bl trao dat trong suat dem do hay khong. MAS6I ‘ *. Vao ban dem, tat ca Ma Sat ma mat va tlm nhang Ma Sat khac. J/ ‘éu khong ca at mo mat, thl Ma Sat con lat dang a gtaa. Ma Sat sé theo phe Sat. Sat at, thac day va ttm nhang Sat khac dt Sat se ma mdt va nhin xung quanh xem at dang ma mat ( ca kha nang se khong co at hat) Sat at, nhdm mdt lat dt Qtyén lot Gan Sat: J/ ‘éu chi ca duy nhat 1 Sat, Sat duoc xem 1 la bat 0 gttza. Qay la dac quyén rteng danh cho Sat khong co dong dot, va cung cdp cho anh chang nhang cong ca hau tch de’ xac dtnh nhtzng nguat chat con lat HEPHANBQI Thttc day ngay sau Sat vao ban dem, J<Ié ? han SBot thac day va nhin xem at la Sat. Suat gtat doan nay, tat ca Sat gto ngan cat len dé lam ddu hteu nhan btét. Sat khong duoc btét at la ? han §Bot. J/ an ? han Got chat va khong ca Sat nao chat, Sat ( va SPhan Qot) thdng. J/ ‘éu khong ca at la Sat, Q’han $01 sé thdng kht ca 1 nguat khac ( khong phat Q’han Qot) chét. J/ han vat nay ca the’ la 1 dong mtnh rat hau tch cho phe Sat. JCé SPhan @391 theo phe Sat. , n_-.
  5. 7. JCé flhan Got at, day dt. Sat gto ngan cat len dé JCé flhan flot nhan dten JCé flhan flat ma mdt. JCé flhan flat va Ma Sat cung thac hten theo hteu lenh Sat at, ha ngan cat xuang. flhan flat at, nham mat lat dt TI-to H6 ” ” JCht dung ‘Ihoj-lb tht phat – luon at 1 cap trong tro chat. ‘Ihojib thuc day vao buat tat va ktém nguot con lat. .Jféu khong thay, ca nghia ta Tho jib can lat dang nam a gtaa. Tho]-I2) theo phe flan. ‘fhoflb at, day dt, hat Tho]-lb nhin nhau dt Tho]-Ib ma mat ra va thac hien chtzc nang theo hteu lenh (ca thé cht ca dug nhat 1 nguot). ‘Ihoflb at, ngu dt! HEN TRI . L -, flan dem, Tten ‘Trt co the’ nhin “. — . . la bat nhan vat caa nguot ‘ C khac hoac la xem 2 la bat a gtaa, nhung khong duoc dt chuyén chung. ‘Iten ‘T rt theo phe flan. ‘Tten ‘T rt at, day dt. flan ca thé xem la nhan vat caa nguot khac hoac 2 la a gtaa ‘T ten ‘Trt ma mat, thuc hten chvtc nang theo hteu lenh T ten 9′ rt at, nga dt! , ago TAG . flan dem, flao tac co the’ chon .1′ -I ‘ -‘ cuap 1 la bat ta nguot khac va dat nhan vat caa minh the’ vao ‘ la bat kta. Sau do 940 tac , _ nhin vao la bat mat caa minh. Nguat mang la flao tac sé theo phe flan. flan tac se theo phe caa la bat duoc lay, duang nhten, anh chang khong duoc thac hten chzzc nang caa nhan vat mat vao ban dem, J/ ‘éu flao tac chon khong cuap la bat ta nhang nguat khac, anh chang van ta flao tac va theo phe flan 4. -‘o flao tac at, day at. flan co the’ nhin bat 1 nguot bat kt, sau do trao bat minh vat nguat do ¥)ao tac ma mat ra va thuc hten chzlc nang theo hteu lenh flao tac at, ngzz dt! RE GAY R61 ‘ Vao ban dem, JCé Gay Rat ca the’ trao bat caa 2 nguat khac it V _ ‘-, – ma khong duoc nhtn nhang la V ” ” nay. J/ hang nguat nhan la _ _ X i 3 bat khac thl bay gto dd Id nhan vat ( va phe) mat theo la vita dat, ddu cho ho khong he btat nhan vat do Id at cho dén kht tro chat két thac. Jcé Gay Rat theo phe flan flao tac at, day at. flan co the’ nhtn bat 1 nguat bat kt, sau do trao bat minh vat nguot do JCé Gay R61′ ma mat ra va thac hten chac nang theo hteu lenh JCé Gay R6: at, ngu dz!


【#10】The Civil Code Law No. 91/2015/qh13

The Civil Code Law No. 91/2015/QH13

CIVIL CODE

Pursuant to the Constitution of the Socialist Republic of Vietnam; The National Assembly promulgates a Civil Code.

PART ONE. GENERAL PROVISIONS

Chapter I. GENERAL PROVISIONS

Article 1. Scope

The Civil Code provides the legal status, legal standards for the conduct of natural and juridical persons; the rights and obligations of natural and juridical person (hereinafter referred to as persons) regarding personal and property rights and obligations in relations established on the basis of equality, freedom of will, independence of property and self-responsibility (hereinafter referred to as civil relations).

Article 2. Recognition, respect, protection and guarantee of civil rights

1. In the Socialist Republic of Vietnam, all civil rights are recognized, respected, protected and guaranteed under the Constitution and law.

2. Civil rights may be limited as pscribed in law in exceptional circumstances that due to national defense and security, social safety and order, social ethics and the community’s health.

Article 3. Basic principles of civil law

1. Every person shall be equal in civil relations, may not use any reason for unequal treatment to others, and enjoy the same protection policies of law regarding moral rights and economic rights.

2. Each person establishes, exercises/fulfills and terminates his/her civil rights and obligations on the basis of freely and voluntarily entering into commitments and/or agreements. Each commitment or agreement that does not violate regulations of law and is not contrary to social ethics shall be bound by contracting parties and must be respected by other entities.

3. Each person must establish, exercise/ fulfill, or terminate his/her civil rights and/or obligations in the principle of goodwill and honesty.

4. The establishment, exercise and termination of civil rights and/or obligations may not infringe national interests, pubic interests, lawful rights and interests of other persons.

5. Each person shall be liable for his/her failure to fulfill or the incorrect fulfillment of any such civil obligations.

Article 4. Application of the Civil Code

1. This Law is a common law that applies to civil relations.

2. Any relevant law that applies to civil relations in specific fields may not be contrary to the basic principle of civil law pscribed in Article 3 of this Law.

3. If another relevant law has no regulation or has regulations that infringe Clause 2 of this Article, the regulations of this Law shall apply.

4. In cases where an international agreement to which the Socialist Republic of Vietnam is a signatory contains provisions different from the provisions of this Code with regard to a same matter, the provisions of such agreement shall apply.

Article 5. Application of practices

1. Practices mean rules of conduct obvious to define rights and obligations of persons in specific civil relations, forming and repeating in a long time, recognized and applying generally in a region, race, or a community or a field of civil.

2. In cases where it is neither provided for by law nor agreed upon by the parties, practices may apply but they must not contravene the principles provided for in Article 3 of this Code.

Article 6. Application of analogy of law

1. In cases where a issue rises under scope of civil law which it is neither provided for by law nor agreed upon by the parties nor, nor applied by practices, analogy of law shall apply.

2. In cases where it is neither provided for by law nor agreed upon by the parties, practices may apply but they must not contravene the principles provided for in Article 3 of this Code.

Article 7. State policies on civil relations

1. The establishment, performance and termination of civil rights and obligations must ensure the pservation of national identities, respect and promote good customs, practices and traditions, solidarity, mutual affection and cooperation, the principle of every inpidual for the community and the community for every inpidual and the noble ethical values of ethnicities living together on Vietnamese soil.

2. In civil relations, the conciliation between contracting parties in accordance with regulations of law shall be encouraged.

Chapter II. ESTABLISHMENT, EXERCISE AND PROTECTION OF CIVIL RIGHTS

Article 8. Bases for establishment of civil rights

Civil rights shall be established on the following bases:

1. Contracts;

2. Unilateral legal acts;

3. Decisions of courts or other competent state agencies as pscribed;

4. Outcomes of labor, production and business; or creation of subjects of intellectual property rights;

5. Possession of property;

6. Illegal use of assets or illegal gain therefrom;

7. Damage caused by an illegal act;

8. Performance of a task without authorization;

9. Other bases specified by law.

Article 9. Exercise of civil rights

1. Each person shall exercise his/her civil on his/her own will in accordance with Article 3 and Article 10 of this Code.

2. The non-exercise of civil rights does not constitute a basis for termination of those rights, unless otherwise pscribed by law.

Article 10. Limitations on exercise of civil rights

1. Each person may not abuse his/her own civil rights to cause damage to other persons or violate his/her own obligations or for other unlawful purposes.

2. If a person fails to comply with Clause 1 of this Article, a court or a competent agency shall, according to the nature and consequences of the violation, either protect part or the whole of his/her rights, compel him/her to given compensation and other sanctions as pscribed by law.

Article 11. Methods for protecting civil rights

If a person has his/her civil rights violated, he/she may protect them himself/herself as pscribed in this Code, other relevant laws or request competent authorities to:

1. Recognize, respect, protect and guarantee of his/her civil rights;

2. Order the termination of the act of violation;

3. Order a public apology and/or rectification;

4. Order the performance of civil obligations;

5. Order compensation for damage;

6. Cancellation of isolated unlawful decision of competent agencies, organizations or persons;

7. Other requirements specified by law.

Article 12. Self-protection of civil rights

The self-protection of a particular civil right must conform to the nature and severity of the violation against such civil right and be not contrary to basic principles of civil law pscribed in Article 3 of this Code.

Article 13. Compensation for damage

Each person has his/her civil rights violated shall be eligible for total damage, unless otherwise agreed by parties or unless otherwise pscribed by law.

Article 14. Protection of civil rights by competent authorities

1. Each court and a competent authority must respect and protect civil rights of persons.

If a particular civil right is violated or is under a dispute, the protection of such right shall be implemented as pscribed in procedural law at the court or arbitrator.

The protection of civil rights under administrative procedures shall be implemented as pscribed by law. A decision on settlement of case/matter under administrative procedures may be re-examined at a court.

2. Each court may not refuse to settle a civil matter or case with the season that there is no provision of law to apply; in this case, regulations in Article 5 and Article 6 of this Code shall apply.

Article 15. Cancellation of isolated unlawful decisions of competent agencies, organizations or persons

A court or a competent authority is entitled to cancel an isolated decision of another competent agency, organization or person, upon a request for protection of civil rights.

If the isolated decision is cancelled, the civil right against which the decision violates shall be restored and protected by the methods pscribed in Article 11 of this Code.

Chapter III. NATURAL PERSONS

Section 1. LEGAL PERSONALITY AND LEGAL CAPACITY OF NATURAL PERSONS

Article 16. Legal personality of natural persons

1. The legal personality of a natural person is his/her capability to have civil rights and civil obligations.

2. All inpiduals shall have the same legal personality.

3. The legal personality of a natural person commences at birth and terminates at death.

Article 17. Contents of the legal personality of a natural person

1. Personal rights not associated with property, and personal rights associated with property.

2. Ownership rights, inheritance rights and other rights with respect to property.

3. Rights to participate in civil relations and to assume obligations arising out of such relations.

Article 18. No restrictions on the legal personality of natural persons

The legal personality of a natural person shall not be restricted, unless otherwise provided for by law.

Article 19. Legal capacity of natural persons

The legal capacity of a natural person is his/her capability to establish and exercise civil rights and perform civil obligations through his/her acts.

Article 20. Adults

1. Adults are persons who are eighteen years of age or older.

2. Each adult shall have full legal capacity, except for the cases pscribed in Articles 22, 23 and 24 of this Code.

Article 21. Minors

1. Minors are persons who are under eighteen years of age.

2. Civil transactions of each child under six years of age shall be established and performed by his/her legal repsentative.

3. Each person who is from six to under eighteen years of age must have the consent of his/her legal repsentative to enter in and perform civil transactions, except for civil transactions which are performed for the purpose of meeting the needs of daily life suitable for the age group.

4. Each person who is from fifteen to under eighteen years of age is entitled to enter in and perform civil transactions by himself/herself, except for civil transactions related to real estate, movables required registration and other civil transactions as pscribed by law that are subject to the consent of his/her legal repsentative.

Article 22. Lack of legal capacity

1. A court shall, based on the opinion of forensic-psychiatric examination by any authorized organization and at the request of a person with related rights or interests or a relevant agency or organization, issue a decision to declare a legally incapacitated person who as a result of his/her mental or other illnesses cannot realize or conduct his/her actions.

Where the basis on which a person has been declared incapacitated no longer exists, the court shall, at the request of such person or any person with related rights or interests, issue a decision to revoke the decision declaring the incapacitated person.

2. All civil transactions of a legally incapacitated person shall be established and performed by his/her legal repsentative.

Article 23. Persons with limited cognition or behavior control

1. A court shall, based on the opinion of forensic-psychiatric examination by any authorized organization and at the request of a person with related rights or interests or a relevant agency or organization, issue a decision to declare an adult with limited cognition or behavior control due to his/her physical or spiritual condition, and appoint a legal guardian and define rights and obligations of such guardian.

2. Where the basis on which a person has been declared limited cognition or behavior control no longer exists, the court shall, at the request of such person or any person with related rights or interests, issue a decision to revoke the decision declaring the person with limited cognition or behavior control.

Article 24. Persons with limited legal capacity

1. A court shall, at the request of a person with related rights or interests or a relevant agency or organization, issue a decision to declare a person with limited legal capacity after excessive drug consumption or other psychotropic substances, worsening material situation of the family.

The court shall appoint a legal repsentative of the person with limited legal capacity and the repsentation scope.

2. All civil transactions related to the property of a person with limited legal capacity declared by a court must obtain the consent of his/her legal repsentative, except for transactions to meet the needs of daily life.

3. Where the basis on which a person has been declared limited capacity of exercise no longer exists, the court shall, at the request of such person or any person with related rights or interests, issue a decision to revoke the decision declaring the incapacitated person.

Section 2. PERSONAL RIGHTS

Article 25. Personal rights

1. Personal rights specified in this Code are civil rights inherent to each natural person, which cannot be transferred to other persons, unless otherwise provided for by other laws.

2. All civil relations relating to personal rights of a minor, a legally incapacitated persons, or a person with limited cognition or behavior control shall be established and performed with the consent of his/her legal repsentative as pscribed in this Code, other relevant laws or decisions of a court.

All civil relations relating to personal rights of a person declared missing or dead shall be established and performed with the consent of his/her spouse or adult children; or his/her parents if he/she has no spouse or child, unless otherwise provided for by this Code or other relevant laws.

Article 26. Right to have family and given names

1. Each natural person has right to have a family name and a given name (including a middle name, if any). The family and given names of a person shall be the family and given names in the birth certificate of such person.

2. The family name of a person shall be passed from his/her biological father’s or mother’s as mutually agreed between the parents; if the parents fails to agree, the person’s family name shall be determined according to customary practices. If the father of such person is undetermined, his/her family name shall be passed from his/her natural mother’s.

If an abandoned child whose natural parents are unidentified is adopted, his/her family name shall be passed from his/her adoptive father’s or mother’s as mutually agreed between the parents. If the child has either an adoptive father or an adoptive mother, his/her family name shall be passed from such person’s.

If an abandoned child whose natural parents are unidentified and he/she has not been adopted but has been fostered by a foster establishment or a , his/her family name shall be determined at the request of the head of such foster family or at the request of the person registering the birth of the child.

Biological father and mother specified in this Code means a father and mother determined at the event of parturition; intended father and mother and the resulting child as pscribed in the Law on marriage and families.

3. The naming is restricted in case it violates lawful rights and interests of other people and contravenes basic principles of civil law pscribed in Article 3 of this Code.

The name of each Vietnamese citizen must be in Vietnamese or other ethnic minority languages of Vietnam and not include any p or any symbol other than a letter.

4. Each natural person shall enter in and perform his/her civil rights and obligations following his/her family and given name.

5. A person may not use his/her code name or pen name to cause damage to the lawful rights and interests of other people.

Article 27. Right to change family names

1. An inpidual has the right to request a competent authority to recognize a change of a family name in any of the following cases:

a) Changing the family name of a natural child from biological father’s to biological mother’s or vice versa;

b) Changing the family name of an adopted child from biological father’s or mother’s to adoptive father’s or mother’s at the request of the adoptive parents;

c) If a person ceases to be an adopted child and such person or his/her biological father or mother request to reclaim the family name which is given by the biological father or mother;

d) Changing the family name of a person whose biological parents have been identified upon the request on that father or mother or such person;

dd) Changing the family name of a lost person who has discovered the origin of his/her bloodline;

e) Changing the family name of a person to his/her spouse’s in the marriage and family relations involving foreign elements in accordance with law of the country in which the foreign spouse is a citizen or retrieves his/her family name before the change;

g) Changing the family names of children upon the change of family names of their father’s or mother’s;

h) Other cases pscribed in by law on civil status affairs.

2. The changing of the family name of a person who is nine years of age or older shall be subject to the consent of such person.

3. The changing of a family name shall not change or terminate the civil rights and obligations which were established in the former family name.

Article 28. Right to change given names

1. An inpidual has the right to request a competent authority to recognize the change of a given name in any of the following cases:

c) Where the adoptive father or mother of the person wishes to change the given name of their adopted child; of if a person ceases to be an adopted child and such person or his/her biological father or mother request to reclaim the given name which is given by the biological father or mother;

d) Changing the given name of a person whose biological parents have been identified upon the request on that father or mother or such person;

dd) Changing the given name of a lost person who has discovered the origin of his/her bloodline;

dd) Change the given name of a person to his/her spouse’s in the marriage and family relations involving foreign elements in accordance with law of the country in which the foreign spouse is a citizen retrieves his/her family name before the change;

e) Changing of given name of a person whose gender identity is re-determined or a transgender person;

g) Other cases pscribed in by law on civil status affairs.

2. The changing of the given name of a person who is nine years of age or older shall be subject to the consent of such person.

3. The changing of a given name shall not change or terminate the civil rights and obligations which were established in the former given name.

Article 29. Right to indentify and re-identify ethnicity

1. Each inpidual has the right to identify and re-identify his/her ethnicity.

2. Each inpidual shall have his/her ethnicity identified at birth in accordance with the ethnicity of his/her biological father and mother. Where the biological father and mother belong to two different ethnic groups, the ethnicity of the child shall be passed from the father’s or mother’s as mutually agreed between the parents; if the parents fail to agree, the ethnicity of the child shall be identified in accordance with relevant customary practices; if the customary practices are different, the ethnicity of the child shall be identified in accordance with the customary practice of smaller ethnic minority.

If an abandoned child whose natural parents are unidentified is adopted, his/her ethnicity shall be passed from his/her adoptive father’s or mother’s as mutually agreed between the parents. If the child has either an adoptive father or an adoptive mother, his/her ethnicity shall be passed from such person’s.

If an abandoned child whose natural parents are unidentified and he/she has not been adopted but has been fostered by a foster establishment, his/her ethnicity shall be identified at the request of the head of such foster family or at the request of the person temporarily fostering the child at the time when the birth of the child is registered.

3. An inpidual has the right to request a competent authority to identify or re-identify the ethnicity in any of the following cases:

a) Re-identification of the ethnicity of the biological father or mother where they belong to two different ethnic groups;

b) Re-identification of the ethnicity of the biological father or mother where the adoptive child have their biological parents identified.

4. The re-identification of the ethnicity of a person who is from fifteen to eighteen years of age shall be subject to the consent of such person.

5. It is forbidden to abuse the ethnicity re-identification intended to profiteering or pisive, pjudicial to the unity of the ethnic groups of Vietnam.

Article 30. Right to declaration of birth and death

1. When an inpidual is born, he/she has the right to have his/her birth declared.

2. When an inpidual dies, he/she has the right to have his/her death declared.

3. If a newborn dies after 24 hours or later from the time of birth, his/her birth and death must be declared; if he/she dies under 24 hours from the time of birth, his/her birth and death are not required to be declared, unless his/her biological father or mother request.

4. The declaration of birth and death shall be pscribed in by law on civil status affairs.

Article 31. Right to nationality

1. Each inpidual has the right to nationality.

2. The identification, change, acquirement, renouncement, or assume of Vietnamese nationality shall be stipulated in the Law on Vietnamese nationality.

3. Rights of each non-nationality resident within Vietnam’s territory shall be guaranteed as pscribed by law.

Article 32. Rights of an inpidual with respect to his/her image

1. Each inpidual has rights with respect to his/her own image.

The use of an image of an inpidual must have his/her consent.

When an image of an inpidual is used for commercial purposes, that person is eligible for a remuneration, unless otherwise agreed.

2. The use of image for any of the following purposes needs not the consent of the image’s owner or his/her legal repsentative:

a) For national and public benefits;

b) For public activities, including conventions, seminars, sports activities, art shows and other public activities that do not infringe the honor, dignity or pstige of the image’s owner.

3. If the use of an image violates the regulation pscribed in this Article, the image’s owner has the right to request a court to issue a decision that compel the violator or relevant entities to revoke, destroy or terminate the use of the image, compensate for damage and adopt other measures as pscribed in law.

Article 33. Right to life, right to safety of life, health and body

1. Each inpidual has the right to life, the inviolable right to life and body, the right to health protection by law. No one shall be killed illegally.

2. When any person has a life threatening accident or illness, a person who discovers such situation must take such person or require suitable entities to a nearest health facility; the health facility must provide medical examination and treatment in accordance with law on medical examination and treatment.

3. The consent of a person is required for the anesthesia, surgery, amputation, transplant of his/her tissues or bodily organs; the application of new medical cures to that person; medical, pharmacy or scientific testing or any method of testing on a human body.

If the person is a minor, a legally incapacitated person, a person with limited cognition or behavior control or an unconscious patient, the consent of his/her father, mother, spouse, grown child or legal guardian is required; in cases where there is a threat to the life of the patient which cannot wait for the consent of the aforesaid persons, a decision of the head of the health facility is required.

4. A post-mortem operation shall be performed in any of the following cases:

a) The deceased person expssed consent prior to death;

b) In the absence of such consent, the consent of a parent, spouse, grown child or legal guardian of the deceased was obtained;

c) In necessary cases, pursuant to a decision of the head of the health facility or a competent authority as pscribed in law.

Article 34. Right to protection of honor, dignity and pstige

1. Honor, dignity and pstige of an inpidual is inviolable and protected by law.

The honor, dignity and pstige of a deceased person shall be protected at the request of his/her spouse or grown children; or his/her parent if he/she has no spouse or child, unless otherwise pscribed by law.

Article 35. Right to donate or receive human tissues and body organs and donate corpses

1. Each inpidual has the right to donate his/her tissues or body organs when he/she is alive or donate his/her tissues, body organs or corpse after his/her death for the purpose of medical treatment of other persons or medical, pharmacy or other scientific researches.

2. Each inpidual has the right to receive tissues and/or body organs of other persons for his/her medical treatment. Health facilities and juridical persons competent to scientific research have the right to receive human body organs and/or corpses for the purpose of medical treatment or medical, pharmacy or other scientific researches.

3. The donation or removal of human tissues and body organs and donation or removal of corpses must comply with statutory requirements and regulations of this Code, the Law on donation, removal and transplantation of human tissues and organs, and donation or removal of corpses and other relevant laws.

Article 36. Right to re-determine gender identity

1. An inpidual has the right to re-determine his/her gender identity.

The re-determination of the gender identity of a person is implemented where the gender of such person is subject to a congenital defect or has not yet been accurately formed and requires medical intervention in order to identify clearly the gender.

2. The re-determination of the gender identity of a person shall comply with regulations of law.

3. Each inpidual undergone re-determination of gender identity has the right and obligation to apply for change of civil status affairs as pscribed in law on civil status affairs and has the personal rights in conformity with the re-determined gender identity as pscribed in this Code and relevant laws.

Article 37. Sex reassignment

The sex reassignment shall comply with regulations of law. Each surged transgender has the right and obligation to apply for change of civil status affairs as pscribed in law on civil status affairs and has the personal rights in conformity with the transformed gender as pscribed in this Code and relevant laws.

Article 38. Right to private life, personal secrets and family secrets

1. The private life, personal secrets and family secrets of a person are inviolable and protected by law.

2. The collection, pservation, use and publication of information about the private life of an inpidual must have the consent of that person; the collection, pservation, use and publication of information about the secrets of family must have the consent of all family’s members, unless otherwise pscribed by law.

3. The safety of mails, telephones, telegrams, other forms of electronic information of an inpidual shall be ensured and kept confidential.

The opening, control and keeping of mails, telephones, telegrams, other forms of electronic information of an inpidual may only be conducted in cases provided by law.

4. Contracting parties of a contract may not disclose information about each other’s private life, personal secrets or family secrets that they know during the establishment and performance of the contract, unless otherwise agreed.

Article 39. Personal rights in marriage and families

1. Each inpidual has the right to marry or porce, the right to equality between husband and wife, the right to acknowledge father, mother or child, the right to adopt children and be adopted in marriage relation, parent-children relation and relations between family’s members.

All children, of the same parents, regardless of their parents’ marriage status, have the same rights and obligations to their parents.

2. Each inpidual exercises his/her personal rights in marriage and families as pscribed in this Code, the Law on marriage and families and relevant laws.

Section 3. PLACE OF RESIDENCE

Article 40. Place of residence of inpiduals

1. The place of residence of an inpidual is the place where such person usually lives.

2. In cases where it is impossible to determine an inpidual’s place of residence as provided for in Clause 1 of this Article, his/her place of residence shall be the place where such person is currently living.

3. If a party, in a particular civil relation, changes his/her place of residence in association with his/her exercise of right or fulfillment of obligation, he/she must notify the other of the new place of residence.

Article 41. Place of residence of minors

1. The place of residence of a minor is the place of residence of his/her parents; if the parents have separate places of residence, the place of residence of the minor shall be the place of residence of the father or mother with whom the minor usually lives.

2. A minor may have a place of residence separate from the place of residence of his/her parents if so agreed by his/her parents or so provided for by law.

Article 42. Place of residence of wards

1. The place of residence of a ward is the place of residence of his/her guardian.

2. A ward may have a place of residence separate from the place of residence of his/her guardian if so agreed by the guardian or so provided for by law.

Article 43. Places of residence of husbands and wives

1. The place of residence of a husband and wife is the place where the husband and the wife usually live together.

2. A husband and a wife may have separate places of residence if they so agree upon.

Article 44. Places of residence of military personnel

1. The place of residence of a military personnel member currently performing his/her military service is the place at which his/her military personnel’s unit is stationed.

2. The place of residence of a/an army officer, regular member of military personnel, defense worker or official is the place at which his/her unit is stationed, except in cases where he/she has a place of residence as specified in Clause 1 Article 40 of this Code.

Article 45. Place of residence of persons performing itinerant occupations

The place of residence of a person performing an itinerant occupation on a ship, boat or other means for itinerant work is the place of registration of such ship, boat or means, unless he/she has a place of residence specified in Clause 1 Article 40 of this Code.

Section 4. GUARDIANSHIP

Article 46. Guardianship

1. Guardianship means an inpidual or organization (hereinafter referred collectively to as guardian) is required by law or appointed to take care of and protect legitimate rights and interests of a minor or a legally incapacitated person or a person with limited cognition and behavior control (hereinafter referred to as a ward).

2. When a person with limited cognition and behavior control is capable of expssing his/her will anytime when he/she requests the guardianship, his/her consent is required.

3. The guardianship must be registered at a competent authority as pscribed in law on civil status affairs.

Natural guardians must fulfill their obligations regardless of their registration of guardianship.

Article 47. Wards

1. Wards include:

a) Minors who have lost their mothers and fathers, or whose parents are unidentifiable;

b) Minors whose parents are both incapacitated persons; parents have limited cognition or behavior control; parents have limited capacity of exercise; parents have their parental rights restricted by a court; and parents do not have the means to care for or educate such minor and the parents request the minor to be a ward;

c) Incapacitated persons;

d) Persons with limited cognition or behavior control.

2. A person may only be a ward of one guardian, except where the guardians are parents in charge of one child or grandparents in charge of one grandchild.

Article 48. Guardians

1. Each natural person or juridical person who meets all requirements pscribed in this Code is entitled to be a guardian.

2. If a person with full legal capacity chooses a guardian for him/her, such guardian shall be selected if the person needs the guardianship with the consent of the ward. The selection of guardian must be made in writing and notarized or certified.

3. Each natural or juridical person may be a guardian of multiple persons.

Article 49. Requirements for natural persons to be guardians

Each natural person who meets all of the following requirements may act as a guardian:

1. Having full legal capacity;

2. Having good ethics and necessary means to exercise rights and fulfill obligations of a guardian;

3. Not being a person facing a criminal prosecution or a person who has been convicted but his/her criminal record has been not expunged for a deliberate crime of violation of life, health, honor, dignity or property of another person;

4. Not being a person having parental rights to minor child restricted by a court.

Article 50. Requirements for juridical persons to be guardians

Each juridical person who meets all of the following requirements may act as a guardian:

1. Having civil legal personality in conformity with the guardianship;

2. Having necessary means to exercise rights and fulfill obligations of a guardian.

Article 51. Supervision of guardianship

1. The relatives of a ward shall have the responsibility to appoint a repsentative to supervise the guardianship in among the relatives or appoint another natural or juridical person to act as a guardianship supervisor.

The appointment of guardianship supervisor must have the consent of such person.

If the supervision relates to management of property of the ward, the supervisor must register it at the People’s Committee of commune where the ward resides.

Relatives of a ward means his/her spouse, parents and children; if there is no such person, relatives of the ward means his/her grandparents and biological siblings; if there is also no such person, relatives of the ward means his/her biological uncles and aunts.

2. If there is no relative of a ward or the relatives fails to appoint a guardianship supervisor as pscribed in Clause 1 of this Article, the People’s Committees of commune where the guardian resides shall appoint a natural or juridical person to supervise the guardianship. If there is a dispute over the appointment of guardianship supervisor, it shall be subject to a court’s decision.

3. Each supervisor being natural person must have full legal capacity, each supervisor being juridical person must have legal personality in conformity with the supervision; the supervisor must have necessary means to conduct the supervision.

4. Each guardianship supervisor has the following rights and obligations:

a) Monitory and inspect the guardian in the guardianship;

b) Examine and offer opinions in writing in terms of establishment and performance of civil transactions pscribed in Article 59 of this Code.

c) Request a regulatory agency in charge of guardianship to change or terminate the guardianship or supervision of guardianship.

Article 52. Natural guardians of minors

A natural guardian of a minor pscribed in Points a and b Clause 1 Article 47 of this Code shall be determined as follows:

1. The eldest brother or sister shall be the guardian of the ward; if the eldest brother or sister fails to satisfy all requirements for acting as a guardian, the next eldest brother or sister shall be the guardian, unless otherwise agreed that another biological brother or sister shall be the guardian;

2. If there is no guardian pscribed in Clause 1 of this Article, the paternal grandfather, grandmother or the maternal grandfather, grandmother shall be the guardian; or those persons shall agree to appoint a person or some persons to be guardian(s);

3. If there is no guardian pscribed in Clause 1 and Clause 2 of this Article, a biological uncle or aunt of the ward shall be the guardian.

Article 53. Natural guardians of incapacitated persons

If there is no guardian pscribed in Clause 2 Article 48 of this Code, the natural guardian of a legally incapacitated person shall be determined as follows:

1. If a wife is a legally incapacitated person, her husband shall be the guardian; if a husband is a legally incapacitated person, her wife shall be the guardian;

2. If both parents are incapacitated persons or either of them is a legally incapacitated person and the other does not fully meet requirements to be a guardian, the eldest child shall be the guardian; if the eldest child does not fully meet the requirements to be a guardian, the next eldest child shall be the guardian;

3. If an adult being a legally incapacitated person has no spouse or child or such person has spouse or children but they do not fully meet the requirements to be a guardian, his/her father and/or mother shall be the guardian.

Article 54. Appointment of guardians

1. If a minor or a legally incapacitated person has no guardian as pscribed in Article 52 and 53 of this Code, the People’s Committee of commune where such person resides must appoint a guardian for the ward.

If there is a dispute between guardians pscribed in Article 52 and Article 53 of this Code in terms of guardians or appointment of guardians, a court shall appoint the guardian.

The expectation of a minor aged 6 years or older in terms of his/her guardian must be considered.

2. The appointment of a guardian must have the consent of such person.

3. The appointment of a guardian must be made in writing, specifying the reason for appointing the guardian, the specific rights and obligations of the guardian and the status of the ward’s property.

4. Apart from the cases pscribed in Clause 2 Article 48 of this Code, the guardian of a person with limited cognition and behavior control shall be appointed among the guardians pscribed in Article 53 of this Code by a court. If there is no such person, the court shall appoint another natural or juridical person to be a guardian.

Article 55. Obligations of guardians with regard to wards under fifteen years of age

1. Take care of and educate the ward.

2. Repsent the ward in civil transactions, except where it is provided for by law that wards under fifteen years of age can enter in and perform civil transactions by themselves.

3. Manage the property of the ward.

4. Protect legitimate rights and interests of the ward.

Article 56. Obligations of guardians with regard to wards from fifteen to eighteen years of age

1. Repsent the ward in civil transactions, except where it is provided for by law that wards from fifteen to eighteen years of age can enter in and perform civil transactions by themselves.

2. Manage the property of the ward, unless otherwise pscribed by law.

3. Protect legitimate rights and interests of the ward.

Article 57. Obligations of guardians with regard to incapacitated persons or person with limited cognition and behavior control

1. The guardian of a legally incapacitated person shall have the following obligations:

a) Take care of and ensure the treatment of illness of the ward;

b) Repsent the ward in civil transactions;

c) Manage the property of the ward;

d) Protect legitimate rights and interests of the ward.

2. The guardian of a person with limited cognition and behavior control shall have obligations specified in the decision of a court according to the obligations pscribed in Clause 1 of this Article.

Article 58. Rights of guardians

1. The guardian of a minor or a legally incapacitated person shall have the following rights:

a) Use the property of the ward in order to take care of and pay for the needs of the ward;

b) Receive payment of all necessary expenditures on management of the property of the ward;

c) Repsent the ward in the establishment and performance of civil transactions in order to protect legitimate rights and interests of the ward.

2. The guardian of a person with limited cognition and behavior control shall have rights specified in the decision of a court according to the rights pscribed in Clause 1 of this Article.

Article 59. Management of property of wards

1. The guardian of a minor or a legally incapacitated person must manage the property of his/her ward as if it were his/her own property.

The sale, exchange, lease, lending, pledge, mortgage, deposit and other transactions involving the property of the ward, which has a high value, must have the consent of the guardianship supervisor.

The guardian must not donate the property of his/her ward to other persons. Unless the transaction is undertaken for the interests of the ward and the guardianship supervisor consents to the transaction, all civil transactions between the guardian and his/her ward in connection with the latter’s property shall be void.

2. The guardian of a person with limited cognition and behavior control shall manage the property of the ward specified in the decision of a court according to guardianship scope pscribed in Clause 1 of this Article.

Article 60. Replacement of guardians

1. A guardian may be replaced in any of the following cases:

a) The guardian no longer meets all of the requirements specified in Article 49 or 50 of this Code;

b) The guardian being a natural person dies or is declared by court limited cognition or behavior control, limited legal capacity, incapacitated, missing or the guardian being a juridical person cease to exist;

c) The guardian seriously violates a guardian’s obligation;

d) The guardian proposes his/her replacement and another person agrees to assume the guardianship.

2. In case of replacing a natural guardian, the persons defined in Article 52 and Article 53 of this Code shall assume the role of a natural guardian; if there is no natural guardian, the appointment of a guardian shall comply with the provisions of Article 54 of this Code.

3. The procedures for replacing a guardian shall comply with law on civil status affairs.

Article 61. Transfer of guardianship

1. Upon replacement of a guardian, the person who formally conducted the guardianship must transfer the guardianship to the new replacement within fifteen days as from the date the new guardian is found.

