StatsCapybara · Cheat sheet
Which Test Should I Use?
Find your outcome and design in the left column; the test is on the right. Sections (like 2.1 ) point to the interactive lesson.
① Compare means · a numeric outcome
Your design Test §
One group vs. a known value One-sample t -test 1.13 Same people, measured twice Paired t -test 1.13 Two separate groups Independent t -test (Welch) 1.14 3+ groups, one factor One-way ANOVA 2.2 3+ groups, two crossed factors Factorial ANOVA 2.4 Same people, 3+ conditions Repeated-measures ANOVA 2.5 Adjust for a baseline covariate ANCOVA 2.15 Several outcomes tested jointly MANOVA 3.4
② Assumptions broken? · swap to ranks
Parametric test Rank-based swap §
Independent t -test Mann–Whitney U 2.1 Paired / one-sample t Wilcoxon signed-rank 2.1 One-way ANOVA Kruskal–Wallis 2.1 Repeated-measures ANOVA Friedman 2.1
Use these when the outcome is ordinal , or badly skewed or full of outliers in a small sample. With a large n , the parametric tests are robust.
③ Relationships & association
Your variables Test §
Two numeric, roughly linear Pearson r 1.17 Two numeric, monotone / outliers Spearman ρ 1.17 Two categorical Chi-square test 1.16
④ Predict / model an outcome
What you're predicting Model §
A number, one predictor Simple linear regression 1.18 A number, several predictors Multiple regression 2.9 Yes / no (binary) Logistic regression 2.16 A count (0, 1, 2, …) Poisson / NB (GLM) 3.9 Nested / clustered / repeated Mixed (multilevel) model 3.10 Time until an event, with dropouts Survival (Kaplan–Meier / Cox) 3.16 Combining many studies Meta-analysis 3.18 Max accuracy, not a hypothesis test Machine-learning models ML 5.1
Before you trust the answer
1 · Check the assumptions. Every test above has them: see the Assumption Checks sheet, or §2.6 .2 · Pick one test, not many. Running several and keeping whichever reaches p < .05 inflates false positives (the garden of forking paths ). Decide before you look.3 · A p-value needs an effect size. Significant ≠ large, so always report one beside it (§3.1 ).4 · One numeric + one categorical variable is a group comparison → box ①.