t-test / One-way ANOVA
Compare means of a continuous variable across two or more groups. Two groups give an independent-samples t-test (Student's with equal variances and Welch's without) plus an optional paired t-test; three or more groups automatically run a one-way ANOVA. Results include the statistic, degrees of freedom and p-value.
① Enter data
One group per line; numbers within a group separated by spaces or commas. 2 groups = t-test, ≥3 groups = ANOVA.
How to use & methodology
When to use a t-test vs ANOVA?
Use a t-test to compare the means of two groups; use one-way ANOVA for three or more. This tool chooses automatically based on how many groups you enter.
Student's vs Welch's?
Use Student's pooled-variance test when the two variances are equal (homogeneous); use Welch's when they are not. In practice a variance-homogeneity test is often run first, and many packages default to the more robust Welch's.
When is a paired t-test appropriate?
For two repeated measurements on the same subjects, or one-to-one matched groups (e.g. left/right side, matched case-control). It requires equal sample sizes in matching order. Tick the box to also get the paired result.
What if data are not normal?
The t-test and ANOVA assume approximately normal data. For markedly skewed data, ordinal grades or small samples, use a non-parametric test instead (e.g. Mann–Whitney U, Kruskal–Wallis).