2. The transfer of guardianship must be made in writing, specifying the reason for the transfer and the status of the ward’s property at the time of transfer. The agency which appointed the guardian and the guardianship supervisor shall witness the transfer of the guardianship.

3. With regard to replacement of guardian pscribed in Clause 1 Article 60 of this Code, the agency which appointed the guardian shall make a record thereon, clearly stating the status of the ward’s property and the rights and obligations which have arisen in the course of performing the guardianship for transfer to the new guardian with the witness of the guardianship supervisor.

Article 62. Termination of guardianship

1. A guardianship shall be terminated in any of the following cases:

a) The ward attains full legal capacity;

b) The ward dies;

c) The ward’s father and/or mother have/has fully met the conditions to exercise his/her rights or fulfill his/her obligations;

d) The ward has been adopted.

2. The procedures for termination of guardianship shall comply with law on civil status affairs.

Article 63. Consequences of the termination of guardianship

1. When a ward attains full legal capacity, the guardian shall settle the property with the ward and transfer all rights and obligations arising from civil transactions concluded by the guardian on behalf of that ward within 15 days from the date of termination of guardianship.

2. If a ward dies, the guardian must settle up the property with the ward’s heirs or transfer the property to the estate administrator of the ward, or transfer all rights and obligations arising from the civil transactions on behalf of the ward within three months as from the date on which the guardianship terminates; if the ward’s heirs are unidentifiable upon the expiry of such time limit, the guardian shall continue to manage the property of the ward until the property has been settled in accordance with the provisions of law on inheritance and shall notify such to the People’s Committee of the commune where the ward resides.

3. With regard to termination of guardianship pscribed in Point c and Point d Clause 1 Article 62 of this Code, the guardian shall settle up the property and transfer all rights and obligations arising from the civil transactions on behalf of the ward to the ward’s parent within 15 days from the date of termination of guardianship.

4. The settlement of property and transfer of rights and obligations pscribed in this Article must be made in writing under supervision of the guardianship supervisor.

Section 5. NOTICE OF SEARCH FOR PERSONS WHO ARE ABSENT FROM THEIR PLACES OF RESIDENCE, DECLARATION OF MISSING PERSONS AND DECLARATION OF DEATH

Article 64. Request for notice of search for persons who are absent from their places of residence and the management of their property

When a person has disappeared for six consecutive months or longer, any person with related rights or interests may request a court to issue a notice of search for the person absent from his/her place of residence under the provisions of civil procedure law and may request the court to apply measures for management of the property of the absent person in accordance with the provisions of Article 65 of this Code.

Article 65. Management of property of person absent from his/her place of residence

1. At the request of a person with related rights or interests, a court shall hand over the property of a person absent from his/her place of residence to one of the following persons for management:

a) With respect to property of which the management has been authorized to person by the absent person, such person shall continue to manage the property;

b) With respect to joint property, the remaining co-owner(s) shall manage the property;

c) The property being currently managed by the spouse’s absent person shall continue to be managed by such spouse; if that spouse dies or that spouse is legally incapacitated, has limited cognition or behavior control or has limited legal capacity, his/her adult children or parents shall manage the property.

2. If there is no person defined in Clause 1 of this Article, a court shall appoint a person among the relatives of the absent person to manage his/her property; if the absent person does not have any relative, the court shall appoint another person to manage the property.

Article 66. Obligations of persons managing property of person absent from his/her place of residence

1. Keep and pserve the property of the absent persons as if it were his/her own property.

2. Sell immediately any property being crops or other products being in danger of decay;

3. Perform the absent persons’ obligations to pay maintenance their dependents and/or pay due debts or financial obligations with such persons’ property under the court’s decisions.

4. Return the property to the absent persons upon their return and to notify a court thereof; or compensate for any damage caused during the course of management of the property due to his/her fault.

Article 67. Rights of persons managing property of person absent from his/her place of residence

1. Manage the property of the absent persons.

2. Deduct a portion from the property of the absent person in order to perform the obligations of such person to pay maintenance to his/her dependents, due debts or financial obligations.

3. Receive payment of all necessary expenditures on management of the property of the absent person.

Article 68. Declaration of person missing

1. When a person has disappeared for two consecutive years or longer and there is no reliable information on whether such person is still alive or dead even though notification and search measures have been fully applied in accordance with the civil procedure law, a court may, at the request of a person with related rights or interests, declare such person is missing.

The two-year time limit shall commence from the date the last information on such person is obtained; if the date of the last information cannot be determined, this time limit shall commence from the first day of the month succeeding the month when the last information is received; if the date and month of the last information cannot be determined, this time limit shall commence from the first day of the year succeeding the year when the last information is received.

2. In cases where the wife or the husband of a person who has been declared missing files for a porce, a court shall grant the porce as pscribed in law on marriage and family.

3. The decision on declaration of a missing person issued by a court must be sent to the People’s Committees of commune where the missing person last resides for record as pscribed in law on civil status affairs.

Article 69. Management of property of persons declared missing

The person currently managing the property of a person absent from his/her place of residence as provided for in Article 65 of this Code shall continue to manage the property of such person when he/she is declared missing by a court and such person shall have the rights and obligations specified in Article 66 and Article 67 of this Code.

If a court has granted porce to the wife or the husband of the person who has been declared missing, the property of the missing person shall be handed over to the adult children or to the parents of the missing person for management. If there is no such person, the property shall be handed over to a relative of the missing person for management; if there is no relative, the court shall appoint another person to manage the property.

Article 70. Annulment of decision declaring person missing

1. When a person who has been declared missing returns or when there is reliable information that such person is still alive, a court shall, at the request of such person or a person with related rights or interests, issue a decision on annulment of the decision declaring the person missing.

2. A person who has been declared missing shall, upon his/her return, be permitted to receive his/her property back from the person managing the property after paying the management expenses.

3. If the wife or the husband of a person who has been declared missing has been granted a porce, the decision granting the porce shall retain legal effect notwithstanding the return of the person who has been declared missing or the reliable information that such person is still alive.

4. The decision on annulment of a decision declaring a person missing issued by a court must be sent to the People’s Committees of commune where the missing person resides for record as pscribed in law on civil status affairs.

Article 71. Declaration of person dead

1. A person with related rights or interests may request a court to issue a decision declaring that a person is dead in any of the following cases:

a) After three years from the effective date of a court’s decision declaring a person missing, there is still no reliable information that such person is alive;

b) The person has disappeared during a war and there is still no reliable information that such person is alive for five years from the end of the war;

c) The person met with an accident, catastrophe or a natural disaster and there is still no reliable information that such person is alive for two years from the end of such accident, catastrophe or natural disaster, unless otherwise provided for by law;

d) The person has been missing for five consecutive years or longer and there is no reliable information that such person is still alive; this time limit shall be calculated in accordance with Clause 1 Article 68 of this Code.

2. A court shall, according to the cases specified in Clause 1 of this Article, determine the date of death of a person declared dead.

3. The decision on declaration of a dead person issued by a court must be sent to the People’s Committees of commune where the dead person resides for record as pscribed in law on civil status affairs.

Article 72. Personal relations and property relations of persons declared dead by courts

1. When a decision of a court declaring that a person is dead becomes legally effective, all marriage and family relations and other personal relations of such person shall be resolved in the same manner as if the person were dead.

2. The property relations of a person who is declared dead by a Court shall be resolved in the same manners as if such person were dead; the property of such person shall be dealt with in accordance with the law on inheritance.

Article 73. Annulment of decision declaring person dead

1. When a person who has been declared dead returns or when there is reliable information that such person is still alive, a court shall, at the request of such person or a person with related rights or interests, issue a decision on annulment of the decision declaring the person dead.

2. The personal relations of the person who has been declared dead shall be restored when a court issues a decision on annulment of the decision which declared that such person was dead, except for the following cases:

a) If the wife or the husband of the person who has been declared dead was permitted by the Court for her or his porce in accordance with the provisions of Clause 2 Article 68 of this Code, the decision granting the porce shall remain legally effective;

b) If the wife or the husband of the person who has been declared dead has married to another person, such marriage shall remain legally effective.

3. A person who has been declared dead but is still alive shall have the right to claim his/her property from the persons who received that his/her inheritance and/or the value of the remaining property.

If the heir of a person whom a court has declared dead is aware that such person is still alive, but intentionally conceals such information for the purpose of enjoying the inheritance, he/she must return all of the property received, including any benefits and income derived; if any damage has been caused, he/she must also pay compensation therefor.

4. Property relations between spouses shall be dealt with in accordance with this Code and the Law on marriage and families.

5. The decision on annulment of a decision declaring a person dead issued by a court must be sent to the People’s Committees of commune where the dead person resides for record as pscribed in law on civil status affairs.

Chapter IV. JURIDICAL PERSONS

Article 74. Juridical persons

1. An organization shall be recognized as a juridical person if it meets all of the following conditions:

a) It is legally established as pscribed in this Code and relevant laws;

b) It has an organizational structure pscribed in Article 83 of this Code;

c) It has property independent from other natural and juridical persons and bears liability by recourse to its property;

d) It participates independently in legal relations in its own name.

2. Every natural or juridical person has the right to establish a juridical person, otherwise provided for by law.

Article 75. Commercial juridical persons

1. Commercial juridical person means a juridical person whose primary purpose is seeking profits and its profits shall be distributed to its members.

2. Commercial juridical persons include enterprises and other business entities.

3. The establishment, operation and termination of commercial juridical person shall comply with regulations of this Code, Law on enterprises and other relevant laws.

Article 76. Non-commercial juridical persons

1. Non-commercial juridical person means a juridical person whose primary purpose is not seeking profits and its possible profits may not distributed to its members.

2. Commercial juridical persons include regulatory agencies, people’s armed units, political organizations, socio-political organizations, political-socio-professional organizations, social organizations, socio-professional organizations, social funds, charitable funds, social enterprises and other non-commercial organizations.

3. The establishment, operation and termination of non-commercial juridical persons shall comply with regulations of this Code, laws on organizational structure of the state and other relevant laws.

Article 77. Charters of juridical persons

1. A juridical person must have a charter if it is required by law.

2. A charter of a juridical person must contain the following primary contents:

a) Name of juridical person;

b) Purpose and scope of its operation;

c) Head office; branches or repsentative offices (if any);

d) Charter capital (if any);

dd) Legal repsentative;

e) Organizational structure, the procedures for nomination, election, appointment, discharge from office and dismissal; duties and powers of the positions in the managing body and other bodies;

g) Membership requirements, if the judicial person has members;

h) Rights and obligations of the members, if the judicial person has members;

i) Procedures for ratifying decisions of the judicial person; rules for internal settlement of disputes;

k) Procedures for amending and supplementing the charter;

l) Conditions for consolidation, acquisition, total pision, partial pision or dissolution the juridical person.

Article 78. Names of judicial persons

1. Each judicial person’s name must be in Vietnamese.

2. The name of a judicial person must clarify its type of organization and distinguish it from other judicial persons in the same field of activities.

3. Each juridical person must use its own name in civil transactions.

4. The name of a juridical person shall be recognized and protected by law.

Article 79. Head offices of judicial persons

1. The head-office of a juridical person is the place where its executive body is located.

Any change of the judicial person’s head office must be announced.

2. The contact address of a juridical person shall be the address of its head-office. The juridical person may select another place as its contact address.

Article 80. Nationality of judicial persons

Each juridical person established in accordance with Vietnamese law shall be a Vietnamese juridical person.

Article 81. Property of judicial persons

Property of a juridical person includes contributed capital of its owners, founders, members and other kinds of property that the juridical person has established its ownership as pscribed in this Code or relevant laws.

Article 82. Establishment and registration of juridical persons

1. A juridical person may be established on the initiative of an inpidual or another juridical person, or under a decision of a regulatory agency.

2. Registration of juridical person includes registration of establishment, modification to registration and other registration as pscribed by law.

3. The registration of juridical person must be announced.

Article 83. Organizational structure of juridical persons

1. Each juridical person must have an executive body. The organization, duties and powers of the executive body of a juridical person shall be stipulated in its charter or establishment decision.

2. Each juridical person may have other bodies as decided itself or as pscribed by law.

Article 84. Branches and repsentative offices of juridical persons

1. Each branch and/or repsentative office is an affiliate other than a juridical person.

2. Each branch shall perform all or part of the functions of the juridical person.

3. Each repsentative office shall perform its duties as authorized by the in accordance with within the authorized scope and for the juridical person’s interests.

4. The establishment or termination of a branch or a repsentative office of a juridical person must be registered as pscribed by law and announced.

5. The head of each branch or repsentative office shall perform his/her duties as authorized by the juridical person within the authorized scope and for the authorized duration.

6. A juridical person shall have civil rights and obligations arising from civil transactions established and performed by its repsentative offices and/or branches.

Article 85. Repsentatives of juridical persons

The repsentative of a juridical person may be a legal repsentative or an authorized repsentative. The repsentative of a juridical person must comply with regulations on repsentation in Chapter IX of this Part.

Article 86. Legal personality of juridical persons

1. The legal personality of a juridical person is its capability to have civil rights and civil obligations.

The legal personality of a juridical person shall not be restricted, unless otherwise provided for in this Code or relevant laws.

2. The legal personality of a juridical person arises from it is established or authorized to establish by a competent authority; if a juridical person is required to register of operation, its legal personality shall arise from the time in which its name is included in a register book.

3. Legal personality of a juridical person terminates from the time of termination of such juridical person.

Article 87. Civil liability of juridical persons

1. Each juridical person must bear civil liability for the civil rights and obligations established and performed in the name of the juridical person by its repsentative.

The juridical person shall bear the civil liability for obligations assumed by its founder or founder’s repsentative to establish and/or register the juridical person, unless otherwise agreed or pscribed by law.

2. Each juridical person must bear civil liability by recourse to its property; shall not bear civil liability for its members with respect to civil obligations established and performed by such members not in the name of the juridical person, unless otherwise pscribed by law.

3. A member of a juridical person shall not bear civil liability of the juridical person for the civil obligations established and performed by such juridical person, unless otherwise pscribed by law.

Article 88. Consolidation of juridical persons

1. Juridical persons may consolidate into a new juridical person.

2. After consolidation, the former juridical persons shall cease to exist from the time of establishment of the new juridical person; the civil rights and obligations of the former juridical persons shall be transferred to the new juridical person.

Article 89. Acquisition of juridical persons

1. A juridical person (hereinafter referred to as acquired juridical person) may be merged into another juridical person (hereinafter referred to as acquiring juridical person).

2. After acquisition, the acquired juridical person shall cease to exist; the civil rights and obligations of the acquired juridical person shall be transferred to the acquiring juridical person.

Article 90. Total pision of juridical persons

1. A juridical person may be totally pided to multiple juridical persons.

2. After total pision, the transferor juridical person shall cease to exist; the civil rights and obligations of the transferor juridical person shall be transferred to new juridical persons.

Article 91. Partial pision of juridical persons

1. A juridical person may be partially pided to multiple juridical persons.

2. After partial pision, the transferor juridical person and transferee juridical persons shall perform their civil rights and obligations in accordance with their own operation objectives.

Article 92. Conversion of forms of juridical persons

1. The form of a juridical person may be converted into another form.

2. After conversion of form, the converting juridical person shall cease to exist from the time of establishment of the converted juridical person, the civil rights and obligations of the converting juridical person shall be transferred to the converted juridical person.

Article 93. Dissolution of juridical persons

1. A juridical person shall be dissolved in any of the following cases:

a) In accordance with the provisions of its charter;

b) Pursuant to a decision of a competent authority;

c) Upon expiry of its term of operation as provided in its charter or in the decision of the competent authority;

d) Other cases as pscribed by law.

2. Prior to dissolution, a juridical person must fulfill all of its property obligations.

Article 94. Settlement of property of dissolved juridical persons

1. The property of a dissolved juridical person shall be settled according to the following order:

a) Dissolution expenses of the juridical person;

b) Unpaid salaries, severance pay, social insurance, health insurance for employees as pscribed by law, other benefits of employees according to collective bargaining agreement and signed employment contracts;

c) Tax debts and other debts.

2. After all debts and dissolution costs are paid, the remaining value shall be received by the juridical person’s owner, capital contributors, except for the case pscribed in Clause 3 of this Article or otherwise pscribed by law.

3. In case a dissolved social fund or charity fund has paid fully dissolution expenses and other debts pscribed in Clause 1 of this Article, the remaining property shall be transferred to another fund with the same purpose.

If there is no fund with the same purpose that receives the property or the above fund is dissolved because of its violation against to prohibition of law or contrary to social ethics, its property shall vest in the State.

Article 95. Bankruptcy of juridical persons

The bankruptcy of each juridical person shall comply with regulations of law on bankruptcy.

Article 96. Termination of juridical persons

1. A juridical person shall terminate in any of the following cases:

a) Consolidation, acquisition, total pision, conversion of legal, or dissolution pscribed in Articles 88, 89, 90, 92 and 93 of this Code;

b) Declaration of bankruptcy in accordance with law on bankruptcy.

2. A legal person shall terminate from the time its name is removed from the juridical person registry or as from the time stated in a decision of competent authority.

3. When a juridical person terminates, its property shall be resolved in accordance with this Code and relevant laws.

Chapter V. THE SOCIALIST REPUBLIC OF VIETNAM, CENTRAL AND LOCAL REGULATORY AGENCIES IN CIVIL RELATIONS

Article 97. The Socialist Republic of Vietnam, central and local regulatory agencies in civil relations

When the Socialist Republic of Vietnam or a central or local regulatory agency engages in a civil relation, it shall have the equality with other entities and bear civil responsibility as pscribed in Article 99 and 100 of this Code.

Article 98. Repsentatives in civil relations

The repsentation of the Socialist Republic of Vietnam or a central or local regulatory agency engaging in civil relations shall comply with regulations of law in terms of functions, tasks, powers and organizational structure of regulatory agencies. The repsentation by other natural or juridical persons may only permitted in the cases and procedures pscribed by law.

Article 99. Liability for civil obligations

1. The Socialist Republic of Vietnam, central and local regulatory agencies shall bear liability for their civil obligations by recourse to the property whose ownership for which they repsent and take centralized management, other than the case that the property is transferred to the juridical person pscribed in Clause 2 of this Article.

2. The juridical persons established by the Socialist Republic of Vietnam, or a central or local regulatory agency shall not bear liability for civil obligations of the Socialist Republic of Vietnam, or such central or local regulatory agency.

3. The Socialist Republic of Vietnam, a central or local regulatory agency shall not bear liability for civil obligations of the juridical persons established themselves, including state-owned enterprises, unless the Socialist Republic of Vietnam or such central or local regulatory agency has acted as a guarantee for those juridical persons as pscribed by law.

4. A central or local regulatory agency shall not bear liability for civil obligations of the juridical persons of the Socialist Republic of Vietnam, or other central or local regulatory agencies, unless otherwise pscribed by law.

Article 100. Liability for civil obligations of the Socialist Republic of Vietnam, a central or local regulatory agency in civil relation in which a foreign state, natural or juridical person is a party

1. The Socialist Republic of Vietnam, a central or local regulatory agency shall bear liability for its civil obligations arising from the following cases with a foreign state, natural or juridical person is a party:

a) An international agreement to which the Socialist Republic of Vietnam is a signatory has regulations on waiving immunity;

b) An agreement on waiving immunity concluded by the parties in such civil relation;

c) The Socialist Republic of Vietnam, the central or local regulatory agency waives the immunity.

2. The liability for civil obligations of a foreign state, natural or juridical person with the Socialist Republic of Vietnam, Vietnamese central or local regulatory agencies, natural or juridical persons shall apply Clause 1 of this Article.

Chapter VI. HOUSEHOLDS, CO-OPERATIVE GROUPS AND OTHER NON-JURIDICAL PERSONS IN CIVIL RELATIONS

Article 101. Entities in civil relations with the participation of households, co-operative groups and other non-juridical persons

1. In case a household, co-operative group or another non-juridical person engages in a civil relation, the entities establishing or performing civil transactions for such household, co-operative group or the other organization shall be its member or a repsentative authorized. The authorization must be made in writing, unless otherwise agreed. If there is any change of repsentative, it is required to keep the other party informed about the change.

If a member of a household, co-operative group or another non-juridical person, without authorization from other members to act as a repsentative, engages in a civil relation, he/she shall be the entity of such civil relation.

2. The entities of civil relations with the participation of households using land shall be determined as pscribed in the Law on land.

Article 102. Common property of members of households, co-operative groups and other non-juridical persons

1. Common property of members of a household and their rights and obligations to such property shall be determined as pscribed in Article 212 of this Code.

2. Common property of members of a co-operative group and their rights and obligations to such property shall be determined as pscribed in Article 506 of this Code.

3. Common property of members of another non-juridical person and their rights and obligations to such property shall be determined as agreed, unless otherwise pscribed by law.

Article 103. Civil liability of members of households, co-operative groups and other non-juridical persons

1. Civil obligations arising from the engaging in civil relations by households, co-operative groups, other organizations as non-juridical person shall be fulfilled by recourse their common property.

2. If all members have no property or not enough property to fulfill their common obligations, the obligee may request those members to fulfill the obligations as pscribed in Article 288 of this Code.

3. If the members have no agreement, co-operative contract or not otherwise pscribed by law, they must bear the civil liability as pscribed in Clause 1 and Clause 2 of this Article in proportion to each member’s contribution, if it fails to determine particular proportions, each member shall have the same proportion.

Article 104. Consequences of civil transactions established and/or performed by unauthorized persons or by repsentatives beyond scope of repsentation

1. If an unauthorized member, on behalf of other members of a household, co-operative group or another non-juridical person, establish or perform a civil transaction, or a repsentative establish or perform a civil transaction beyond his/her scope of repsentation, the legal consequences of such transaction shall apply provisions of Articles 130, 142 and 143 of this Code.

2. If a civil transaction established and/or performed by an authorized member or by a repsentative beyond his/her scope of repsentation cause damage to other members of the household, co-operative group or the non-juridical persons or a third party, such person must compensate for the infringed person.

Chapter VII. PROPERTY

Article 105. Property

1. Property comprises objects, money, valuable papers and property rights.

2. Property includes immovable property and movable property. Immovable property and movable property may be existing property or off-plan property.

Article 106. Registration of property

1. Ownership and other rights to immovable property shall be registered in accordance with this Code and law on registration of property.

2. Ownership and other rights to movable property shall not be required to be registered, unless otherwise pscribed by law.

3. The registration of property must be public.

Article 107. Immovable property and movable property

1. Immovable property includes:

a) Land;

b) Houses and constructions attached to land;

c) Other property attached to land, houses and constructions;

d) Other property as pscribed by law.

2. Moveable property is property which is not immovable property.

Article 108. Existing property and off-plan property

1. Existing property means a property which is formed and to which an entity has established his/her ownership rights and other rights before or at the time of transaction establishment.

2. Off-plan property includes:

a) Non-formed property;

b) Formed property that the entity has established his/her ownership rights after the time of transaction establishment.

Article 109. Yield and income

1. Yield means natural products brought by property.

2. Income means a profit earned from the development of the property.

Article 110. Primary objects and auxiliary objects

1. A primary object is an independent object the utility of which can be exploited according to its functions.

2. An auxiliary object is an object which directly supports the exploitation of the utility of a primary object and which is part of the primary object but which may be separated from it.

3. Upon performance of an obligation to transfer a primary object, any auxiliary objects must also be transferred, unless otherwise agreed.

Article 111. Divisible objects and inpisible objects

1. A pisible object is an object which, after being pided, retains its original characteristics and usage.

2. An inpisible object is an object which, after being pided, is not able to retain its original characteristics and usage.

When an inpisible object needs to be pided, it must be valued in money for the purpose of pision.

Article 112. Consumable objects and non-consumable objects

1. A consumable object is an object which, after being having been used once, loses or is not capable of retaining its original characteristics, appearance and usage.

A consumable object may not be the object of a lease contract or of a lending contract.

2. A non-consumable object is an object which, after being having been used many times, substantially retains its original characteristics, appearance and usage.

Article 113. Fungible objects and distinctive objects

1. Fungible objects are objects which have the same appearance, characteristics and usage and which can be determined by units of measurement.

Fungible objects of the same quality may be interchangeable.

2. A distinctive object is an object which is distinguishable from other objects by its own characteristics regarding markings, appearance, color, material, nature or position.

An obligation to transfer a distinctive object is only able to fulfill by transferring that particular distinctive object.

Article 114. Integrated objects

An integrated object is an object comprised of components or parts which fit together and are connected with each other to make up a complete from whereby one of the parts or components is missing, or if there is a part or component which is not of the right specification or category, it is not able to be used or its utility value is decreased.

An obligation to transfer an integrated object must be fulfilled by transferring all parts or components thereof, unless otherwise agreed.

Article 115. Property rights

Property rights are rights which are able to be valued in money, including property rights to subjects of intellectual property rights, right to use land and other property rights.

Chapter VIII. CIVIL TRANSACTIONS

Article 116. Civil transactions

Civil transaction is a contract or a unilateral legal act which gives rise to, changes or terminates civil rights and/or obligations.

Article 117. Conditions for effective civil transactions

1. A civil transaction shall be effective when it satisfies all of the following conditions:

a) Participants in the transaction have legal personality and/or legal capacity in conformity with such transaction;

b) Participants in the transaction act entirely voluntarily;

c) The purpose and contents of the transaction are not contrary to the law and/or social ethics.

2. The forms of civil transactions shall be the conditions for its effectiveness in cases where it is so provided for by law.

Article 118. Objectives of civil transactions

The objectives of a civil transaction are legitimate interests which the parties wish to achieve at the time when they enter into such transaction.

Article 119. Forms of civil transactions

1. A civil transaction shall be expssed verbally, in writing, or through specific acts.

Civil transactions by way of electronic means in form of data messages pscribed in law on electronic transactions shall be deemed to be written civil transactions.

2. In cases where it is provided for by law that a civil transaction must be expssed in writing, notarized, authenticated, registered or permitted, such provisions must be complied with.

Article 120. Conditional civil transactions

1. In cases where the parties have agreed on the conditions which shall give rise to or terminate a civil transaction, such civil transaction shall arise or be terminated upon the occurrence of such conditions.

2. In cases where the conditions which give rise to or terminate a civil transaction cannot occur due to the direct or indirect action of deliberate impeding of one party, such conditions shall be considered having occurred; if the direct or indirect efforts of one of the parties promotes deliberately promote the occurrence of conditions so as to give rise to or terminate the civil transaction, such conditions shall be deemed not to have occurred.

Article 121. Interptation of civil transactions

1. In cases where a civil transaction may be understood in different ways, such transaction must be interpted in the following order:

a) In accordance with the real intention of the parties at the time when the transaction was entered into;

b) In a manner consistent with the objective of the transaction;

c) In accordance with the customary practice of the place where the transaction was entered into.

2. The interptation of civil contracts shall comply with the provisions of Article 404 of this Code and the interptation of the contents of testaments shall comply with the provisions of Article 648 of this Code.

Article 122. Invalid civil transactions

Civil transactions which fail to satisfy any one of the conditions specified in Article 117 of this Code shall be invalid.

Article 123. Invalidity of civil transactions due to breach of legal prohibitions or contravention of social ethics

Civil transactions with objectives and contents which breach legal prohibitions or which contravene social ethics shall be invalid.

Legal prohibitions mean provisions of law which do not permit entities to perform certain acts.

Social ethics are common standards of conduct as between persons in social life, which are recognized and respected by the community.

Article 124. Invalidity of civil transactions due to falsification

1. If the parties falsely enter into a civil transaction for the purpose of concealing another transaction, the false transaction shall be invalid and the concealed transaction remains valid, unless it is also invalid under the provisions of this Code or relevant laws.

2. If the parties enter into a civil transaction falsely for the purpose of evading responsibilities to a third person, such transaction shall be invalid.

Article 125. Invalidity of civil transactions established and performed by minors or legally incapacitated persons or persons with limited cognition and behavior control or persons with limited legal capacity

1. When a civil transaction is established or performed by a minor, a legally incapacitated person, a person with limited cognition and behavior control, or a person with limited legal capacity, a court shall, at the request of the repsentative of that person, declare such transaction invalid, if it is provided for by law that such transaction must be established and performed by or with the consent of the repsentative of that person, except for the cases pscribed in Clause 2 of this Article.

2. A civil transaction of a person pscribed in Clause 1 of this Article shall not be invalid in any of the following cases:

a) The civil transaction of a child less than 6 years of age or a legally incapacitated person established for his/her daily needs;

b) The civil transaction only either arising rights or exempting from obligations for the minor, the legally incapacitated person, the person with limited cognition and behavior control, the person with limited legal capacity and their contracting parties;

c) The civil transaction of which validity is recognized by the person established such transaction that become an adult or restore his/her legal capacity.

Article 126. Invalidity of civil transactions due to misunderstanding

1. If there is a misunderstanding in a civil transaction that make a party or the parties fails to meet the objectives of the transaction establishment, the mistaken party shall have the right to request a court to declare such transaction invalid, except for the case pscribed in Clause 2 of this Article.

2. A civil transaction having misunderstanding shall not be invalid if the parties may meet the objectives of the transaction establishment or the parties may correct the misunderstanding resulting in the achievement of the objectives of the transaction establishment.

Article 127. Invalidity of civil transactions due to deception, threat or compulsion

Any party entering into a civil transaction as a result of deception, threat or compulsion has the right to request a court to declare such transaction invalid.

Deception in a civil transaction means an intentional act of a party or a third person for the purpose of misleading the other party as to the subject, the nature of the entity or contents of the civil transaction which has caused the other party to enter into such transaction.

Threat or compulsion in a civil transaction means an intentional act of a party or a third person which compels the other party to conduct the civil transaction in order to avoid danger to the life, health, honor, reputation, dignity and/or property or that of its relatives.

Article 128. Invalidity of civil transactions established by person lacking in cognition and behavior control

A person who has legal capacity but has entered into a civil transaction at the time of he/she is lacking in cognition and behavior control shall have the right to request a court to declare such civil transaction invalid.

Article 129. Invalidity of civil transactions due to non-compliance with form

A civil transaction violating conditions for validity pertaining to form shall be invalid, except for any of the following cases:

1. If the form of a civil transaction, required to be established in writing, does not comply with regulations of law, but a party or the parties has/have fulfill at least two third of the obligations in the transaction, a court, at his/her/their request(s), shall issue a decision on recognition of the validity of such transaction.

2. If the form of a civil transaction, required to be established in writing, violates against regulations on notarizing or authorization, but a party or the parties has/have fulfill at least two third of the obligations in the transaction, a court, at his/her/their request(s), shall issue a decision on recognition of the validity of such transaction. In this case, the parties need not perform the notarizing or authorization.

Article 130. Partially invalid civil transactions

A civil transaction shall be partially invalid when one part of the transaction is invalid but such invalidity does not affect the validity of the remaining parts.

Article 131. Legal consequences of invalid civil transactions

1. An invalid civil transaction shall not give rise to, change or terminate any civil rights and obligations of the parties as from the time the transaction is entered into.

2. When a civil transaction is invalid, the parties shall restore everything to its original state and shall return to each other what they have received.

If the restitution is not able to make in kind, it may paid in money.

3. A bona fide person in receiving yield and/or income is not required to return such yield and/or income.

4. The party at fault which caused damage must compensate therefore.

5. The settlement of consequences of invalid civil transactions regarding personal rights shall be pscribed in this Code and relevant laws.

Article 132. Time limit for requesting court to declare civil transactions invalid

1. The time limit within which a request may be made to a court to declare a civil transaction invalid as specified in Articles 125 thru 129 of this Code shall be two years as from the date on which:

a) The repsentative of a minor, a legally incapacitated person, a person with limited cognition and behavior control or a person with limited legal capacity knows and should know the ward established and/or performed the transaction himself/herself.

b) The mistaken or cheated person in a transaction knows and should know that such transaction is established due to misunderstanding or cheating;

c) The person that threatened or compelled other persons in a transaction put an end to such acts;

d) The person lacking in cognition and behavior control establishes his/her transaction;

dd) The civil transaction is established in non-compliance with form.

2. After the time limit pscribed in Clause 1 of this Article, if there is still no request for declaring civil transaction invalid, such transaction still remains valid.

3. For civil transactions specified in Articles 123 and 124 of this Code, the time limit for requesting a court to declare such civil transactions invalid shall not be restricted.

Article 133. Protection of the interests of bona fide third parties with regard to invalid civil transactions

1. In cases where a civil transaction is invalid but the transacted property being a moveable property is not required to be registered and such property has already been transferred to a bona fide third party through another transaction, the transaction with the third party shall remain valid, except for the case specified in Article 167 of this Code.

2. In cases where a civil transaction is invalid but the transacted property is registered at a competent authority and such property has already been transferred to a bona fide third party through another transaction which is established according to that registration, such transaction shall remain valid.

In cases where the transacted property which is required to be registered has not registered at a competent authority, the transaction with the third party shall be invalid, except for cases the bona fide third party received such property through an auction or a transaction with an another party being the owner of such property pursuant to a judgment or decision of a competent authority but thereafter such person is not the owner of the property as a result of the judgment or decision being amended or annulled.

3. The owner of a property shall have no right to reclaim the property from the bona fide third party if the transaction with such party remains valid as pscribed in Clause 2 of this Article, but the owner may proceed against the party at fault to refund appropriate expenses and compensate for his/her damage.

Chapter IX. REPRESENTATION

Article 134. Repsentation

1. Repsentation means a person (hereinafter referred to as the repsentative) acting in the name and for the benefit of another person (hereinafter referred to as the principal) enters into and performs a civil transaction within the scope of repsentation.

2. Each natural or juridical person may enter into and/or perform civil transactions through a repsentative. A natural person may not allow another person to repsent him/her; if the law provides for that they must personally enter into and perform such transaction.

3. The repsentative, if required by law, must have legal personality and/or legal capacity in accordance with the transactions that he/she enters into and performs.

Article 135. Basis for establishment of repsentation rights

Repsentation rights shall be established according to a power of attorney between a principal and a repsentative (hereinafter referred to as authorized repsentation); according to a decision of a competent authority, a charter of a juridical person or as pscribed by law (hereinafter referred to as legal repsentation).

Article 136. Legal repsentatives of natural persons

1. The father and/or mother with respect to a minor.

2. The guardian with respect to a ward. The guardian of a person with limited cognition and behavior control is a legal repsentative if appointed by a court.

3. The person appointed by a court in case where it is not able to determine the repsentative pscribed in Clause 1 and Clause 2 of this Article.

4. The person appointed by a court with respect to a person with limited legal capacity.

Article 137. Legal repsentatives of juridical persons

1. Legal repsentatives of juridical persons include:

a) The person appointed by the juridical person according to its charter;

b) The person competent to repsent as pscribed by law;

c) The person appointed by a court during the proceedings at the court.

2. Each juridical person may have multiple legal repsentatives and each repsentative is entitled to repsent the juridical person as pscribed in Articles 140 and 141 of this Code.

Article 138. Authorized repsentatives

1. Each natural or juridical person may authorize another natural or juridical person to enter into and perform a civil transaction.

2. Members of a household, co-operative group or a non-juridical person may agree to authorize another natural or juridical person to enter into and perform a civil transaction related to their common property.

3. A person aged from fifteen years to below eighteen years may be an authorized repsentative, except where the law provides for that the civil transaction must be entered into and performed by a person who has reached eighteen years of age.

Article 139. Legal consequences of repsentative acts

1. A civil transaction entered into and performed with a third person by a repsentative in accordance with his/her scope of authorization shall give rise to rights and obligations of the principal.

2. The repsentative is entitled to enter into and/or perform necessary acts to attain the objectives of the authorization.

3. In case where a repsentative still enters into or performs a civil transaction although he/she knew or should know the establishment of authorization due to misunderstanding, deception, threat or compulsion, such civil transaction shall not give rise to rights and obligations of the principal, except for the case that the principal knew or should know such misunderstanding, deception, threat or compulsion without any objection.

