For data with observations on variable Y restricted to consecutive and untied ranks 1,2... n, and with observations on X restricted to nominal classification into group 1 or 2, the Mann-Whitney U-test provides a test of group mean difference. It is shown below that the rank-biserial correlation coefficient rrb is a linear function of the U-statistic, so that a test of group mean difference is equivalent to a test of zero correlation for the rank-biserial coefficient. This provides a distribution theory for sample values of rrb when ρrb = 0.
Get full access to this article
View all access options for this article.
References
1.
Cureton, E. G.Rank-biserial correlation. Psychometrika, 1956, 21, 287-290.
2.
Cureton, E. G.Rank biserial correlation when ties are present. EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT , 1968, 28, 77-79.
3.
Glass, G. V.Note on rank-biserial correlation . EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT, 1966, 26, 623-631.
4.
Mann, H. B. and Whitney, D. R.On a test of whether one or two random variables is stochastically larger than the other. Annals of Mathematical Statistics, 1947, 18, 50-60.
5.
Milton, R. C.An extended table of critical values for the Mann-Whitney (Wilcoxon) two-sample statistic. Journal of the American Statistical Association, 1964, 59, 925-934.
6.
Siegel, S.Nonparametric statistics for behavioral sciences. New York: McGraw-Hill , 1956.
7.
Stanley, J.An important similarity between biserial r and the Brogden-Cureton-Glass biserial r for ranks. EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT, 1968, 28, 249-253.
