A non-parametric test used to assess the hypothesis that there is no difference between the distributions of two related samples (before and after some treatment).