Mann–Whitney U test when variances are unequal

  title={Mann–Whitney U test when variances are unequal},
  author={Eiiti Kasuya},
  journal={Animal Behaviour},
  • E. Kasuya
  • Published 1 June 2001
  • Mathematics
  • Animal Behaviour
The Mann–Whitney U test is one of the most widely used statistical tests in studies of behaviour. In 1999, 42 of 151 articles surveyed (27.8%) in volume 56 of Animal Behaviour used this test, and another eight used its extension to more than two samples: the Kruskal–Wallis test. The Mann–Whitney U test, like many other nonparametric tests based on rank, assumes equal variances in the two populations (in this commentary, I use the term population only in the statistical sense) from which the two… 

Tables from this paper

The unequal variance t-test is an underused alternative to Student's t-test and the Mann–Whitney U test
The aim in this forum article is to argue for the greater use of the last of these tests, the t-test for unequal variances, which is not commonly used.
Distribution-free two-sample comparisons in the case of heterogeneous variances
This work proposes the use of Brunner and Munzel’s generalized Wilcoxon test followed by randomization to allow for small sample sizes and provides a recipe for carrying out a statistical test of the null hypothesis that this probability is 50% and demonstrates the effectiveness of this technique for sample sizes typical in behavioral ecology.
The Mann-Whitney U: A Test for Assessing Whether Two Independent Samples Come from the Same Distribution
Nadim Nachar Université de Montréal It is often difficult, particularly when conducting research in psychology, to have access to large normally distributed samples. Fortunately, there are
A note on the use of the non-parametric Wilcoxon-Mann-Whitney test in the analysis of medical studies
The non-parametric test was more common in case of non-continuous data, in high-impact journals, in studies in humans, and when the statistical software is specified, in particular when SPSS was used.
The Chen-Luo test in case of heteroscedasticity
A Bayesian analysis for the Mann-Whitney statistic
  • R. Chechile
  • Mathematics
    Communications in Statistics - Theory and Methods
  • 2019
Abstract A Bayesian analysis is developed for the Mann-Whitney (U) statistic. The analysis is focused on an Ω parameter, which is defined as the integrated proportion of the population in a generic
Mann-Whitney U ile Student’s t testinin I.Tip Hata ve Güç bakımından Karşılaştırılması: Monte Carlo Simülasyon Çalışması (011302) (5-11)
The nonparametric Mann-Whitney U test is an alternative to parametric Student’s t-test, when the assumptions of t-test cannot be respected. However as one understood, it is more reliable to use the


Randomization tests and the equality of variance assumption when comparing group means
Results are provided from a series of simulations showing that when the population variances differ, the use of a randomization test described by Adams & Anthony (1996) to compare group mean differences often falsely rejects a true null hypothesis at a rate sometimes far greater than the level of significance chosen for the test.
THE third part of Biometrika, published in April, contains several important contributions, the first of which is by Prof. Karl Pearson, who describes “a systematic method of curve-fitting by
Variance in male mating success and female choice for persuasive courtship displays
It is concluded that sexual selection via female choice favours persuasive displays because they confer high mating success on the males that perform them at the highest frequencies.