POWER ANALYSIS ON REPEATED MEASURES DESIGNS

A simulation study comparing powers of the multivariate analysis (PROC GLM) with using a mixed model MIXED) variety covariance structures for various treatment, time, and interaction affect sizes in repeated measures design is conducted. Type 1 errors are estimated. Powers estimated when actual structure AR (1). It was found that treatment effect were all very similar PROC MIXED correct having largest powers. When testing time effect, it doubt better to use simpler structure. The generally higher this case than more complex