Abstract
A scheme for conducting two-sample profile analysis with Hotelling's T[SUP2] tests and univariate t tests is described. Formulas are provided and a decision tree is presented to guide users in selecting among the tests. Two example problems are worked to illustrate the analysis. The procedure evaluates virtually the same hypotheses as two-factor analysis of variance designs that have a repeated measures factor. However, the procedure has less restrictive assumptions than do repeated measures designs.