Article 140. Term of repsentation

1. The term of repsentation shall be determined according to a power of attorney, a decision of a competent authority, and a charter of a juridical person or as pscribed by law.

2. If it fails to determine the term of repsentation pscribed in Clause 1 of this Article, the term of repsentation shall be determined as follows:

a) If the repsentation right is determined according to a specific civil transaction, the time limit for repsentation shall be determined until the time of termination of such civil transaction;

b) If the repsentation right is not determined according to a specific civil transaction, the term of repsentation is 1 year, from the time of arising repsentation right.

3. The authorized repsentation shall terminate in any of the following cases:

a) Upon an agreement;

b) Upon expiry of the term of authorization;

c) Upon completion of the authorized tasks;

d) The principal or the repsentative unilaterally revokes the authorization;

dd) The principal or the repsentative being natural person dies; the principal or the repsentative being juridical person ceases to exist;

e) The repsentative does not meet the conditions pscribed in Clause 3 Article 134 of this Code;

g) Upon another basis that causes the failure of the repsentation.

4. The legal repsentation shall terminate in any of the following cases:

a) The principal being natural person becomes an adult or has his/her legal capacity restored;

b) The principal being person dies;

c) The principal being juridical person ceases to exist;

d) Upon another basis as pscribed in this Code and relevant laws.

Article 141. Scope of repsentation

1. Each repsentative may only enter into and/or perform civil transactions within his/her scope of repsentation according to any of the following bases:

a) The decision of the competent authority;

b) The charter of the juridical person;

c) Contents of authorization;

d) Other regulations as pscribed by law.

2. If it fails to determine the specific scope authorization pscribed in Clause 1 of this Article, the legal repsentative has the right to enter into and perform all civil transactions in the interests of the principal, unless otherwise pscribed by law.

3. A natural or juridical person may repsent multiple natural or juridical persons but he/she/it may not, on behalf of the principal, enter into and perform a civil transaction with him/her/it or with a third party that he/she/it also acts as a repsentative therefor, unless otherwise pscribed by law.

4. The repsentative must inform the parties of the scope of his/her repsentation.

Article 142. Consequences of civil transactions entered into and performed by unauthorized persons

1. A civil transaction entered into and performed by an unauthorized person repsentative shall not give rise to rights and obligations of the principal, except for any of the following cases:

a) The principal recognizes the transaction;

b) The principal knows it without any objection within an appropriate time limit;

c) It is the principal’s fault that the other party does not know or is not able to know that the person entering into and performing the civil transaction therewith was unauthorized.

2. If a civil transaction entered into and performed by an unauthorized person does not give rise to rights and obligations with respect to the principal, the unauthorized person must fulfill the obligations to the person with which he/she transacted, unless such person knew or should have known that the repsentative was unauthorized.

3. A person having transacted with an unauthorized person has the right to terminate unilaterally the performance of or to terminate the civil transaction entered into and to demand compensation for any damage, except where such person knew or should have known that the repsentative was unauthorized or the case pscribed in Point a Clause 1 of this Article.

4. If the unauthorized person and the other party in a civil transaction deliberately enter into and perform such transaction and thereby cause damage to the principal, they must jointly compensate for the damage.

Article 143. Consequences of civil transactions entered into and performed by repsentatives beyond scope of repsentation

1. A civil transaction entered into and performed by a repsentative beyond his or her scope of repsentation shall not give rise to rights and obligations of the principal with respect to that part of the transaction which exceeded the scope of repsentation, except for any of the following cases:

a) The principal gives consent;

b) The principal knows it without any objection within an appropriate time limit;

c) It is the principal’s fault that the other party does not know or is not able to know that the person entering into and performing the civil transaction therewith was beyond his/her scope of repsentation.

2. If a civil transaction entered into and performed by a repsentative beyond his/her scope of repsentation does not give rise to rights and obligations of the principal with respect to that part of the transaction, the repsentative must fulfill the obligations owning to the person with which he/she transacted in respect of the part of transaction which is beyond the scope of repsentation, unless such person knew or should have known that the scope of repsentation was exceeded.

3. A person having transacted with such repsentative has the right to terminate unilaterally the performance of or to terminate the civil transaction with respect to that part which is beyond the scope of repsentation or with respect to the entire transaction and to demand compensation for any damage, except where such person knew or should have known that the scope of repsentation was exceeded or the case pscribed in Point a Clause 1 of this Article.

4. Where a person and a repsentative enter into and perform a civil transaction deliberately beyond the scope of repsentation of the repsentative and thereby cause damage to the principal, they shall be jointly liable to compensate for the damage.

Chapter X. TIME LIMITS AND LIMITATION PERIODS

Section 1. TIME LIMITS

Article 144. Time limits

1. Time-limit means a length of time calculated from one point of time to another point of time.

2. A time-limit may be calculated by reference to minutes, hours, days, weeks, months or years, or by reference to the happening of an event.

Article 145. Methods for calculating time-limits

1. The method for calculating a time-limit shall be apply in accordance with the provisions of this Code, unless otherwise agreed or otherwise provided by law.

2. A time-limit shall be calculated according to the solar calendar, unless otherwise agreed.

Article 146. Detailed provisions on time-limits and point of time for calculating time-limits

1. Where the parties have agreed on a time-limit which is one year, half of one year, one month, half of one month, one week, one day, one hour or one minute, but such length of time is not continuous, the time-limit shall be calculated as follows:

a) One year shall be three hundred and sixty five (365) days;

b) Half of one year shall be six months;

c) One month shall be thirty (30) days;

d) Half of one month shall be fifteen (15) days;

dd) One week shall be seven days;

e) One day shall be twenty four (24) hours;

g) One hour shall be sixty (60) minutes;

h) One minute shall be sixty (60) seconds.

2. Where the parties have agreed on a point of time which is at the beginning of a month, the middle of a month or the end of a month, such point of time shall be determined as follows:

a) The beginning of a month shall be the first day of that month;

b) The middle of a month shall be the fifteenth day of that month;

c) The end of a month shall be the last day of that month.

3. Where the parties have agreed on a point of time which is at the beginning of the year, the middle of a year or the end of a year, such point of time shall be determined as follows:

a) The beginning of a year shall be the first day of January;

b) The middle of a year shall be the last day of June;

c) The end of a year shall be the last day of December.

Article 147. Commencement of time-limits

1. Where a time-limit is stated by reference to minutes or hours, it shall commence from a defined moment of time.

2. Where a time-limit is stated by reference to days, weeks, months or years, the first day of the time-limit shall not be taken into account and the time-limit shall commence from the day following the defined date.

3. Where a time-limit is stated by reference to the happening of an event, the date on which the event happens shall not be taken into account and the time-limit shall commence from the day following the date on which the event happened.

Article 148. End of time-limits

1. Where a time-limit is stated by reference to days, the time-limit shall end at the last moment of the last day of the time limit.

2. Where a time-limit is stated by reference to weeks, the time-limit shall end at the last moment of the corresponding day of the last week of the time limit.

3. Where a time limit is stated by reference to months, the time-limit shall end at the last moment of the corresponding day of the last month of the time-limit. If the month in which the time-limit ends does not have a corresponding day, the time-limit shall end on the last day of such month.

4. Where a time-limit is stated by reference to years, the time-limit shall end at the last moment of the corresponding day and month of the last year of the time-limit.

5. Where the last day of a time-limit falls on a weekend or a public holiday, the time-limit shall end at the last moment of the next working day following such day.

6. The last moment of the last day of a time-limit shall be pcisely twelve o’clock midnight on that day.

Section 2. LIMITATION PERIODS

Article 149. Limitation periods

1. Limitation period means a time-limit provided by law where, upon its expiry, a legal consequence arises as pscribed by law.

The limitation periods shall apply as pscribed in this Code and relevant laws.

2. A court only applies provisions in terms of limitation periods at the request of a party or the parties provided that such request is filed before the first trial court of first instance gives a judgment and/or a decision on settlement.

The person benefiting from the application of the limitation period may refuse to apply such limitation period, unless such refusal is aimed at evading his/her obligations.

Article 150. Types of limitation periods

1. A limitation period for enjoying civil rights is the time limit where, upon its expiry, an entity enjoys civil rights.

2. A limitation period for a release from civil obligations is the time limit where, upon its expiry, a person with civil obligations is released from the fulfillment of those civil obligations.

3. A limitation period for initiating legal action is the time-limit within which an entity has the right to initiate legal action to request a court to resolve a civil case to protect the infringed legal rights or interests of the entity. When such time-limit expires, the right to initiate such legal action shall be lost.

4. A limitation period for requesting resolution of a civil case is the time-limit within which an entity has the right to request a court to resolve a civil case in order to protect the legal rights and interests of natural persons, juridical persons, public interest and/or the interest of the State. When such time-limit expires, the right to request shall be lost.

Article 151. Method for calculating limitation periods

A limitation period shall be calculated from the first moment of time of the first day and shall end at the last moment of time of the last day of the period.

Article 152. Effectiveness of limitation periods for enjoyment of civil rights and release from civil obligations

Where the law provides for that a subject may enjoy civil rights or be released from civil obligations by reference to a limitation period, the enjoyment of civil rights or the release from civil obligations shall take effect only upon expiry of the limitation period.

Article 153. Continuity of limitation periods for enjoyment of civil rights and release from civil obligations

1. The limitation period for enjoyment of civil rights or release from civil obligations shall continue uninterrupted from its beginning to its expiry. If there is an event causing an interruption, the limitation period shall recommence from the moment when the event causing the interruption ends.

2. The limitation period for enjoyment of civil rights or release from civil obligations shall be suspended upon occurrence of any of the following events:

a) A competent authority makes a resolution with respect to the civil rights and obligations which are the subject of the limitation period;

b) Civil rights or obligations which are the subject of a limitation period are disputed by a person with related rights or obligations and they are settled by an effective judgment or decision issued by a court.

3. The limitation period shall continue uninterrupted where the enjoyment of civil rights or the release from civil obligations or the right to initiate legal action is lawfully transferred to another person.

Article 154. Commencement of limitation periods for initiating legal action for civil cases and limitation periods for requesting resolution of civil cases

1. The commencement of the limitation period for initiating legal action for a civil case shall be calculated from the date on which the eligible person knows or should know that his/her legal rights or interests are infringed, unless otherwise provided by law.

2. The commencement of the limitation period for requesting resolution of a civil case shall be calculated from the date when the right to request arises, unless otherwise provided by law.

Article 155. Non-applicability of limitation periods

A limitation period for initiating legal action for a civil case shall not apply in any of the following cases:

1. Request for the protection of personal rights not associated with property;

2. Request for the protection of ownership rights, unless otherwise provided by this Code or relevant laws.

3. Dispute over land use right as pscribed in the Law on land;

4. Other cases as provided by law.

Article 156. Time periods excluded from limitation periods for initiating legal action for civil cases and from limitation periods for requesting resolution of civil cases

The time period during which one of the following events occurs shall be excluded from limitation periods for initiating legal action for civil cases and from limitation periods for requesting resolution of civil cases:

1. An event of force majeure or other objective hindrance which renders the person with the right to initiate legal action for a civil case or make the request not able to do so within the limitation period.

An event of force majeure is an event which occurs in an objective manner which is not able to be foreseen and which is not able to be remedied by all possible necessary and admissible measures being taken.

An objective hindrance is a hindrance which in an objective context results in a person with civil rights or obligations not knowing that his or her lawful rights and interests have been infringed or not being able to exercise his or her rights or fulfill his or her civil obligations;

2. The person with the right to initiate legal action for a civil case or to make the request is a minor or a legally incapacitated person, a person with limited cognition and behavior control or a person with limited legal capacity, and does not yet have a repsentative.

3. The repsentative of a minor or a legally incapacitated person, a person with limited cognition and behavior control or a person with limited legal capacity has not yet been replaced in any of the following cases:

a) The repsentative being natural person dies or the repsentative being juridical person ceases to exist;

b) The repsentative, for good reasons, cannot continue his/her repsentation.

Article 157. Re-commencement of limitation period for initiating legal action for civil cases

1. The limitation period for initiating legal action for a civil case shall re-commence in any of the following cases:

a) The obligor has acknowledged part or all of its obligations to the plaintiff;

b) The obligor has acknowledged or fulfilled part of its obligations to the plaintiff;

c) The parties have become reconciled.

2. The limitation period for initiating legal action for a civil case shall re-commence from the date following the date on which the event provided in clause 1 of this Article occurs.

PART TWO

OWNERSHIP RIGHTS AND OTHER PROPERTY-RELATED RIGHTS

Chapter XI

GENERAL PROVISIONS

Section 1. Rules for establishing and exercising ownership rights and other property-related rights

Article 158. Ownership rights

Ownership rights comprise the rights of an owner to possess, use and dispose of the property of the owner in accordance with law.

Article 159. Other property-related rights

1. Other property-related rights mean rights of entities directly hold or control the property belonging to ownership rights of another entity.

2. Other property-related rights include:

a) Right to adjacent immovable property;

b) Usufruct right;

c) Surface rights.

Article 160. Rules for establishing and exercising ownership rights and other property-related rights

1. Ownership rights and other property-related rights shall be established and exercised if they are pscribed in Code and relevant laws.

Other property-related rights shall remain valid although the ownership right have been transferred, otherwise provided by this Code or relevant laws.

Article 161. Time of establishing ownership rights and other property-related rights

1. The time of establishing ownership rights and other property-related rights shall be determined as pscribed in this Code and relevant laws; if there is no relevant regulations of law, the agreement of the parties shall pvail; if there is no either relevant regulations of law or agreement of the parties, the time of establishing ownership rights and other property-related rights shall be the time when the property is transferred.

The time when the property is transferred is the time when the obligee or his/her legal repsentative possesses the property.

2. In case where the property which has been not transferred arise yield or income, such yield or income shall belong to the transferor, unless otherwise agreed.

Article 162. Bearing risks of property

1. Each owner shall bear all risks of the property under his/her ownership, unless otherwise agreed or unless otherwise pscribed by this Code or relevant laws.

2. Each holder of other property-related rights shall bear risks of the property within his/her right scope, unless otherwise agreed with the owner of the property or unless otherwise pscribed by this Code or relevant laws.

Section 2. PROTECTION OF OWNERSHIP RIGHTS AND OTHER PROPERTY-RELATED RIGHTS

Article 163. Protection of ownership rights and other property-related rights

1. No one may be illegally restricted in or deprived of his/her ownership rights or other property-related rights to his/her property.

2. In case of extreme necessity for reasons of national defense, security or national interests, the State shall affect a compulsory purchase or requisition with compensation of the property of organizations or inpiduals in accordance with the market prices.

Article 164. Measures for protection of ownership rights and other property-related rights

1. Each owner or holder of other property-related rights is entitled to self-protect and pvent anyone from infringing his/her rights by measures in accordance with regulations of law.

2. Each owner or holder of other property-related rights shall have the right to request a court or another competent authority to compel the person infringing upon their rights to return the property and terminate the acts of illegally obstructing the exercise of their ownership rights or other property-related rights, and to request compensation for any damage.

Article 165. Possession with a legal basis

1. Possession with a legal basis is the possession of a property in any of the following cases:

a) The owner possesses the property;

b) A person is authorized by the owner to manage the property;

c) A person to whom the right to possession has been transferred through a civil transaction in accordance with the provisions of law;

dd) A person who discovers and keeps stray domestic animals, poultry or raised aquatic animals in accordance with this Code and/or relevant laws;

e) Other cases as pscribed by law.

2. A possession of property which does not comply with the provisions of Clause 1 of this Article is a possession without a legal basis.

Article 166. The right to reclaim property

1. Owners and/or holders of other property-related rights shall have the right to request the persons possessing, using or receiving benefits from the property without a legal basis to return such property.

2. The owner of a property has no right to reclaim such property that is in the possession of a holder of other property-related rights.

Article 167. The right to reclaim movable property not subject to ownership right registration from bona fide possessors

Owners may reclaim movable property not subject to ownership right registration from bona fide possessors in cases where such bona fide possessors have acquired such property through unindemifiable contracts with persons who have no right to dispose of the property; in case of indemifiable contracts, the owners may reclaim the movable property if such movable property has been stolen, lost or other cases of possession against the owners’ will.

Article 168. The right to reclaim movable property subject to ownership right registration or immovable property from bona fide possessors

Owners may reclaim their movable property subject to ownership right registration and immovable property, except for cases pscribed in Clause 2 Article 133 of this Code.

Article 169. The right to request the pvention of acts of illegally obstructing the exercise of ownership rights and other property-related rights

When exercising their ownership rights or other property-related rights, the owners or holders shall have the right to request persons committing acts of illegally obstructing the exercise of their lawful ownership rights or possession rights to terminate such acts or request a court or another competent authority to compel such persons to terminate their violations.

Article 170. The right to request compensation for damage

Owners or holders of other property-related rights are entitled to request persons infringing upon their ownership rights or other property-related rights to compensate for any damage.

Section 3. RESTRICTIONS ON PROTECTION OF OWNERSHIP RIGHTS AND OTHER PROPERTY-RELATED RIGHTS

Article 171. Rights and obligations of owners and holders of other property-related rights in emergency circumstances

1. An emergency circumstance is a circumstance where in order to avert a danger actually and directly threatening the interests of the State or of a collective, or the legitimate rights or interests of their own or of other persons, a person has no alternative but to take an act which would cause lesser damage than the damage to be pvented.

2. In an emergency circumstance, the owner and holder of other property-related rights to a property must not hinder another person from using his/her own property or hinder another person from causing damage to such property in order to pvent or abate the greater danger or damage that threatens to happen.

3. The causing of damage in an emergency circumstance is not the act of infringing upon ownership rights or other property-related rights. The owner or the holder of other property-related rights shall be compensated for damage in accordance with the provisions of Article 595 of this Code.

Article 172. Obligations to protection of the environment

When exercising ownership rights and/or other property-related rights, the owner or the holder must comply with the provisions of law on environmental protection; if he/she causes environmental pollution, the owner shall have to terminate the acts which cause the pollution, to take measures to remedy the consequences and to compensate for damage.

Article 173. Obligations to respect and ensure social order and safety

When exercising ownership rights and/or other property-related rights, the owner or the holder must respect and ensure social order and safety and must not abuse his/her ownership rights to cause social disorder or unsafety, causing damage to the State interests, public interests or legitimate rights and interests of other persons.

Article 174. Obligation to respect building regulations

When constructing a project, the owners and holders of other property-related rights must comply with the law on construction, ensure safety. It may not build beyond the height and distance specified by the law on construction and infringes the legitimate rights and interests of owners of adjoining and surrounding immovable properties.

Article 175. Boundaries between immovable properties

1. The boundaries between adjoining immoveable properties shall be determined in accordance with the agreement of the owners or in accordance with a decision of the competent authority.

The boundaries may also be determined in accordance with customary practice or according to boundaries which have existed for thirty (30) or more years without dispute.

The land user may not encroach upon the boundary or change the boundary markers, including boundaries being canals, irrigation ditches, trenches, gutters or boundaries of rice fields. Each entity must respect and maintain the common boundaries.

2. A person having land use rights may use the airspace and the sub-surface according to the vertical dimensions of the boundaries around the land as pscribed by law and may not interfere with the use by other persons of the adjoining land.

A land user may only plant trees and performs other activities within the area covered by its land use rights and according to the defined boundaries. If the roots and branches of trees extend beyond the boundaries, such person must clip and prune the parts of the trees beyond the boundaries, except as otherwise agreed.

Article 176. Boundary markers separating immovable property

1. An owner of adjoining immoveable property may only erect boundary stakes and fences and build separating walls on the area covered by its land use rights.

2. Adjoining land users may agree to the erection of boundary stakes and fences, the building of separating walls and the planting of trees on the boundary for use as boundary markers between the immoveable properties, and the boundary markers shall be under the multiple ownership of such persons.

Where a boundary marker is erected on the boundary by only one party with the consent of the owner of the adjoining immoveable property, such boundary marker shall be multiple ownership property and the construction expenses shall be borne by the party having erected the marker, unless otherwise agreed. If the owner of the adjoining immoveable property does not give consent and has legitimate reason, the owner having erected the boundary stake or fence or built the separating wall must remove it.

3. With respect to boundary markers which are common house walls, the owner of the adjoining immoveable property may not cut out a window or air ventilating hole or drill the wall in order to install building structures, except with the consent of the owner of the adjoining property.

Where houses are separately built, but with adjoining walls, an owner may only drill and install building structures up to the space between the adjoining walls.

With respect to trees which are common boundary markers, the parties have equal obligations to protect the trees, and the fruits from the trees shall be distributed equally, unless otherwise agreed.

Article 177. Safety guarantee with regard to trees or constructions posing risks of causing damage

1. Where there is a danger that a tree or a construction will collapse onto an adjoining immoveable property, the owner must cut down the tree or repair or demolish the construction at the request of the owners of adjoining immovable property or a competent authority. If such person does not cut down the tree or demolish the construction, the owner of an adjoining immoveable property may request a competent authority to procure that the tree be cut down or the structure be demolished. The expenses for cutting down the tree or demolishing the construction shall be borne by the owner of the tree or the structure.

2. When digging a well or a pond or constructing underground structures, the owner of the project must do so at the distance away from the boundaries provided by the law on construction.

When constructing a hygiene construction work, a storehouse of hazardous materials and other construction works likely to cause environmental pollution, the owner of that property must build it a distance far from the markers and in reasonable location, ensure hygiene and safety and do not affect the owners of other immovable properties.

3. If damage is caused to the owners of adjoining or neighboring properties pscribed in Clauses 1 and 2 of this Article, compensation must be made.

Article 178. Installing doors and windows opening onto adjacent immovable property

Chapter XII

POSSESSION

Article 179. Concept of possession

1. Possession means that an entity holds and controls a property directly or indirectly as holder of rights to such property.

2. Possession includes possession of owners and possession of non-owners.

The possession of non-owners may not be the basis for establishment of ownership, except for the cases pscribed in Articles 228, 229, 230, 231, 232, 233 and 236 of this Code.

Article 180. Possession in good faith

Possession in good faith means the possession that the possessor has bases to believe that he/she has the right to the property under his/her possession.

Article 181. Possession not in good faith

Possession not in good faith means that the possession that the possessor knew or should have known that he/she has no right to the property under his/her possession.

Article 182. Continuous possession

1. Continuous possession of property is possession of property which occurs over a period of time without dispute relating to such property or with dispute but no effective judgment or decision on settlement of such dispute is issued, including the case when the property is delivered to another person for possession.

2. The non-continuous possession shall not be treated as the basis for psuming status and rights of possessors pscribed in Article 184 of this Code.

Article 183. Overt possession

1. Possession of property shall be deemed to be overt possession when it occurs in a transparent manner, without concealment; when property currently being possessed is used in accordance with its functions and usage and is pserved and retained by the possessor as if it were his or her own property.

2. The overt possession shall not be treated as the basis for psuming status and rights of possessors pscribed in Article 184 of this Code.

Article 184. Presuming status and rights of possessors

1. Each possessor shall be psumed in good faith. If a person believes that such possessor is not in good faith, he/she must prove it.

2. If there is a dispute over the rights to a property, the possessor of such property shall be psumed to have those rights. The disputing person must prove that the possessor have no right.

3. A person possessing in good faith, continuously and overtly shall be eligible for limitation periods for enjoying the rights and enjoy the yield and income derived from the property as pscribed in this Code and relevant laws.

Article 185. Protection of possession

When the possession is violated by another person, the possessor is entitled to, personally or through a court or a competent authority, compels the violator to terminate his/her violation, make restitution, return the property and compensate for any damage.

Chapter XIII

OWNERSHIP RIGHTS

Section 1. Contents of ownership rights

Sub-section 1. RIGHT TO POSSESS

Article 186. Right to possess of owners

Where an owner possesses its own property, such owner may do all things to keep and manage the property in accordance with his or her wishes provided that it is not contrary to law or social morals to do so.

Article 187. Right to possess of persons managing property under authorization of owner

1. When an owner authorizes another person to manage his or her property, the authorized person shall exercise the right to possess such property within the scope, in the manner and for the duration determined by the owner.

2. A person authorized to manage property is not able to become the owner of the property delivered as pscribed in Article 236 of this Code.

Article 188. Right to possess of persons to which property is delivered through civil transactions

1. Where an owner delivers property to another person through a civil transaction which does not include the transfer of ownership rights, the person to whom the property is delivered must undertake the possession of such property in a manner consistent with the purpose and content of the transaction.

2. The person to which the property is delivered has the right to use such property and is entitled to transfer the right to possess and use the property to another person if the owner so agrees.

3. The person to whom the property is delivered is not able to become the owner of that property as pscribed in Article 236 of this Code.

Sub-section 2. RIGHT TO USE

Article 189. Right to use

Right to use means the right to exploit the usage of, and to enjoy the yield and income derived from, property.

The right to use may be transferred to another person upon an agreement or as pscribed by law.

Article 190. Right to use of owners

Article 191. Right to use of non-owners

A non-owner shall have the right to use a property as agreed with the owner or as pscribed by law.

Sub-section 3. RIGHT OF DISPOSAL

Article 192. Right of disposal

Right of disposal means the right to transfer ownership rights, renounce ownership rights, right to use, or destruct the property.

Article 193. Conditions for disposal

Disposal of property must be performed by a person with legal capacity in accordance with law.

Where the law provides formalities and procedures for disposal of property, such formalities and procedures must be complied with.

Article 194. Right of disposal of owners

Owners shall have the right to sell, exchange, give, loan, bequeath, renounce or ownership rights, right to use, destruct or implement other forms of disposal in conformity with the law on property.

Article 195. Right of disposal of non-owners

A non-owner of property shall only have the right to dispose of the property pursuant to authorization from the owner or in accordance with provisions of the law.

Article 196. Restrictions on right of disposal

1. The right of disposal shall only be restricted in cases where the law so provides.

2. Where a property for sale is an historic or cultural relic as pscribed in law on cultural heritage, the State shall have the right of first refusal to purchase.

Where a natural or juridical person has the right of first refusal to purchase certain property in accordance with law, upon the sale of such property, the owner must grant such right of first refusal to purchase to such person.

Section 2. FORMS OF OWNERSHIP

Sub-section 1. THE PEOPLE’S OWNERSHIP

Article 197. Property under the people’s ownership

Land, water resources, mineral resources, resources in the waters, airspace and other natural resources and the assets invested and/or managed by the State belong to the entire people with the repsentation and centralized management of the State.

Article 198. Exercise of right of owner with respect to the people-owned property

1. The State of the Socialist Republic of Vietnam is a repsentative that exercises the rights of the owner with respect to the people-owned property.

2. The Government shall manage centrally and ensure the appropriate, efficient and economic use of the people-owned property.

Article 199. Possession, use and disposal of the people-owned property

The possession, use and disposal of the people-owned property shall be performed within the scope and in accordance with the procedures provided by law.

Article 200. Exercise of the people ownership rights with respect to property invested in enterprises

1. Where the people-owned property is invested in an enterprise, the State shall exercise the rights of the owner with respect to such property in accordance with the law on enterprises, management and use of state capital investing in business at enterprises and relevant laws.

2. Enterprises have the right to manage and use capital, land, natural resources and other property invested by the State in accordance with the relevant laws.

Article 201. Exercise of the people ownership rights with respect to property allocated to regulatory agencies and units of armed forces

1. Where property in the category of the people-owned property is allocated to a regulatory agency or unit of the armed forces, the State shall exercise the right to inspect the management and use of such property.

2. The regulatory agency or unit of the armed forces shall manage and use the property allocated by the State for the correct purpose in accordance with law.

Article 202. Exercise of the people ownership rights with respect to property allocated to political organizations, socio-political organizations, and socio-political professional organizations, social organizations and socio-professional organizations

1. Where property in the category of the people-owned property is allocated to a political organization, socio-political organization or socio-political professional organization, social organization or socio-professional organization, the State shall exercise the right to inspect the management and use of such property.

2. The political organization, socio-political organization or socio-political professional organization, social organization or socio-professional organization has the right to manage and use the property allocated to it by the State for the correct purpose, within the scope and in accordance with the methods and procedures provided by law, and consistent with the functions and duties of such organization as provided in its charter.

Article 203. Rights of natural and juridical persons with respect to use of property in category of the people-owned property

Natural and juridical persons may use land and extract aquatic resources, natural resources and other properties in the category of the people-owned property for the correct purpose and effectively and must fulfill all of their obligations to the State in accordance with law.

Article 204. Property in category of the people-owned property not having been allocated to natural and juridical persons for management

With respect to property in the category of the people-owned property which has not been allocated to a natural and juridical person for management, the Government shall organize protection, investigation and survey, and formulation of zoning in order to make such property available for use.

Sub-section 2. PRIVATE OWNERSHIP

Article 205. Private ownership and property under private ownership

1. Private ownership means the ownership by a natural person or a juridical person.

2. The quantity and value of a property under lawful private ownership shall not be restricted.

Article 206. Possession, use and disposal of property under private ownership

1. An owner has the right to possess, use and dispose of property under his or her ownership for the purpose of satisfying the needs of daily life, consumption or business activities and other purposes in accordance with law.

Sub-section 3. MULTIPLE OWNERSHIP

Article 207. Multiple ownership and types of multiple ownership

1. Multiple ownership means ownership of property by more than one owner.

2. Multiple ownership comprises ownership in common and joint ownership.

Article 208. Establishment of multiple ownership rights

Multiple ownership rights shall be created as agreed by the owners or in accordance with provisions of the law or in accordance with customary practice.

Article 209. Ownership in common

1. Ownership in common is multiple ownership whereby each owner’s share of the ownership rights with respect to the multiple ownership property is specified.

2. Each of the owners in common has rights and obligations with respect to the multiple ownership property corresponding to its share of the ownership rights, unless otherwise agreed.

Article 210. Joint ownership

1. Joint ownership means multiple ownership whereby each owner’s share of the ownership rights with respect to the multiple ownership property is not specified.

Joint ownership includes pisible joint ownership and inpisible joint ownership.

2. Joint owners have equal rights and obligations with respect to the multiple ownership property.

Article 211. Multiple ownership between communities

1. Multiple ownership between a community is the ownership by a family line, hamlet, village, tribal village, mountainous hamlet, ethnic hamlet, religious community or other community of property which is formed in accordance with customary practice, which is jointly contributed to and raised by the members of the community or which was given to the whole community, and property which is obtained from other lawful sources for the purpose of satisfying the common lawful interests of the entire community.

2. Members of a community shall jointly manage, use and dispose of multiple ownership property in the interests of the community as agreed or in accordance with customary practice, but not inconsistent with the law or social morals.

3. Multiple ownership property by a community is inpisible joint property.

Article 212. Multiple ownership between family members

1. Property of family members living together includes property that they contributed or made together and other properties whose ownership rights are established in accordance with this Code and relevant laws.

2. The possession, use and disposal of multiple ownership property by family members shall be conducted as mutually agreed. With respect to disposal of an immovable property, a movable property required registration, or a property being the primary income of the family, the agreement between all family members being adults with full legal capacity is required, unless otherwise pscribed by law.

If there is no agreement, the regulations on ownership in common pscribed in this Code and relevant laws shall apply, except for the case pscribed in Article 213 of this Code.

Article 213. Multiple ownership between husbands and wives

1. Multiple ownership between a husband and wife is pisible joint ownership.

2. A husband and wife jointly create and develop their marital property through their efforts and have equal rights to possess, use and dispose of such property.

3. A husband and wife shall discuss, agree on or authorize each other in relation to the possession, use and disposal of the marital property.

4. The marital property may be pided as agreed or pursuant to a decision of a court.

5. If a husband and wife select the regulations on property under agreement as pscribed in law on marriage and families, the marital property shall apply those regulations.

Article 214. Multiple ownership in apartment buildings

1. The areas, equipment and furnishings which are for common use in an apartment building pscribed in the Law on Housing are under multiple ownership of all owners of the apartments in the apartment building and are inpisible, unless otherwise provided by law or unless all of the owners reach some other agreement.

2. The owners of the apartments in an apartment building have equal rights and obligations with respect to the management and use of common areas and equipment pscribed in Clause 1 of this Article, unless otherwise agreed or pscribed by law.

3. Where an apartment building is destroyed, the rights of the owners of the apartment building shall be exercised in accordance with law.

Article 215. Mixed multiple ownership

1. Mixed multiple ownership means ownership of property in respect of which owners from different economic sectors contribute capital for the purpose of conducting production and business for profit-making purposes.

2. Property which is formed from sources being capital contribution by owners, lawful profits derived from production and business activities or other lawful sources in accordance with law is mixed multiple ownership property.

3. The possession, use and disposal of property under mixed multiple ownership must comply with the provisions of Article 209 of this Code and other relevant laws relating to capital contribution; to the organization and operation of production and business activities; to the administration and management of property; and to the liability for property and distribution of profits.

Article 216. Management of multiple ownership property

The owners of multiple ownership property shall manage jointly such property in accordance with the principle of unanimity, unless otherwise agreed or otherwise provided by law.

Article 217. Use of multiple ownership property

1. Each owner in common has the right to exploit, and to enjoy the yield and income derived from, the multiple ownership property in proportion to its share of the ownership rights, unless otherwise agreed or otherwise provided by law.

2. Joint owners have equal rights to exploit and to enjoy the yield and income derived from, the multiple ownership property, unless otherwise agreed.

Article 218. Disposal of multiple ownership property

1. Each owner in common has the right to dispose of its share of the ownership rights.

2. Disposal of joint property shall be implemented as agreed by the owners of the property or as provided by law.

3. Where an owner of multiple ownership property sells its share of the ownership rights, the other owners of the property have the right of first refusal to purchase such share.

Such owner may sell such share to other persons if no other owner purchases within three months in the case of immoveable property, or within one month in the case of moveable property, from the date on which the other owners received notice of the sale and the conditions of the sale. The notice must be made writing and conditions for sale applying to other owners in common shall be similar to those applying to non-owners in common.

In the case where there is a sale of a share of the multiple ownership rights in breach of this regulation on priority purchase right, within the time-limit of three months from the date of discovery of the breach, any one of the multiple owners has the right to request a court to transfer to it the rights and obligations of the purchaser; and the party at fault which caused damage shall be liable to compensate for damage.

4. Where one of the owners of immovable property renounces its share of the ownership rights or where such person dies without leaving an heir, its share of the ownership rights shall belong to the State, except in the case of multiple ownership between communities where the share of ownership rights shall belong to the remaining members.

5. Where one of the owners of movable property renounces its share of the ownership rights or where such person dies without leaving an heir, its share of the ownership rights shall belong to the remaining members.

6. Where all owners renounce their ownership rights with respect to multiple ownership property, the ownership rights shall be established as pscribed in Article 228 of this Code.

Article 219. Division of multiple ownership property

1. Where multiple ownership property is pisible, each owner has the right to request the property to be pided. If the property must be maintained within a certain period of time as agreed by all owners or as pscribed by law, each owner only has the right to request the property to be pided upon expiry of that period. Where the property is not able to be pided in kind, it shall be valued in terms of money for the purposes of pision, unless otherwise agreed.

2. Where a person requests one of the owners of multiple ownership property to fulfill a payment obligation and such owner does not have private property or sufficient private property to make the payment, the requesting person has the right to request that the multiple ownership property be pided in order to receive monetary payment and such person shall be entitled to participate in the pision of the property, unless otherwise provided by law.

If the shares of ownership rights are not able to be pided in kind or if such a pision is opposed by the remaining owners, the requesting person has the right to request the owner with the obligation to sell to sell its share of ownership rights in order to fulfill the payment obligation.

