Nonparametric tests, also referred to as distribution-free tests, do not require stringent assumptions of parametric tests and are especially attractive when the underlying population is markedly nonnormal. Also, while parametric tests require data of interval or ratio scale, nonparametric tests can be performed on data of nominal or ordinal scale.

However, a nonparametric test ignores useful information since it often focuses on the rank rather than the magnitude of sample values. Therefore, in situations when the parametric assumptions are valid, the nonparametric test is less powerful (more prone to Type II error) than its parametric counterpart. In general, when the assumptions for a parametric test are met, it is preferable to use a parametric test rather than a nonparametric test. Since the normality assumption for parametric tests is less stringent in large samples, the main appeal of rank-based nonparametric tests tends to be with relatively small samples.