New: Cookbooks and AI ExplanationsStep-by-Step recipes to solve problems connected to Roadmaps and Cheat Sheets. Need more details? Use AI buttons for structured and simple explanations with concrete examples throughout the whole platform.Take a look
My data is full of outliers and skewed values, and the usual t-test rules feel like they don't apply anymore.
What you'll have at the end
A Mann-Whitney or Wilcoxon result you can trust instead of a t-test that data doesn't support
You need
Two lists of numeric measurements, one per group, pulled the same way you'd feed a t-test, plus a clear read on whether the groups are separate people or the same people measured twice.
Not covered
Comparing three or more groups at once, which needs the rank-based ANOVA alternative instead of this two-group test.
Compare two groups' averages with a t-test
your two groups actually look roughly normal with no heavy outliers, so the plain t-test is the better first move
An A/B test isn't significant, and the team still needs a call
your Mann-Whitney result also comes back not significant and you still need to give your manager a next step
Checked 26 Aug 2026
Part of the Statistics Fundamentals cookbook