The Wilcoxon-Mann-Whitney is a good test of medians. Not!"


April 13, 2010
A Presentation by George Divine

As pointed out by Ralph O'Brien, but contrary to assertions by a large number of statisticians and textbooks, the Wilcoxon-Mann-Whitney (WMW) test need _not_ be a good test of medians. The virtues and shortcomings of the WMW as a test of medians will be discussed and illustrated with (counter-) examples. Finally, a critique of a somewhat muddled recent contribution to the literature on this issue will be presented.

George Divine, Ph.D. is a Senior Research Biostatistician in the Department of Biostatistics and Research Epidemiology (BRE) at the Henry Ford Health System (HFHS). He is past head for the BRE Division of Biostatistics. He has been involved in a range of research projects from Stroke to Cancer and including Health Services Research. He is the current head of the WSU/HFHS CTSA Workgroup for Biostatistics. His research interests include multiple comparisons and sample size calculation methods.  

Scott Millis giving George Divine a Certificate of Appreciation from the Detroit Chapter