ABSTRACT

Whether one should test for Judge * Sausage Type interaction or not may depend on the kinds of judges one is using. If the judges are chosen on the basis of their expertise (i.e., they are experts), then it is always important to test for Judge * Sausage Type interaction. This is true because the judges should be evaluating the same characteristics of the samples, and they should agree on the relative differences in these characteristics from sample to sample. For example, if the judges are evaluating the samples for saltiness, they should generally agree on which samples are most salty, which samples are least salty, etc. One is generally not interested in whether they score the samples exactly the same or not, but in whether they agree on the relative comparisons between the samples. If the judges cannot agree on the relative comparisons between the samples, they could hardly be called experts. Thus if most of the judges are in agreement with each other except for perhaps one or two judges, then the judges not in agreement with the majority should probably be removed and the data reanalyzed.