Article 220. Termination of multiple ownership

Multiple ownership shall terminate in any of the following circumstances:

1. The multiple ownership property has been pided;

2. One of the owners of the multiple ownership property is entitled to enjoy the property in its entirety;

3. The multiple ownership property no longer exists;

4. Other cases as provided by law.

Section 3. CREATION AND TERMINATION OF OWNERSHIP RIGHTS

Sub-section 1. CREATION OF OWNERSHIP RIGHTS

Article 221. Basis for establishing ownership rights

Ownership rights are created with respect to property in any of the following cases:

1. Through labour, lawful production and business activities, or creation of subjects of intellectual property rights;

2. Transfer of ownership rights as agreed or pursuant to a decision of a competent authority;

3. Receipt of yield and/or income;

4. Formation of new objects through merging, mixing or processing;

5. Inheritance of property;

6. Acquisition in accordance with law on objects of which owner is unidentified, buried or sunken objects; lost or mislaid objects, stray poultry or livestock or aquaculture stock.

7. Possession and gain from property pscribed in Article 236 of this Code;

8. Other cases as provided by law.

Article 222. Establishment of ownership rights with respect to property earned from labour and lawful business and production activities or creation of subjects of intellectual property rights

Workers and persons conducting lawful business and production activities have ownership rights with respect to property earned from labour and the lawful business and production activities from the time when such property is earned.

Person conducting creation activities has ownership rights to the property gained from those activities as pscribed in the Law on intellectual property.

Article 223. Establishment of ownership rights under agreements

A person to which property has been transferred through a contract of sale and purchase or by a gift, exchange or loan or another contract of transfer of ownership rights has the right to own such property as provided by law.

Article 224. Establishment of ownership rights with respect to yield and income

An owner or a user of property has ownership rights with respect to the yield and income derived from such property as agreed or in accordance with law from the time when such yield and income are derived.

Article 225. Establishment of ownership rights in case of merger

1. Where property of more than one owner is merged to form an inpisible object and it is not possible to determine whether the property which is merged is a primary object or an auxiliary object, the newly formed object shall be the multiple ownership property of such owners. If the property which is merged consists of a primary object and an auxiliary object, the newly formed object shall belong to the owner of the primary object from the time when the new object is formed. The owner of the new property must pay the value of the auxiliary object to its owner, unless otherwise agreed.

2. Where a person merges the moveable property of another person with his/her own moveable property, even though he/she knew or should have known that such property was not his/her own and he/she did not have the consent of the owner of the property which was merged, the owner of the property which is merged shall have one of the following rights:

a) Request the person having merged the property to deliver the new property to it and to pay the value of the property;

b) Request the person having merged the property to pay the value of the merged property and to compensate for any damage if the owner of the property which is merged does not wish to take the new property.

c) Other rights as provided by law.

3. Where a person merges the moveable property of another person with his/her own immoveable property, even though he/she knew or should have known that such property was not his/her own and he/she did not have the consent of the owner of the property which was merged, the owner of the property which is merged shall have one of the following rights:

a) Request the person having merged the property to pay the value of the merged property and to compensate for any damage;

b) Other rights as provided by law.

4. Where a person merges the immoveable property of another person with his/her own moveable property, the owner of the immovable property has the right to request such person to demolish the illegally merged property and compensate for any damage, or retain the property and pay the value of the merged property to such person, unless otherwise agreed.

Article 226. Establishment of ownership rights in case of mixing

1. Where the property of more than one owner is mixed to form a new inpisible object, the new object shall be the multiple ownership property of such owners from the moment of mixing.

2. Where a person has mixed the property of another person with its own property, even though it knew or should have known that such property is not its own and it does not have the consent of the owner of the property which has been mixed, the owner of the property which has been mixed may:

a) Request the person having mixed the property to deliver the new property to it and pay such person the value of the property of such person;

b) If the owner of the property which has been mixed does not wish to take the new property, request the person having mixed the property to pay the value of the property of the owner and to compensate for any damage.

Article 227. Establishment of ownership rights in case of processing

1. An owner of raw materials which are processed to form a new object is also the owner of the newly formed object.

2. A person using raw materials under the ownership of another person for processing who acts in good faith shall become the owner of the new property, but must pay the value of the raw materials to the owner and compensate it for any damage.

3. Where a person processes raw materials not in good faith, the owner of the raw materials has the right to request that the new object be delivered to it. Where the raw materials are owned by more than one person, such persons shall become the owners of the newly formed object in proportion to the value of the raw materials owned by each person. The owners of the raw materials processed not in good faith may request the person carrying out the processing to compensate for any damage.

Article 228. Ownership rights are established with respect to abandoned objects and objects the owner of which is not able to be identified

1. An abandoned object is an object in respect of which the owner has renounced its ownership rights.

A person finding an abandoned object which is moveable property shall have the right to own such property, unless otherwise pscribed by law. If the found object is immoveable property, it shall belong to the State.

2. A person finding an object the owner of which is not able to be identified must inform or deliver the object to the people’s committee or police station of the nearest commune in order that a public announcement may be made notifying the owner to reclaim the object.

The delivery of the object must be recorded, specifying the surnames, given names and addresses of the finder and the receiver and the condition, quantity and volume of the property delivered.

The people’s committee or police station of commune which received the object must notify the finder of the results of their inquiries in order to determine the owner.

Where the object, the owner of which is not able to be identified, is moveable property, if the owner of the object is still not able to be identified after one year from the date of the public announcement, such property shall be under the ownership of the finder in accordance with law.

Where the object is immoveable property, if the owner is still not able to be identified after five years from the date of the public announcement, such property shall belong to the State. The finder shall be entitled to enjoy a monetary reward in accordance with law.

Article 229. Establishment of ownership rights with respect to buried or sunken objects which are found

1. A person finding an object which is buried or sunken must notify and return to the owner; if the owner is not able to be identified, he/she must inform or deliver the object to the people’s committee or police station of the nearest commune or a competent authority in accordance with regulations of law.

2. Ownership rights with respect to a buried or sunken object which is found, but which has no owner or the owner of which is not able to be identified, shall be determined, after deducting search and maintenance expenses, as follows:

a) A found object which is an historic or cultural relic shall belong to the State as pscribed in Law on cultural heritage and the finder shall be entitled to enjoy a monetary reward in accordance with law.

b) A found object which is not an historic or cultural relic as pscribed in Law on cultural heritage, and which has a value equivalent up to ten-month base salary provided for by the State, shall belong to the finder; if the value of the found object is more than the equivalent of ten-month base salary provided for by the State, the finder shall be entitled to the value of ten-month base salary plus fifty (50) per cent of the remaining value of the object in excess of the ten-month base salary provided for by the State, with the remaining value belonging to the State.

Article 230. Establishment of ownership rights with respect to objects which other persons have lost or mislaid

1. A person finding an object which another person has lost or mislaid and being aware of the address of the person having lost or mislaid the object must inform or return the object to such person. If the finder is not aware of the address of the person having lost or mislaid the object, it must inform or deliver the object to the people’s committee or police station of the nearest commune in order that a public announcement may be made notifying the owner to reclaim the object.

The people’s committee or police station of commune which received the object must notify the finder of the results of their inquiries in order to determine the owner.

2. If, after one year from the date of the public announcement of the object having being found, the owner of the object is still not able to be identified or the owner does not claim the object, the ownership rights with respect to such property shall be determined as follows:

a) If the value of lost or mislaid object is up to ten-month base salary provided for by the State, it shall belong to the finder as pscribed in this Code and relevant laws; if the value of the found object is more than the equivalent of ten-month base salary provided for by the State, the finder shall be entitled to the value of ten-month base salary, deducted from pservation expenses, and plus fifty (50) per cent of the remaining value of the object in excess of the ten-month base salary provided for by the State, with the remaining value belonging to the State.

b) A lost or mislaid object which is an historic or cultural relic as pscribed in the Law on cultural heritage shall belong to the State. The finder shall be entitled to enjoy a monetary reward in accordance with law.

Article 231. Establishment of ownership rights with respect to stray domestic livestock

1. A person capturing a stray domestic livestock must take care of it and notify the people’s committee of the commune in which such person resides in order that a public announcement may be made notifying the owner to reclaim the stray domestic livestock. After 6 months or after 1 year, with regard to domestic livestock allowed to roam according to customary practice, from the date of the public announcement, the ownership rights with respect to domestic livestock and any offspring born thereof shall belong to the capturer.

2. If the owner reclaims the stray domestic livestock, he/she must pay care remuneration and other expenses for the capturer. During the period of feeding and taking care of the stray domestic livestock, the capturer shall be entitled to half of or 50% of value any offspring born. Such person must compensate for any damage if it intentionally causes the death of the stray domestic livestock.

Article 232. Establishment of ownership rights with respect to stray domestic poultry

1. Where the domestic poultry of a person is lost and captured by another person, the person having captured the stray domestic poultry must make a public announcement notifying the owner to reclaim such poultry. If no one reclaims the stray domestic poultry after one month from the date of the public announcement, it shall be under the ownership of the person having captured it.

2. An owner reclaiming the stray poultry must remunerate the person having captured it for feeding and taking care of the stray domestic poultry and any other expenses incurred. During the period of feeding and taking care of the stray domestic poultry, the person having captured it shall enjoy the benefits from the stray domestic poultry. Such person must compensate for any damage if it intentionally causes the death of the stray domestic poultry.

Article 233. Establishment of ownership rights with respect to aquaculture stock

Where the aquaculture stock of a person moves naturally into the field, pond or lake of another person, the stock shall be under the ownership of the person having such field, pond or lake. Where the aquaculture stock has special marks which make it possible to determine that it is not under the ownership of the person having such field, pond or lake, such person must make a public announcement notifying the owner to reclaim the stock. If no one reclaims the stock after one month from the date of the public announcement, it shall be under the ownership of the person having such field, pond or lake.

Article 234. Establishment of ownership rights due to inheritance

An heir shall have ownership rights with respect to inherited property in accordance with Part Four of this Code.

Article 235. Establishment of ownership rights in accordance with judgment or decision of court or in accordance with decision of another competent authority

Ownership rights may also be created on the basis of an effective judgment or decision of a court or an effective decision of another competent authority.

Article 236. Establishment of ownership rights resulting from limitation periods with respect to possession or deriving benefits from property unlawfully

A person unlawfully but in good faith possessing, or deriving benefits from, property continuously and in an overt manner for ten (10) years with respect to moveable property, and for thirty (30) years with respect to immoveable property, shall become the owner of such property from the moment of commencement of possession, unless otherwise pscribed by this Code and relevant laws.

Sub-section 2. TERMINATION OF OWNERSHIP RIGHTS

Article 237. Bases for terminating ownership rights

Ownership rights terminate in any of the following cases:

1. The owners transfers his or her ownership rights to another person;

2. The owner renounces his or her ownership rights;

3. The property is consumed or destroyed;

4. The property is realized in order to fulfill the obligations of the owner;

5. The property is requisitioned;

6. The property is confiscated;

7. Other persons have established ownership rights with respect to property in accordance with this Code;

8. Other bases as provided by law.

Article 238. Transfer of ownership rights by owner

Where an owner transfers its ownership rights to another person through a contract for sale and purchase, by exchange, gift or loan, or through inheritance, the ownership rights of the owner with respect to the property shall terminate from the time when the ownership rights of the transferee arise.

Article 239. Renunciation of ownership rights

An owner may terminate ownership rights with respect to its property by publicly declaring, or by performing certain acts evidencing, its renunciation of the right to possess, use and dispose of such property.

With respect to property the renunciation of which may harm social order or security or cause environmental pollution, the renunciation of ownership rights must comply with the law.

Article 240. Property in respect of which other persons have established ownership rights

When a person has, in accordance with Article 228 through 233 of this Code, lawfully established ownership rights with respect to an object the owner of which is not able to be identified; a buried or sunken objects; a lost or mislaid object or stray domestic livestock, poultry or aquaculture stock, the ownership rights of the person formerly having the property shall terminate.

When the ownership rights of a possessor or a person benefiting from property have been created in accordance with Article 236 of this Code, the ownership rights of the person who formally had the property shall terminate.

Article 241. Realization of property in order to fulfill obligations of owner

1. Ownership rights with respect to property shall terminate when such property is realized in order to fulfill the obligations of the owner pursuant to a decision of a court or another competent authority, unless otherwise provided by law.

2. Property which the law provides is not able to be seized may not be realized in order to fulfill the obligations of the owner.

3. The ownership rights with respect to property realized in order to fulfill the obligations of the owner shall terminate at the time when the ownership rights of the recipient of such property arise.

4. The realization of land use rights shall be carried out in accordance with the law on land.

Article 242. Destroyed property

When property is destroyed, the ownership rights with respect to such property shall terminate.

Article 243. Property which is compulsorily acquired

Where property is compulsorily acquired as pscribed by law, the ownership rights of the owner shall terminate from the time when the decision of the competent authority becomes legally effective.

Article 244. Confiscated property

Where property of an owner is confiscated and paid into the State Budget due to the owner committing a crime or an administrative offence, the ownership rights of the owner with respect to such property shall terminate from the time when the judgment or decision of the court or the decision of the competent authority becomes legally effective.

Chapter XIV

OTHER PROPERTY-RELATED RIGHTS

Section 1. Right to adjoining immovable property

Article 245. Right to adjoining immovable property

Right to adjoining immovable property means a right to be exercised on an immovable property (hereinafter referred to as obliged immovable property) to serve the exploitation of another immovable property under ownership of another person (hereinafter referred to as entitled immovable property).

Article 246. Bases for establishment of right to adjoining immovable property

The right to adjoining immovable property shall be established according to natural terrain, as pscribed by law, according to agreement or will.

Article 247. Effect of right to adjoining immovable property

The right to adjoining immovable property shall take effect to every natural and juridical person and it is transferred concurrently with the transferred immovable property, unless otherwise pscribed by law.

Article 248. Rules for exercising right to adjoining immovable property

The right to adjoining immovable property shall be exercised as agreed by the parties. If the parties fail to agree, the rules below must be followed:

1. Ensure the appropriate exploitation of the entitled immovable property in conformity with the use purposes of both entitled and obliged immovable property;

2. Do not abuse the right to the obliged immovable property;

3. Do not obstruct or hassle the exercising of right to the entitled immovable property.

Article 249. Change of exercising right to adjoining immovable property

In case the change of use or exploitation of the obliged immovable property leading the change of exercising of right to the entitled immovable property, the owner of the former immovable property must notify the owner of the latter immovable property within an appropriate period. The owner of the obliged immovable property must enable the owner of the entitled immovable property to adapt to such change.

Article 250. Obligation of owners relating to draining of rainwater

An owner of house or construction works must install water drain pipes in order that the rainwater from its roof does not run onto any adjoining immoveable properties.

Article 251. Obligation of owners relating to draining of waste water

An owner of house or construction work must install underground drains or water drainage channels to discharge waste water to the pscribed location in order that the waste water does not run and spill onto any adjoining immoveable properties or onto public streets or public places.

Article 252. Rights relating to supply and drainage of water through adjoining immoveable property

Where, due to the natural location of immoveable property, the supply and drainage of water must pass through another immoveable property, the owner of the immoveable property through which the water flows must provide an appropriate channel for the supply and drainage of water and may not hinder or pvent the flow of water.

The person using the water supply and drainage channel must minimize to the lowest possible extent any damage to the owner of the immoveable property through which the water flows when installing the water channel. If damage is caused, compensation must be made. Where water flows naturally from a higher position to a lower position and causes damage to the owner of the property through which the water flows, the person using the water supply and drainage channel shall not be liable to compensate for any damage.

Article 253. Rights relating to irrigation and water drainage in cultivation

A person having the right to use land for cultivation has the right to request neighboring land users to provide a reasonable and convenient water channel for irrigation and drainage. A person having been so requested has the obligation to grant such request. If the person using such water channel causes damage to neighboring land users, compensation must be made.

Article 254. Right of passage

1. An owner of immoveable property which is surrounded by immoveable properties of other owners such that there is no exit has the right to request one of the owners of adjoining immoveable properties to provide it with a passage to a public road on their land.

The passage shall be opened in the adjoining immoveable property which is deemed to be the most convenient and reasonable, taking into consideration the special characteristics of the location, the interests of the immoveable property which does not have an exit, and what will cause the least damage to the immoveable property through which the passage is created.

The owner of the immovable property eligible for the passage must compensate for the obliged immovable property, unless otherwise agreed.

2. The location and the length, width and height of the passage shall be agreed by the owners in order to ensure convenient passage and minimize inconvenience to the parties. If there are any disputes regarding the passage, the parties may request the authorized State body to resolve.

3. Where immoveable property is pided into more than one portion for different owners or users, upon pision, necessary passages must be provided, without compensation, to persons in the interior as provided in Clause 2 of this Article.

Article 255. Right to install electricity transmission cables and communication cables through other immoveable properties

An owner of immoveable property has the right to install electricity transmission cables and communication cables in a reasonable manner through the immoveable property of other owners, but must ensure the safety and convenience of such owners. If damage is caused, compensation must be made.

Article 256. Termination of easements over adjoining immoveable property

An easement over adjoining immoveable property shall terminate in the following circumstances:

1. The entitled immovable property and the obliged immovable property belong to ownership rights of a person;

2. The use and exploitation of the immovable property do not arise the needs of enjoying rights;

3. Upon agreement of contracting parties;

4. Other bases as provided by law.

Section 2. USUFRUCT RIGHT

Article 257. Usufruct right

Usufruct right means the right to use a property, under ownership of another entity, and enjoy its yield or income in a specific period of time.

Article 258. Bases for establishment of usufruct right

The usufruct right shall be established as pscribed by law, according to agreement or will.

Article 259. Effect of usufruct right

The usufruct right shall be established from the time of transfer of the property, unless otherwise agreed or otherwise pscribed by law.

The established usufruct right shall take effect to every natural and juridical person, unless otherwise pscribed by law.

Article 260. Time limit of usufruct right

1. The time limit of usufruct right shall be agreed by the parties or pscribed by law provided that its maximum length is the full life of the first usufructary being natural person or the period of time for which the first usufructary being juridical person exists as long as it does not exceed 30 years.

2. The usufructary has the right to lease the usufruct right within a specific period of time pscribed in Clause 1 of this Article.

Article 261. Rights of usufructary

1. On his/her own or permit another person exploit, use and enjoy yield and/or income from the object of the usufruct right.

2. Request the owner of the property perform obligation to repair the property as pscribed in Clause 4 Article 263 of this Code; if the usufructary performs the obligation on behalf of the owner of property is entitled to request the owner to refund the expenses.

3. Lease the usufruct right to the property.

Article 262. Obligations of usufructary

1. Receive property under current conditions and register it if required by law.

2. Exploit the property for appropriate purposes.

3. Preserve property as if it is his/her own property.

4. Maintain and repair property periodically to ensure the normal use; restore the status of the property and remedy the bad consequences of property due to his/her poor performance of obligations in line with technical requirements or by custom of property pservation.

5. Return the property to the owner of the expiration of usufruct time limit.

Article 263. Rights and obligations of property owner

1. Dispose property without any change of the usufruct right which has been established.

2. Request a court to deprive usufruct right from a usufructary who seriously breaches his/her obligations.

3. Do not obstruct or hassle or otherwise violate the legitimate rights and interests of the usufructary.

4. Perform obligation to repair property to ensure that there is no significant decline leading the property cannot be used or lost all its utility and value.

Article 264. Right to enjoy yield and income

1. Each usufructary has ownership right to the yield and income derived from the property being the object of the usufruct right during its effective period of time.

2. If the usufruct right cease to exist before the harvest time of yields and income, the usufructary shall, upon the harvest time, be entitled to enjoy the value of yield and income received corresponding the time that person is entitled to such usufruct right.

Article 265. Termination of usufruct right

The usufruct right shall terminate in any of the following cases:

1. The time limit of usufruct right has expired;

2. As agreed by the parties;

3. The usufructary becomes the owner of the property being the subject of the usufruct right;

4. The usufructary waives or fails to exercise the usufruct right during a time limit pscribed by law;

5. The property being subject of the usufruct right no longer exists;

6. Pursuant to a decision of a court;

7. Pursuant to other provisions of law.

Article 266. Returning property upon termination of usufruct right

The property being subject of usufruct right must be returned to the owner upon the termination of usufruct right, unless otherwise agreed or otherwise pscribed by law.

Section 3. SURFACE RIGHTS

Article 267. Surface rights

Surface rights mean an entity’s rights to the ground, water surface, space thereon and earth bowel of the land whose land use rights belong to another entity.

Article 268. Bases for establishment of surface rights

Surface rights shall be established by law, according to agreement or will.

Article 269. Effect of surface rights

Surface rights shall take effect from the point of time when the holder of land use rights transfer ground, water surface, space thereon and earth bowel of the land to the holder of surface rights, unless otherwise agreed or otherwise pscribed by law.

Surface rights shall take effect to every natural and juridical person, unless otherwise pscribed by relevant laws.

Article 270. Time limit of surface rights

1. The time limit of surface rights shall be established by law, according to agreement or will provided that it does not exceed the time limit of land use rights.

2. If the agreement or will does not mention the time limit of surface rights, each party is entitled to terminate any time provided that it provides a written notification to the other party within 6 months.

Article 271. Contents of surface rights

1. Each holder of surface rights has the right to exploit and use ground, water surface, space thereon, the water and the earth bowel of the land whose land use rights belong to another entity for construction, planting or cultivation provided that it is not contrary to the provisions of this Code, the law on land, construction, planning, resources, minerals and other provisions of relevant laws.

2. The holder of surface rights has the ownership rights to every property derived as pscribed in Clause 1 of this Article.

3. If part of the whole of surface rights is transferred, the transferee shall inherit the surface rights according to conditions and within the scope in proportion to the part or the whole transferred surface rights.

Article 272. Termination of surface rights

The surface rights shall terminate in any of the following cases:

1. The time limit of surface rights has expired;

2. The holder of surface rights and the holder of land use rights shall be the same;

3. The holder of surface rights waives his/her rights;

4. Surface rights of land use rights are appropriated as pscribed in law on land;

5. As agreed by the parties or as pscribed by law.

Article 273. Realization of property upon termination of surface rights

1. Upon the termination of surface rights, its holder must return ground, water surface, space thereon and earth bowel of the land to the holder of land use rights, unless otherwise agreed or otherwise pscribed by law.

2. The holder of surface rights must realize the property under ownership upon its termination, unless otherwise agreed.

If the holder of surface rights must realize the property under ownership upon its termination, the ownership of such property shall be transferred to the holder of land use rights from the termination time, unless the latter holder refuse such property.

If the holder of land use rights refuses the property while the property is required to be realized, the holder of surface rights must pay the property realization expenses.

PART THREE

OBLIGATIONS AND CONTRACTS

Chapter XV.

GENERAL PROVISIONS

Section 1. Bases for giving rise to and subject matter of obligations

Article 274. Obligations

Obligations means acts whereby one or more entities (hereinafter referred to as obligors) must transfer objects, transfer rights, pay money or provide valuable papers, perform other acts or refrain from performing certain acts in the interests of one or more other subjects (hereinafter referred to as obligees).

Article 275. Bases for giving rise to obligations

Obligations arise from the following bases:

1. Contracts;

2. Unilateral legal acts;

3. Unauthorized performance of acts;

4. Unlawful possession or use of or receipt of benefits from property;

5. Causing damage through unlawful acts;

6. Other bases as provided by law.

Article 276. Subject matter of obligations

1. The subject matter of an obligation may be property or acts which must be performed or acts which must not be performed.

2. The subject matter of an obligation must be defined pcisely.

Section 2. Performance of obligations

Article 277. Places for performing obligations

1. The place for performing an obligation shall be agreed by the parties.

2. Where the parties do not have an agreement, the place for performance of the obligation shall be:

a) The location of the immoveable property, if the subject matter of the obligation is immoveable property;

b) The place of residence or head office of the obligee, if the subject matter of the obligation is not immoveable property.

Where the obligee changes its place of residence or head office, it must notify the obligor of the change and must bear any increase in expenses resulting from the change in residence or head office, unless otherwise agreed.

Article 278. Time-limits for performing obligations

1. The time-limit for performing an obligation shall be as agreed by the parties or as provided by law.

2. The obligor must perform the obligation strictly in accordance with the relevant time-limit, unless otherwise pscribed by this Code or relevant laws.

If the obligor performs the obligation prior to the time-limit and the obligee accepts such performance, the obligation shall be deemed to have been fulfilled on time.

3. Where the parties do not have an agreement and the time-limit for the performance of a civil obligation is not identifiable pscribed in Clause 1 of this Article, a party may fulfill the obligation or demand the fulfillment of the obligation as the case may be at any time, but must give reasonable prior notice to the other party.

Article 279. Performance of obligations to deliver objects

1. A person having the obligation to deliver an object must take care of and pserve the object until the time of delivery.

2. Where an object to be delivered is a distinctive object, the obligor must deliver that particular object in the same condition as agreed. If the object is a fungible object, it must be delivered in accordance with the quality and quantity agreed. If there is no agreement as to the quality, the object delivered must be of average quality. If the object is an integrated object, the whole integrated object must be delivered.

3. An obligor must bear all expenses related to the delivery of an object, unless otherwise agreed.

Article 280. Performance of obligations to pay money

1. An obligation to pay money shall be performed in full, strictly on time, at the place and by the method as agreed.

2. The obligation to pay money shall include the payment of interest on principal, unless otherwise agreed.

Article 281. Performance of obligations to perform acts or not to perform acts

1. Obligation to perform an act means an obligation whereby the obligor must perform that particular act.

2. Obligation not to perform an act means an obligation whereby the obligor must not perform that particular act.

Article 282. Performance of obligations in stages

An obligation may be performed in stages if so agreed or so provided by law or pursuant to a decision of a competent authority.

The late performance of one stage of an obligation shall be deemed to be late performance of the obligation.

Article 283. Performance of obligations through third parties

With the consent of the obligee, an obligor may authorize a third person to perform an obligation on behalf of the obligor provided that the obligor shall be liable to the obligee if the third person fails to perform or performs incorrectly the obligation.

Article 284. Conditional performance of obligations

1. Where the parties have agreed on conditions for the performance of a civil obligation or where the law provides certain conditions for the performance of an obligation, the obligor must perform the obligation when such conditions are satisfied.

2. If the conditions do not occur or occur resulting from the influence of a party, Clause 2 Article 120 of this Code shall apply.

Article 285. Performance of obligations having optional subject matters

1. Obligation having an optional subject matter means an obligation the subject matter of which is one of several different items of property or acts from which the obligor may select at its discretion, except where it is agreed or provided by law that the right to select is reserved to the obligee.

2. The obligor must notify the obligee of the property or act selected in order to perform the obligation. In the case where the obligee has fixed a time-limit for performance of the obligation with a selected subject matter, the obligor must fulfill the obligation on time.

3. Where there remains only one property or one act to select, the obligor must deliver that particular property or perform that particular act.

Article 286. Performance of substitutable civil obligations

Substitutable obligation means an obligation whereby if the obligor fails to perform the original obligation, it may perform a different obligation as agreed by the obligee as a substitute for the original obligation.

Article 287. Performance of severable obligations

Where more than one person jointly performs an obligation and each person has a clearly defined share of the obligation which is severable from that of the other person, each person must perform only its own share of the obligation.

Article 288. Performance of joint obligations

1. Joint obligation means an obligation which must be performed by more than one person and which the obligee may request any one of the obligors to perform in its entirety.

2. When one person has performed an obligation in its entirety, such person may require the other joint obligors to make payment for their respective shares of the joint obligation to such person.

3. Where an obligee designates one person from amongst the joint obligors to perform an entire obligation and later releases that person, the other obligors shall also be released from performing the obligation.

4. Where an obligee releases one of the joint obligors from its share of the joint obligation, the other obligors must, nevertheless, perform jointly their respective shares of the obligation.

Article 289. Performance of obligations for joint obligees

1. Civil obligation for joint obligees means an obligation whereby each joint obligee may require the obligor to perform the obligation in its entirety.

2. An obligor may perform its obligation with respect to any one of the joint obligees.

3. Where one of the joint obligees releases the obligor from performing the share of the obligation owed to such joint obligee, the obligor must, nevertheless, perform the remaining shares of the obligation owed to the other joint obligees.

Article 290. Performance of pisible obligations

1. Divisible obligation means an obligation the subject matter of which is a pisible object or an act which is able to be pided into portions for the purpose of performance.

2. An obligor may perform the obligation in stages, unless otherwise agreed.

Article 291. Performance of inpisible obligations

1. Inpisible obligation means an obligation the subject matter of which is an inpisible object or an act which must be performed in its entirety at the one time.

2. Where several persons must perform an inpisible obligation, they must perform the obligation in its entirety at the same time.

Section 3. SECURITY FOR PERFORMANCE OF OBLIGATIONS

Sub-section 1. GENERAL PROVISIONS

Article 292. Types of security for performance of obligations

Types of security for the performance of obligations comprise the following:

1. Pledge of property;

2. Mortgage of property;

3. Deposit;

4. Security collateral;

5. Escrow deposit;

6. Title retention;

7. Guarantee;

8. Fidelity guarantees;

9. Lien on property.

Article 293. Scope of security for performance of obligations

1. An obligation may be fully or partly secured, as agreed or as provided by law. If there is no agreement on or if the law does not provide, the scope of the security, the obligation, including the obligation to pay interest and to compensate for any damage, shall be deemed to be fully secured.

2. Secured obligations may comprise current obligations, future obligations and conditional obligations.

3. With respect to a future obligation which is going to arise within a guaranteed time limit, it shall be the secured obligation, unless otherwise agreed.

Article 294. Security for performance of future obligations

1. With respect to a future obligation, the parties may agree on the scope of the secured obligation and the deadline by which the secured obligation must be performed, unless otherwise pscribed by law.

2. When the future obligation arises, the parties are not required to re-establish the security for such obligation.

Article 295. Collateral

1. Collateral must be under the ownership rights of the securing party, except for the cases of lien on property or title retention.

2. Collateral may be described generally but must be identified.

3. Collateral may be existing property or off-plan property.

4. The value of collateral may be greater, equal or smaller than the value of the secured obligation.

Article 296. Single item of property used as security for performance of several obligations

1. A single item of property may be used as security for performance of several obligations if, at the time of establishment of the security transaction, the value of such property is greater than the total aggregate value of the secured obligations, unless otherwise agreed or otherwise provided by law.

2. Where a single item of property is used as security for performance of several obligations, the securing party must notify the later secured party that the security property is being used as security for performance of other obligations. The provision of security on each occasion must be made in writing.

3. Where the security property must be realized in order to satisfy one obligation which has fallen due, the other obligations which have not yet fallen due shall also be deemed due and all secured parties shall be entitled to take part in the realization. The secured party which provided notice of realization of the property shall be responsible for realizing the property, unless otherwise agreed by the secured parties.

If the parties wish to continue to fulfill the obligations which have not yet fallen due, they may agree that the securing party will use other property as security for performance of the obligation which has fallen due.

Article 297. Effectiveness against third parties

1. Security shall take effect against a third party from the time of registration of such security or the secured party keeps or possess the collateral.

2. When the security takes effect against a third party, the secured party is entitled to reclaim the collateral and the payment pscribed in Article 308 of this Code and relevant laws.

Article 298. Registration of security

1. Security shall be registered as agreed by the parties or provided by law.

The registration shall be the condition for a secured transaction become valid only the case as pscribed by law.

2. A registered security shall take effect against third party from the time of registration.

3. The registration of security shall comply within regulations of law on registration of security.

Article 299. Cases of realization of collateral

1. An obligator fails to perform or perform not as agreed an obligation when it falls due.

2. An obligator must perform the secured obligation before time limit due to his/her violation against the obligation as agreed or pscribed by law.

3. Other cases as agreed by the parties or pscribed by law.

Article 300. Notification of realization of collateral

1. Before a collateral is realized, a secured party must notify the securing party and other secured parties of the realization of the collateral within a reasonable time limit.

If the collateral at risk of being damaged resulting in diminished value or lose the entire value, a secured party may realize it immediately and notify the securing party and other secured parties of the realization of such asset.

2. If the secured party does not notify the realization of collateral as pscribed in Clause 1 of this Article that cause damage to the securing party and/or other secured parties, compensation must be made.

Article 301. Giving collateral for realization

The holder of collateral is obliged to give it to the secured party for realization in any of the cases pscribed in Article 299 of this Code.

If the holder of collateral fails to give the asset, the secured party is entitled to request a court for settlement, unless otherwise pscribed by relevant laws.

Article 302. Right to reclaim collateral

The securing party may reclaim the collateral if, before the realization of the collateral, it completely performs its obligations and pay all expenses incurred for the late performance of obligations, unless otherwise pscribed by law.

Article 303. Methods of realizing collateral

1. The securing party and the secured party may agree any of the following methods of realizing collateral:

a) Put collateral up for an auction;

b) The secured party sells collateral itself;

c) The secured party accepts the collateral as substitutions for the performance of obligations of the securing party;

d) Other methods.

2. If there is no agreement on methods of realizing collateral as pscribed in Clause 1 of this Article, the collateral shall be put up for auction, unless otherwise pscribed by law.

Article 304. Selling collateral

1. The collateral shall be put up for auction as pscribed by law on property auction.

2. The collateral sold by the secured party must comply with the regulations on property sale in this Code and the regulations below:

a) The payment amount derived from the realization of collateral shall comply with Article 307 of this Code;

b) The owner of collateral and the person competent to realize the collateral, upon the completion of the sale, shall comply with procedures for transfer of ownership rights to the buyer.

Article 305. Acceptance of the collateral as substitution for the performance of obligations of the securing party

1. The secured party may accept the collateral as substitution for the performance of obligations of the securing party if agreed by the parties.

2. If there is no agreement pscribed in Clause 1 of this Article, the secured party may only accept the collateral as substitution for the performance of obligations of the securing party with the written consent of the securing party.

3. Where the value of the collateral is greater than the value of the secured obligation, the secured party must pay the difference amount to the securing party; where the value of the collateral is less than the value of the secured obligations then the unpaid obligations become unsecured obligations.

4. The securing party is obliged to follow the procedures for transfer of ownership rights to the secured party as pscribed by law.

Article 306. Valuation of collateral

1. The securing party and the secured party may agree on collateral prices or have the collateral valuated by an asset valuation organization upon the realization of the collateral.

If there is no agreement mentioned above, the collateral shall be valuated by an asset valuation organization.

2. The valuation of the collateral must be objective and in conformity with market price.

3. The asset valuation organization must compensate for any damage to the securing party and/or the secured party during the process of valuation due to its legal violations.

Article 307. Payment of the sum of money obtained from the realization of collateral

1. The sum of money obtained from the realization of the collateral after deducting from the cost of pservation, capture and realization of the collateral shall be paid in order of priority specified in Article 308 of this Code.

2. Where the sum of money obtained from the realization of the collateral, after deducting from the cost of pservation, seizure and realization of the collateral is greater than the value of secured obligations, the difference amount must be paid to the securing party.

3. Where the sum of money obtained from the realization of the collateral, after deducting from the cost of pservation, seizure and realization of the collateral is less than the value of secured obligations, part of the unpaid obligations are defined as unsecured obligations, unless the parties otherwise agree additional collateral. The secured party may request the obligor to perform the unpaid secured obligations.

Article 308. Order of priority for payment between joint secured parties

1. When an asset is used to secure the performance of many obligations, payment priority order between the joint secured parties shall be determined as follows:

a) If all types of security take effect against a third party, the order of priority for payment shall be determined according to the order of effect against the third party;

b) If there are some types of security take effect against a third party while some types of security do not take effect against the third party, the payment of obligations with security taking effect against the third party shall be given priority;

c) If all types of security do not take effect against a third party, the order of priority for payment shall be determined according to the order of establishment of types of security.

2. The order of priority for payment pscribed in Clause 1 of this Article may be changed as agreed by the parties. The subrogating party of the right to priority of payment shall only be given priority within the secured extent of the subrogated party.

Sub-section 2. PLEDGE OF PROPERTY

Article 309. Pledge of property

Pledge of property means the delivery by one party (hereinafter referred to as the pledgor) of property under its ownership to another party (hereinafter referred to as the pledgee) as security for the performance of an obligation.

Article 310. Effectiveness of pledge of property

1. Agreement on pledge of property shall take effect from the time of concluding, unless otherwise agreed or pscribed by law.

2. Pledge of property shall take effect against third party from the time at which the pledgee keeps the pledged property.

