Location and scale testing in a mixed design

Abstract

There are several situations in scientific research where the researcher is interested in either changes in means or variances if new treatments are applied and compared to a control. Recently, Khalawi and Magel proposed and compared nonparametric tests for the simple tree alternative for testing differences in means or variances in a mixed design (completely randomized (CRD) and a randomized complete block (RCBD)) under symmetric distributions. However, there are times in which the data come from a right-tailed skewed distribution. In this paper, we estimate and compare these nonparametric tests under an asymmetric distribution via Monte Carlo simulation.