ABSTRACT
This chapter describes methods for repeated measures analyses that have been traditionally used, but which are appropriate only in special situations, such as fixed measurement times, complete data, specific pattern of variances and covariances, and normally distributed measures. It focuses on methods for assessment of change, from baseline to endpoint, and for testing and estimation of group differences in longitudinal studies. This is done via methods for analysis of independent data such as t-test, ANOVA, or ANCOVA. The chapter also focuses on between-group comparisons of other summary measures using the same set of statistical approaches for independent data (t-test, ANOVA, ANCOVA). It then focuses on the assumptions of the methods and explains the disadvantages of these approaches, especially when some data are missing. More appropriate approaches for longitudinal and clustered data, are repeated measures analysis of variance (rANOVA), and repeated measures multivariate analysis of variance (rMANOVA).