If an immovable property is the subject of pledge as pscribed in law, the pledge on immovable property shall take effect against third party from the time of registration.

Article 311. Obligations of pledgors

1. Deliver the pledged property to the pledgee as agreed.

2. Notify the pledgee of any third person rights with respect to the pledged property. In the case of failure to provide such notice, the obligee shall have the right to cancel the contract of pledge of property and demand compensation for damage or the right to maintain the contract and agree on the rights of the third person with respect to the pledged property.

3. Pay the pledgee reasonable expenses for taking care of and pserving the pledged property, unless otherwise agreed.

Article 312. Rights of pledgors

1. Require the pledgee to suspend use of the pledged property in cases provided in Clause 3 of Article 314 of this Code if the pledged property is in danger of losing its value or depciating in value as a result of such use.

2. Require the pledgee to hold the pledged property to return the pledged property and related documents after the obligation secured by the pledge has been fulfilled.

3. Require the pledgee to compensate for any damage caused to the pledged property.

4. Sell, substitute, exchange, or give the pledged property to other property if so agreed by the pledgee or pscribed by law.

Article 313. Obligations of pledgees

1. Take care of and pserve the pledged property; if the pledgee loses or damages the pledged property, the pledgee must compensate the pledgor for the damage.

2. Do not sell, exchange, give or use the pledged property as security for the performance of another obligation.

3. Do not lease, lend, exploit the yield or income derived from, the pledged property, unless otherwise agreed.

4. Return the pledged property and related documents upon fulfillment of the secured obligation or where the pledge is substituted with another security.

Article 314. Rights of pledgees

1. Require a person unlawfully possessing or using the pledged property to return the property.

2. Demand the realization of the pledged property in accordance with the methods as agreed or as provided by law.

3. Lease, lend, exploit, and to enjoy the yield and income derived from, the pledged property if so agreed.

4. Receive reimbursement of reasonable expenses incurred in taking care of the pledged property upon returning the pledged property to the pledgor.

Article 315. Termination of pledges on property

A pledge of property shall terminate in any of the following cases:

1. The obligation secured by the pledge has terminated;

2. The pledge has been cancelled or substituted with another security;

3. The pledged property has been realized;

4. As agreed by the parties.

Article 316. Return of pledged property

Where a pledge of property is terminated in accordance with Clause 1 or Clause 2 of Article 315 of this Code or as agreed by parties, the pledged property and documents evidencing the ownership rights with respect to the property shall be returned to the pledgor. Any yield and income derived from the pledged property shall also be returned to the pledgor, unless otherwise agreed.

Sub-section 3. MORTGAGES ON PROPERTY

Article 317. Mortgage of property

1. Mortgage of property means the use by one party (hereinafter referred to as the mortgagor) of property under the ownership of the obligor as security for the performance of an obligation to the other party (hereinafter referred to as the mortgagee) without transferring such property to the mortgagee.

2. The mortgaged property shall be held by the mortgagor. The parties may agree to deliver the mortgaged property to a third person to hold.

Article 318. Mortgaged property

1. Where entire immoveable property or moveable property having auxiliary objects is mortgaged, such auxiliary objects shall also form part of the mortgaged property, unless otherwise agreed.

2. Where a portion of immoveable property or moveable property having auxiliary objects is mortgaged, such auxiliary objects shall also form part of the mortgaged property, unless otherwise agreed by the parties.

3. With respect to mortgage on land use rights that property on land is owned by the mortgagor, such property shall also part of the mortgaged property, unless otherwise agreed.

4. Where mortgaged property is insured, the mortgagee must notify the insurer that the insured property is being mortgaged. The insurer shall pay the insured sum directly to the mortgagee upon occurrence of an insured event.

If the mortgagee failed to notify the insurer that the insured property was mortgaged, the insurer shall pay the insured sum in accordance with the insurance contract and the mortgagor shall be obliged to make payment to the mortgagee.

Article 319. Effectiveness of mortgage of property

1. Agreement on mortgage of property shall take effect from the time of concluding, unless otherwise agreed or pscribed by law.

2. The mortgage of property shall take effect against third party from the time of registration.

Article 320. Obligations of mortgagor

1. Transfer documents related to the mortgaged property, unless otherwise agreed or pscribed by law.

2. Take care of and pserve the mortgaged property.

3. If the mortgaged property is in danger of losing its value or depciating in value due to its exploitation, to take necessary remedial measures, including ceasing the exploitation of the mortgaged property.

4. When the mortgaged property is damaged, the mortgagor is obligated to, within a reasonable period, repair or substitute another property with equivalent value, unless otherwise agreed.

5. Provide information about the actual condition of the mortgaged property to for the mortgagee.

6. Deliver the mortgaged property to the mortgagee for realization in one of the cases pscribed in Article 299 of this Code.

7. Notify the mortgagee of any third person rights with respect to the mortgaged property (if any). In the case of failure to provide such notice, the mortgagee shall have the right to cancel the contract of mortgage of property and demand compensation for damage or the right to maintain the contract and agree on the rights of the third person with respect to the mortgaged property.

8. Do not sell, exchange or give the mortgaged property, except in the cases provided in Clauses 4 and 5 of Article 321 of this Code.

Article 321. Rights of mortgagor

1. Exploit, and to enjoy the yield and income derived from, the property, except where the yield and income also form part of the mortgaged property as agreed.

2. Invest in order to increase the value of the mortgaged property.

3. Recover the mortgaged property and related documents held by a third person when the obligation secured by the mortgage is terminated or is substituted by other security.

4. Sell or replace mortgaged property being goods rotating during the production and business process. In the case of a sale of mortgaged property being goods rotating during the production and business process, the right to require the purchaser to pay money, the proceeds received or the assets formed from the proceeds received shall form the mortgaged property in substitution for the property which was sold.

When a warehouse is mortgaged, the mortgagor may substitute goods in the warehouse but must ensure the value of the goods in the warehouse remains the value agreed.

5. Sell, exchange or give mortgaged property not being goods rotating during the production and business process with the consent of the mortgagee or as pscribed by law.

6. Lease or lend the mortgaged property provided that notice must be provided to the lessee and the borrower that the property is being mortgaged and that the mortgagee must also be notified that such notice has been provided.

Article 322. Obligations of mortgagees

1. Where the parties agree that the mortgagee will hold the documents relating to the mortgaged property, to return to the mortgagor such documents upon termination of the mortgage.

2. Follow procedures for realization of mortgaged property in accordance with regulations of law.

Article 323. Rights of mortgagees

1. Examine and inspect directly the mortgaged property provided that such examination and inspection does not hinder or cause difficulty to the use and exploitation of the mortgaged property.

2. Require the mortgagor to provide information on the current status of the mortgaged property.

3. Require the mortgagor to apply necessary measures to pserve the property and the value of the property if there is a danger that use and exploitation of the mortgaged property will cause loss of value or depciation in value of the property.

4. Conduct the registration of mortgage as pscribed by law.

5. Require the mortgagor or a third person holding the mortgaged property to deliver it to the mortgagee for realization if, upon expiry of the term for fulfillment of the obligation, the obligor has failed to perform or performed incorrectly the obligation.

6. Hold documents related to mortgaged property as agreed by parties, unless otherwise pscribed by law

7. Follow procedures for realization of mortgaged property as pscribed in Article 299 of this Code.

Article 324. Rights and obligations of third parties holding mortgaged property

1. A third person holding mortgaged property has the following rights:

a) Exploit the property if so agreed;

b) Receive remuneration and be reimbursed for expenses incurred in taking care of and pserving the mortgaged property, unless otherwise agreed.

2. A third person holding mortgaged property has the following obligations:

a) Take care of and pserve the mortgaged property, and to compensate for any damage if the third person loses the mortgaged property or causes the mortgaged property to lose its value or depciate in value;

b) Cease the exploitation of the property if it is in danger of losing its value or depciating in value;

c) Return the mortgaged property to the mortgagee or mortgagor as agreed or pscribed by law.

Article 325. Mortgage on land use rights without mortgage of property on land

1. With respect to mortgage on land use rights without mortgage property on that land but the land user is also the owner of the property on land; such property shall also part of the realized property, unless otherwise agreed.

2. With respect to mortgage on land use rights that the land user is not also the owner of the property on land, such owner may keep using such land within his/her rights and obligations during the realization of the land use rights. The rights and obligations of the mortgagor in relation with the owner of the property on land shall be transferred to the transferee of the land use rights, unless otherwise agreed.

Article 326. Mortgage of property on land without mortgage on land use rights

1. With respect to mortgage of property on land without mortgage on land use rights but the owner of the property on land is also the land user, such land use rights shall also part of the realized property, unless otherwise agreed.

2. With respect to mortgage of property on land without mortgage on land use rights that the owner of the property on land is not also the land user, the transferee of property on land may keep using such property within the transferred rights and obligations from the owner of the property on land during the realization of the land use rights, unless otherwise agreed.

Article 327. Termination of property mortgages

A mortgage of property shall terminate in any of the following cases:

1. The obligation which is secured by the mortgage has terminated;

2. The mortgage of the property has been cancelled or substituted with another security;

3. The mortgaged property has been realized;

4. As agreed by the parties.

Sub-section 4. DEPOSIT, SECURITY COLLATERAL, ESCROW DEPOSIT

Article 328. Deposit

1. Deposit is an act whereby one party (hereinafter referred to as the depositor) transfers to another party (hereinafter referred to as the depositary) a sum of money or pcious metals, gemstones or other valuable things (hereinafter referred to as the deposited property) for a period of time as security for the entering into or performance of a contract.

2. Upon a contract being entered into or performed, any deposited property shall be returned to the depositor, or deducted from the amount of an obligation to pay money. If the depositor refuses to enter into or perform the contract, the deposited property shall belong to the depositary. If the depositary refuses to enter into or perform the contract, it must return the deposited property and pay an amount equivalent to the value of the deposited property to the depositor, unless otherwise agreed.

Article 329. Security collateral

1. Security collateral is an act whereby a lessee of a movable property transfers a sum of money or pcious metals, gems or other valuable things (hereinafter referred to as security collateral property) to the lessor for a specified time limit to secure the return of the leased property.

2. In cases where the leased property is returned, the lessee shall be entitled to reclaim the security collateral property after pay the rental; if the lessee does not return the leased property, the lessor shall be entitled to reclaim the leased property; if the leased property is no longer available for the return, the security collateral property shall belong to the lessor.

Article 330. Escrow deposit

1. Escrow deposit is an act whereby an obligor deposits a sum of money, pcious metals, gems or valuable papers into an escrow account at a credit institution to secure the performance of an obligation.

2. In cases where the obligor has failed to perform or has improperly performed an obligation, the obligee shall be entitled to receive payment and compensation for damage caused by the obligor from the bank where the escrow deposit is affected, after deducting the bank service charges.

3. The procedures for making deposits and making payments shall be as provided by the law.

Sub-section 5. TITLE RETENTION

Article 331. Title retention

1. In a sale contract, the ownership of property of the seller may remain until the buyer pays the purchase price in full.

2. Title retention must be made in a separate document or included in the sale contract.

3. The title retention shall take effect against third party from the time of registration.

Article 332. Right to reclaim property

If the buyer fails to fulfill the payment obligation for the seller as agreed, the seller is entitled to reclaim the property. The seller shall refund the paid amount by the buyer deducted from the depciated value due to use. Where the buyer lost or damaged property, the seller has the right to claim damages.

Article 333. Rights and obligations of the buyer

1. Using the property and enjoying the yield and income derived therefrom within the effective term of title retention.

2. Facing the risks of the property within the effective term of the title retention, unless otherwise agreed.

Article 334. Termination of title retention

The title retention shall terminate in any of the following cases:

1. Payment obligation fulfilled completely by the buyer;

2. The seller receives the property under title retention back;

3. As agreed by the parties.

Sub-section 6. GUARANTEES

Article 335. Guarantees

1. Guarantee means an undertaking made by a third person (hereinafter referred to as the guarantor) to an obligee (hereinafter referred to as the creditor) to perform an obligation on behalf of an obligor (hereinafter referred to as the principal debtor) if the obligation falls due and the principal fails to perform or performs incorrectly the obligation.

2. The parties may agree that the guarantor shall only be obliged to perform the obligation if the principal debtor is incapable of performing it.

Article 336. Scope of guarantees

1. A guarantor may guarantee an obligation in whole or in part on behalf of a principal debtor.

2. A guaranteed obligation includes interest on the principal, penalties and compensation for any damage and interest on late payment, unless otherwise agreed.

3. The parties may agree on using security as property to secure the performance of guaranteed obligation.

4. If the obligation to guarantee is an obligation arising in the future, the scope of guarantee is exclusive of any obligations arising after the guarantor being natural person dies or the guarantor being juridical person ceases to exist.

Article 337. Remuneration

The guarantor shall be entitled to receive remuneration if so agreed by the guarantor with the principal debtor.

Article 338. Joint guarantors

When more than one person guarantee an obligation, those persons must perform jointly the guarantee, except where it is agreed or provided by law that the guarantee comprises separate portions. The obligee may require any of the joint guarantors to perform the obligation in its entirety.

Where one of the joint guarantors has performed the entire obligation on behalf of the principal debtor, the guarantor may require the other guarantors to perform their respective portions of the obligation with respect to that guarantor.

Article 339. Relationship between guarantors and creditors

1. If the principal fails to perform or performs incorrectly the obligation, the creditor is entitled to request the guarantor to fulfill the guaranteed obligation , unless contracting parties has agreed that the guarantor only be required to perform the obligation on behalf of the principal debtor in case of the failure to perform obligation by the principal debtor.

2. A creditor may not require a guarantor to perform an obligation on behalf of the principal debtor until the obligation falls due.

3. Where a guarantor is able to offset an obligation with a principal debtor, a guarantor does not have to perform the guaranteed obligation.

Article 340. Rights to require of guarantors

Each guarantor may require the principal debtor to indemnify the guarantor to the extent of the guarantee, unless otherwise agreed.

Article 341. Discharge from guaranteed obligations

1. Where the guarantor must perform the guaranteed obligation but the creditor discharges the guarantor from an obligation, the principal debtor is discharged from performance of the obligation with respect to the creditor, except where it is agreed or provided by law.

2. Where one person from amongst the joint guarantors is discharged from the performance of its portion of the guaranteed obligation, the other joint guarantors must, nevertheless, perform their portion of the guaranteed obligation.

3. Where one person from amongst the joint creditors discharge the guarantor from the performance of its portion of the guaranteed obligation, the guarantor must, nevertheless, perform their portion of the guaranteed obligation with respect to remaining joint creditors.

Article 342. Civil liability of guarantor

1. If the principal debtor fails to perform or perform incorrectly the obligation, the guarantor is obligated to perform such obligation.

2. If the guarantor performs incorrectly the guaranteed obligation, the creditor is entitled to request the guarantor to pay the value of the breached obligation and compensate for any damage.

Article 343. Termination of guarantees

A guarantee shall terminate in any of the following cases:

1. The obligation secured by the guarantee terminates;

2. The guarantee is cancelled or is substituted by another security;

3. The guarantor has satisfied the guaranteed obligation;

4. As agreed by the parties.

Sub-section 7. FIDELITY GUARANTEES

Article 344. Fidelity guarantees provided by socio-political organizations

A socio-political organization at the grassroots level may provide a fidelity guarantee in order that poor inpiduals and households are able to borrow sums from banks or other credit institutions for purposes of production, business or provision of services in accordance with the regulations of law.

Article 345. Formalities and contents of fidelity guarantees

A loan guaranteed by a fidelity guarantee must be made in writing with certification of a socio-political organization in terms of conditions and circumstances of the borrower.

The agreement on fidelity guarantee must specify the loan amount, the purpose of loan, the term of loan, the interest rate, and the rights, obligations and responsibilities of the borrower, the lending bank or credit institution and the guarantor organization.

Sub-section 8. LIEN ON PROPERTY

Article 346. Lien on property

Lien on property means that the obligee (hereinafter referred to as the lienor) who is legally possessing the property being an object of a bilateral contract is entitled to retain the property when the obligor fails to perform the obligations or has performed the obligations not strictly as agreed upon.

Article 347. Establishment of lien on property

1. Lien on property shall arise from the due time for performance of obligation that the obligor failed to perform or perform incorrectly the obligation.

2. Lien on property shall take effect against third party from the time of possession of the possessor.

Article 348. Rights of lienors

1. Request the obligor to fulfill completely the obligations arising from a bilateral contract.

2. Require the obligor to pay expenses necessary for taking care of and keeping such property.

3. Exploit the property to obtain yield and income therefrom with the consent of the obligor.

The value of benefits from the exploitation of the property shall be offset against the value of the obligation of the obligor.

Article 349. Obligations of lienors

1. Take care and pserve the property

2. Do not change the status of the property.

3. Do not transfer or use the property without the consent of the obligor.

4. Return the property upon the complete performance of the obligation.

5. Compensate for lost or damaged property.

Article 350. Termination of lien on property

A lien on property shall terminate in any of the following cases:

1. The lienor actually no longer retains the property;

2. Contracting parties shall agree on another security instead of retain on property;

3. upon the complete performance of the obligation;

4. The property ceases to exist;

5. As agreed by the parties.

Section 4. CIVIL LIABILITY

Article 351. Civil liability arising from breach of civil obligations

1. An obligor which fails to perform or performs incorrectly an obligation has civil liability to the obligee.

Breach of obligations means that the obligor fails to perform the obligations on time, perform the obligations incompletely or incorrectly.

2. Where an obligor is not able to perform a civil obligation due to an event of force majeure, it shall not have civil liability, unless otherwise agreed or otherwise provided by law.

3. An obligor shall not have civil liability if it is able to prove that failure to perform an obligation is due entirely to the fault of the obligee.

Article 352. Responsibility for continuing performing obligations

When an obligor perform its obligations improperly, the obligee is entitled to request the obligor to continue perform such obligations.

Article 353. Late performance of civil obligations

1. Late performance of a civil obligation is the failure to have performed the civil obligation in whole or in part as at the expiry of the time-limit for the performance of such obligation.

2. The party being late in performance of a civil obligation must notify immediately the obligee about the failure to have performed the civil obligation in a timely manner.

Article 354. Postponement of performance of civil obligations

1. When it is not possible to perform a civil obligation on time, the obligor must inform immediately the obligee and may suggest postponement of performance of the civil obligation.

In the case of failure to notify the obligee, the obligor must compensate for any damage arising, unless otherwise agreed or unless it was impossible to provide notification due to objective reasons.

2. The obligor may postpone the performance of the obligation only if the obligee consents. The performance of the civil obligation in this case of postponement shall be deemed to be performance in a timely manner.

Article 355. Late acceptance of performance of civil obligations

1. The late acceptance of the performance of a civil obligation is where the time-limit for the fulfillment of the civil obligation has expired and the obligor has already fulfilled the civil obligation as agreed but the obligee does not accept the performance of such obligation.

2. When the subject matter of late acceptance of performance of a civil obligation is property, the obligor may hand over the property to a bailee must or take the necessary measures to take care of the property and is entitled to demand reimbursement of reasonable expenses. If the property is kept by a bailee, the obligor must notify the obligee.

3. The obligor has the right to sell property which is in danger of being damaged or of deteriorating, and shall pay the proceeds of sale of such property to the obligee after deducting necessary expenses for the pservation and sale of the property.

Article 356. Civil liability for failure to perform obligations to deliver objects

1. Where an obligor fails to deliver a distinctive object, the obligee has the right to require the obligor to deliver that particular object. If the object no longer exists or is damaged, the obligor must pay the value of the object.

2. Where an obligor fails to deliver a fungible object, the obligee has the right to require the obligor to deliver another fungible object. If there is no fungible object, the obligor must pay the value of the object.

3. Where an obligor fails to perform an obligation as provided in clauses 1 and 2 of this article and causes damage to the obligee, the obligor must compensate for any damage.

Article 357. Liability for late performance of the obligation to pay

1. Where the obligor makes late payment, then it must pay interest on the unpaid amount corresponding to the late period.

2. Interest arising from late payments shall be determined by agreement of the parties, but may not exceed the interest rate specified in paragraph 1 of Article 468 of this Code; if there no agreement mentioned above, the Clause 2 of Article 468 of this Code shall apply.

Article 358. Civil liability for failure to perform obligations to perform acts or not to perform acts

1. Where an obligor fails to perform an act which it must perform, the obligee has the right to request the obligor to perform the act, or the obligee may perform the act or assign the performance of the act to another person and to require the obligor pay reasonable expenses incurred and compensate for any damage.

2. Where a person has an obligation not to perform an act but, nevertheless, performs such act, the obligee has the right to require the obligor to cease performing the act, make restitution and compensate for any damage.

Article 359. Liability for late acceptance of performance of civil obligations

An obligee which is late in accepting the performance of a civil obligation, and thereby causes damage to the obligor, must compensate the obligor for any damage and shall accept all risks arising from the time when acceptance fell due, unless otherwise agreed or otherwise provided by law.

Article 360. Liability for compensation due to breach of obligations

With respect to damage caused by breach of an obligation, the obligor must compensate for the whole damage, unless otherwise agreed or pscribed by law.

Article 361. Damage caused by breach of obligations

1. Damage caused by breach of obligations comprises physical damage and spiritual damage.

2. The physical damage means those actual physical losses, comprising loss of property, reasonable expenses to pvent, mitigate or restore damage, and the actual loss or reduction of income.

3. Spiritual damage means losses related to life, health, honor, dignity or reputation and other personal benefits of an entity.

Article 362. The obligation to pvent or limit damage

The obligee must adopt the necessary and reasonable measures to pvent or limit its damage.

Article 363. Compensation for damages in case of the aggrieved party at fault

Where the breach of the obligations and damage incurred due to part of the fault of the aggrieved party, the violating party only be required to pay damages corresponding to its degree of fault.

Article 364. Fault in civil liability

Fault in civil liability includes intentional fault and unintentional fault.

Intentional fault means that a person is fully aware that its act will cause damage to another person but, nevertheless, performs the act and, irrespective of whether or not it so wishes, allows the damage to occur.

Unintentional means that a person does not foresee that its act is capable of causing damage, even though it knows or should know that the damage will occur, or where it does foresee that such act is capable of causing damage but believes that the damage will not occur or will be able to be pvented.

Section 5. TRANSFER OF RIGHT TO DEMAND AND TRANSFER OF CIVIL OBLIGATIONS

Article 365. Transfer of right to demand

1. A party having the right to demand the performance of a civil obligation may transfer such right to demand to a subrogatee of the obligee as agreed, except in the following cases:

a) The right is the right to demand support or the right to demand compensation for any damage resulting from harm to life, health, honor, dignity or reputation;

b) The obligee and the obligor agree that the right to demand may not be transferred;

2. Where a person having a right to demand transfers such right to a subrogatee, the subrogatee of the obligee shall become the person having the right to demand. The transfer of right to demand does not require the consent of the obligor.

A person transferring a right must notify the obligor in writing of the transfer of the right to demand, unless otherwise agreed. If the person transferring the right fails to notify the obligor thereby the obligor incurs expenses, the person transferring the right must pay for those expenses.

Article 366. Obligation to provide information and documents

1. A person transferring a right to demand must provide the necessary information and the relevant documents to the subrogatee of the obligee.

2. A person transferring a right to demand and breaching the provisions in Clause 1 of this Article, thereby causing damage, must compensate for such damage.

Article 367. No liability after transfer of right to demand

A person transferring a right to demand shall not be liable for the capability of the obligor to perform the obligation, unless otherwise agreed.

Article 368. Transfer of right to demand performance of secured civil obligations

Where a right to demand the performance of a secured civil obligation is transferred, the transfer of the right to demand shall include the security.

Article 369. Right to refuse of obligors

1. Where the obligor is not notified of the transfer of the right to demand or where the subrogatee of the obligee does not prove the authenticity of the transfer of the right to demand, the obligor has the right to refuse to perform the obligation with respect to the subrogatee of the obligee.

2. Where the obligor is not notified of the transfer of the right to demand and has already fulfilled the obligation with respect to the person having transferred the right to demand, the subrogatee of the obligee may not demand the obligor to perform the obligation with respect to that subrogatee.

Article 370. Transfer of civil obligations

1. An obligor may transfer a civil obligation to a subrogatee of the obligor with the consent of the obligee, except where the obligation is personal to the obligor or where the law provides that the obligation may not be transferred.

2. Upon a transfer of the obligation, the subrogatee of the obligor shall become the obligor.

Article 371. Transfer of secured civil obligations

Where a secured civil obligation is transferred, the security shall terminate, unless otherwise agreed.

Section 6. TERMINATION OF CIVIL OBLIGATIONS

Section 372. Termination of civil obligations

A civil obligation shall terminate in any of the following cases:

1. The obligation is fulfilled;

2. The parties so agree;

3. The obligee waives performance of the obligation;

4. The obligation is substituted by another civil obligation;

5. The obligation is offset;

6. The obligee and the obligor merge;

7. The limitation period for a release from the civil obligation has expired;

8. The obligor being a natural person dies, or the obligor being a juridical person ceases to exist, and the obligation must be performed by that particular natural person or juridical person;

9. The obligee being a natural person dies and the right to demand does not form part of the bequeathed estate, or the obligee being a juridical person ceases to exist and the right to demand is not able to be transferred to another juridical person;

10. A distinctive object which is the subject matter of the civil obligation no longer exists and is substituted by another civil obligation.

11. Other cases as provided by law.

Article 373. Fulfillment of civil obligations

The civil obligation shall be deemed to be have been fulfilled when the obligor has performed the obligation in its entirety, or has performed a portion of the obligation and the obligee waives any further performance.

Article 374. Fulfillment of civil obligations where obligees are late in accepting subject matter of obligations

When an obligee is late in accepting the subject matter of an obligation which is an object, the obligation to deliver an object shall be fulfilled at the moment when the object is deposited for bailment as pscribed in Clause 2 Article 355 of this Code.

Article 375. Termination of civil obligations by agreement

Parties may agree to terminate a civil obligation at any time but must not cause damage to the interests of the State or the public or the legal rights or interests of other persons.

Article 376. Termination of civil obligations due to waiver

1. A civil obligation shall terminate when the obligee waives the obligation of the obligor, unless otherwise provided by law.

2. When a secured civil obligation is waived, the security arrangement shall also terminate.

Article 377. Termination of civil obligations by substitution

1. Where parties agree to substitute an original civil obligation with another civil obligation, the original civil obligation shall terminate.

2. A civil obligation shall also terminate if the obligee has accepted another property or the performance of another act as a substitute for the property or act pviously agreed.

3. Where a civil obligation is an obligation to support others or to compensate for any damage due to harm to life, health, honor, dignity or reputation, or another personal obligation which is not able to be transferred to other persons, such obligation may not be substituted with another obligation.

Article 378. Termination of civil obligations where obligations are offset

1. Where parties have reciprocal obligations with respect to fungible objects, when both obligations fall due, the parties shall not be required to perform their obligations to each other, and the obligations shall be deemed to have terminated, unless otherwise provided by law.

2. Where the values of properties or acts are not equivalent, the parties shall settle with each other the difference in value.

3. Objects having monetary value may be used to offset an obligation to pay money.

Article 379. Cases where civil obligations may not be offset

A civil obligation may not be offset in the following cases:

1. The civil obligation is in dispute;

2. The obligation is to compensate for harm to life, health, dignity, honor or reputation;

3. The obligation is to support others;

4. Other obligations as provided by law.

Article 380. Termination of civil obligations upon merger of obligor and obligee

A civil obligation of an obligor shall terminate when the obligor becomes the obligee with respect to that particular obligation.

Article 381. Termination of civil obligations due to expiry of duration of waiver of civil obligation

Upon expiry of the duration of waiver of a civil obligation, the obligation shall terminate.

Article 382. Termination of civil obligations when obligor being natural person dies or when obligor being juridical person ceases to exist

Where parties have agreed or the law provides that an obligation must be performed by a particular obligor, when such natural person dies or such juridical person ceases to exist, the obligation shall terminate.

Article 383. Termination of civil obligations when distinctive objects no longer exist

An obligation to deliver a distinctive object shall terminate when such distinctive object no longer exists.

Parties may agree on the substitution of such object with another object or on compensation for any damage.

Article 384. Termination of civil obligations in cases of bankruptcy

In cases of bankruptcy, civil obligations shall terminate in accordance with the Law on bankruptcy.

Section 7. CONTRACTS

Sub-section 1. ENTERING INTO CIVIL CONTRACTS

Article 385. Definition of civil contract

Civil contract means an agreement between parties in relation to the establishment, modification or termination of civil rights and obligations.

Article 386. Offers to enter into civil contracts

1. Offer to enter into a contract means a clear expssion by the offeror of its intention to enter into a contract and to be bound by such offer made to another specific party or the public (hereinafter referred to as the offeree).

2. Where an offer to enter into a contract has specified the time for reply and the offeror enters into a contract with a third person during the time-limit for reply by the offeree, if the offeror fails to enter into the contract with the offeree and the offeree suffers damage, the offeror must compensate the offeree for such damage.

Article 387. Information in entering into contracts

1. Each party must notify the other party of any piece of information affecting the acceptance of offer to enter into the contract by the latter party.

2. When a party receives any secret information from the other party during the process of entering into the contract, it must protect that information and may not use it for its own purposes or other illegal purposes.

3. Any party violating Clause 1 or Clause 2 of this Article thereby causes damage must compensate for it.

Article 388. Time-limit within which offer to enter into contract remains effective

1. The time-limit within which an offer to enter into a contract remains effective shall be determined as follows:

a) Where an offeror has specified such time-limit;

b) Where an offeror has not specified the time-limit, the offer to enter into the contract is effective as from the time the offeree receives the offer.

2. The following cases shall be deemed to be receipt of an offer to enter into a contract:

a) The offer is delivered to the place of residence if the offeree is a natural person, or the offer is delivered to the head office if the offeree is a juridical person;

b) The offer is placed into the official information system of the offeree;

c) When the offeree knows about the offer to enter into a contract by way of other means.

Article 389. Modification or withdrawal of offers to enter into contracts

1. An offeror may modify or withdraw an offer to enter into a contract in the following cases:

a) If the offeree receives notice of modification or withdrawal of the offer prior to or at the same time as receipt of the offer;

b) The offeror clearly specified the circumstances in which the offer could be modified or withdrawn and such circumstances have in fact arisen.

2. When the offeror modifies the contents of the offer, that offer shall be deemed to be a new offer.

Article 390. Rescission of offers to enter into contracts

If the offeror exercises the right to rescind the offer to enter into a contract on the ground that such right was specified in the offer, the offeror must notify the offeree and such notice shall only be effective if the offeree receives the notice prior to the offeree providing its acceptance of the offer to enter into the contract.

Article 391. Termination of offers to enter into contracts

An offer to enter into a civil contract shall terminate in the following cases:

1. The offeree replies that the offer is accepted.

2. The offeree replies that the offer is not accepted;

3. The time-limit for acceptance has expired;

4. When notice of modification or withdrawal of the offer becomes effective;

5. When notice of rescission of the offer becomes effective;

6. As agreed by the offeror and the offeree within the time-limit within which the offer to enter into a contract remains effective.

Article 392. Amendment of offer proposed by offeree

When an offeree accepts the offer to enter into a contract but specifies conditions or amendments to the offer, the offeree shall be deemed to have made a new offer.

Article 393. Acceptance of offers to enter into contracts

1. Acceptance of an offer to enter into a contract means a reply by the offeree to the offeror accepting the entire contents of the offer.

2. The silence of the offeree shall not mean an acceptance of the offer to enter into the contract, unless it is agreed upon or habit established by the parties.

Article 394. Time-limits for acceptance of offers to enter into civil contracts

1. Where an offeror has specified a time limit for reply, a reply accepting shall only be effective if it is made within that time limit. If the offeror receives an acceptance after the time-limit has expired, such acceptance shall be deemed to be a new offer from the party which is late in replying.

When the offeror does not specify the time limit for reply, the reply accepting shall only be effective if it is made within reasonable period.

2. If a notice of acceptance of an offer to enter into a contract arrives late for objective reasons which the offeror knows or should know, such notice shall still be effective, unless the offeror immediately replies that it does not agree with such acceptance by the offeree.

3. Where the parties communicate directly, including conversations by telephone or other means of communication, the offeree must reply immediately as to whether or not it will accept, except where there is an agreement on the time-limit for reply.

Article 395. Cases where offeror dies or lacks of legal capacity or has limited cognition and behavior control

Where the offeror dies or lacks of legal capacity or has limited cognition and behavior control after the offeree has replied accepting the offer, the offer to enter into a contract shall still be valid, unless the contents of contract is associated with the personal identity of the offeror.

Article 396. Cases where offeree dies or lacks of legal capacity or has limited cognition and behavior control

Where the offeree dies or lacks of legal capacity or has limited cognition and behavior control after the offeree has replied accepting the offer, the offer to enter into a contract shall still be valid, unless the contents of contract is associated with the personal identity of the offeror.

Article 397. Withdrawal of notice of acceptance to enter into contract

The offeree may withdraw notice of acceptance to enter into a contract if such notice arrives prior to or at the same time as the offeror receives the reply accepting the offer to enter into a contract.

Article 398. Contents of contracts

1. The contracting parties may agree on the contents of a contract.

2. A contract may have the following contents:

a) Subject matter of the contract;

b) Quantity and quality;

c) Price and method of payment;

d) Time-limit, place and method of performing the contract;

dd) Rights and obligations of the parties;

e) Liability for breach of contract;

g) Methods of settlement of disputes.

Article 399. Places for entering into contracts

The place where a contract is entered into shall be as agreed by the parties; if there is no agreement, such place shall be the residence of the inpidual, or the head office of the legal entity, having made the offer to enter into the contract.

Article 400. Time when contracts are entered into

1. A contract is entered into at the time when the offeror receives the reply accepting to enter into the contract.

2. If the parties have agreed that silence shall constitute an acceptance within a time limit, the contract shall also be deemed to be entered into when such time-limit has expired.

3. The time when an oral contract is entered into is the time when the parties have reached agreement on the contents of the contract.

4. The time when a written contract is entered into shall be the time when the last party signs the contract or by other forms of written acceptance.

If a contract is entered into orally and then it is made in writing, the time when the contract is entered into shall be determined as pscribed in Clause 3 of this Article.

Article 401. Effectiveness of contracts

1. A contract legally entered into shall take effect from the time when it is entered into, unless otherwise agreed or otherwise provided by law.

2. From the effective date of the contract, contracting parties must mutually exercise rights and perform obligations as agreed. A contract may be amended or terminated as agreed by the parties or pscribed by law.

Article 402. Principal types of contracts

Contracts comprise the following principal types:

1. A bilateral contract is a contract whereby each party has an obligation to the other;

2. A unilateral contract is a contract whereby only one party has an obligation;

3. A principal contract is a contract the effectiveness of which does not depend on another contract;

4. An ancillary contract is a contract the effectiveness of which depends on a principal contract;

5. A contract for the benefit of a third person is a contract whereby contracting parties must perform obligations for the benefit of a third person and the third person enjoys benefits from such performance;

6. A conditional contract is a contract the performance of which depends on the occurrence, modification or termination of a specified event.

Article 403. Appendices to contracts

1. Appendices providing details on certain terms and conditions of a contract may be attached to the contract. The appendices shall have the same effectiveness as the contract. The contents of the appendices shall not contradict the contents of the contract.

2. If the terms and conditions of the appendices contradict the terms and conditions of the contract, such terms and conditions of the appendices shall be ineffective, unless otherwise agreed. If the parties agree that the terms and conditions of the appendices contradict the terms and conditions of the contract, the terms and conditions of the contract which are contradicted shall be deemed to have been amended.

Article 404. Interptation of contracts

1. Where a contract contains terms and conditions which are unclear, the interptation of such terms and conditions shall be based not only on the wording of the contract but also on the mutual intentions of the parties during the process prior to and after the time of establishment and performance of the contract.

2. Where a term of a contract may be interpted in different ways, it shall be interpted in the way which, when effective, will best benefit the parties.

