R. V. Hopkins (1982) has criticized the use of means as the unit of analysis in situations where intact groups, such as classes, rather than individuals have been randomly assigned to various treatment conditions. Instead, Hopkins advocated the use of certain analysis of variance (ANOVA) models that, as far as test for treatment effects are concerned, yield results that are equivalent to those that would be obtained if class means were used as the unit of analysis. It is pointed out that, because of the non-robustness of the sample mean as an estimator of location, use of the class mean as the unit of analysis or of the ANOVA models advocated by Hopkins can lead to larger than necessary Type II error rates in tests of significance for treatment effects. The ways in which, in non-normal population situations, use of summary statistics other than the mean can lead to significant increases in the power of tests for treatment effects are discussed. It is also suggested that the pooling options offered by Hopkins should be viewed with caution. Two data tables are included. (Author/SLD)