AbstractThe principles underlying the correct interpretation of the judgments of Tasting Panels are discussed with special reference to the use of an analysis of variance. Certain assumptions inherent in present methods for assigning palatability scores are examined critically. There is some risk of erroneous conclusions when a series of scores, each for one desirable quality, is to be combined to obtain a single score for general palatability; suggestions arc made for improving the correlation between such scores and consumers' actual verdicts.