Abstract Statistical tests in the ANOVA and correlation families (e.g., t ‐test, Pearson's correlation, multiple regression, path analysis) require the distribution underlying the dependent variables to be normally distributed. Because the normal curve is defined by two parameters (the mean and standard deviation), such tests are referred to parametric . Another class of tests is called nonparametric , or more properly, distribution‐free , because they do not make any assumptions about the parameters of the population from which the sample(s) are drawn. However, while they are distribution‐free, they are not assumption‐free. Many of the nonparametric tests have the same assumptions as parametric ones, such as that the observations are independent; that they are at a certain level of measurement (e.g., nominal or ordinal); and with some, that the distributions are similar across groups (i.e., if they are skewed, then they are skewed in the same direction for all groups).