3. Where the wording of a contract may be interpted in different ways, such wording shall be interpted in the way most appropriate to the nature of the contract.

4. Where a contract contains a term or wording which is difficult to understand, such term or wording shall be interpted in accordance with the customary practice of the place where the contract was entered into.

5. Where there is a conflict between the mutual intentions of the parties and the wording used in the contract, the mutual intentions of the parties shall be used in order to interpt the contract.

Article 405. Standard form contracts

1. Standard form contract means a contract containing terms and conditions which are ppared by a party based on a standard form requiring the other party to reply within a reasonable period of time. If the offeree accepts, it shall be deemed to have accepted the entire contract provided by the offeror.

The standard form contract must be public in order for the parties to know or should know the contents of the contract.

Procedures for announcement of standard form contract shall comply with regulations of law.

2. Where a standard form contract contains terms and conditions which are unclear, such terms and conditions shall be interpted in a manner favoring the offeree.

3. Where a standard form contract contains provisions exempting the party providing such standard form contract from liability, or increasing the liability of or waiving legitimate interests of the other party, such provisions shall be ineffective, unless otherwise agreed.

Article 406. General trading conditions in concluding contracts

1. General trading conditions are stable terms announced by a party to apply to the offeree; if the offeree accepts the contract is then deemed to accept these terms.

2. General trading conditions shall be effective only with the parties as long as these conditions have been publicly in order for the parties to know or should know them.

The procedures for announcement of general trading conditions shall comply with regulations of law.

3. The general trading conditions must ensure equality between the parties. If the general trading conditions contain provisions on discharge of liability from the party giving the general trading conditions, increase of responsibility or removal of the legitimate interests of the other party, these provisions do not take effect, unless otherwise agreed.

Article 407. Invalid civil contracts

1. The provisions on invalid civil transactions in Articles 123 to 138 inclusive of this Code shall also govern invalid contracts.

2. Invalidity of a principal contract shall terminate an ancillary contract, unless the parties agree that the ancillary contract replaces the principal contract. This provision shall not apply with respect to security for the performance of civil obligations.

3. Invalidity of an ancillary contract shall not terminate the principal contract, unless the parties agree that the ancillary contract is an inseparable part of the principal contract.

Article 408. Invalidity of civil contracts due to impossibility of performing subject matter

1. If, immediately as from the time a contract is signed, it is impossible to perform the subject matter of the contract for objective reasons, the contract shall be invalid.

2. If, when entering into a contract, one party knew or should have known that it was impossible to perform the subject matter of the contract for objective reasons but failed to notify the other party which entered into the contract, the former party must compensate the latter party for damage, unless the latter party knew or should have known that it was impossible to perform the subject matter of the contract.

3. The provision in Clause 1 and Clause 2 of this Article shall also apply to a contract containing one or more parts with subject matter which is impossible to perform, but the remaining parts of the contract shall remain valid.

Sub-section 2. PERFORMANCE OF CONTRACTS

Article 409. Performance of unilateral contracts

With respect to a unilateral contract, the obligor must perform the obligation strictly as agreed. The obligor may only perform the obligation prior to or after the time-limit with the consent of the obligee.

Article 410. Performance of bilateral contracts

1. With respect to a bilateral contract, where the parties have agreed on a time limit for the performance of an obligation, each party must perform its obligation when the obligation falls due. One party may not postpone performance by reason of the other party not having performed the obligations owed to the former party, except in the cases provided in Articles 411 and 413 of this Code.

2. Where the parties have no agreement on which party will perform its obligation first, the parties must perform their obligations concurrently; where obligations are not able to be performed concurrently, the obligation the performance of which will take longer shall be performed first.

Article 411. Right to postpone performance of civil obligations in bilateral contracts

1. The party which is required to perform its obligation first has the right to postpone the performance of such obligation, if the property of the other party has substantially decreased in value such that its obligation is not able to be performed as undertaken, until the other party is able to perform its obligation or has a guarantor.

2. The party which is required to perform its obligation last has the right to postpone the performance of such obligation when it falls due if the party which was required to perform its obligation first failed to do so when such obligation fell due.

Article 412. Lien on property in bilateral contracts

If the obligor fails to perform his/her obligations, the obligee shall establish the right to lien on property of the obligor as pscribed in Article 346 to Article 350 of this Code.

Article 413. Obligations not able to be performed due to fault of obligee

With respect to a bilateral contract, when one party is not able to perform its obligations due to the fault of the other party, the former party has the right to demand the latter party to continue to perform its obligation with respect to the former party or has the right to cancel the contract and demand compensation for damage.

Article 414. Failure to perform obligations not due to fault of parties

With respect to a bilateral contract, when one party is not able to perform its obligations but there is no fault of any party, the party not being able to perform does not have the right to demand the other party to perform its obligation with respect to the former party. When one party has performed part of its obligations, such party has the right to demand the other party to perform its corresponding obligation with respect to the former party.

Article 415. Performance of contracts for benefit of third parties

Where a contract is performed for the benefit of a third person, the third person has the right to demand personally the obligor to perform the obligations with respect to such third person. If there is a dispute between the parties over the performance of the contract, the third person does not have the right to demand performance until the dispute is resolved.

An obligee also has the right to demand the obligor perform a contract for the benefit of a third person.

Article 416. Right to waive of third persons

1. Where a third person waives its right to benefits prior to the performance of an obligation by an obligor, the obligor shall not be required to perform the obligation but must notify the obligee, the contract shall be deemed to be cancelled, and each party shall return anything it has received from the other party.

2. If a third person waives its must compensate if the failure to notify causes damage.

Article 424. Cancellation of the contract due to late performance of obligations

1. Where the obligor fails to perform the obligations that the obligee requests in a reasonable period of time but the obligor still fails to perform, the obligee may cancel the contract.

2. If, due to the nature of the contract or by the will of the parties, the contract will not achieve the objective if it is not performed within a certain time limit, but the obligor fails to perform that contract upon the expiry date of such time limit, the obligee has the right to cancel the contract without adherence to Clause 1 of this Article.

Article 425. Cancellation of the contract due to inability to perform

Where the obligor cannot perform part or all of its obligations to make the purpose of the obligee may not be reached, the obligee party can cancel the contract and claim damages.

Article 426. Cancellation of the contract in the case of lost or damaged property

Where a party losses or causes damage to property being the subject of a contract that cannot be refunded or compensated by other property or cannot be repaired or replaced with the same type of property, the other party may cancel contract.

The violating party shall compensate in cash equal to the value of lost or damaged property, unless otherwise agreed or stipulated in Clause 2, Clause 3, Article 351 and Article 363 of this Code.

Article 427. Consequences of cancellation of contracts

1. When a contract is canceled, the contract is void from the time of signing; the parties do not have to fulfill the obligations agreed upon, except for agreement on fines against violations, compensation and settlement of disputes.

2. The parties must return to each other what they have received after deducting from the reasonable costs of contract performance and cost of pservation and development of property.

The refund is made in kind. In case it cannot be returned in kind, it is worth the money to repay.

Where the parties are jointly obliged to refund, the refund must be made at the same time, unless otherwise agreed or otherwise provided by law.

3. The aggrieved party shall be compensated due to breach of obligations of the other party.

4. The settlement of the consequences of the cancellation of the contract relating to personal rights shall comply with this Code and other relevant law provisions.

5. In case of canceling the contract without grounds specified in Articles 423, 424, 425 and 426 of this Code, the cancellation of the contract is determined as the violating party to perform its obligations and responsibilities to comply with its obligations under the provisions of this Code, other relevant laws.

Article 428. Unilateral termination of performance of contracts

1. A party has the right to terminate unilaterally the performance of a contract without any compensation for damage when a party violates its obligations seriously if so agreed by the parties or so provided by law.

2. A party terminating unilaterally the performance of a contract must notify the other party immediately of its termination of the contract and must compensate if the failure to notify causes damage.

3. Where the performance of a contract is terminated unilaterally, it shall terminate from the time when the other party is notified of the termination. In such case, the parties are not required to continue to perform their obligations, except for agreement on fines for violations, compensation for damage and settlement of disputes. A party which has already performed its obligation may demand the other party to make payment for the performed obligation.

4. The aggrieved party shall receive a compensation for damage caused by the improper performance of obligation by the violating party.

5. If a contract is terminated unilaterally without any basis pscribed in Clause 1 of this Article, the party terminating unilaterally the performance of the contract shall be deemed to be the violating party and must perform civil liability as pscribed in this Code and relevant laws.

Article 429. Limitation period for initiating legal action with respect to contracts

The limitation period for initiating legal action to request a court to resolve a dispute relating to a contract is three years from the date on which the party entitled to request knows or should know that their lawful rights and interests are infringed.

Chapter XVI

COMMON CONTRACTS

Section 1. SALE CONTRACTS OF PROPERTY

Article 430. Sale contract of property

Sale contract means an agreement between parties whereby a seller is obligated to transfer the ownership rights of property to the purchaser and the purchaser is obligated to make a payment to the seller.

Sale contracts of houses or sale contracts of houses for other purposes shall comply with this Code, the Law on Housing and relevant laws.

Article 431. Subject matter of sale contracts

1. Each property pscribed in this Code may be the subject matter of a sale contract. If a property is banned or restricted from transfer as pscribed by laws, it shall become a subject matter of a sale contract if it complies with the regulations of such laws.

2. The property is under ownership of the seller and the seller has the right to sell it.

Article 432. Quality of objects for sale

1. The quality of an object for sale and purchase shall be as agreed by the parties.

2. Where the quality of an object has been proclaimed or is provided by a competent authority, the quality of the object agreed by the parties shall not be lower than the quality proclaimed standard or the stipulations of the competent authority.

3. Where parties have not agreed on or agree unclearly on the quality of the object for sale, its quality shall conform to requirements pertaining to quality of the object proclaimed or pscribed by a competent authority or by industry standards.

If there is no quality standard, regulations of a competent authority and industry standard in terms of an object for sale, its quality shall be determined according to normal standards or separate standards in conformity with the purposes of entering into contract and as pscribed in the Law on consumers’ right protection.

Article 433. Price and method of payment

1. Price shall be as agreed by the parties or as determined by a third person at the request of the parties. With respect to property in a transaction for which a competent authority has provided price and method of payment, the parties shall reach an agreement in accordance with such regulations.

2. Where parties reach no agreement or reach an agreement with unclear terms about price and method of payment, the price shall be determined according to the market price and the method of payment shall be determined according to the customary practice at the time and place of entering into the contract.

Article 434. Time limits for performance of sale contracts

1. The time limit for performance of a sale contract shall be as agreed by the parties. The seller must deliver the property to the purchaser at the agreed time. The seller may only deliver the property prior to the time limit with the consent of the purchaser.

3. The purchaser shall make the payment according to the agreed time. Where the parties have no agreement or have an unclear agreement on the time-limit for payment, the purchaser must pay immediately upon receipt of the property or documents proving the ownership of the property.

Article 435. Place for delivery of property

The place for delivery of the property shall be as agreed by the parties. If there is no agreement, Clause 2 of Article 277 of this Code shall apply.

Article 436. Method for delivery of property

1. Property shall be delivered by the method as agreed by the parties. If there is no agreement on the method for delivery of the property, the property shall be delivered at one time directly to the purchaser.

2. If the parties agreed that the seller shall deliver property to the purchaser many times, but the seller violates obligation of delivery in a certain time, the purchaser may cancel the part of contract related to such violation and claim compensation.

Article 437. Liability in respect of delivery of objects in incorrect quantities

1. Where a seller delivers objects in a quantity which is more than that agreed, the purchaser has the right to accept or not to accept the excess. If it accepts the excess, payment shall be made in accordance with the agreement on the excess.

2. Where a seller delivers objects in a quantity which is less than that agreed, the purchaser has one of the following rights:

a) Accept the amount delivered and set a time-limit for the seller to deliver the amount outstanding;

b) Accept the amount delivered and demand compensation for damage;

c) Cancel the contract and demand compensation for damage if the violation pvents the purchaser from achieving the purposes of enter into the contract.

Article 438. Liability in respect of delivery of incomplete integrated objects

1. Where an integrated object is delivered incomplete, thereby rendering the object unusable, the purchaser has one of the following rights:

a) Accept the object and demand the seller to deliver the remaining parts, demand compensation for damage, and postpone payment in respect of the parts received until the missing parts are delivered;

b) Cancel the contract and demand compensation for damage.

2. Where a purchaser has paid for, but not yet accepted, the delivery of an incomplete integrated object, the purchaser shall be paid interest on the amount p-paid as agreed by the parties provided it does not exceed the interest rate pscribed in Clause 1 Article 468 of this Code. If the parties do not agree the interest rate, Clause 2 Article 468 of this Code shall apply and the purchaser may demand the seller to compensate for damage due to the delivery of the incomplete integrated object from the time when the contract is required to be performed to the time when the complete integrated object is delivered.

Article 439. Liability in respect of delivery of objects of incorrect type

Where an object delivered is of an incorrect type, the purchaser has one of the following rights:

1. Accept the object and pay the agreed price;

2. Demand delivery of an object of the correct type and compensation for damage;

3. c) Cancel the contract and demand compensation for damage if the delivery of incorrect type pvents the purchaser from achieving the purposes of enter into the contract.

With regard to an object including many types, if the seller fails to deliver it in conformity with the agreement, the purchaser may cancel the part of contract related to such object and claim compensation.

Article 440. Obligation to make payment

1. A purchaser must pay the full price at the agreed place and time.

2. If the parties only agree on time limit for delivery of object, the time limit for payment shall be determined equivalent to the time limit for delivery of object. If the parties do not agree on time limit for delivery of object and payment, the purchaser must make payment upon the receipt of the object.

3. If the purchaser fails to make payment, he/she/it must pay interest on the late payment as pscribed in Article 357 of this Code.

Article 441. Transfer of risks

1. The seller shall bear all risks of the property until the property is delivered to the purchaser, the purchaser shall bear all risks of the property from the time of acceptance of the property, unless otherwise agreed or pscribed by law.

2. Where the law requires that ownership rights with respect to property which is the subject matter of a contract for sale and purchase must be registered, the seller bear all risks until the completion of the registration procedures and the purchaser bear all risks from the completion of the registration procedures, unless otherwise agreed.

Article 442. Transport costs and costs related to transfer of ownership rights

1. Transport costs and costs related to transfer of ownership rights shall be agreed by the parties, unless otherwise pscribed by law.

2. Where the parties do not reach an agreement or reach an unclear agreement on transport costs and costs related to transfer of ownership rights, those costs shall be determined according to the costs proclaimed or pscribed by a competent authority or industry standards.

3. If there is no basis pscribed in Clause 1 and Clause 2 of this Article, the transport costs and costs related to transfer of ownership rights shall be determined according to normal standards or separate standards in conformity with the purposes of entering into contract.

4. Where the parties have not agreed on and the law does not provide transport costs and costs related to transfer of ownership rights, the seller shall be liable for the costs of transportation to the place of delivery of the property and the costs related to the transfer of the ownership rights.

Article 443. Obligation to provide information and instructions for use

A seller has the obligation to provide a purchaser with necessary information on the property for sale and instructions on the use of the property. If the seller fails to perform this obligation, the purchaser has the right to require the seller to perform such obligation within a reasonable time limit and, if the seller still fails to perform such obligation that pvents the purchaser from achieving the purposes of entering into the contract, the purchaser has the right to cancel the contract and demand compensation for damage.

Article 444. Assurances as to ownership rights of purchasers with respect to purchased property

1. A seller has the obligation to assure that the ownership rights with respect to the property sold to a purchaser are not disputed by a third person.

2. Where ownership rights with respect to property are disputed by a third person, the seller must support the purchaser in protecting the interests of the purchaser. If a third person has the ownership rights with respect to all or part of the property for sale and purchase, the purchaser has the right to cancel the contract and require the seller to compensate for damage.

3. Where a purchaser knows or should know that property for sale and purchase is under the ownership of a third person but, nevertheless, purchases the property, the purchaser must return the property to the owner and does not have the right to demand compensation for damage.

Article 445. Assurances as to quality of objects for sale

1. A seller must assure the utility value or the characteristics of the object for sale and purchase. If, after having purchased an object, a purchaser discovers defects which cause the object to lose its value or diminish its utility value, the purchaser must notify the seller immediately of such defects and has the right to require the seller to repair or replace the defective object with another object, to reduce its price and to compensate for damage, unless otherwise agreed.

2. A seller must assure that an object for sale corresponds to descriptions on any package, to any trademark or to any sample selected by the purchaser.

3. A seller shall not be liable for any defect of an object in the following cases:

a) Where the purchaser knew or should have known of the defect at the time of purchase;

b) Where the object was sold at an auction or a second-hand shop;

c) Where the purchaser was at fault for causing the defect.

Article 446. Warranty obligation

If agreed by parties or provided by law, a seller has the obligation to provide a warranty for the object for sale and purchase for a certain period, hereinafter referred to as the warranty period.

The warranty period shall be calculated from the time when the purchaser has the obligation to accept the object.

Article 447. Right to claim on warranty

If a purchaser discovers a defect in a purchased object during the warranty period, it has the right to require the seller to repair the object free of charge, or reduce its price or replace it with another object, or it has the right to return the object in exchange for a refund.

Article 448. Repairs of objects during warranty periods

1. A seller must repair a defective object and assure that it satisfies the quality standards or characteristics as undertaken.

2. A seller shall pay the costs for repairing a defective object and for transporting it from the place of residence or head office of the purchaser to the place of repair and vice versa.

3. A purchaser has the right to require the seller to complete the repairs within a time limit agreed by the parties or within a reasonable time. If the seller is not able to make or complete the repairs within such time, the purchaser has the right to demand a price reduction or replacement of the defective object with another object, or it has the right to return the object in exchange for a refund.

Article 449. Compensation for damage during warranty periods

1. In addition to demanding the performance of warranty obligations, a purchaser has the right to require the seller to compensate for damage caused during the warranty period due to technical defects of the object.

2. A seller shall not be liable to compensate for damage if it is able to prove that the damage was caused due to the fault of the purchaser. The seller shall be entitled to a reduction in the amount of compensation for the damage where the purchaser has failed to take all necessary measures available to it to mitigate the damage.

Article 450. Sale of property rights

1. Where property rights are sold, a seller must deliver the relevant documents and complete the procedures for transferring the ownership rights to the purchaser and the purchaser must pay the seller.

2. Where property rights are the right to demand payment of a debt and the seller has guaranteed the ability to pay of the debtor, the seller must be jointly liable for payment if the debt falls due and the debtor fails to pay.

3. The time when ownership rights with respect to property rights are transferred is the time when a purchaser receives documents evidencing the ownership rights with respect to the property rights, or the time when the transfer of the ownership rights is registered if so provided by law.

Article 451. Auctions

Property may be sold at an auction as required by the owner or as provided by law. A sale by auction of multiple ownership property must have the consent of all owners, unless otherwise agreed or otherwise provided by law.

The auction must be conducted in conformity of principles of objective, public and transparent, assurance of lawful rights and interests of the participants and in accordance with law on auction.

Article 452. Purchases after trial use

1. Parties may agree on the trial use of purchased property by the purchaser for a period, hereinafter referred to as the trial use period. During the trial use period, the purchaser may inform the seller whether or not the purchaser wishes to make the purchase. If the purchaser fails to inform the seller prior to the expiry of the trial use period, the purchaser shall be deemed to have accepted the purchase on the terms agreed prior to the property being received for trial use.

If the parties do not reach an agreement and reach an unclear agreement on the trial use period, such period shall be determined according to the customary practice of the transactions with the same subject matter.

2. During the trial use period, property shall remain under the ownership of the seller. The seller shall bear all risks of the property, unless otherwise agreed. During the trial use period, the seller may not sell, give, lease, exchange, mortgage or pledge the property until the purchaser informs the seller whether or not it wishes to make the purchase.

3. Where a prospective purchaser informs the seller that it does not wish to make the purchase, the prospective purchaser must return the property to the seller and must compensate the seller if the prospective purchaser loses or damages the trial property. The prospective purchaser shall not be liable for normal wear and tear caused by the trial use and shall not have to return any yield derived from the trial use.

Article 453. Purchases by deferred payment or payment in instalments

1. Parties may agree that the purchaser may purchase by deferred payment or payment in instalments of the purchase price within a time limit after delivery of the purchased property. The seller has title retention of the property sold until the purchaser has paid the purchase price in full, unless otherwise agreed.

2. Contracts for purchase by deferred payment or payment in instalments shall be made in writing. The purchaser may use the property purchased by deferred payment or payment in instalments and bear all risks during the period of use, unless otherwise agreed.

Article 454. Buy-back of property sold

1. A seller and purchaser of property may agree that the seller has the right to buy-back the property within a period of time following the sale (hereinafter referred to as the buy-back period).

The buy-back period for property shall be as agreed by the parties but shall not exceed one year in respect of moveable property, and five years in respect of immoveable property, from the time of delivery of the property, unless otherwise pscribed by law. During this period, the seller may buy back the property at any time provided that reasonable prior notice is given to the purchaser. The buy-back price shall be the market price at the time when and place where the buy-back occurs, unless otherwise agreed.

2. During the buy-back period, a purchaser may not sell, exchange, give, lease, mortgage or pledge the property, and shall bear all risks with respect to the property, unless otherwise agreed.

Section 2. CONTRACTS FOR EXCHANGE OF PROPERTY

Article 455. Contracts for exchange of property

1. Contract for the exchange of property means an agreement between parties whereby they deliver property, and transfer the ownership rights thereto, to each other.

2. A contract for the exchange of property must be made in writing, and must be notarized, certified or registered if so provided by law.

3. Where one party exchanges with another party property which it does not own or property in respect of which it has no authorization from the owner, the other party may cancel the contract and demand compensation for damage.

4. Each party shall be deemed to be the seller of the property delivered to the other party and the purchaser of the property delivered to it. The provisions on contracts for sale and purchase in articles 430 to 439 inclusive and articles 449 to 454 inclusive of this Code shall also apply to contracts for the exchange of property.

Article 456. Settlement of differences in value

Where exchanged property differs in value, the parties must settle that difference between themselves, unless otherwise agreed or otherwise provided by law.

Section 3. CONTRACTS FOR GIFTS OF PROPERTY

Article 457. Contracts for gifts of property

Contract for a gift of property means an agreement between parties whereby the giver delivers its property and transfers its ownership rights to the recipient without requiring compensation and the recipient agrees to accept the gift.

Article 458. Gifts of moveable property

1. A contract for a gift of moveable property shall take effect when the recipient accepts the property, unless otherwise agreed.

2. Where the law requires the ownership rights with respect to such moveable property to be registered, the contract shall take effect from the time of registration.

Article 459. Gifts of immoveable property

1. A gift of immoveable property must be recorded in writing and notarized or certified, and must be registered if the law on immoveable property requires registration of ownership.

2. A contract for a gift of immoveable property shall take effect from the time of registration. In the case of immoveable property for which no registration of ownership rights is required, the gift contract shall take effect from the time when the property is delivered.

Article 460. Liability in respect of intentional gift of property not under one’s ownership

Where a giver intentionally gives property which is not under its ownership and the recipient does not know or is not able to know, such giver must reimburse the recipient for expenses incurred by the recipient in increasing the value of the property at such time as it is reclaimed by the owner.

Article 461. Disclosure of defects in gifts

A giver has the obligation to notify the recipient of any defects in a gift. If the giver knows about defects in a gift but fails to provide notice thereof, the giver must be liable to compensate for damage caused to the recipient; but if the giver does not know about defects in a gift, the giver shall not be liable to compensate for damage.

Article 462. Conditional gifts of property

1. A giver may require a recipient to perform one or several civil obligations prior to or after the giving of a gift. The conditions for giving a gift must not contravene the law or social morals.

2. Where a recipient performs an obligation required to be performed as a condition to the giving of a gift and the giver fails to deliver the gift, the giver must pay for the obligation already performed by the recipient.

3. Where a recipient fails to perform an obligation required to be performed after the giving of a gift, the giver may reclaim the gift and demand compensation for damage.

Section 4. CONTRACTS FOR LOAN OF PROPERTY

Section 463. Contracts for loan of property

Contract for the loan of property means an agreement between parties whereby a lender delivers property to a borrower. When the loan falls due, the borrower must repay the lender property of the same type in accordance with the correct quantity and quality, and must pay interest if so agreed or so provided by law.

Article 464. Ownership rights with respect to property lent

A borrower shall become the owner of borrowed property from the time of delivery of the property.

Article 465. Obligations of lenders

1. Deliver the property to the borrower in full, strictly in accordance with the quality and quantity, and at the time and place, agreed.

2. Compensate the borrower for any damage where the lender knows that the property is not of the agreed quality but fails to notify the borrower, unless the borrower accepts the property with knowledge that the property is not of the agreed quality.

3. Do not demand the borrower to return the property prior to the due date, except in the cases provided in article 470 of this Code or relevant laws.

Article 466. Obligations of borrowers to repay loans

1. Where the property lent is a sum of money, the borrower must repay the lender the loan in full when due. If the property is an object, the borrower must deliver to the lender an object of the same type, quantity and quality, unless otherwise agreed.

2. Where a borrower is not able to deliver an object, it may, with the consent of the lender, repay the value of the borrowed object, in cash, as at the time and place of delivery.

3. The place for repayment of a loan shall be the place of residence or head office of the lender, unless otherwise agreed.

4. If a borrower fails to repay all or any instalment of an interest-free loan, in whole or in part, when payment falls due, the borrower must, if the parties so agree, pay interest on the overdue amount from the due date until the date on which payment is made, at the basic interest rate pscribed in Clause 2 Article 468 of this Code, unless otherwise agreed or otherwise pscribed by law.

5. If a borrower fails to repay, in whole or in part, a loan with interest, the borrower must pay:

a) Interest on the principal as agreed in proportion to the overdue loan term and interest at the rate pscribed in Clause 2 Article 468 in case of late payment;

b) Overdue interest on the principal equals one hundred and fifty (150) per cent of the interest rate in proportion to the late payment period, unless otherwise agreed.

Article 467. Use of borrowed property

Parties may agree that borrowed property may only be used for the agreed purpose of the loan. The lender may check the use of the property and may demand its early return if, despite warning, the borrower continues to use the property contrary to the agreed purpose.

Article 478. Interest rates

1. The rate of interest for a loan shall be as agreed by the parties.

The rate of interest for a loan agreed by the parties may not exceed 20% per year, unless otherwise pscribed by law. According to actual conditions and at the proposal of the Government, the Standing Committee of National Assembly shall adjust the above interest and send report to the National Assembly at the latest session.

If the agreed interest exceeds the maximum interest pscribed in this Clause, the agreed interest shall become invalid.

2. Where parties agree that interest will be payable but fail to specify the interest rate, or where there is a dispute as to the interest rate, the interest rate for the duration of the loan shall equal 50% of the maximum interest pscribed in Clause 1 of this Article at the repayment time.

Article 469. Performance of contracts for loans without fixed term

1. With respect to a contract for an interest-free loan without a fixed term, the lender may reclaim the property, and the borrower may repay the debt, at any time provided that each party gives reasonable prior notice to the other party, unless otherwise agreed.

2. With respect to a contract for a loan with interest without a fixed term, the lender may reclaim the property at any time, subject to giving reasonable prior notice to the borrower, and shall be paid interest until the time when the property is returned. The borrower may also return the property at any time, subject to giving reasonable prior notice to the lender, in which case the borrower shall pay interest only up to the date on which repayment is made.

Article 470. Performance of contracts for fixed term loans

1. With respect to a contract for a fixed term interest-free loan, the borrower may return the property at any time, subject to giving reasonable prior notice to the lender. The lender may reclaim the property prior to the due date, subject to the consent of the borrower.

2. With respect to a contract for a fixed term loan with interest, the borrower may return the property prior to the due date, but must pay interest for the entire term, unless otherwise agreed or otherwise pscribed by law.

Article 471. “Họ, hụi, biêu, phường”

1. “Họ, hụi, biêu, phường” (hereinafter referred to as “ho”) means a form of transaction regarding property in accordance with customary practice on the basis of an agreement reached by a group of people who assemble together and jointly determine the number of people, the term, the amount of money or other property, the form of contribution and payment of “ho”, and the rights and obligations of the members of the group.

2. “Ho” is aimed at mutual assistance of citizens and shall be implemented in accordance with law.

3. If “ho” is organized with interest, it shall comply with this Code.

4. It shall be strictly prohibited to organize “ho” in the form of lending at high interest rates.

Section 5. CONTRACTS FOR LEASE OF PROPERTY

Sub-section 1. GENERAL PROVISIONS ON CONTRACTS FOR LEASE OF PROPERTY

Article 472. Contracts for lease of property

Contract for lease of property means an agreement between parties whereby a lessor delivers property to a lessee for use during a fixed term and the lessee is required to pay rent.

Lease contracts of houses or lease contracts of houses for other purposes shall comply with this Code, the Law on Housing and relevant laws.

Article 473. Rent

1. Rent for a lease of property shall be as agreed by the parties or determined by a third party at the request of the parties, unless otherwise pscribed by law.

2. Where parties reach no agreement or reach an agreement with unclear terms about the rent, it shall be determined according to the market price at the time and place of entering into the contract.

Article 474. Lease term

1. The term of a lease shall be as agreed by the parties. If there is no agreement, the term of the lease shall be determined according to the purpose of the lease.

2. Where parties have not agreed on the term of a lease or where the term of a lease is not able to be determined according to the purpose of the lease, the lease shall terminate when the lessee has achieved the purpose of the lease.

Article 475. Sub-leases

A lessee may sub-lease leased property with the consent of the lessor.

Article 476. Delivery of leased property

1. A lessor must deliver property to the lessee strictly in accordance with the agreed quantity, quality, type and condition and at the agreed place and time, and must provide information necessary for use of the property.

2. Where a lessor is late in delivering property, the lessee may extend the time for the delivery of the property or may cancel the contract and demand compensation for damage. If the leased property is not of the quality agreed, the lessee has the right to require the lessor to repair the property or reduce the rent, or to cancel the contract and demand compensation for damage.

Article 477. Obligation to assure utility value of leased property

1. A lessor must assure that leased property is in the condition agreed and is suitable for the purpose of the lease contract for the entire term of the lease. The lessor must repair any damage to or defect in the leased property, with the exception of minor damage which the lessee must repair in accordance with customary practice.

2. Where the utility value of leased property decreases otherwise than due to the fault of the lessor, the lessee has the right to demand the lessor:

a) Repair the property;

b) Reduce the rent;

c) Replace the property or, if the leased property is beyond repair and the purpose of the lease is not able to be achieved as a result or if the property has defects which the lessee did not know of, the lessee may terminate unilaterally the performance of the contract and demand compensation for damage.

3. Where a lessee demands repair but the lessor fails to repair in time or at all, provided that the lessee informs the lessor, the lessee has the right to repair personally the leased property and has the right to require the lessor to reimburse the costs of repair.

Article 478. Obligation to assure right of lessees to use property

1. A lessor must assure the right of a lessee to uninterfered use of the property.

2. In the event of a dispute as to the ownership rights with respect to leased property, which interferes with use of that property by the lessee, the lessee has the right to terminate unilaterally the performance of the contract and demand compensation for damage.

Article 479. Obligation to take care of leased property

1. A lessee shall take care of leased property as if it were its own and shall carry out minor repairs and maintenance. If the lessee causes any loss of or damage to the property, it must compensate the lessor.

The lessee shall not be liable for normal wear and tear due to the use of the leased property.

2. A lessee may, with the consent of the lessor, carry out repairs and improvements to leased property which increase its value and may require reimbursement from the lessor for reasonable costs incurred.

Article 480. Obligation to use leased property strictly in accordance with utility and purpose

1. A lessee must use leased property strictly in accordance with its utility and the agreed purpose.

2. Where a lessee fails to use leased property strictly in accordance with its utility and purpose, the lessor has the right to terminate unilaterally the performance of the contract and to demand compensation for damage.

Article 481. Payment of rent

1. A lessee must pay rent in full and on time as agreed. If there is no agreement on the time for payment of rent, the time shall be determined in accordance with the customary practice at the place of payment. If it is not possible to determine the time of payment in accordance with such customary practice, the lessee must make payment when the lessee returns the leased property.

2. Where the parties have agreed on payment of rent in instalments, if the lessee fails to make payment for three consecutive instalments, the lessor has the right to terminate unilaterally the performance of the lease contract, unless otherwise agreed or otherwise provided by law.

Article 482. Return of leased property

1. A lessee must return leased property in the same condition in which it was received, normal wear and tear excepted, or in the condition agreed. If the value of the leased property has decreased in comparison with its condition at the time it was received, the lessor has the right to demand compensation for any damage, normal wear and tear excepted.

2. Where leased property is moveable property, the place for returning the leased property shall be the place of residence or head office of the lessor, unless otherwise agreed.

3. Where leased property is livestock, the lessee must return both the leased livestock and any offspring born during the term of the lease, unless otherwise agreed. The lessor must reimburse the lessee for expenses incurred in caring for the offspring.

4. Where a lessee is late in returning leased property, the lessor has the right to require the lessee to return the leased property and to pay rent for the period of delay and the lessee must compensate for damage. The lessee must pay a penalty for the late return of the leased property if so agreed.

5. The lessee must bear the risk in relation to the leased property during the period of delay.

Sub-section 2. CONTRACTS FOR “THUE KHOAN” OF PROPERTY

Article 483. Contracts for “thue khoan” of property

Contract for “thue khoan” of property means an agreement between parties whereby a “thue khoan” lessor delivers the property to a “thue khoan” lessee for the exploitation of its utility and for the enjoyment of the yield and income derived from such property and the lessee has the obligation to pay rent.

Article 484. Subject matter of contracts for “thue khoan” of property

The subject matter of a contract for “thue khoan” may be land, forest, and water surfaces which have not been exploited, livestock, production and business facilities, and other means of production together with the equipment required to exploit the utility of such property and to enjoy the benefits and income derived therefrom, unless otherwise provided by law.

Article 485. Terms of “thue khoan”

The term of a “thue khoan” shall be agreed by parties. If the parties do not reach an agreement and reach an unclear agreement on the term of “thue khoan”, such period shall be determined according to the production or business cycle appropriate to the nature of the subject matter of the “thue khoan”.

Article 486. Rent in respect of “thue khoan”

Rent in respect of a “thue khoan” shall be as agreed by the parties. If a “thue khoan” is awarded by tender, the rent in respect of such “thue khoan” shall be determined in the bidding process.

Article 487. Delivery of “thue khoan” property

Upon delivery of “thue khoan” property, parties must record their assessment of the condition and value of the property.

If the parties are not able to determine the value, a third person shall be invited to determine the value. Such determination shall be made in writing.

Article 488. Payment of “thue khoan” rent and method of payment

1. “Thue khoan” rent may be paid in kind or money, or by performance of acts.

2. A “thue khoan” lessee must pay “thue khoan” rent in full even where the lessee does not exploit the utility of the “thue khoan” property.

3. When entering into a contract for “thue khoan”, parties may agree on conditions for a reduction in rent. If at least one third of the benefits or income is lost due to an event of force majeure, a “thue khoan” lessee may demand a reduction of or exemption from rent, unless otherwise agreed.

4. Where a “thue khoan” lessee must pay in kind according to the season or cycle in the exploitation of the utility of the “thue khoan” property, payment must be made at the end of such season or cycle, unless otherwise agreed.

5. Where a “thue khoan” lessee is required to perform an act, the “thue khoan” lessee must perform that particular act.

6. Repayment term of “thue khoan” shall be agreed by the parties. If the parties have no agreement, the lessee has to pay on the last day of each month; in case the “thue khoan” follows the business cycle, the payment must be made at the end of the business cycle at the latest.

Article 489. Exploitation of “thue khoan” property

A “thue khoan” lessee must exploit “thue khoan” property strictly in accordance with the agreed purpose and must inform the “thue khoan” lessor periodically of its condition and its exploitation. If the “thue khoan” lessor demands or requires information at any other time, the “thue khoan” lessee must provide such information promptly. If the “thue khoan” lessee does not exploit the “thue khoan” property strictly in accordance with the agreed purpose, the “thue khoan” lessor has the right to terminate unilaterally the performance of the contract and demand compensation for damage.

