MANOVA method for analyzing repeated measures designs: an extensive primer.

This article teaches the multivariate analysis of variance (MANOVA) method for repeated measures analysis to researchers who are already familiar with regular analysis of variance (ANOVA) methods. Repeated measures designs are traditionally analyzed with mixed-model ANOVAS. However, sphericity violations markedly affect the true Type I error rates and power for the mixed-model tests. However, tests based on the MANOVA approach are free of sphericity assumptions, and with modem computing software, the MANOVA approach is straightforward to use, even for complex designs and nontraditional hypotheses. We present a general strategy for implementing the MANOVA approach within statistical computing packages, and illustrate how regular (preplanned) tests and Scheff~-type simultaneous (post hoc) tests are conducted for a variety of designs and hypotheses.