Your ANOVA table
| Source | SS | df |
|---|---|---|
| Between groups | 27.8973 | 2 |
| Within groups | 17.452 | 12 |
| Total | 45.3493 | 14 |
| Group | Values | Mean |
|---|---|---|
| Group 1 | 5 | 5.34 |
| Group 2 | 5 | 7.72 |
| Group 3 | 5 | 8.56 |
What is one-way analysis of variance (ANOVA)?
One-way analysis of variance (ANOVA) checks whether three or more groups have the same mean, using one test instead of many. It compares two estimates of the spread: how far the group means are from each other (between groups) and how far the values are from their own group’s mean (within groups).
If the group means are really equal, the two estimates should be about the same and F, their ratio, should be close to 1. A large F, with a p-value below your significance level, means the groups probably don’t all share one mean. The test is always right-tailed.
Source OpenStax · Introductory Statistics 2e, 13.1 One-Way ANOVA ↗OpenStax · Introductory Statistics 2e, 13.2 The F Distribution and the F-Ratio ↗
How to calculate one-way ANOVA
Here it is with your numbers:
- There are 3 groups and 15 values, so df between is 3 − 1 and df within is 15 − 3.
- Overall mean x̄ = 7.2067
- SS between = Σ n × (group mean − x̄)²
- SS between = 27.8973
- SS within = Σ (value − its group mean)²
- SS within = 17.452
- MS between = 27.8973 ÷ 2
- MS between = 13.948667
- MS within = 17.452 ÷ 12
- MS within = 1.454333
- F = 13.948667 ÷ 1.454333
- F = 9.5911
- p = 0.0032, the area above 9.5911 under the F distribution with 2 and 12 degrees of freedom
It means at least one group’s mean is different from the others. ANOVA doesn’t say which one: compare the group means below.
Source NIST/SEMATECH e-Handbook of Statistical Methods · 7.4.3.3 The ANOVA table and tests of hypotheses about means ↗NIST/SEMATECH e-Handbook of Statistical Methods · 7.4.3.4 One-way ANOVA calculations ↗
ANOVA examples
| Data | F | p-value |
|---|---|---|
| Resistors at 3 temperatures (NIST) | 9.5911 | 0.0032 |
| Weight loss on 3 diet plans | 0.3769 | 0.6991 |
| Tomato yield under 5 mulches | 4.481 | 0.0248 |
| Mean grades in 4 sororities | 2.2303 | 0.1241 |
The resistor readings are NIST’s example; the diet plans, tomato yields and sorority grades are OpenStax’s.
Source OpenStax · Introductory Statistics 2e, 13.3 Facts About the F Distribution ↗
FAQs
What is a one-way ANOVA test used for?
It tests whether the means of several groups are all equal, for example the yields of plants grown under 5 different mulches. A small p-value says at least one group mean is different. It assumes each group comes from a normal population, the groups are independent, and they have equal variances.
Source OpenStax · Introductory Statistics 2e, 13.1 One-Way ANOVA ↗
Why use an ANOVA instead of a t-test?
With more than two groups, running a t test for every pair raises the chance of a false positive (a Type I error); one ANOVA tests them all at once. With exactly two groups, the ANOVA F is the square of the pooled two-sample t, and the p-values match.
Source OpenStax · Introductory Statistics 2e, 13.2 The F Distribution and the F-Ratio ↗
What is the formula for calculating ANOVA?
F = MS between ÷ MS within, where MS between = SS between ÷ (k − 1) and MS within = SS within ÷ (N − k), for k groups and N values in all. For NIST’s resistor data, MS between is 13.9487 and MS within is 1.4543, so F is 9.59.
How to do ANOVA on a calculator?
On a TI-83 or TI-84, put each group in a list (STAT, EDIT: L1, L2, L3 …). Then press STAT, arrow to TESTS, choose ANOVA(, enter L1, L2, L3 and press ENTER. The screen shows F, the p-value and the factor and error rows of the table.
Source OpenStax · Introductory Statistics 2e, 13.3 Facts About the F Distribution ↗
Sources
- NIST/SEMATECH e-Handbook of Statistical Methods: 7.4.3.3 The ANOVA table and tests of hypotheses about means
- NIST/SEMATECH e-Handbook of Statistical Methods: 7.4.3.4 One-way ANOVA calculations
- OpenStax: Introductory Statistics 2e, 13.1 One-Way ANOVA
- OpenStax: Introductory Statistics 2e, 13.2 The F Distribution and the F-Ratio
- OpenStax: Introductory Statistics 2e, 13.3 Facts About the F Distribution