Power and Sample Size Calculations for Fixed Effects in Repeated Measures Designs

• In Event: Division D: Multiple Linear Regression: The General Linear Model SIG

Abstract

Power analyses for repeated measures are generally more complicated than those for univariate general linear models (GLMs). The focus in this paper is to employ the R2 statistics for Linear Mixed Models proposed by Edwards et al. (2008) to perform power and sample size estimations in two “reversible” repeated measures analyses and demonstrate their similarity with standard GLM power calculations. For approaches that are not one of the two “reversible” analyses, simulation studies will be conducted to demonstrate that the proposed R2 statistics can be used to perform accurate power and sample size estimates.

Author

• T. Mark Beasley, The University of Alabama - Birmingham