Kruskal-Wallis H Test 

Different name: Kruskal-Wallis one-way ANOVA by ranks test

  • Assumptions to check:

    • Data has to be at least ordinal scale 

    • The group independence 

  • Report design

    • For each factor, specify the name and level of the factor and state whether the factor was within-subjects (independent) factor or between subjects (dependent) factor. 

    • Report on post hoc test. 

    • Name the statistical package or program used in the analysis

  • Report statistics:

    • Report the H-statistics, degrees of freedom and exact p-value 

    • Example:

      • Methods section: We used the Kruskal- Wallis H test to test differences in body weight (dependent variable) with drug consumption as independent factors (levels: drug, placebo, no drug and no placebo). 

      • Results section: We used Kruskal-Wallis test to test the difference in body weight with drug consumption (levels: drug, placebo, no drug and no placebo). 

        The differences between the rank totals of  34.91 (drug) , 30.71 (placebo) and 46.43 (no drug and no placebo) were significant, H (2, n = 73) = 6.75, p = .034.

        Post hoc comparisons were conducted using Mann-Whitney Tests with a Bonferroni adjusted alpha level of .016 (0.05 ÷ 3). The difference between “drug” and Group “no drug and no placebo” was statistically significant (U (NGroup B = 24, NGroup C = 22) = 145, z = 2.62, p = .009). None of the other comparisons were significant after the Bonferroni adjustment (ps > .017).