Wallis Flashcards
Test involving more than two independent Samples (Kruskal-Wallis H test)
For more than two independent samples and given that the distribution of the samples is unknown, the appropriate nonparametric test is .
Kruskal-Wallis H
What is the test
Kruskal-Wallis H
Also called the H test, the Kruskal Wallis test is a generalisation of the Mann Whitney test. It provides the non-parametric alternative to the test involving several population means (one-way ANOVA).
Let n_i be the size of the i^th sample
For given sets of data of sizes n_1, n_2, n_3, . . . n_k
we rank the k sets of data collectively
identify the ranks of n_1, n_2, n_3, . . . n_k and
sum the ranks of each of the groups n_1, n_2, n_3, . . . n_k
k being the number of samples. Then, the Kruskal Wallis (H) test is given by:
Formula
H calculated
=12. Ri.^2
—— ∑. ——. -3(n+1)
n(n+1). nj
H = 12. Ri.^2
—— ∑. ——. -3(n+1)
n(n+1). nj