Online Program

Independent Means T-test or Robust Alternatives: A Guide to Selecting the Best Tool for Inferences (302997)

Keywords: heteroscedasticity, non-normality, Independent means t-test, Satterthwaite approximate t-test, trimmed means, simulation

While the independent means t-test is popular for testing the equality of two population means, it is sensitive to violations of the population normality and homogeneity of variance assumptions. In such cases, Satterthwaite’s approximate t-test and Yuen’s trimmed t-test are recommended as robust alternatives, which relax those assumptions; choice between the t-test and the alternatives may be conditioned on a preliminary test of the assumption of homogeneity of variance. A simu- lation study was conducted to explore type I error rates and statistical power of the independent means t-test, the Satterthwaite’s approximate t-test, the trimmed t-test, and the conditional tests under normal and non-normal distributions (the two alternative tests were conducted both unconditionally and conditionally on the preliminary test of variances). Power comparisons (for conditions in which type I error control was adequate) were used to identify the most powerful test among the set.This presentation provides the results of the comparison of the performance of the t-test and its alternatives and presents a SAS macro to facilitate easy computation of Yuen’s symmetric trimmed t-test.