Table : TABLE E.4 Category 4 Statistics When at least one variable is ordinal, we have several choices. If both variables are ordinal, or can be converted to ordinal, and we are interested in quantifying the relation between them, we use the Spearman rank-order correlation coefficient. If the independent variable is nominal and the dependent variable is ordinal, we choose the correct nonparametric test based on the research design and the number of levels of the independent variable.

Type of Independent Variable (And Number of Levels, If Applicable)	Research Design	Question to Be Answered	Hypothesis Test
Ordinal	Not applicable	Are two variables related?	Spearman rank-order correlation coefficient
Nominal (two levels)	Within-groups	Are two groups different?	Wilcoxon signed-rank test
Nominal (two levels)	Between-groups	Are two groups different?	Mann–Whitney U test
Nominal (three or more levels)	Between-groups	Are three or more groups different?	Kruskal–Wallis H test