Statistical methods that make no assumptions about the precise form of the frequency distribution from which the data are sampled. They are mainly of use for hypothesis testing using the information in the rank order within each sample. For example, the Mann Whitney U-test may be used to test whether two samples are drawn from the same distribution, and the rank correlation coefficient may be used to test whether two variables are independent.

These methods can be contrasted with parametric techniques, which require specific models with parameters to be estimated.