StatsTest.com is making great progress. I just finished the 16th statistical method. All 16 of these methods fall under difference, meaning you want to see whether two groups are significantly different on your variable of interest, rather than relationship or prediction. All 16 of these methods also fall under continuous, meaning the variable that you care about is continuous, or can take on basically any value.

Since the Difference -> Continuous methods are finished, I’m now moving on to Proportional methods, and will then tackle the relationship/prediction side of the tree.

Building out StatsTest has been a fun exercise in creating a domain map, or mapping out a high-level course domain into its sub-components. Each of these methods has distinguishing features and assumptions about them, and grouping them together by their assumptions and use cases results in a fairly interpretable series of questions that can guide someone to the correct statistical test.