Article 490. Taking care, maintenance and disposal of “thue khoan” property

1. During the period of exploitation of the “thue khoan” property, the “thue khoan” lessee must take care of and pserve the “thue khoan” property and any related equipment at its own expense, unless otherwise agreed. If the “thue khoan” lessee loses or damages the “thue khoan” property or causes any loss or reduction in its value, the “thue khoan” lessee must compensate for damage. The “thue khoan” lessee shall not be liable for normal wear and tear due to the use of the “thue khoan” property.

2. A “thue khoan” lessee may replace or improve “thue khoan” property if the parties so agree, but must pserve its value.

A “thue khoan” lessor must reimburse a “thue khoan” lessee for the reasonable expenses incurred in replacing or improving the “thue khoan” property as agreed.

3. A “thue khoan” lessee may not sub-lease “thue khoan” property without the consent of the “thue khoan” lessor.

Article 491. Enjoyment of benefits and liability for damage with respect to “thue khoan” livestock

During the term of a “thue khoan” of livestock, the “thue khoan” lessee shall be entitled to enjoy half of the number of offspring born and shall be liable for half of any damage of the “thue khoan” livestock caused by an event of force majeure, unless otherwise agreed.

Article 492. Unilateral termination of performance of contracts for “thue khoan”

1. Where a party terminates unilaterally the performance of a “thue khoan” contract, that party must give reasonable prior notice to the other party. If the “thue khoan” is based on a season or cycle of exploitation, the period of prior notice must conform to such season or cycle.

2. Where a “thue khoan” lessee breaches an obligation and the exploitation of the “thue khoan” property is the sole means of livelihood of the lessee and continuation of the “thue khoan” would not seriously affect the interests of the “thue khoan” lessor, the “thue khoan” lessor may not terminate unilaterally the performance of the contract. In such case, the “thue khoan” lessee must undertake to the “thue khoan” lessor not to commit further breaches of the contract.

Article 493. Return of “thue khoan” property

Upon termination of a “thue khoan” contract, the “thue khoan” lessee must return the “thue khoan” property in a condition which takes into account the agreed depciation. If the “thue khoan” lessee has caused any reduction in the value of the “thue khoan” property, the “thue khoan” lessee must compensate for damage.

Section 6. CONTRACTS FOR BORROWING PROPERTY

Article 494. Contracts for borrowing property

Contract for borrowing property means an agreement between parties whereby a lender delivers property to a borrower for use free of charge for a period of time and the borrower returns the property at the end of the period of time or when the purpose of the borrowing has been achieved.

Article 495. Subject matter of contracts for borrowing property

Any non-consumable object may be the subject matter of a contract for borrowing property.

Article 496. Obligation of borrowers of property

1. Take care of and pserve the borrowed property and not to change the condition thereof at the volition of the borrower. The borrower must repair any normal damage to the property.

2. Do not on-lend the property to any other person without the consent of the lender.

3. Return the borrowed property on the due date. If there is no agreement on the time for returning the property, the borrower must return the property immediately after the purpose of the borrowing has been achieved.

4. Compensate for damage where the borrower causes damage to or loss of the borrowed property.

5. The borrower must bear the risk in relation to the borrowed property during the period of late return.

Article 497. Rights of borrowers of property

1. Use the borrowed property strictly in accordance with its utility and agreed purpose.

2. Require the lender to reimburse reasonable expenses incurred in carrying out repairs or improvements to the borrowed property which increase its value if so agreed.

3. Do not bear liability for normal wear and tear of the borrowed property.

Article 498. Obligations of lenders of property

1. Provide necessary information on the use of the property and its defects, if any.

2. Reimburse the borrower for expenses incurred in carrying out repairs or improvements to the borrowed property which increase its value if so agreed.

3. Where the lender knows but fails to notify the borrower of a defect in the property which results in damage to the borrower, to compensate the borrower for such damage, except where the borrower knows or should know of such defect.

Article 499. Rights of lenders of property

1. Reclaim the property immediately after the borrower has achieved its purpose where there is no agreement on the borrowing period. If the lender suddenly and urgently needs to use the borrowed property, the lender may reclaim it upon giving reasonable prior notice to the borrower, even if the borrower has not yet achieved its purpose.

2. Reclaim the property where the borrower fails to use it strictly in accordance with the agreed purpose, utility, or method or where the borrower on-lends the property without the consent of the lender.

3. Demand compensation for damage to the property caused by the borrower.

Section 7. CONTRACTS OF LAND USE RIGHTS

Article 500. Contract of land use rights

The contract of land use rights means the agreement between the parties that the land user convert, transfer, lease, sublease, donate, mortgage and contribute land use rights as capital or exercise other rights to the other party as pscribed in the Law on land; and the other party shall exercise rights and perform obligations according to the contract to the land user.

Article 501. Contents of contract of land use rights

1. General provisions on contracts and the content of common contracts related in this Code shall also apply to contracts on land use rights, unless otherwise provided by law.

2. Contents of the contract of land use rights are not contrary to the provisions of the purpose of use, the duration of land use, zoning, land use planning and the rights and obligations stipulated by law on land and other provisions of relevant laws.

Article 502. Forms and procedures for performing contract of land use rights

1. Contracts relating to land use rights must be made in writing in the form consistent with the provisions of this Code, the law on land and other provisions of relevant laws.

2. The performance of the contract for the land use rights must follow the procedures pscribed by the law of the land and other provisions of relevant laws.

Article 503. Effect of the transfer of land use rights

The transfer of land use rights shall be effect from the date of registration under the provisions of the law on land.

Section 8. COOPERATION CONTRACT

Article 504. Cooperation contract

1. A cooperation contract means an agreement between natural and/or juridical persons regarding the property contribution, effort to perform certain jobs, the same benefit and mutual responsibility.

2. Each cooperation contract must be made in writing.

Article 505. Contents of cooperation contract

Each cooperation contract shall contain the major contents below:

1. Purpose and duration of cooperation;

2. Full name and place of residence of natural person; name and headquarters of juridical person;

3. Contributed property (if any);

4. Contributed labor (if any);

5. Method of distributing the yield and/or income;

6. Rights and obligations of cooperative members;

7. Rights and obligations of repsentatives (if any);

8. Conditions for participation and withdrawal from the cooperation contract (if any);

9. Conditions for termination of cooperation.

Article 506. Common property of the cooperative members

1. Property contributed and created by the members and other property as pscribed by law shall be considered as common property part of the cooperative members.

Where there is agreement on the contribution that members fail to contribute money on schedule, they must pay interest on the unpaid portion of the money under the provisions of Article 357 of this Code and must pay damages.

2. The disposition of the property is land use rights, housing, factories, other production materials must have a written agreement of all the members; the disposal of other property shall be decided by the repsentatives of the members, unless otherwise agreed.

3. The common property may not be pided before the termination of the cooperation contract, unless otherwise agreed by all members.

The pision of common property shall not change or terminate the rights and obligations established before the pision time.

Article 507. Rights and obligations of cooperative members

1. Enjoy the yield and income gained from cooperation activities.

2. Participate in decisions involving issues of performance of cooperation contract and co-operation monitoring.

3. Compensate for damages to other cooperative members caused by their faults.

4. Perform other rights and obligations under the contract.

Article 508. Establishing and performing civil transactions

1. Where the cooperative members appoint repsentatives, they shall establish and perform civil transactions.

2. Where the cooperative members do not appoint any repsentative, they shall jointly establish and perform civil transactions, unless otherwise agreed.

3. Civil transactions that are be established and performed by entities specified in Clauses 1 and 2 of this Article shall give rise to rights and obligations of all cooperative members.

Article 509. Civil liability of cooperative members

The cooperative members are responsible for general civil liability by their common property; if the common property is insufficient to fulfill the obligation, the members shall be held responsible by their own property corresponding to its contribution, unless otherwise specified by the cooperation contract or otherwise pscribed by law.

Article 510. Withdrawal from the cooperation contract

1. A member has the right to withdraw from the cooperation contract in any of the following cases:

a) Satisfy the conditions specified in the cooperation contract;

b) There is a good reason and the consent of more than half the total cooperative members.

2. The member withdrawing from the cooperation contract has the right to reclaim contributed property, pided part of the property in the common property and must pay all obligations under the agreement. Where the pision of property in kind affects the cooperation activities, the property are worth the money to pide.

The withdrawal of a member from the cooperation contract does not terminate his/her rights and obligations that are established or performed before the withdrawal time.

3. The withdrawal of a member from the contract in a case other than Clause 1 of this Article shall be defined as the breach of contract and should execute the civil liability as pscribed in this Code and other relevant laws.

Article 511. Joining cooperation contract

Unless the cooperation contract otherwise specifies, a natural or juridical person shall become a new member of the contract with the consent of more than half the total cooperative members.

Article 512. Termination of cooperation contract

1. A cooperation contract shall terminate in any of the following cases:

a) As agreed by cooperative members;

b) The time limit mentioned in the cooperation contract has expired;

c) The purpose of cooperation has been achieved;

d) Pursuant to a decision of a competent authority;

dd) Other cases pscribed in this Code or other relevant laws.

2. Upon the termination of the cooperation contract, the debts arising from the contract must be paid; if the common property is insufficient to repay it to private assets of cooperative members to pay under the provisions of Article 509 of this Code.

Where the debt was repaid and the common property still exists, it shall be pided by the cooperative members in proportion to the contribution of each person, unless otherwise agreed.

Section 9. CONTRACTS FOR SERVICES

Article 513. Contracts for services

Contract for services means an agreement between parties whereby a service provider performs an act for a client which pays a fee for that act.

Article 514. Subject matter of contracts for services

The subject matter of a contract for services must be an act which is capable of being performed, which is not prohibited by law and which does not contravene social morals.

Article 515. Obligations of clients

1. Supply the service provider with the information, documentation and facilities necessary for the performance of the act if so agreed or required for the performance of the act.

2. Pay a fee to the service provider as agreed.

Article 516. Rights of clients

1. Require the service provider to perform the act strictly in accordance with the agreement on quality, quantity, time, location and other matters.

2. Where a service provider commits a serious breach of its obligations, the client may terminate unilaterally performance of the contract and demand compensation for damage.

Article 517. Obligation of service providers

1. Perform the act strictly in accordance with the agreement on quality, quantity, time, location and other matters.

2. Do not assign the act to another person for performance on its behalf without the consent of the client.

3. Take care of, and to return to the client after completion of the act, the documents and facilities provided to it.

4. Notify the client promptly of any inadequacy in the information or documents and any failure of the facilities to satisfy the quality required for the completion of the act.

5. Keep confidential any information of which it has had knowledge during the period of providing the service as agreed or as provided by law.

6. Compensate the client for damage where the service provider causes any loss of or damage to the documents or facilities supplied or discloses confidential information.

Article 518. Rights of service providers

1. Require the client to provide information, documents and facilities.

2. Amend the terms of service in the interests of the client without necessarily asking for the opinion of the client where waiting for such opinion would cause damage to the client provided that the service provider promptly informs the client thereof.

3. Require the client to pay the fee.

Article 519. Payment of fees

1. A client must pay the agreed fee for services.

2. If, upon entering a contract, there is no agreement on the service fee rate or on the method for fixing the fee for services and there are no other instructions on fees, the service fee rate shall be fixed on the basis of market fees for services of the same type at the time when and place where the contract was entered into.

3. A client must pay the fee for services at the place where the service is provided and at the time of its completion, unless otherwise agreed.

4. Where the services provided fail to meet the terms of the agreement or the act is not completed in time, the client has the right to reduce the fee for services and demand compensation for damage.

Article 520. Unilateral termination of performance of contracts for services

1. Where the continued provision of services does not benefit the client, the client has the right to terminate unilaterally the performance of the contract but must provide reasonable prior notice to the service provider, in which case the client must pay a fee according to the portion of services already provided and will endanger the passenger or other persons during the journey;

c) In order to pvent the spad of contagious diseases.

Article 526. Obligations of passengers

1. Pay in full the passenger transport fare and fares for the transport of luggage in excess of the pscribed limit, and to take care of his or her hand-luggage.

2. Present at the place of departure at the agreed time.

3. Respect and comply strictly with the regulations of the carrier and all other regulations ensuring traffic safety.

Article 527. Rights of passengers

1. Request to be transported by the agreed means of transport, in the class commensurate with the value of the ticket and in accordance with the agreed route.

2. Be exempt from transport fares for check-in luggage and hand-luggage within the limits as agreed or as provided by law.

3. Demand reimbursement of expenses incurred or compensation for any damage if the carrier is at fault in failing to transport according to the agreed time schedule and destination.

4. Receive a refund of all or part of the transport fare in the cases provided for in Points b and c Clauses 2 of Article 525 of this Code and in other cases as agreed or as provided by law.

5. Receive the luggage at the agreed destination in accordance with the agreed time and route.

6. Request temporary interruption of the journey for the duration and in accordance with the procedures provided by law.

Article 528. Liability to compensate for damage

1. In case of loss of life of or damage to the health or luggage of a passenger, the carrier must compensate for any damage in accordance with law.

2. Unless otherwise provided by law, a carrier shall not be liable to compensate for loss of life of or damage to the health and luggage of a passenger in the case where such loss or damage is entirely due to the fault of the passenger, unless otherwise pscribed by law.

3. A passenger which breaches the agreed terms for transport or the transport regulations, thereby causing damage to the carrier or a third person, must compensate.

Article 529. Unilateral termination of contracts for transport of passengers

1. A carrier has the right to terminate unilaterally the performance of a contract in the cases provided in Clause 2 of Article 525 of this Code.

2. A passenger has the right to terminate unilaterally the performance of a contract where the carrier breaches the obligations provided in Clauses 1, 3 and 4 of article 524 of this Code.

Sub-section 2. CONTRACTS FOR TRANSPORT OF PROPERTY

Article 530. Contracts for transport of property

Contract for transport of property means an agreement between parties whereby a carrier has the obligation to transport property to an agreed destination and to deliver it to the authorized recipient, and the customer has the obligation to pay the freight charges.

Article 531. Formalities for contracts for transport of property

1. A contract for transport of property may be entered into orally or in writing or a specific act.

2. A bill of lading or equivalent source document of transport shall be evidence of the entering into of a contract by the parties.

Article 532. Delivery of property to carriers

1. A customer has the obligation to deliver property to a carrier at the agreed time and place and pack the property in accordance with the agreed specifications. The customer must bear the costs f loading the property onto and unloading the property from the means of transport, unless otherwise agreed.

2. Where a customer is late in delivering property to the agreed place, the customer must reimburse the carrier for the costs associated with the delay and pay the freight charges for transporting the property to the place agreed in the contract, or must pay a penalty for breach of the contract.

If the carrier is late in accepting the property at the agreed at the agreed place, it shall be liable for the costs incurred due to such late acceptance.

Article 533. Freight charges

1. The rate of freight charges shall be as agreed by the parties. If the law regulates freight charges, charges shall apply as regulated.

2. A customer must pay freight charges in full after the property is loaded onto the means of transport, unless otherwise agreed.

Article 534. Obligations of carriers

1. Transport the property in its entirety and safely to the agreed destination at the agreed time.

2. Deliver the property to the person entitled to receive it.

3. Bear all costs related to the transport of the property, unless otherwise agreed.

4. Purchase civil liability insurance as required by law.

5. Compensate the customer for damage where the loss of or damage to the property is caused by the fault of the carrier, unless otherwise agreed or otherwise provided by law.

Article 535. Rights of carriers

1. Check the authenticity of the property and the bill of lading or equivalent source document of transport.

2. Refuse to transport property which is different from that agreed in the contract.

3. Demand the full and timely payment of the freight charges by the customer.

4. Refuse to transport property if the carrier knows or should know that the transacting of such property is prohibited or the property is of a dangerous or toxic nature.

Article 536. Obligations of customers

1. Pay in full the freight charges to the carrier, at the time and by the method of payment as agreed.

2. Provide necessary information about the transported property to ensure its safety.

3. Take care of the property during transport if so agreed. Where the customer takes care of the property and it is lost or damaged, the customer shall not be entitled to compensation.

Article 537. Rights of customers

1. Demand the carrier to transport the property to the agreed destination at the agreed time.

2. Receive directly the property which has been transported, or appoint a third person to receive it.

Article 538. Delivery of property to recipients

1. A recipient of property may be the customer or a third person appointed by the customer to receive the property.

2. A carrier must deliver the property to a recipient in full, at the time and place and by the method as agreed.

3. Where property has been delivered to the point of delivery on time but there is no recipient of the property, the carrier may deposit such property at a place of bailment and must notify immediately the customer or the recipient of the property. The customer or recipient of the property must bear the reasonable expenses incurred in relation to the bailment of the property.

The obligation to deliver property shall be completed upon bailment of the property in compliance with the agreed terms and when the customer or the recipient of the property has been notified about the bailment.

Article 539. Obligation of recipients of property

1. Produce to the carrier the bill of lading or other equivalent source document of transport, and to receive the property at the agreed time and place.

2. Bear the costs for loading and unloading the transported property, unless otherwise agreed or otherwise provided by law.

3. Reimburse the carrier for reasonable costs incurred due to late acceptance of the property.

4. Notify the customer of the acceptance of the property and provide other necessary information required by the customer if the recipient is a third party appointed by the customer.

Article 540. Rights of recipients of property

1. Verify the quantity and quality of the delivered property.

2. Accept the delivered property.

3. Require the carrier to reimburse reasonable costs incurred due to any delay by the carrier in delivering the property.

4. Require the carrier compensate for loss of or damage to the property.

Article 541. Liability to compensate for damage

1. Where a carrier is responsible for the loss of or damage to property, the carrier must compensate the customer for damage, except in the case provided in Clause 3 of Article 536 of this Code.

2. A customer must compensate a carrier and any third parties for damage caused by the transport of dangerous or toxic property which is not safely packaged or the safety of which is not otherwise ensured.

3. A carrier shall not be liable to compensate for damage in the event of force majeure causing loss or deterioration of or damage to the property during transport, unless otherwise agreed or otherwise provided by law.

Section 11. PROCESSING CONTRACTS

Article 542. Processing contracts

Processing contract means an agreement between parties whereby a processor carries out work to create products at the request of a supplier, and the supplier receives the products and pays fees.

Article 543. Subject matter of processing contracts

The subject matter of a processing contract shall be items which are specified by samples, the standard of which is agreed by the parties or provided by law.

Article 544. Obligation of suppliers

1. Supply raw materials to the processor strictly in accordance with the agreed quantity, quality, time and place, unless otherwise agreed by the parties; and to provide necessary documents relating to the processing.

2. Provide the processor with instructions as to how to perform the contract.

3. Pay agreed fees.

Article 545. Rights of suppliers

1. Accept the processed products in accordance with the agreed quantity, quality, manner, time and place.

2. Terminate unilaterally performance of the contract and demand compensation for damage if the processor commits a serious breach of the contract.

3. Where the products are not of the agreed quality and the supplier accepts the products but requests repairs, but the processor is not able to perform the repairs within the agreed time, the supplier has the right to cancel the contract and demand compensation for damage.

Article 546. Obligations of processors

1. Take care of the raw materials supplied by the supplier.

2. Notify the supplier to replace any raw materials supplied which are not of the agreed quality; to refuse to perform the processing if the use of the raw materials may create products which pose a danger to society.

3. Deliver the products to the supplier strictly in accordance with the agreed quantity, quality, method, time and place.

4. Keep confidential all information relating to the processing and the products.

5. Bear liability for the quality of the products, unless the lack of quality is due to the raw materials supplied by the supplier or due to the unreasonable instructions of the supplier.

6. Return any leftover raw materials to the supplier after completing performance of the contract.

Article 547. Rights of processors

1. Require the supplier to deliver the raw materials strictly in accordance with the agreed quality, quantity, time and place.

2. Refuse to comply with unreasonable instructions of the supplier where the processor is of the view that reimburse that person for reasonable expenses incurred in performing such act, even where the performance has failed to achieve the results desired by the beneficiary.

2. If a person has performed an act properly for the benefit of a beneficiary, the beneficiary must remunerate the person having performed the act, unless the person having performed the act refuses to accept the remuneration.

Article 577. Obligation to compensate for damage

1. If a person performing an act without authorization intentionally causes damage to the beneficiary while performing the act, such person must compensate for such damage.

2. If a person performing an act without authorization unintentionally causes damage to the beneficiary while performing the act, the compensation by such person may be reduced on the basis of the circumstances in which the act was performed.

Article 578. Termination of performance of acts without authorization

The performance of acts without authorization shall terminate in the following cases:

1. The beneficiary so requests;

2. The beneficiary, or its heir or repsentative, takes over the acts;

3. The person performing the acts without authorization becomes not able to continue performance in accordance with Clause 5 Article 575 of this Code;

4. The person performing the acts without authorization dies with regard to natural person or juridical person ceases to exist with regard to juridical person.

Chapter XIX

OBLIGATIONS TO RETURN PROPERTY DUE TO UNLAWFUL POSSESSION OR USE OF PROPERTY OR DERIVING OF BENEFITS FROM PROPERTY

Article 579. Obligation to return property

1. A person possessing or using property of another unlawfully must return the property to its owners and holders of other property-related rights. If the lawful owners and holders of other property-related rights of such property are not able to be found, the property must be delivered to a competent authority, except in the case provided in article 236 of this Code.

2. A person deriving benefits from property unlawfully, thereby causing damage to another person, must give such benefits to such other person, except in the case provided in Article 236 of this Code.

Article 580. Property to be returned

1. A person possessing or using property unlawfully must return the whole of such property.

2. Where the property to be returned is a distinctive object, that particular object must be returned and, if such distinctive object is lost or damaged, monetary compensation must be paid, unless otherwise agreed.

3. If the property to be returned is a fungible object which has been lost or damaged, an object of the same type must be returned or monetary compensation must be paid, unless otherwise agreed.

4. A person deriving benefits from property unlawfully must return, either in kind or in money, the benefits derived from the property to any person having suffered loss of such benefits.

Article 581. Obligation to return yield and income

1. A person possessing or using property, or a person deriving benefits from property, unlawfully and not in good faith, must return yield and income derived from the property during the time of unlawful possession or use of, or deriving benefits from, the property.

2. A person possessing or using property, or a person deriving benefits from property, unlawfully but in good faith, must return any yield and income derived from the property from the time when it knew or should have known that the possession or use of, or deriving benefits from, the property was unlawful, except in the case provided in Article 236 of this Code.

Article 582. Right to require third person to return property

Where a person unlawfully possessing or using property transfers the property to a third person, the third person must return the property if so demanded by the owner and holders of other property-related rights, unless this Code contains some other provision. If money or compensation has been paid for such property, the third person has the right to demand the party having transferred the property to the third person to compensate for damage.

Article 583. Obligation to pay

Upon taking back property, an owner or holders of other property-related rights or an aggrieved person must reimburse the person having taken possession of or used the property, or having derived benefits from the property, unlawfully but in good faith, for the necessary expenses such person has incurred for taking care of the property and increasing its value.

Chapter XX

LIABILITY FOR COMPENSATION FOR NON-CONTRACTUAL DAMAGES

Section 1. GENERAL PROVISIONS

Article 584. Grounds giving rise to liability to compensate for damage

1. A person intentionally or unintentionally harming the life, health, honor, dignity, reputation, property, or other legal rights or interests of a person, must compensate for such damage, unless otherwise pscribed in this Code or relevant laws.

2. The person who causes damage shall be discharged from liability for compensation in case where the damage incurs due to force majeure events or at entire fault of the aggrieved person, unless otherwise agreed or otherwise pscribed by law.

3. If a property causes damage, its owner or possessor must compensate for the damage, except for the damage pscribed in Clause 2 of this Article.

Article 585. Principles of compensation for damage

1. Actual damage must be compensated in full and promptly. Unless otherwise provided by law, parties may agree on the amount of compensation; on the form of compensation, which may be money, in kind or the performance of an act; lump sum payment or payment in instalments; and on the method of compensation.

2. The compensation payable by a person having caused damage may be reduced if such damage was caused unintentionally and is very large in comparison to the financial positions of such person.

3. If the amount of compensation determined becomes unrealistic, the aggrieved person, or the person having caused damage, has the right to request a court or another competent authority to change the amount of compensation.

4. If the aggrieved party is partly his/her fault for causing the damage, that part of damage shall not be compensated.

5. The party having rights and interests infringed shall not be compensated if such damage incurs due to his/her failure to adopt necessary measures to pvent the damage.

Article 586. Capacity of inpiduals for liability to compensate for damage

1. A person of eighteen years of age or older who causes damage shall be personally liable to compensate.

2. Where a minor under fifteen years of age causes damage, his or her parents, if any, must compensate for the total damage. If the parents have insufficient property to compensate and the minor who has caused the damage has property of his or her own, such property shall be used to satisfy the outstanding amount of compensation, except in the cases provided in Article 599 of this Code.

Where a person who is between fifteen and eighteen years of age causes damage, such person must compensate by recourse to his or her own property. If such person has insufficient property to compensate, the parents of such person must satisfy the outstanding amount by recourse to their own property.

3. Where a minor, legally incapacitated person, person with limited cognition and behavior control, causes damage but there is a guardian, such guardian shall use the property of the ward to compensate. If the ward has no or insufficient property to compensate, the guardian must do so by recourse to the property of the guardian. If the guardian is able to prove that he or she was not at fault with respect to guardianship, the guardian shall not be required to use its property to compensate.

Article 587. Compensation for damage caused jointly by several persons

Where several persons jointly cause damage, they must jointly compensate any aggrieved person. Liability for compensation of each person having jointly caused the damage shall be determined in proportion to the degree of fault of each person. If the degree of fault is not able to be determined, the persons causing damage must compensate in equal shares.

Article 588. Limitation period for initiating legal action claiming compensation for damage

The limitation period for initiating legal action claiming compensation for damage shall be two years from the date on which the legal rights or interests of an inpidual, legal entity or other subject were infringed.

Section 2. ASSESSMENT OF DAMAGE

Article 589. Damage caused by infringement of property

In the event of an infringement of property, the compensable damage shall comprise:

1. Property which was lost, destroyed or damaged;

2. Interests associated with the use and exploitation of the property was lost or declined;

3. Reasonable costs for the pvention, mitigation and remedy of the damage;

4. Other damage as pscribed by law.

Article 590. Damage caused by harm to health

1. Damage caused by harm to health shall comprise:

a) Reasonable costs for treating, nursing and rehabilitating health, and functional losses and impairment of the aggrieved person;

b) Loss of or reduction in the actual income of the aggrieved person. If the actual income of the aggrieved person is irregular and is not able to be determined, the average income level for the type of work performed by the aggrieved person shall be applied;

c) Reasonable costs and actual income losses of the carers of the aggrieved person during the period of treatment. If the aggrieved person loses his or her ability to work and requires a permanent carer, the damage shall also include reasonable costs for taking care of the aggrieved person.

d) Other damage as pscribed by law.

2. A person causing harm to the health of another person must pay the items provided in Clause 1 of this Article together with an amount of money as compensation for mental suffering of the aggrieved person. The amount of compensation for mental suffering shall be as agreed by the parties; if the parties are not able to agree, the maximum sum shall not exceed fifty-month base salary pscribed by the State.

Article 591. Damage caused by harm to life

1. Damage caused by harm to life shall comprise:

a) Damage caused by harm to life pscribed in Article 590 of this Code;

b) Reasonable funeral costs;

c) Support for the dependants of the aggrieved person;

d) Other damage as pscribed by law.

2. A person causing death to another person must pay compensation for damage as provided in Clause 1 of this Article together with an amount of money as compensation for mental suffering of the closest relatives in the first line of succession to the deceased. If there are no such relatives, this sum shall be paid to the persons who were directly reared by the deceased or to the persons who directly reared the deceased. The amount of compensation for mental suffering shall be as agreed by the parties; if the parties are not able to agree, the maximum sum shall not exceed one-hundred-month base salary pscribed in by the State.

Article 592. Damage caused by harm to honor, dignity or reputation

1. Damage caused by harm to the honor, dignity or reputation shall comprise:

a) Reasonable costs for mitigating and remedying the damage;

b) Loss of or reduction in actual income;

c) Other damage as pscribed by law.

2. A person causing harm to the honor, dignity or reputation of another person must pay compensation for damage as provided in Clause 1 of this Article together with another amount of money as compensation for mental suffering of the aggrieved person. The amount of compensation for mental suffering shall be as agreed by the parties; if the parties are not able to agree, the maximum sum shall not exceed ten-month base salary pscribed by the State.

Article 593. Period of entitlement to compensation for damage caused by harm to health or resulting from loss of life

1. Where an aggrieved person loses totally the ability to work, the aggrieved person shall receive compensation until the time of his or her death, unless otherwise agreed.

2. Where the aggrieved person dies, his or her dependants shall be entitled to receive support for the following durations:

a) A child of the deceased, whether living or conceived prior to his or her death, shall be entitled to compensation until the age of eighteen years, except a child between fifteen and eighteen years of age who is employed and earns sufficient income to look after himself or herself;

b) An adult who is not able to work shall be entitled to receive support until his or her death.

3. With regard to the conceived child of the deceased, the compensation shall be paid from the time he/she is born and alive.

Section 3. COMPENSATION FOR DAMAGE IN A NUMBER OF SPECIFIC CASES

Article 594. Compensation for damage by persons exceeding limits of reasonable self-defense

A person causing damage while acting in reasonable self-defense shall not be liable to compensate any aggrieved person.

A person causing damage while not acting in reasonable self-defense must compensate any aggrieved person.

Article 595. Compensation for damage by persons exceeding requirements of emergency situation

1. A person causing damage as a result of exceeding the requirements of an emergency situation must compensate any aggrieved person for that part of the damage which resulted from exceeding the requirements of an emergency situation.

Article 596. Compensation for damage caused by persons using stimulants

1. A person who, due to the consumption of alcohol or the use of other stimulants, becomes incapable of being aware of or controlling his or her acts, thereby causing damage to another person, must compensate such person.

2. A person who intentionally causes another person to take alcohol or stimulants, thereby causing such person to become incapable of being aware of or controlling his or her acts, must compensate any person aggrieved thereby.

Article 597. Compensation for damage caused by persons belonging to juridical persons

A juridical person must compensate for any damage caused by any person belonging to the juridical person during the performance of duties assigned by it to such person. If a juridical person has compensated for any damage, it has the right to demand the person at fault for causing the damage to reimburse it an amount of money in accordance with law.

Article 598. Compensation for damage caused by law enforcers

The State must compensate for damage caused by law enforcers as pscribed in the Law on compensation liability of the State.

Article 599. Compensation for damage caused by persons under fifteen years of age or persons having lost capacity for civil acts and under direct supervision of school, hospital or other organization

1. Where a person under fifteen years of age causes damage during school hours, the school must compensate for the damage.

2. If a legally incapacitated person causes damage to another person while under the direct supervision of a hospital or another juridical person, such hospital or the juridical person must compensate for the damage.

3. If, in the cases provided in Clauses 1 and 2 of this Article, the school, hospital or another juridical person proves that it was not at fault with respect to supervision, the parents or guardian of the person under fifteen years of age or of the legally incapacitated person must compensate.

Article 600. Compensation for damage caused workers and trainees

A natural person or juridical person must compensate for any damage caused by any worker or trainee belonging to it during the performance by the employee or trainee of his or her assigned duties. The natural person or juridical person has the right to demand such worker or trainee reimburse it an amount of money in accordance with law.

Article 601. Compensation for damage caused by sources of extreme danger

1. Sources of extreme danger comprise motorized means of transport, power transmission systems, operating industrial plants, weapons, explosives, inflammable substances, toxic substances, radioactive substances, dangerous animals and other sources of extreme danger as provided by law.

An owner of a source of extreme danger must comply strictly with the regulations on taking care of, pserving, transporting and using sources of extreme danger in accordance with law.

2. An owner of a source of extreme danger must compensate for damage caused by such source. If the owner has transferred possession or use of the source of extreme danger to another person, such other person must compensate unlawfully must compensate for damage.

Where an owner, or a person to which an owner has transferred possession or use, of a source of extreme danger is at fault by allowing the unlawful possession or use of the source of extreme danger, the owner, or the person to which the owner has transferred possession or use, of the source of extreme danger as the case may be must compensate jointly for the damage.

Article 602. Compensation for damage caused by environmental pollution

Any entity polluting the environment, thereby causing damage, must compensate in accordance with the law, including when the entity polluting the environment was not at fault.

Article 603. Compensation for damage caused by livestock

1. An owner of livestock must compensate for damage caused to another person by such livestock. The possessor or user of livestock must compensate during the period of possession or using, unless otherwise agreed.

2. Where a third person is entirely at fault in causing livestock to cause damage to another person, the third person must compensate for the damage. If both the third person and the owner are at fault, both of them must compensate jointly for the damage.

3. Where livestock which is possessed or used unlawfully causes damage, the unlawful possessor or user must compensate for the damage. When the owner, possessor or user of livestock is at fault leading the livestock is possessed or used unlawfully thereby causes damage, they must jointly compensate for damage.

4. Where livestock which is allowed to roam according to customary practice causes damage, its owner must compensate according to customary practice provided that such compensation does not contravene the law or social morals.

Article 604. Compensation for damage caused by trees

An owner, a possessor or a person in charge of trees must compensate for damage caused by the trees.

Article 605. Compensation for damage caused by houses and other construction works or buildings

An owner or a possessor of a house or another construction work, or a person to which the owner has assigned the management or use thereof, must compensate for damage if such house or construction causes damage to another person.

If the executor of the house or construction work is partly fault that such house or construction work causes damage, he/she must jointly compensate for such damage.

Article 606. Compensation for damage caused by infringement of corpses

1. Each natural person or juridical causing damage to a corpse must compensate.

2. Damage caused by infringement of a corpse shall include reasonable costs for mitigating and remedying the damage.

3. A person causing damage to a corpse must pay an amount of money as provided in Clause 2 of this Article together with another amount of money as compensation for mental suffering of the closest relatives in the first line of succession to the deceased. If there are no such relatives, this sum shall be paid to the persons who directly reared the deceased. The amount of compensation for mental suffering shall be as agreed by the parties; if the parties are not able to agree, the maximum sum shall not exceed thirty-month base salary pscribed by the State.

Article 607. Compensation for damage caused by infringement of graves

1. Each natural person or juridical person causing damage to the grave of another must compensate.

2. Damage caused by infringement of a grave shall include reasonable costs for mitigating and remedying the damage.

3. A person causing damage to a grave must pay an amount of money as provided in Clause 2 of this Article together with another amount of money as compensation for mental suffering of the closest relatives in the first line of succession to the deceased. If there are no such relatives, this sum shall be paid to the persons who directly reared the deceased. The amount of compensation for mental suffering shall be as agreed by the parties; if the parties are not able to agree, the maximum sum for each damaged grave shall not exceed ten-month base salary pscribed by the State.

Article 608. Compensation for damage caused by infringement of consumer interests

A natural person or juridical person carrying out production or business and failing to ensure the quality of goods, thereby causing damage to consumers, must compensate for such damage.

PART FOUR

INHERITANCE

Chapter XXI

GENERAL PROVISIONS

Article 609. Rights of inheritance

A natural person may make a will to dispose of his or her estate, may leave his or her property to an heir in accordance with law, or may inherit an estate left to him or her under a will or in accordance with law.

An heir not being natural person has the right to inherit estate under a will.

Article 610. Equality of inpiduals with respect to rights of inheritance

All natural persons are equal with respect to rights to bequeath their property to others and to inherit estates under wills or in accordance with law.

Article 611. Time and place of commencing inheritance

1. The time of commencement of an inheritance shall be the time when the deceased dies. Where a court declares that a person is dead, the time of commencement of the inheritance shall be the date provided in Clause 2 of Article 81 of this Code.

