Abstract
Performing analyses of variance on data obtained from experimental designs with repeated measures is a common practice in psychophysiological research. However, investigators often find difficulty in selecting appropriate statistical methods. The present article describes statistically sound and easily performed procedures for conducting repeated-measures analyses of variance of psychophysiological data. Topics include comparison between univariate and multivariate analyses of variance, selection of error terms, tests for simple effects and interaction contrasts, multiple comparison, and effect size. A typical sequence of statistical tests is illustrated using a numerical example. (Japanese Journal of Physiological Psychology and Psychophysiology, 22 (3) : 275-290, 2004.)
