Browse > Article

Sensory Difference Testing: The Problem of Overdispersion and the Use of Beta Binomial Statistical Analysis  

Lee, Hye-Seong (Department of Food Science and Technology, University of California)
O'Mahony, Michael (Department of Food Science and Technology, University of California)
Publication Information
Food Science and Biotechnology / v.15, no.4, 2006 , pp. 494-498 More about this Journal
Abstract
An increase in variance (overdispersion) can occur when a binomial statistical analysis is applied to sensory difference test data in which replicate sensory evaluations (tastings) and multiple evaluators (judges) are combined to increase the sample size. Such a practice can cause extensive Type I errors, leading to serious misinterpretations of the data, especially when traditional simple binomial analysis is applied. Alternatively, the use of beta binomial analysis will circumvent the problem of overdispersion. This brief review discusses the uses and computation methodology of beta binomial analysis and in practice evidence for the occurrence of overdispersion.
Keywords
sensory evaluation; sensory difference tests; binomial statistics; overdispersion; beta binomial test; gamma;
Citations & Related Records
Times Cited By KSCI : 1  (Citation Analysis)
Times Cited By Web Of Science : 2  (Related Records In Web of Science)
연도 인용수 순위
1 Brockhoff PB, Schlich P. Handling replications in discrimination tests. Food Qual. Prefer. 9: 303-312 (1998)   DOI   ScienceOn
2 Harries JM, Smith GL. The two-factor triangle test. J. Food Technol. 17: 153-162 (1982)   DOI
3 Bi J, Ennis DM. The power of sensory discrimination methods used in replicated difference and preference tests. J. Sens. Stud. 14: 289-302 (1999a)   DOI
4 O'Mahony M. Sensory Evaluation of Food. Statistical Methods and Procedures. Marcel Dekker, Inc., New York, NY, USA. pp.408-413 (1986)
5 Braun V, Rogeaux M, Schneid N, O'Mahony M, Rousseau B. Corroborating the 2-AFC and 2-AC Thurstonian models using both a model system and sparkling water. Food Qual. Prefer. 15: 501-507 (2004)   DOI   ScienceOn
6 Delwiche J, Ligget R. Accounting for between-subject variance in discrimination and preference tasks. J. Sens. Stud. 20: 48-61 (2005)   DOI   ScienceOn
7 Ennis DM, Bi J. The beta-binomial model: accounting for inter-trial variation in replicated difference and preference tests. J. Sens. Stud. 13: 389-412 (1998)   DOI   ScienceOn
8 Brockhoff PB. The statistical power of replications in difference tests. Food Qual. Prefer. 14: 405-417 (2003)   DOI   ScienceOn
9 Angulo O, Lee HS, O'Mahony M. Sensory difference tests: Over-dispersion and warm-up. Food Qual. Prefer. 17: in press (2006)
10 O'Mahony M, Thieme U, Goldstein LR. The warm-up effect as a measure of increasing the discriminability of sensory difference tests. J. Food Sci. 53: 1848-1850 (1988)   DOI
11 Thieme U, O'Mahony M. Modifications to sensory difference test protocols: the warmed up paired comparison, the single standard duo-trio and the A-not A test modified for response bias. J. Sens. Stud. 5: 159-176 (1990)   DOI
12 Bi J, Ennis DM. Beta-binomial tables for replicated difference and preference tests. J. Sens. Stud. 14: 347-368 (1999b)   DOI
13 Lee H-S, O'Mahony M. Sensory difference testing: Thurstonian models. Food Sci. Biotechnol. 13: 841-847 (2004)
14 Dacremont C, Sauvageot F, Duyen TH. Effect of assessors expertise on efficiency of warm-up for triangle tests. J. Sens. Stud. 15: 151-152 (2000)   DOI   ScienceOn
15 Bi J, Templeton-Janik L, Ennis JM, Ennis DM. Replicated difference and preference tests: how to account for inter-trial variation. Food Qual. Prefer. 11: 269-273 (2000)   DOI   ScienceOn
16 Rousseau B, O'Mahony M. Investigation of the effect within-trial retasting and comparison of the dual-pair, same-different, and triangle paradigms. Food Qual. Prefer. 11: 457-464 (2000)   DOI   ScienceOn
17 Rousseau B, O'Mahony M. Investigation of the dual pair method as a possible alternative to the triangle and same-different tests. J. Sens. Stud. 16: 161-178 (2001)   DOI   ScienceOn
18 Hautus MJ, Irwin RJ. Two models for estimating the discriminability of foods and beverages. J. Sens. Stud. 10: 203-215 (1995)   DOI   ScienceOn
19 Bi J. Ennis DM. A Thurstonian variant of the beta-binomial model for replicated difference tests. J. Sens. Stud. 13: 461-466 (1998)   DOI   ScienceOn