2. The place of commencement of the inheritance shall be the last place of residence of the owner of the estate. If the last place of residence is not able to be determined, the place of commencement of the inheritance shall be the place at which all or most of the estate is located.

Article 612. Estates

An estate comprises property which the deceased owned and property which the deceased jointly owned with other persons.

Article 613. Heirs

If an heir is an inpidual, such person must be alive at the time of commencement of the inheritance or, if such person is born and alive after the commencement of inheritance, must have been conceived prior to the time when the deceased dies. Where an heir under a will is a body or organization, it must be in existence at the time of commencement of the inheritance.

Article 614. Time when rights and obligations of heirs arise

From the time of commencement of an inheritance, the heirs have the property rights and obligations left by the deceased.

Article 615. Performance of property obligations left by deceased

1. A person entitled to an inheritance has the responsibility to perform the property obligations within the scope of the estate left by the deceased, unless otherwise agreed.

2. Where an estate has not yet been pided, the property obligations left by the deceased shall be performed by the administrator of the estate as agreed by the heirs.

3. Where an estate has already been pided, each heir shall perform those property obligations left by the deceased corresponding to, but not exceeding, that part of the estate that the heir has inherited, unless otherwise agreed.

4. Where the heir inheriting an estate under a will is not a natural person, it must perform the property obligations left by the deceased in like manner as a natural person.

Article 616. Administrators of estates

1. Administrator of an estate means the person who is appointed in the will or by agreement of the heirs.

2. Where a will fails to appoint, and the heirs have not yet appointed, an administrator, any person currently possessing, using or managing property within the estate at the time of the commencement of the inheritance shall continue its administration until the heirs have appointed an administrator.

3. Where an heir has not yet been determined and there is not yet an administrator of the estate as pscribed in Clauses 1 and 2 of this Article, the estate shall be administered by a competent authority.

Article 617. Obligations of administrators of estates

1. An administrator of an estate as provided in Clauses 1 and 3 of article 616 of this Code has the following obligations:

a) Make a list of the property within the estate and collect any property belonging to the estate of the deceased which is possessed by others, unless otherwise provided by law;

b) Take care of the estate and do not sell, exchange, give, pledge, mortgage or otherwise dispose of property within the estate without the written consent of the heirs;

c) Notify the heirs of the estate;

d) Compensate for any damage if the administrator breaches any of its obligations, thereby causing damage;

dd) Deliver back the estate at the request of the heirs.

2. A person possessing, using or managing property within an estate as provided in Clause 2 of Article 638 of this Code has the following obligations:

a) Take care of the estate and do not sell, exchange, give, pledge, mortgage or otherwise dispose of property within the estate;

b) Notify the heirs of the estate;

c) Compensate for any damage if the administrator breaches any of its obligations, thereby causing damage;

d) Deliver back the estate as agreed with the deceased in a contract or at the request of the heirs.

Article 618. Rights of administrators of estates

1. An administrator of an estate as provided in clauses 1 and 3 of article 616 of this Code has the following rights:

a) Repsent the heirs in dealings with any third parties in relation to the estate of inheritance;

b) Receive remuneration as agreed with the heirs;

c) Receive payment of costs of estate pservation.

2. A person possessing, using or managing property within an estate as provided in Clause 2 of Article 616 of this Code has the following rights:

a) Continue to use the estate as agreed with the deceased in a contract or with the consent of the heirs;

b) Receive remuneration as agreed with the heirs;

c) Receive payment of costs of estate pservation.

3. If the estate administrator fails to reach an agreement on the remuneration with the heirs, he/she shall be entitled to receive an appropriate remuneration.

Article 619. Inheritance by persons entitled to inherit each other’s estate but dead at same time

Where persons who are entitled to inherit each other’s estate die at the same time or are deemed to have died at the same time because it is impossible to determine who of them died first (hereinafter referred to as simultaneous death), they do not have the right to inherit each other’s estate and the estate of each of the deceased shall be inherited by their respective heirs, except in the case of inheritance pursuant to Article 652 of this Code.

Article 620. Disclaimer of inheritance

1. An heir may disclaim an inheritance, unless such disclaimer is for the purpose of avoiding the performance of its property obligations to other persons.

2. A disclaimer of an inheritance must be made in writing. A person disclaiming must notify the other heirs and the person authorized to distribute the estate.

3. The disclaimer of an estate must be expssed before the time of inherit distribution.

Article 621. Persons not entitled to inherit

1. The following persons are not entitled to inherit:

a) Persons convicted of having intentionally caused the death of or harmed the health of the deceased, of having seriously mistreated or tortured the deceased, or of having harmed the honor or dignity of the deceased;

b) Persons having seriously breached their duty to support the deceased;

c) Persons convicted of having intentionally caused the death of another heir in order to obtain all or part of the entitlement of such other heir to the estate;

d) Persons deceiving, coercing or obstructing the deceased with respect to the making of the will, or forging, altering or destroying the will in order to obtain all or part of the estate contrary to the wishes of the deceased.

2. Persons provided in Clause 1 of this Article may, nevertheless, inherit the estate if the deceased was aware of such acts but, nevertheless, allowed them to inherit the estate under the will.

Article 622. Estates which no one inherits

Where there is no heir under a will and at law, or where there is an heir but such heir is not entitled to inherit the estate or disclaims the inheritance of the estate, the residual estate for which there is no heir shall, after fulfillment of property obligations, belong to the State.

Article 623. Limitation periods with respect to inheritance

1. The limitation period with respect to a claim of an heir for distribution of an estate shall be thirty years regarding immovable property or ten years regarding movable property from the time of commencement of the inheritance. Upon the expiry date of the aforesaid period, the estate shall belong to the estate administrator. In case where there is no estate administrator, the estate shall be dealt with as follows:

a) It shall belong to the person possessing it as pscribed in Article 236 of this Code;

b) It shall belong the State if there is no possessor pscribed in Point a of this Clause.

2. The limitation period with respect to a claim of an heir for a declaration of right of inheritance of the requester or to disallow the claim to inheritance of another shall be ten years from the time of commencement of the inheritance.

3. The limitation period with respect to a claim for an heir to fulfill property obligations of the deceased shall be three years from the time of commencement of the inheritance.

Chapter XXII. NHERITANCE UNDER WILLS

Article 624. Wills

Will means an expssion of the wishes of a natural person, made in order to bequeath his or her property to others after his or her death.

Article 625. Testators

1. An adult satisfying conditions pscribed in Point a Clause 2 Article 630 of this Code may make a will to dispose his/her property.

2. A person who is between fifteen and eighteen years of age may make a will with the consent of his or her parents or guardian.

Article 626. Rights of testators

A testator has the following rights:

1. Appoint heirs or to deprive an heir of the right to inherit the estate;

2. Determine those parts of the estate which each heir is entitled to;

3. Reserve part of the estate as a gift or for worship purposes;

4. Designate heirs to perform obligations;

5. Appoint a custodian of the will, an administrator of the estate, and a distributor of the estate.

Article 627. Formalities for wills

A will must be made in writing. If it is not able to be made in writing, it may be made orally.

Article 628. Written wills

Written wills comprise:

1. Unwitnessed written wills;

2. Witnessed written wills;

3. Written wills which are notarized;

4. Written wills which are certified.

Article 629. Oral wills

1. Where a person is likely to die due to illness or any other reason and it is not possible for him or her to make a written will, such person may make an oral will.

2. If the testator is alive and is of sound mind three months after he or she has made an oral will, such will shall automatically become invalid.

Article 630. Lawful wills

1. A will must satisfy the following requirements in order to be lawful:

a) The testator was of sound mind when he or she made the will; and he or she was not deceived, threatened or coerced into making the will;

b) The contents of the will are not contrary to law or social morals and the will complies with legal formalities.

2. A will made by a person between fifteen and eighteen years of age must be made in writing and with the consent of the parents or guardian of such person.

3. A will made by a person who is incapacitated or illiterate must be made in writing by a witness and must be notarized or certified.

4. A written will which is not notarized or certified shall be deemed lawful only if it satisfies the requirements provided in Clause 1 of this Article.

5. An oral will shall be deemed lawful only if the testator orally expssed his or her last wishes before at least two witnesses who immediately thereafter recorded those wishes in writing and signed or fingerprinted the document. Such will must be notarized or certified within five working days of the date on which the testator orally expssed his or her last wishes.

Article 631. Contents of written wills

1. A will must specify clearly the following:

a) The date on which the will is made;

b) The full name and place of residence of the testator;

c) The full names of the persons and the bodies or organizations entitled to inherit the estate;

d) The estate to be bequeathed and its location.

2. Apart from the contents pscribed in Clause 1 of this Article, the will may have other contents.

3. A will may not be written using abbreviations or other symbols. If a will consists of several pages, each page must be numbered and bear the signature or fingerprint of the testator.

Where a will has erasure or correction, the testator or the testament witness must sign beside erasing and corrected place.

Article 632. Witnesses to making of will

Any person may act as a witness to the making of a will, except the following persons:

1. Persons who are heirs of the testator under the will or at law;

2. Persons with property rights or obligations which relate to the will;

3. Minors, legally incapacitated persons, persons with limited cognition and behavior control.

Article 633. Unwitnessed written wills

A testator must write a will by his or her own hand and must sign it.

The drawing up of a written will without witnesses must comply with article 631 of this Code.

Article 634. Witnessed written wills

Where a testator is not able to write a will by his or her own hand, the testator may request another person to write the will, but there must be at least two witnesses. The testator must sign or fingerprint the will in the psence of the witnesses; the witnesses shall acknowledge the signature or fingerprint of the testator and sign the will.

The will must be made in compliance with articles 631 and 632 of this Code.

Article 635. Wills which are notarized or certified

A testator may request that the will be notarized or certified.

Article 636. Procedures for pparation of wills at notary office or people’s committee of commune

The pparation of a will at a notary office or the people’s committee of the commune must comply with the following procedures:

1. The testator shall declare the contents of his or her will to a notary public officer or a member of the people’s committee of the commune who has the authority to certify it. The notary public officer or the person having the authority to certify must record the wishes stated by the testator. The testator shall sign or fingerprint the will after acknowledging that it has been recorded accurately and that it expsses faithfully the intentions of the testator. Thereafter, the notary public officer or the member of the people’s committee of the commune shall sign the will;

2. Where the testator is not able to read or hear the will or not able to sign or fingerprint it, there must be a witness who must acknowledge the will by signing it before a notary public officer or a member of the people’s committee of the commune who has the authority to certify it. The notary public officer shall notarize the will, or the member of the people’s committee of the commune who has the authority to certify the will shall certify it, in the psence of the testator and the witnesses.

Article 637. Persons not permitted to notarize or certify wills

A notary public officer or a member of the people’s committee of the commune who has authority shall not be permitted to notarize or certify a will if such person is:

1. An heir of the testator under the will or at law;

2. A person whose father, mother, wife, husband or child is an heir under the will or at law;

3. A person having property rights or obligations relating to the will.

Article 638. Written wills valid as though notarized or certified

1. A written will made by a serving soldier who is not able to request a notarization or certification of his or her will provided that such will is certified by the head of his or her unit having the rank of a company commander or higher.

2. A written will made by a person travelling on a seagoing vessel or aircraft provided that the will is certified by the captain of the vessel or aircraft.

3. A written will made by a person undergoing medical treatment in a hospital or other medical establishment or sanatorium provided that the will is certified by the person in charge of such hospital or establishment or sanatorium.

4. A written will made by a person conducting surveys, explorations or research in mountainous areas, forests or offshore islands provided that the will is certified by the person in charge of the unit.

5. A written will made by a Vietnamese citizen residing abroad provided that the will is certified by a Vietnamese consular or diplomatic repsentative mission in that country.

6. A written will made by a person held in temporary detention, serving a prison sentence or administrative penalty, or at an educational or medical facility provided that the will is certified by the person in charge of such facility.

Article 639. Wills ppared by notary public officers at places of residence of testators

1. A testator may request a notary public officer to visit his or her place of residence in order to ppare a will.

2. Such will shall be ppared in accordance with the procedures for the pparation of wills at a State notary public provided in article 636 of this Code.

Article 640. Amendment of, addition to, replacement or revocation of wills

1. A testator may amend, add to, replace or revoke his or her will at any time.

2. If a testator adds to his or her will, the original will and the codicil shall have equal validity. If a part of the original will and the codicil conflict with each other, the codicil shall pvail.

3. Where a testator replaces a will with a new will, the pvious will shall be deemed to have been revoked.

Article 641. Custody of wills

1. A testator may request a notary office or another person to keep custody of the will of the testator.

2. Where a will is kept in custody by a notary office, it must be taken care of and looked after in accordance with the law on notaries.

3. An inpidual keeping custody of a will has the following obligations:

a) Keep the contents of the will confidential;

b) Take care of and look after the will. If the will is lost or damaged, the person must notify immediately the testator;

c) Upon the death of the testator, to deliver the will to his or her heirs or to the person authorized to announce the will. The delivery of the will must be recorded in writing and signed by the person delivering the will, and by the person receiving it, in the psence of two witnesses.

Article 642. Loss and damage of wills

1. If, from the commencement of the inheritance, the will is lost or damaged to the extent that it is incapable of indicating clearly the wishes of the testator and there is no evidence of the true wishes of the testator, it shall be deemed that no will exists and inheritance at law shall apply.

2. Where the will is found prior to distribution of the estate, the estate shall be distributed according to the will.

3. Within the limitation periods for requesting estate distribution, if a will is found after the distribution of the estate, the estate shall be distributed according to the will at the request of the heir under will.

Article 643. Legal effectiveness of wills

1. A will shall become legally effective from the time of commencement of the inheritance.

2. All or part of a will shall be legally ineffective in any of the following cases:

a) An heir under the will dies prior to or at the same time as the testator dying;

b) A body or organization named as an heir no longer exists at the time of commencement of the inheritance.

Where there are several heirs under a will and one of them dies prior to or at the same time as the death of the testator or one of the bodies or organizations named as an heir under the will no longer exists at the time of commencement of the inheritance, only that part of the will which relates to the inpidual, body or organization no longer existing shall be legally ineffective.

3. A will shall not be legally effective if the estate left to the heirs no longer exists at the time of commencement of the inheritance. If only part of the estate left to the heirs remains, only that part of the will which relates to such part of the estate shall be legally effective.

4. Where a will contains provisions which are unlawful but such provisions do not affect the effectiveness of the remainder of the will, only such provisions shall be legally ineffective.

5. Where a person leaves behind more than one will with respect to certain property, only the most recent of such wills shall be legally effective.

Article 644. Heirs notwithstanding contents of wills

1. Where a testator does not grant any of the following persons an inheritance, or grants any such person an inheritance which is less than two-thirds of the share that person would have received if the estate had been distributed according to law, such person shall be entitled to a share of the estate equivalent to two-thirds of the share that he or she would have received if the estate had been distributed in accordance with law:

a) Children who are minors, father, mother, wife or husband of the testator;

b) Children who are adults but who are incapable of working.

2. Clause 1 of this Article shall not apply to persons who have disclaimed their inheritance as pscribed in Article 620 or person who are not entitled to inherit as pscribed in Clause 1 Article 621 of this Code.

Article 645. Estates used for worship purposes

1. Where a testator designates part of his or her estate for worship purposes, such part of the estate shall not be distributed among the heirs and shall be delivered to the person appointed in the will to manage for worship purposes. If such appointee fails to implement strictly the will or the agreement of the heirs, the heirs have the right to appoint another person to manage for worship purposes.

Where the testator fails to appoint a person to manage that part of his or her estate which is designated for worship purposes, the heirs shall appoint a person to manage such part of the estate.

Where all heirs under a will have died, that part of the estate which is designated for worship purposes shall belong to the person managing that part of the estate for worship purposes provided that he or she is an heir at law.

2. Where the entire estate of the deceased is insufficient to satisfy all property obligations of the deceased, no part of the estate may be designated for worship purposes.

Article 646. Testamentary gifts

1. A testator may designate part of his or her estate as a testamentary gift to another person. The testamentary gift must be expssly stated in the will.

2. The grantee of the testamentary gift must be alive at the time of commencement of the inheritance or he/she must bear and alive after the time of commencement of the inheritance he/she must be conceived before the death of the estate leaver. If the grantee of the testamentary gift is not a natural person, it must exist at the time of commencement of the inheritance.

3. The grantee of a gift shall not be required to fulfill property obligations with respect to that part of the estate granted as a gift, unless the whole estate is insufficient to satisfy all property obligations of the grantor, in which case the part of the estate granted as a gift shall also be applied towards satisfying the remainder of the obligations of the grantor.

Article 647. Announcement of wills

1. Where a written will is kept by a notary office, the notary officer shall be the person announcing the will.

2. Where a testator has appointed a person to announce the will, such person shall announce the will. If the testator fails to appoint a person or has appointed a person but the appointee refuses to announce the will, the heirs shall agree on the appointment of a person to announce the will.

3. After the time of commencement of an inheritance, the person announcing the will must send copies of the will to all persons with an interest in the contents of the will.

4. A recipient of a copy of a will has the right to verify the copy against the original.

5. Where a will has been ppared in a foreign language, it must be translated into Vietnamese and notarized.

Article 648. Interptation of contents of wills

Where the contents of a will are unclear and may be interpted in different ways, the person announcing the will and the heirs must interpt jointly the contents of the will based on the true wishes of the deceased, taking into consideration the relationship of the deceased with the heirs under the will. If such persons fail to agree on the interptation of the contents of the will, they have the right to request a court for settlement.

Where part of the contents of a will is not able to be interpted but the remainder of the will is not affected, only that part which is not able to be interpted shall not be legally effective.

Chapter XXIII. INHERITANCE AT LAW

Article 649. Inheritance at law

Inheritance at law means inheritance in accordance with the order of priority of inheritance and the conditions and procedures of inheritance provided by law.

Article 650. Cases of inheritance at law

1. Inheritance at law shall apply in the following cases:

a) There is no will;

b) The will is unlawful;

c) All heirs under the will died prior to or at the same time as the testator dying, or the bodies or organizations which are entitled to inherit under the will no longer exist at the time of commencement of the inheritance;

d) The persons appointed as heirs under the will do not have the right to inherit or disclaimed the right to inherit.

2. Inheritance at law shall also apply to the following parts of an estate:

a) Parts of an estate in respect of which no disposition has been made in the will;

b) Parts of an estate related to an ineffective part of the will;

c) Parts of an estate related to heirs under the will not having the right to inherit, having disclaimed the right to inherit, or having died prior to or at the same time as the testator dying; and parts of an estate related to bodies or organizations entitled to inherit under the will but no longer existing at the time of commencement of the inheritance.

Article 651. Heirs at law

1. Heirs at law are categorized in the following order of priority:

a) The first level of heirs comprises: spouses, biological parents, adoptive parents, offspring and adopted children of the deceased;

b) The second level of heirs comprises: grandparents and siblings of the deceased; and biological grandchildren of the deceased;

c) The third level of heirs comprises: biological great-grandparents of the deceased, biological uncles and aunts of the deceased and biological nephews and nieces of the deceased.

2. Heirs at the same level shall be entitled to equal shares of the estate.

3. Heirs at a lower level shall be entitled to inherit where there are no heirs at a higher level because such heirs have died, or because they are not entitled to inherit, have been deprived of the right to inherit or have disclaimed the right to inherit.

Article 652. Succeeding heirs

Where a child of a testator died prior to or at the same time as the testator, the grandchildren of the testator shall inherit that part of the estate which their father or mother would have been entitled to inherit had such father or mother still been alive. If the grandchildren also died prior to or at the same time as the testator, the great-grandchildren of the testator shall inherit that part of the estate which their father or mother would have been entitled to inherit had such father or mother still been alive.

Article 653. Inheritance relations between adopted children and their adoptive parents and biological parents

An adopted child and his or her adoptive parents may inherit each other’s estates and may also inherit in accordance with articles 651 and 652 of this Code.

Article 654. Inheritance relations between stepchildren and their stepparents

If a stepchild and his or her stepparents care for and support each other as though they were biologically related, they may inherit each other’s estates and may also inherit in accordance with articles 652 and 653 of this Code.

Article 655. Inheritance where wives and husbands have pided multiple ownership property, have applied for porce or have remarried

1. Where a wife and husband have pided their multiple ownership property while they are still married and one of them subsequently dies, the surviving spouse shall still be entitled to inherit the estate of the deceased.

2. Where a wife and husband have applied for but not yet obtained a legally effective porce pursuant to a judgment or decision of a court, or they have obtained such a porce but the judgment or decision of the court is not yet effective, and one of them dies, the surviving spouse shall, nevertheless, be entitled to inherit the estate of the deceased.

3. A person who is the wife or husband of the deceased at the time when his or her spouse dies shall be entitled to inherit the estate of the deceased even if that person subsequently remarries.

Chapter XXIV. SETTLEMENT AND DISTRIBUTION OF ESTATES

Article 656. Meeting of heirs

1. After being notified of the commencement of an inheritance, or after a will has been announced, the heirs may meet to agree on the following matters:

a) If the testator has failed to appoint an administrator of the estate or a distributor of the estate, or has not determined the powers and obligations of such persons, the appointment of such persons and the determination of their powers and obligations, as the case may be;

b) Method of distributing the estate.

2. All agreements by the heirs must be made in writing.

Article 657. Distributors of estates

1. A distributor of an estate may also be the administrator of the estate appointed in the will or by agreement of the heirs.

2. A distributor of an estate must distribute it strictly in accordance with the will or the agreement of the heirs at law.

3. A distributor of the estate may receive remuneration if so allowed by the testator in the will or if so agreed by the heirs.

Article 658. Order of priority of payment

Property obligations and expenses related to an inheritance shall be paid in the following order of priority:

1. Reasonable funeral expenses in accordance with customary practice;

2. Outstanding support payments;

3. Expenditures on pservation of estate;

4. Allowances for dependants of the deceased;

5. Wages;

6. Monetary compensation for any damage;

7. Taxes and other liabilities owed to the State;

8. Other liabilities owed to other natural persons or juridical persons;

9. Fines;

10. Other expenses.

Article 659. Distribution of estates in accordance with wills

1. An estate shall be distributed in accordance with the wishes of the testator. If the will fails to specify the share of each heir, the estate shall be pided equally between the persons named in the will, unless otherwise agreed.

2. Where a will provides for the distribution in kind of an estate, each of the heirs shall be entitled to receive his or her share in kind, plus the benefits and income derived therefrom, or must bear the depciation in value of such share in kind up to the time when the estate is distributed. If the property which is the subject of a share in kind has been destroyed due to the fault of another person, the heir has the right to demand compensation for damage.

3. Where a will provides for the distribution of an estate according to certain proportions of the total value of the estate, such proportions shall be calculated on the basis of the value of the estate at the time of distribution.

Article 660. Distribution of estates in accordance with law

1. If, at the time of distribution, an heir has been conceived but not yet born, a part of the estate equal to the share of another heir at the same level of heirs shall be set aside for the unborn heir. If the heir is born alive, he or she shall inherit such part of the estate. If the heir does not survive his or her birth, the other heirs at the same level of heirs shall be entitled to his or her share.

2. The heirs have the right to demand the estate to be distributed in kind. If the estate is not able to be equally distributed in kind, the heirs may agree that the property shall be valued and may agree on which heirs shall be entitled to receive which particular items of property. Failing such agreement, the assets in kind shall be sold for distribution.

Article 661. Limited distribution of estates

Where it was the wish of a testator, or where the heirs agree, that an estate is to be distributed only after a certain period of time, it shall be distributed only after such period of time has expired.

Article 662. Distribution of estates where new heir or where right of heir to inherit has been disallowed

1. Where a new heir appears after an estate has been distributed, the estate shall not be re-distributed in kind but the heirs which have received at the time of distribution of the estate in proportion to the , unless otherwise agreed.

2. Where the right of an heir to inherit is disallowed after an estate has been distributed, such heir must return the inheritance or pay to the other heirs a sum equivalent to the value of the inheritance received at the time of distribution of the estate, unless otherwise agreed.

PART FIVE. CIVIL RELATIONS INVOLVING FOREIGN ELEMENTS

Chapter XXV. GENERAL PROVISIONS

Article 663. Scope

1. This Part provides for applied law to civil relations involving foreign elements.

If any regulation of law providing for applied law to civil relations involving foreign elements complies with Article 664 through Article 671 of this Code, it shall pvail; if it does not comply with those Articles, Part Five of this Code shall pvail.

2. Civil relation involving a foreign element means any of the following civil relations:

a) There is at least one of the participating parties is a foreign natural person or juridical person;

b) The participating parties are Vietnamese natural persons or juridical persons but the basis for the establishment, modification or termination of such relation arose in a foreign country;

c) The participating parties are Vietnamese natural persons or juridical persons but the subject matter of such civil relation is located in a foreign country.

Article 664. Determination of law applying to civil relations involving foreign elements

1. The international agreements to which the Socialist Republic of Vietnam is a signatory or Vietnamese law shall apply to civil relations involving foreign elements.

2. In case the international agreements to which the Socialist Republic of Vietnam is a signatory or a Vietnamese law stipulates that contracting parties have the right to select applied law, the law applied to civil relations involving foreign elements shall be determined according to the selection of the contracting parties.

3. If it fails to determine the applied law as pscribed in Clause 1 and Clause 2 of this Article, the applied law is the law of the country that closely associates with the civil relations involving foreign elements.

Article 665. Application of international treaty in terms of civil relations involving foreign elements

1. In case an international treaty to which the Socialist Republic of Vietnam is a signatory contains regulations on rights and obligations of contracting participants in the civil relations involving foreign elements, such international treaty shall pvail.

2. In case an international treaty to which the Socialist Republic of Vietnam is a signatory contains provisions different from those in this Code and other laws in terms of applied law on civil relations involving foreign elements, such international treaty shall pvail.

Article 666. Application of international customary practices

Contracting parties may select international customary practices in the case pscribed in Clause 2 Article 664 of this Code. If the application of such international customary practices is contrary to the basic principles of the laws of the Socialist Republic of Vietnam, Vietnamese laws shall pvail.

Article 667. Application of foreign laws

Where foreign law applies to a civil relation with different interptations, the application must follow the interptation of the competent authority in that country.

Article 668. Scope of the law to be referred to

1. The law to be referred to include regulations on determination of applied law and regulations on rights and obligations of participants in civil relations, other than the case pscribed in Clause 4 of this Article.

2. Where the Vietnamese law is referred to, regulations of Vietnamese law on rights and obligations of participants of a civil relation shall apply.

3. Where law of a third country is referred to, regulations of that law on rights and obligations of participants of a civil relation shall apply.

4. With respect to regulations on Clause 2 Article 664 of this Code, the law selected by the contracting parties shall include rights and obligations of participants of the civil relation, and not include regulations on applied law.

Article 669. Application of law of countries having multiple legal systems

If the law of a country having multiple legal systems is referred to, the applied law shall be determined according to the rules pscribed by such country’s law.

Article 670. Non-application of foreign laws

1. The foreign law, notwithstanding being referred to, shall not apply in any of the following cases:

a) The consequences of its application are not inconsistent with the fundamental principles of the law of the Socialist Republic of Vietnam;

b) The contents of foreign law are not identifiable regardless of the adoption of necessary measures pscribed by procedural law.

2. In case the foreign law is not applied as pscribed in Clause 1 of this Article, Vietnamese law shall apply.

Article 671. Limitation periods

Limitation periods for civil relations involving foreign elements shall be determined according to the law applying to such civil relations.

Chapter XXVI. THE LAW APPLIED TO NATURAL PERSONS AND JURIDICAL PERSONS

Article 672. Bases for choice of law applicable to stateless persons and to foreigners with multiple foreign nationalities

1. Where the law referred to is the law of a country of which a foreigner is a citizen but he/she is a stateless person, the applied law shall be the law of the country of residence of such person at the time when the civil relations involving foreign elements were established. If that person has multiple residences or his residence is unidentifiable at the time of establishing the civil relation involving foreign elements, the applied law shall be the law of the country with which such person closely associates.

2. Where the law referred to is the law of a country of which a foreigner is a citizen but he/she has multiple foreign nationalities, the applied law shall be the law of which he/she holds nationality and where he/she resided at the time when the civil relations involving foreign elements were established. If that person has multiple residences or his residence is unidentifiable or his/her residence is different from the country of which he/she holds nationality at the time when the civil relations involving foreign elements were established, the applied law shall be the law of the country with which such person closely associates.

Where the law referred to is the law of a country of which a foreigner is a citizen but he/she has multiple foreign nationalities, including Vietnamese nationality, the applied law shall be Vietnamese law.

Article 673. Legal personality of natural persons

1. Legal personality of a natural person shall be determined according to the law of the country of which he/she holds nationality.

2. A foreigner in Vietnam shall have legal personality in the same manner as a Vietnamese citizen, unless otherwise provided by the law of Vietnam.

Article 674. Legal capacity of natural persons

1. Legal capacity of a natural person shall be determined according to the law of the country of which he/she holds nationality, other than the case pscribed in Clause 2 of this Article.

2. Where a foreigner establishes or performs civil transactions in Vietnam, his/her legal capacity shall be determined in accordance with the law of Vietnam.

3. The determination of a legally incapacitated person, a person with limited cognition or behavior control or a person with limited legal capacity in Vietnam shall be in accordance with the law of Vietnam.

Article 675. Determination of persons disappeared or died

1. A determination that a person has disappeared or died must comply with the law of the country of which such person held nationality at the point of time prior to the last information about such disappearance or death, except for the case pscribed in Clause 2 of this Article.

2. The determination of a disappeared or dead person in Vietnam shall be in accordance with the law of Vietnam.

Article 676. Juridical persons

1. The nationality of a juridical person shall be determined according to the law of the country in which such juridical person was established.

2. Legal personality, name, legal repsentatives, organization, restructuring, dissolution of a juridical person; relations between a juridical person and its members; responsibilities of a juridical person and its members pertaining to its obligations shall be determined in accordance with the law of the country of which such juridical person holds nationality, other than the case pscribed in Clause 3 of this Article.

3. Where a foreign juridical person establishes or performs civil transactions in Vietnam, its legal personality shall be determined in accordance with the law of Vietnam.

Chapter XXVII. THE LAW APPLIED TO PROPERTY RELATIONS AND PERSONAL RELATIONS

Article 677. Classification of property

The classification between moveable and immoveable property shall be made in accordance with the laws of the country in which such property is located.

Article 678. Ownership rights and other property-related rights

1. The establishment, exercise, operation and termination of ownership rights and other property-related rights shall be determined in accordance with the laws of the country in which the property is located, except in the cases provided in Clause 2 of this Article.

2. Ownership rights with respect to moveable property in transit shall be determined in accordance with the law of the country of destination, unless otherwise agreed.

Article 679. Intellectual property rights

The intellectual property rights shall be determined in accordance with the laws of the country in which the objects of the intellectual property rights are required to be protected.

Article 680. Inheritance

1. Inheritance must comply with the law of the country of which the person who bequeathed the assets held nationality prior to his or her death.

2. The right to inherit immoveable property must comply with the law of the country where such immoveable property is located.

Article 681. Wills

1. The capacity to create a will, and to alter or rescind a will, must comply with the law of the country of which the testator is a citizen.

2. The form of a will must comply with the law of the country in the place where the will is created. The form of a will shall be also recognized in Vietnam if it complies with the laws of any of the following countries:

a) The country in which the testator resides at the time when the will is created or the testator dies;

b) The country of which the testator holds nationality at the time when the will is created or the testator dies;

c) The country where the inheritance being immovable property is located.

Article 682. Guardianship

The guardianship shall be determined in accordance with the law of the country where the ward resides.

Article 683. Contracts

1. Contracting parties in a contract may agree to select the applied law for the contract, other than regulations of Clauses 4, 5 and 6 of this Article. In case the contracting parties fail to agree the applied law, the law of the country with which such contract closely associates shall apply.

2. The laws of any of the following countries shall be treated as the law of the country with which such contract closely associates:

a) The law of the country where the seller being natural person resides or the seller being juridical person is established in terms of sale contracts;

b) The law of the country where the provider being natural person resides or the provider being juridical person is established in terms of service contracts;

c) The law of the country where the transferee being natural person resides or the seller being juridical person is established in terms of contracts of transferring rights to use or intellectual property rights;

d) The law of the country where employees frequently perform do jobs in terms of labor contracts. If an employee frequently does jobs in multiple countries or the country in which the employee frequently does his/her job is unidentifiable, the law of the country with which his/her labor contract closely associates shall be the law of the country where the employer being natural person resides or the employer being juridical person is established.

dd) The law of the country where consumers resides in terms of consume contract.

3. If there is evident that the law of a country other than the country pscribed in Clause 2 of this Article associates with the contract more closely than the latter, the law of the former country shall pvail.

4. If the object of a contract is an immovable property, the law applied to transfer of its ownership rights and/or other property-related rights, lease of immovable property or using the immovable property as the guarantee for performance of obligations shall be the law of the country where the immovable property is located.

7. Form of a contract shall be determined in accordance with the law applied to such type of contract. In case where the form of a contract does not comply with the form of the law applied to such contract but it comply with the form of the law of the country where the contract is entered into or the law of Vietnam, such form of contract shall be recognized in Vietnam.

Article 684. Unilateral acts

The law applied to unilateral acts of a person shall be determined in accordance with the law of the country of residence with regard to natural person or the law of the country of establishment with regard to juridical person.

Article 685. Obligation to refund property possessed, used or derived unlawfully

The obligation to refund property possessed, used or derived unlawfully shall be determined according to the law of the country where the property is possessed, used or derived unlawfully.

Article 686. Performance of acts without authorization

Contracting parties may agree to select the law applied to the performance of acts without authorization. If the contracting parties fail to agree to select the applied law, the law of the country where the acts without authorization are performed shall pvail.

Article 687. Compensation for non-contractual damage

1. Contracting parties may agree to select the law applied to the compensation for non-contractual damage, except for the case pscribed in Clause 2 of this Article. If the contracting parties fail to agree to select the applied law, the law of the country where the consequences of such acts arise shall pvail.

2. Where the party causing damage and the aggrieved party being natural persons have residence or being juridical person place of establishment in the same country, the law of such country shall pvail.

PART SIX. IMPLEMENTATION

Article 688. Transitional regulations

1. With respect to civil transactions established before the effective date of this Code, the law shall be implemented as follows:

a) The parties of non-performed civil transactions whose contents and forms are different from this Code shall keep complying with regulations of the Civil Code No. 33/2005/QH11 and legislative documents on guidelines for the Civil Code No. 33/2005/QH11, unless the parties agree to amend the contents or forms of the transactions in accordance with this Code.

a) The parties of being-performed civil transactions whose contents and forms are different from this Code shall keep complying with regulations of the Civil Code No. 33/2005/QH11 and legislative documents on guidelines for the Civil Code No. 33/2005/QH11;

b) The parties of non-performed or being-performed civil transactions whose contents and forms are conformable to this Code shall comply with regulations of this Code;

c) The parties of civil transactions that are completely performed before the effective date of this Code but arise dispute shall keep complying with regulations of the Civil Code No. 33/2005/QH11 and legislative documents on guidelines for the Civil Code No. 33/2005/QH11;

d) Limitation periods shall comply with this Code.

2. This Code shall not apply to a case settled by a court in accordance with law on civil before the effective date of such Code for the purpose of appeal under cassation procedures.

Article 689. Effect

This Code comes into force from January 1, 2021.

The Civil Code No. 33/2005/QH11 shall be annulled from the effective date of this Code.

This Code is passed by the 13th National Assembly of Socialist Republic of Vietnam during the 10th session on November 24, 2021.


Bạn đang xem chủ đề Luật Chơi Cờ Cá Ngựa trên website Athena4me.com. Hy vọng những thông tin mà chúng tôi đã chia sẻ là hữu ích với bạn. Nếu nội dung hay, ý nghĩa bạn hãy chia sẻ với bạn bè của mình và luôn theo dõi, ủng hộ chúng tôi để cập nhật những thông tin mới nhất. Chúc bạn một ngày tốt lành!