top of page
  • Black Facebook Icon
  • Black Twitter Icon
  • Black Instagram Icon

Exploiting the Elementary

Wilcoxon-Mann-Whitney Parameters

(under development)

Ralph G. O'Brien
S. Paul Wright

The Wilcoxon-Mann-Whitney (WMW) method compares two groups with respect to an ordered categorical variable, Y. Letting Y1 and Y2 be the Y values for the two groups, the fundamental WMW parameter is
          WMWprob = Prob[Y1 > Y2] + Prob[Y1 = Y2]/2
or, using odds scaling,
          WMWodds = WMWprob/(1 - WMWprob).

​

A major R function, WMW(), computes the estimates and confidence intervals for WMWprob and WMWodds. Optionally, setting a specific null hypothesis begets a p-value congruent with the CI.

​

A proposed Bayesian approach yields counterparts to (frequentist) confidence intervals and p-values.

​

WMW() returns Qscores (transforms of Y1 and Y2), which enable plotting the individual data values in a manner consistent with WMWprob.

fgfdgfdg-600x350.png

Daniel Lunsford’s Latest Releases

"I'm a paragraph. Click here to add your own text and edit me. I’m a great place for you to tell a story and let your users know a little more about you."

“I'm a review. Click to edit me and add text from a critic who has evaluated you and your work.”

Seattle Daily

Reviews

“I'm a review. Click to edit me and add text from a critic who has evaluated you and your work.”

The Washington Paper

“I'm a review. Click to edit me and add text from a critic who has evaluated you and your work.”

T.O.M Magazine

bottom of page