The equivalence of the t test on the ranks of original scores to the Wilcoxon Rank-Sum test on original scores was noted. However, this study demonstrated that, unlike the Wilcoxon, the t on ranks is not competitive with the usual t for normally distributed data. The explanation is due to the lack of a correction for ties and continuity most computer statistical packages automatically employ in the Wilcoxon routine.
Get full access to this article
View all access options for this article.
References
1.
BlairR. C. (1987) RANGEN. Boca Rotan, FL: IBM Corp.
2.
BlairR. C.HigginsJ. J. (1980a) A comparison of the power of the t test and the Wilcoxon statistics when samples are drawn from a certain mixed normal distribution. Evaluation Review, 4, 645–656.
3.
BlairR. C.HigginsJ. J. (1980b) A comparison of the power of the Wilcoxon's rank-sum statistics to that of Student's t statistic under various non-normal populations. Journal of Educational Statistics, 5, 309–335.
4.
CohenJ. (1988) Statistical power analysis for the behavioral sciences. (2nd ed.) Hillsdale, NJ: Erlbaum.
5.
MicceriT. (1989) The unicorn, the normal curve, and other improbable creatures. Psychological Bulletin, 105, 156–166.
6.
SawilowskyS. S.BlairR. C.MicceriT. (1990) A PC FORTRAN subroutine library of psychology and education data sets. Psychometrika, 155, 729.
7.
ZimmermanD. W.ZumboB. D. (1990) Effects of outliers on the relative power of parametric and nonparametric statistical tests. Perceptual and Motor Skills, 71, 339–349.
