Handling replications in discrimination tests
暂无分享,去创建一个
We give in this paper an easy to use and statistically sound way of handling replications in discrimination tests based on a concept called overdispersion. The overdispersion is easily calculated by hand for any discrimination testing setup. And since the method simply amounts to correcting the total number of observations, and similarly the overall response, by the overdispersion, we can use any standard technique, table or software on the corrected numbers to perform a statistical test for difference, for similarity, power calculations, controlling risk I and risk II, etc. We also include a general discussion of the replicated discrimination testing situation, and some examples to illustrate how to use the method. The SAS® macro REPRISKS that handles all computations is available by e-mail: schlich@arome.dijon.inra.fr.
[1] Hans-Georg Müller,et al. Random Effect Threshold Models for Dose–Response Relationships with Repeated Measurements , 1997 .
[2] P. Schlich. Risk tables for discrimination tests , 1993 .
[3] A. W. MacRae,et al. Confidence intervals for the triangle test can give reassurance that products are similar , 1995 .
[4] Herbert Stone,et al. Sensory Evaluation Practices , 1985 .
[5] P. McCullagh,et al. Generalized Linear Models , 1992 .