non-parametric method for testing whether samples originate from the same distribution.[1][2][3] It is used for comparing two or more independent samples of equal or different sample sizes. It extends the Mann–Whitney U test when there are more than two groups.