Some Results on Estimating a Wilcoxon–Mann–Whitney Measure of Effect Size When There are two Covariates

Abstract

A basic method for characterizing the difference between two independent groups is π\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\pi $$\\end{document}, the probability that a randomly sampled value from the first group is less than a randomly sampled value from the second group. Recently, a method for making inferences about π\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\pi $$\\end{document} was proposed that deals with a covariate. A goal here is to report simulation results when there are two covariates. Included is a suggestion for picking the covariate points. The method is illustrated with data dealing with depressive symptoms among older adults.