Mann-Whitney U test

Use - Compare two groups.  It is a non-parametric ver of the t test.

Note the Mann-Whitney U test is essentially the same as the  Wilcoxon rank sum test.

Requirements / Assumptions:

• 8-10 observations in each group.

• Data distribution - no requirement

• Variances - Some sources say equal / similar variances are required or state that the data come from a common population.. while others do not discuss it.  Some authors recommend if the variances are large to use the Brunner–Munzel test (add references and summary)

Useful resources:

Mann-Whitney U test by DataTab