2 levels of onetreatment to the differences in ranks within those levels.After calculating U-values, you should use the larger of the two numbers to determine thesignificance of your test for a...

Extracted text: Which of the following statements about non-parametric tests are correct? O Kruskal-Wallace test compares the differences in the ranks of data across >2 levels of one treatment to the differences in ranks within those levels. After calculating U-values, you should use the larger of the two numbers to determine the significance of your test for a given alpha. If your data violate the assumptions of GLM, they are not comparable to a generic normal distribution for the purposes of significance testing. The Kruskal-Wallace test can be used if data from a 2-way factorial experiment violate the assumptions of the GLM. Spearman correlations are more powerful than Pearson correlations.