Statistical Tests
#Math
How to choose a statistical test
Broader Categories
A way to determine if groups are different from each other. More precisely, it's a way to test whether measurements between two or more groups indicate that their population means are significantly different
An n-way or n-factor statistical test compares n independent variables
Compare Means
Assume variances between groups are equal
2 Groups (e.g. Control vs. Drug)
Same Subjects Measured Twice (Before and After)
Pairwise Differences Normal
(i.e. )
- Paired t-test (parametric)
Pairwise Differences Non-Normal
- Wilcoxon Matched Pairs Signed Rank Test (non-parametric)
Each Subject Measured Once
Each Group Normal and Variances Equal
Each Group Normal But Variances Not Equal
- Welch's t-test
- Unpaired t-test with Welch's Correction (Parametric Test)
One or Both Groups Non-Normal
- Mann-Whitney Test (Nonparametric Test)
>2 Groups (e.g. Drug A vs. Drug B vs Drug C)
One Factor (Drug)
Each Subject Measured More Than Once
Pass Normality
Assume Sphericity
- ANOVA (Missing Data)
Don't Assume Sphericity
Reject Normality
- Friedman Test (Non-parametric Test)
Each Subject Measured Once
Pass Normality and Variances Equal
Pass Normality But Variances Not Equal
Pass Normality and Variances Equal
Two Factors (Drug, Diet)
- Same as One Factor except if there's no normality (so you can't use a parametric method), then you have to consider a transformation. I.e. there's no non-parametric methods
Three Factors (Drug, Diet, Time)
- Same as Two Factors