To report a one-way ANOVA in APA 7 format, write the result as F(df_between, df_within) = F-value, p = exact p-value, followed by an effect size such as η² or partial η² — for example, F(2, 42) = 4.31, p = .019, η² = .17. If you've just run your analysis and you're staring at an output table wondering which numbers to copy and how to italicise them, this is the exact formatting you need — down to the dropped leading zero on p.
Key Takeaways
- The core APA 7 ANOVA line is F(df1, df2) = value, p = value, effect size — italicise F and p, and drop the leading zero on p (write .019, not 0.019).
- Degrees of freedom go in parentheses: the first number is between-groups df (number of groups − 1), the second is within-groups df (total sample − number of groups).
- Always report an effect size (η² or partial η²) alongside p — APA 7 requires it, and p alone tells you nothing about magnitude.
- If your ANOVA is significant, report post hoc comparisons (e.g., Tukey HSD) with mean differences and confidence intervals.
- Report exact p values to three decimals (e.g., p = .019); only use p < .001 when the value is smaller than .001.
What does an APA 7 ANOVA result actually look like?
A complete APA 7 ANOVA sentence names the test, the variables, the F statistic with its two degrees of freedom, the exact p value, and an effect size. A full example in prose:
A one-way between-subjects ANOVA revealed a significant effect of study method on exam score, F(2, 42) = 4.31, p = .019, η² = .17.
Notice three formatting rules that trip people up: F and p are italicised, there's no leading zero before the decimal on p, and the degrees of freedom sit in parentheses immediately after F. These conventions are what reviewers and journals check first, so getting them right signals you know what you're doing.
What do the degrees of freedom in parentheses mean?
The two numbers in F(2, 42) are your between-groups and within-groups degrees of freedom. The first (2) equals the number of groups minus one — here, three study methods − 1 = 2. The second (42) equals your total sample minus the number of groups — here, 45 students − 3 groups = 42. These two values are not optional decoration: they let any reader reconstruct your design and verify the test was run correctly.
A worked example: reporting a one-way ANOVA step by step
Imagine you compared exam scores across three study methods — flashcards, re-reading, and practice testing — with 15 students in each group (N = 45).
Your statistics software gives you this output:
- Between-groups df = 2, within-groups df = 42
- F = 4.31
- p = .019
- η² = .17
- Group means: flashcards M = 72.4, SD = 8.1; re-reading M = 70.9, SD = 7.6; practice testing M = 79.8, SD = 7.9
Step 1 — Write the omnibus result. State what you tested and the overall F:
A one-way between-subjects ANOVA examined the effect of study method on exam score. There was a significant effect, F(2, 42) = 4.31, p = .019, η² = .17.
Step 2 — Interpret the effect size. An η² of .17 means study method explained 17% of the variance in exam scores — a large effect by Cohen's benchmarks (.01 = small, .06 = medium, .14 = large). Never report p without this context.
Step 3 — Report post hoc comparisons. A significant omnibus F tells you at least one group differs, not which. Follow up with a Tukey HSD:
Tukey HSD post hoc comparisons indicated that practice testing (M = 79.8, SD = 7.9) produced significantly higher scores than re-reading (M = 70.9, SD = 7.6), p = .018, 95% CI [1.4, 16.4]. Flashcards did not differ significantly from either group.
That confidence interval is what APA 7 increasingly expects — it shows the plausible range of the true mean difference, not just whether it cleared the .05 line.
How do I report a non-significant ANOVA in APA?
Report it the same way, with the real numbers — never write "the ANOVA was not significant" and stop. For example:
A one-way ANOVA found no significant effect of study method on exam score, F(2, 42) = 1.08, p = .349, η² = .05.
Reporting the exact F, p, and effect size for null results prevents the file-drawer problem and lets meta-analysts use your data. A "non-significant" result with η² = .05 may still be a meaningful medium effect your sample was underpowered to detect.
How do I report a two-way (factorial) ANOVA?
A two-way ANOVA has three F statistics to report: two main effects and one interaction. Report each on its own line with its own df and effect size:
A 2 × 3 between-subjects ANOVA revealed a significant main effect of study method, F(2, 84) = 5.22, p = .007, partial η² = .11, and a significant main effect of sleep condition, F(1, 84) = 9.41, p = .003, partial η² = .10. The interaction was not significant, F(2, 84) = 1.33, p = .270, partial η² = .03.
Use partial η² rather than η² for factorial designs, because it accounts for variance attributable to other factors — most software (and StatRyx) reports partial η² by default for multi-factor models.
APA 7 ANOVA reporting at a glance
| Element | Format | Example |
|---|---|---|
| Test statistic | Italic F | F |
| Degrees of freedom | (between, within) in parentheses | (2, 42) |
| p value | Italic, no leading zero, 3 decimals | p = .019 |
| Very small p | Use the threshold | p < .001 |
| Effect size (one-way) | η² (eta squared) | η² = .17 |
| Effect size (factorial) | partial η² | partial η² = .11 |
| Post hoc | Mean difference + 95% CI | 95% CI [1.4, 16.4] |
| Means | M and SD, italicised | M = 79.8, SD = 7.9 |
Can a tool format my ANOVA in APA 7 automatically?
Yes — this is exactly what StatRyx was built for. StatRyx runs the ANOVA, checks assumptions (normality and homogeneity of variance), calculates η² or partial η², runs the correct post hoc test, and outputs a copy-paste APA 7 sentence with the italics and dropped leading zeros already applied. Compared with running the analysis in SPSS — which costs roughly $100+ per month for a subscription and leaves you to format the write-up yourself — StatRyx hands you the finished reporting line. It's a free, AI-powered alternative to SPSS, R, and JASP that's aimed squarely at researchers who want the test chosen and the write-up done correctly.
If you're not sure a one-way ANOVA is even the right test for your data, see our companion guide on choosing between ANOVA, the t-test, and the Kruskal–Wallis test.
Stop calculating this by hand — run it free in StatRyx → Try StatRyx