Comparative strength of t-test and wilcoxon two-sample test for customary patterns of psychological and educational data

Abstract EN

The current study attempts to investigate the efficient test that were employed to answer the research questions with two randomized group design under the violation of parametric assumptions.

A computer program simulated power functions of independent t and Wilcoxon two-sample tests under combinations of variance heterogeneity, sample size, and multimodality conditions.

Based on Type I error and power results, the Wilcoxon two-sample test will be selected under the extreme violation of normality assumption, where parametric t-test is recommended under the other conditions.