In: Statistics and Probability
Detail the steps in computing the Kruskal-Wallis H-test with a brief
discussion specific to the important aspects of each step.
The kruskal-Wallis H-test is the non-parametric alternative of one way ANOVA which means the normality assumption used in the ANOVA is not assumed.
The hypothesis for the test is
H0 : The samples are from identical population.
H1 : Atleast one of the samples is from a different population.
The steps for testing this hypothesis is given below.