The Kruskal-Wallis H Test is used to find out whether more than two independent groups or populations can be considered statistically different on an outcome variable or not. Two variables are needed: a categorical variable with more than two groups and an outcome variable measured on the ordinal level or higher. The Kruskal-Wallis H Test is the preferred test when your data violates the assumptions of a One-Way ANOVA. Besides having independent observations in the groups, there are no statistical assumptions to be tested for the Kruskal-Wallis H Test. When significant, the Kruskal-Wallis H Test is usually followed by several Mann-Whitney U Tests to determine which groups actually differ from each other. This will also be included in our analyses.


There are no reviews yet.

Be the first to review “Kruskal-Wallis H Test”

Your email address will not be published. Required fields are marked *