Brief Descriptions of Statistical Methods

This page contains brief descriptions of statistical tests to help analysts determine what types analyses may be applicable to their study. This page not meant to be a complete guide on the appropriateness of any given test, but rather general guidelines to the types of variables that can be used with each type of statistical analysis.  Analyses are in alphabetical order.

Analysis Dependent / Response variable Independent Variable(s) Notes & Assumptions
Independent Two Samples t-test Continuous Categorical (with 2 levels) The response variable should be normally distributed, minimal outliers
Kruskal-Wallis test Continuous Categorical (with >2 levels) The response variable is not normally distributed, and/or you have signficant outliers. Analogous to a nonparametric one-way ANOVA
Mann-Whitney U test Continuous Categorical (with 2 levels) The response variable is not normally distributed, and/or you have signficant outliers. Analogous to a nonparametric independent sample t-test
One-way ANOVA Continuous Categorical (with >2 levels) The response variable should be normally distributed, minimal outliers, equal variances
Paired Samples t-test Continuous Categorical (with >2 levels) The difference between groups (matched pairs) should be normally distributed, minimal outliers
Wilcoxon signed-rank test Continuous Categorical (with 2 levels) The difference between groups (matched pairs) are not normally distributed, and/or you have signficant outliers. Analogous to a nonparametric paired samples t-test