A MONTE-CARLO COMPARISON OF 7 EPSILON-ADJUSTMENT PROCEDURES IN REPEATED-MEASURES DESIGNS WITH SMALL SAMPLE SIZES

The purpose of this study was to evaluate seven univariate procedures for testing omnibus null hypotheses for data gathered from repeated me asures designs. Five alternate approaches are compared to the two more traditional adjustment procedures (Geisser and Greenhouse's epsilon a nd Huynh and Feldt's epsilon), neither of which may be entirely adequa te when sample sizes are small and the number of levels of the repeate d factors is large. Empirical Type I error rates and power levels were obtained by simulation for conditions where small samples occur in co mbination with many levels of the repeated factor. Results suggested t hat alternate univariate approaches were improvements to the tradition al approaches. One alternate approach in particular was found to be mo st effective in controlling Type I error rates without unduly sacrific ing power.