Question

Describe the Kruskal-Wallis Test.

Answer 1

The Kruskal Wallis test is the non parametric alternative to the One Way ANOVA. Non parametric are the test in which you don’t assume that data comes from a particular distribution. We use this test when the assumption for ANOVA is not meeting. It is also called the one-way ANOVA on ranks, as the ranks of the data values are used in the test instead of the actual data points.

In this test we check whether the medians of two or more groups are different. Similarly to other test we calculate the test statistics and do the comparison with cut off point. Test statistics here is called H statistics. Hypotheses are:

H0: population medians are equal.

H1: population medians are not equal.

With the Kruskal Wallis test we can find out if there is significant difference between groups, but it won’t help to known which groups are different. For that Post Hoc test is used.