Skip to content
K Knidox Search…
Statistics · Inference

ANOVA Calculator

One-way analysis of variance with the full F-table and p-value.

One group per line. Separate values with commas or spaces.
F-statistic
21.6p < 0.05

df = (2, 9) · p = 0.0004

SourceSSdfMS
Between72236
Within1591.667
Total8711

One-way ANOVA compares three or more group means with F = MS_between ÷ MS_within. For three groups with means 2, 5, and 8, MS_between = 27 and MS_within = 1, so F = 27 on (2, 6) df — far above the 5.14 critical value, so at least one mean differs significantly.

Comparing three or more groups

A t-test compares two means; one-way ANOVA generalizes it to any number of groups in a single test, avoiding the inflated error rate of running many pairwise t-tests. It splits the total variation into a between-groups part and a within-groups part, then asks whether the spread between group means is large relative to the spread within the groups.

F = MS_between ÷ MS_within

MS = SS ÷ df; df_between = k − 1, df_within = N − k

Worked example

Three groups of three — A: 1, 2, 3 (mean 2); B: 4, 5, 6 (mean 5); C: 7, 8, 9 (mean 8). Grand mean 5, N = 9, k = 3.

  1. 1
    Between-groups mean square. SS_between = 3(2−5)² + 3(5−5)² + 3(8−5)² = 27 + 0 + 27 = 54; df = k − 1 = 2, so MS_between = 54 ÷ 2 = 27.
  2. 2
    Within-groups mean square. Each group’s deviations (−1, 0, +1) sum to 2; SS_within = 2 + 2 + 2 = 6; df = N − k = 6, so MS_within = 6 ÷ 6 = 1.
  3. 3
    Form the F-ratio. F = 27 ÷ 1 = 27 on (2, 6) df. The 0.05 critical F is 5.14, so 27 ≫ 5.14 — clearly significant.

Critical F-values (α = 0.05)

Numerator df across the top, denominator df down the side; reject when F exceeds the value.

df_within ↓ / df_between →123
65.995.144.76
104.964.103.71
204.353.493.10
304.173.322.92

Reading the table and the assumptions

What a significant F means. A large F with a small p-value says at least one group mean differs from the rest — but not which one. Follow up with post-hoc comparisons (e.g. Tukey’s HSD) to locate the difference.

Why not many t-tests. Running a t-test on every pair inflates the chance of a false positive. ANOVA tests all groups at once at your chosen α, keeping the overall error rate in check.

The assumptions. One-way ANOVA assumes independent observations, roughly normal groups, and similar variances across groups (homogeneity of variance). It is fairly robust when group sizes are equal.

Entering data. Put one group per line, with values separated by commas or spaces. The tool reports the full F-table — sums of squares, degrees of freedom, mean squares, F, and the p-value.

What does a significant ANOVA tell me?
That at least one group mean differs — but not which one. Follow up with post-hoc comparisons (e.g. Tukey’s HSD) to locate the difference.
How do I enter my data?
One group per line, with values separated by commas or spaces. Add as many lines as you have groups.
What are the assumptions?
Independent observations, roughly normal groups, and similar variances across groups. ANOVA is fairly robust when the group sizes are equal.
Why not just run several t-tests?
Each extra t-test raises the chance of a false positive. ANOVA compares all groups in one test at your chosen significance level, controlling the overall error rate.
What’s the difference between MS_between and MS_within?
MS_between measures how much the group means vary; MS_within measures the average spread inside the groups. Their ratio is F — large when the groups truly differ.
How many groups do I need for ANOVA?
At least three to make it worthwhile — with only two groups a t-test gives the same answer. There is no upper limit; df_between is simply k − 1 for k groups.