THE ANALYSIS OF REPEATED MEASUREMENTS - A QUANTITATIVE RESEARCH SYNTHESIS

Meta-analytic methods were used to summarize the results of Monte Carl o studies investigating the Type I error and power properties of vario us univariate and multivariate procedures for testing within-subjects effects in split-plot repeated measures designs. Results indicated tha t all test procedures were generally robust to violations of the multi variate normality assumption, but varied in terms of their Type I erro r control when the sphericity assumption was not satisfied. For balanc ed designs, the usual F and (e) over cap adjusted F tests (Greenhouse & Geisser, 1959) were generally robust to moderate degrees of covarian ce heterogeneity, whereas the multivariate procedures were slightly mo re affected by departures from this assumption. When the design was un balanced, however, all procedures were sensitive to the presence of he terogeneous covariance matrices, particularly when testing the within- subjects interaction effect. Power rates varied little as a function o f assumption violations. However, this finding may be due to the restr icted range of many of the variables included in the meta-analysis of the power data as well as the strong and overshadowing relationship be tween the degree of non-centrality and power rates. For balanced desig ns, the use of either an (e) over cap-adjusted univariate or a multiva riate approach is recommended; for unbalanced designs, researchers sho uld consider adopting one of several robust alternatives that have rec ently been suggested in the literature.