| The ANOVA F is a widely used statistic in psychological research despite its shortcomings when assumptions of variance heterogeneity and normality are violated. A Monte Carlo investigation compared Type I error and power rates of the ANOVA F, Alexander-Govern, Welch-James, BDM, Welch, and Welch on ranked data. The results suggest that the BDM or the Welch on ranked data offer the best balance of empirical Type I error control and statistical power under several conditions of variance heterogeneity and nonnormality. Recommendations are made to incorporate either of these procedures into applied research. |
