Strategies for the Analysis of Repeated and Multiple Measures Designs

Differentiation is made relative to repeated and multiple measure analyses. A flowchart is provided to help the researcher determine the appropriate analysis for the design of interest. Conservative tests and a multivariate approach are suggested for repeated measures designs. For designs with multiple dependent measures, the following suggestions are provided: rank reduction, Bonferroni approach, and MANOVA. Post hoc tests are also suggested once a significant multivariate F is obtained. The post hoc tests are univariate F tests, two-group breakdown, simultaneous confidence intervals, discriminant analysis, step-down analysis, and contrasts on the classification variable. It is suggested that inclusion of these more appropriate analyses will increase the internal validity of research published in the Research Quarterly.