Menu

Free Statistics Calculator

ANOVA Calculator

Compare three or more groups at once. Paste your data and get the F-statistic, p-value, and Tukey HSD pairwise comparisons - with an automatic fall-back to the Kruskal-Wallis test when your data is not normal - plus a paste-ready methods sentence.

No sign-up requiredRuns on scipy - exact, reproducibleFree to use

Enter your groups

0 values

0 values

0 values

Test (two groups)
Error bars

One-way ANOVA, then post-hoc

When you have three or more groups, running multiple t-tests inflates your false-positive rate. A one-way ANOVA instead asks a single question - are all the group means equal? - and reports an F-statistic and p-value. Enter each group below, choose Auto-pick, and the calculator checks the assumptions before deciding between ANOVA and its non-parametric counterpart.

A significant omnibus result means at least one group stands out, but not which one. The Tukey HSD post-hoc table answers that by comparing every pair of groups with family-wise-error-corrected p-values, so you can mark exactly which differences are significant on your figure.

ANOVA vs Kruskal-Wallis

ANOVA assumes each group is roughly normal with similar variance. When those assumptions do not hold - common with small samples or skewed measurements - the Kruskal-Wallis test compares the groups by rank instead and makes no normality assumption. Auto-pick chooses it for you and runs Holm-corrected pairwise comparisons as the post-hoc. Only comparing two groups? Use the t-test calculator.

Frequently asked questions

What does a one-way ANOVA test?

A one-way ANOVA tests whether the means of three or more independent groups are all equal. A significant result (p < 0.05) tells you at least one group differs, but not which - that is what the Tukey HSD post-hoc comparisons answer.

Why did it run Kruskal-Wallis instead of ANOVA?

On Auto-pick, the calculator runs a parametric ANOVA only when every group is approximately normal (Shapiro-Wilk) and variances are homogeneous (Levene). If an assumption fails, it falls back to the Kruskal-Wallis test, the non-parametric equivalent, with pairwise Mann-Whitney U comparisons (Holm-corrected) as the post-hoc.

What is a Tukey HSD post-hoc test?

After a significant ANOVA, Tukey's Honestly Significant Difference test compares every pair of groups while controlling the family-wise error rate, so the pairwise p-values are already corrected for multiple comparisons.

Can I compare just two groups here?

You can, but for exactly two groups the t-test calculator is the better fit. Use the ANOVA calculator when you have three or more groups to compare at once.

Related tools

From numbers to a figure

Plotivy runs the test, picks the right chart, and annotates significance bars on a publication-ready figure - all from your data.

Open the Analyze tool