ANOVA Calculator - One-Way F Test With Table

About the ANOVA Calculator

Analysis of variance (ANOVA) tests whether the means of several groups are all equal. Do three teaching methods produce the same average test score? Do four fertilisers give the same average crop yield? Do customers spend the same on average in five regions? A t-test compares two means; one-way ANOVA extends the idea to any number of groups in a single test, avoiding the inflated false-positive rate that comes from running many t-tests.

This ANOVA calculator runs a one-way ANOVA. Enter each group's values on its own line, and it gives the F statistic with its degrees of freedom, the p-value, each group mean, the full ANOVA table of sums of squares and mean squares, and eta squared (η²), the share of variation explained by the groups.

How to Use the ANOVA Calculator

Type or paste each group on its own line, with the values separated by commas or spaces. Groups can have different sizes, but each needs at least two values.

The result appears below, with the ANOVA table and effect size.

The Formulas

  SSB (between) = Σ nᵢ × (x̄ᵢ − x̄)²          df = k − 1
  SSW (within)  = Σ Σ (x − x̄ᵢ)²              df = N − k
  MSB = SSB ÷ (k − 1),   MSW = SSW ÷ (N − k)
  F = MSB ÷ MSW
  η² = SSB ÷ (SSB + SSW)

Here k is the number of groups, N the total number of values, x̄ᵢ each group's mean and x̄ the grand mean.

Step-by-Step Example

Test scores for three teaching methods:

  Method 1:  85, 90, 88, 92     mean 88.75
  Method 2:  78, 82, 80, 84     mean 81
  Method 3:  90, 95, 93, 91     mean 92.25
  Grand mean: 1,048 ÷ 12 = 87.333

  SSB = 4 × (1.417² + 6.333² + 4.917²) = 265.17     df = 2
  SSW = 26.75 + 20 + 14.75 = 61.5                    df = 9
  MSB = 132.58,  MSW = 6.833
  F = 132.58 ÷ 6.833 = 19.40
  p = 0.0005,  η² = 265.17 ÷ 326.67 = 0.81

The group means differ far more than the variation within groups would explain. At least one method produces a different average score, and 81 percent of the variation in scores is associated with the method used.

The ANOVA Table

  Source     SS       df    MS       F       p
  Between    265.17    2    132.58   19.40   0.0005
  Within      61.50    9      6.83
  Total      326.67   11

The between and within sums of squares always add up to the total sum of squares around the grand mean. This is the "analysis" in analysis of variance: total variation is split into a part explained by the groups and a part left over.

How F Works

If the groups really have the same mean, the between-groups mean square and the within-groups mean square both estimate the same underlying variance, so F should be close to 1. When the group means differ, MSB grows while MSW does not, and F rises. The p-value gives the probability of an F at least this large if all the true means were equal. With 2 and 9 degrees of freedom, F must exceed about 4.26 to be significant at 5 percent; 19.40 is far beyond that.

Why Not Several T-Tests?

Comparing three groups in pairs needs three t-tests. If each has a 5 percent chance of a false positive, the chance that at least one gives a false positive is about 14 percent; with five groups and ten tests it is about 40 percent. ANOVA keeps the overall false-positive rate at 5 percent by testing all the means at once. After a significant ANOVA, post-hoc tests such as Tukey's HSD compare pairs of groups while still controlling the overall error rate.

Two Groups: ANOVA and the T-Test

With only two groups, one-way ANOVA and the equal-variance t-test are the same test: F equals t squared, and the p-values match exactly. ANOVA's real value appears with three or more groups. It also extends naturally to more complex designs — two-way ANOVA studies two factors at once, such as teaching method and class size, and can reveal whether their effects interact.

Effect Size

Eta squared (η²) shows how much of the total variation the groups account for. Rough benchmarks are 0.01 small, 0.06 medium and 0.14 large. The example's 0.81 is very large, which is typical of small, tidy textbook data; real studies often find much smaller effects that are still important.

Assumptions

One-way ANOVA assumes the observations are independent, the values in each group are roughly normally distributed, and the groups have similar variances. It is fairly robust to non-normality, especially with equal group sizes. If the variances are very different, Welch's ANOVA is safer; if the data is strongly skewed or ordinal, the Kruskal–Wallis test is a non-parametric alternative.

Understanding Your Result

The headline gives F with its degrees of freedom and the p-value.

The decision line says whether the means differ significantly at 5 percent.

The group means line shows each group's mean and size.

The ANOVA table line gives the sums of squares, degrees of freedom and mean squares.

The effect size line gives η² as a proportion and a percentage.

When Should You Use This Calculator?

Use it to compare three or more group averages at once.

Use it for experiments with several treatments or conditions.

Use it to compare sales, scores or measurements across regions or teams.

Use it for statistics coursework on analysis of variance.

Common Mistakes

Running many t-tests instead. The false-positive rate climbs quickly.

Stopping at a significant F. Use post-hoc tests to find which groups differ.

Using it for repeated measurements on the same subjects. That needs repeated- measures ANOVA.

Ignoring very unequal variances. Consider Welch's ANOVA.

Reading a non-significant result as proof the means are equal. Small samples can miss real differences.

Mixing up groups and lines. Put each group on its own line, or the groups will be merged.