The Mann-Whitney U test or the Wilcoxon Rank Sum test is sensitive to unequal variances. Is there an equation that adjusts the standard deviation of this statistic, much like in the Welch-corrected t-test for independent samples?
Asked
Active
Viewed 113 times
0
-
Who says it "is sensitive to unequal variances"? – gung - Reinstate Monica Dec 31 '16 at 18:12
-
By sensitive, I mean the false positive rate increases with increases differences in variances between to two populations. I'm not sure who says this, but it's true. – nick michalak Jan 01 '17 at 21:47
-
How do you know it's true? – gung - Reinstate Monica Jan 01 '17 at 23:40