You have run your ANOVA, the output is staring back at you, and now comes the part many students dread: writing it up. Reporting ANOVA results in APA style is not difficult once you know the formula. It is the same few pieces of information, in the same order, every single time. This guide walks you through that formula step by step, with a worked example you can copy the structure of, plus the mistakes examiners notice most.
What ANOVA Results Actually Tell You
An analysis of variance (ANOVA) tests whether the means of three or more groups differ from each other more than you would expect by chance. The F statistic compares the variation between the groups with the variation within the groups. A large F value means the groups differ more than random noise alone would explain. The p value then tells you whether that difference is statistically significant, and the effect size tells you how big the difference actually is. A good APA report gives the reader all three: the test statistic, the significance level, and the effect size, plus the descriptive statistics behind them.
Before You Write: Check These First
APA style assumes the test you ran was appropriate, so check the assumptions before you report anything. Your dependent variable should be continuous, your observations independent, and the scores in each group roughly normally distributed with similar variances. A quick check of the Shapiro-Wilk test and Levene’s test in your output covers the last two. If an assumption is seriously violated, say so honestly and report the alternative you used, such as Welch’s ANOVA or a Kruskal-Wallis test, rather than quietly reporting a standard ANOVA anyway. Examiners respect a transparent limitation far more than a hidden one.
The APA Format for a One-Way ANOVA
Here is the template. Every number in it has a job, and the order never changes:
F(df between groups, df within groups) = F value, p = p value, effect size = value.
A complete example looks like this:
A one-way ANOVA revealed a significant effect of study technique on test scores, F(2, 57) = 5.84, p = .005, η² = .17.
Notice the details that matter in APA 7th edition. The letters F and p are italicized, as is the Greek letter eta. There is no zero before the decimal point for p values and effect sizes, because they can never exceed 1, so you write p = .005, not p = 0.005. Report the exact p value to two or three decimal places whenever your software gives it, and write p < .001 only when the value is smaller than .001. Degrees of freedom go inside the parentheses, with the between-groups degrees of freedom first.
Step-by-Step: Reporting Your ANOVA
Step 1: Report the descriptive statistics
Before the F statistic, give the reader the group means and standard deviations, either in a sentence or in a small table. Means and standard deviations are written as M = 78.40, SD = 8.20. Report means to two decimal places, and always pair a mean with its standard deviation so the reader can judge the spread of scores. If you have more than three groups, a table is cleaner than a long sentence.
Step 2: Report the F statistic
Write F followed by the two degrees of freedom in parentheses, then the value to two decimal places. The first degrees of freedom is the number of groups minus one; the second is the total sample size minus the number of groups. So with three groups and 60 participants, you get F(2, 57).
Step 3: Report the exact p value
Report the exact p value from your output, such as p = .005. Use p < .001 only for extremely small values, and never write p = .000, because a p value is never exactly zero. If your result is not significant, write the exact value anyway, for example p = .214, rather than the vague “p > .05”.
Step 4: Report the effect size
APA 7th edition expects an effect size alongside every test. For a one-way ANOVA, eta squared (η²) is the simplest choice: it tells you what proportion of the total variation in scores is explained by group membership. As a rough guide, .01 is a small effect, .06 is a medium effect, and .14 is a large effect. If you ran a factorial or repeated-measures ANOVA, report partial eta squared (ηp²) instead, and label it clearly so the reader knows which one you used.
Step 5: Report post hoc tests (if needed)
A significant ANOVA tells you that at least one group differs, but not which ones. That is what post hoc tests are for. Name the test you used, usually Tukey’s HSD for comparing all pairs of groups, and report the significant pairwise differences with their p values. If no pairwise difference reached significance, or if you used a planned contrast instead, say that plainly.
Step 6: Put it together in a sentence
The results paragraph usually reads like this: first the descriptive statistics, then the ANOVA result, then the post hoc findings. One clean paragraph is almost always enough. If you add a table of means and standard deviations, the paragraph can stay short, because the table carries the detail.
Worked Example
Imagine a student compares three study techniques: spaced practice, massed practice, and rereading. Twenty students use each technique, then all take the same test out of 100. The descriptive statistics are:
| Study technique | n | Mean | SD |
|---|---|---|---|
| Spaced practice | 20 | 78.40 | 8.20 |
| Massed practice | 20 | 71.60 | 9.10 |
| Rereading | 20 | 69.80 | 8.70 |
Here is how the results paragraph could read:
Mean test scores differed across the three study techniques. Students using spaced practice (M = 78.40, SD = 8.20) scored higher than those using massed practice (M = 71.60, SD = 9.10) and rereading (M = 69.80, SD = 8.70). A one-way ANOVA confirmed a significant effect of study technique on test scores, F(2, 57) = 5.84, p = .005, η² = .17, a large effect. Tukey’s HSD post hoc tests showed that spaced practice scores were significantly higher than rereading scores (p = .004), while the difference between spaced practice and massed practice was not significant (p = .089), nor was the difference between massed practice and rereading (p = .761).
That paragraph gives the reader everything: the descriptives, the test, the significance, the size of the effect, and exactly which groups differ. Notice that non-significant post hoc results are reported too. Leaving them out would give a misleading picture.
Reporting Two-Way and Repeated-Measures ANOVA
The same formula applies, with one extra step. In a two-way ANOVA, report three F tests: one for each main effect and one for the interaction, each with its own degrees of freedom, p value, and partial eta squared. For example: “There was a significant main effect of teaching method, F(1, 76) = 8.42, p = .005, ηp² = .10, and a significant interaction between teaching method and gender, F(1, 76) = 4.91, p = .030, ηp² = .06.” In a repeated-measures ANOVA, state that the test was repeated-measures, and if sphericity was violated, report the Greenhouse-Geisser corrected degrees of freedom and say that you did so.
Common Mistakes to Avoid
First, reporting only the p value. A p value without an effect size tells the reader nothing about whether the difference matters in practice. Second, writing p = .000 instead of p < .001. Third, forgetting the degrees of freedom or swapping their order. Fourth, reporting a significant ANOVA and then never saying which groups differ. Fifth, rounding means to whole numbers while reporting the F value to two decimal places, which looks careless. And finally, describing a non-significant result as “no difference” rather than “no statistically significant difference”. The first claims something the test cannot prove; the second states exactly what was found.
Frequently Asked Questions
Do I report the F value if my ANOVA is not significant?
Yes. Report the full result, F(df1, df2) = value, p = value, and the effect size, exactly as you would for a significant result. Non-significant findings are still findings, and hiding them distorts the literature.
Should I use eta squared or partial eta squared?
For a one-way ANOVA, eta squared is standard. For factorial or repeated-measures designs, use partial eta squared, because it is what most software reports by default and what readers of those designs expect. Whichever you choose, label it clearly.
Where do post hoc results go?
In the same paragraph, directly after the ANOVA result, or in a short follow-up sentence. If you have many pairwise comparisons, a compact table of mean differences and p values is easier to read than a long sentence.
Do I need a table for a one-way ANOVA?
Not always, but a small table of group means and standard deviations is good practice, especially in a thesis. It lets the reader check your claims at a glance and makes your results section look professional.
Quick Checklist Before You Submit
Run through this list for every ANOVA you report: group means and standard deviations given, F statistic with both degrees of freedom, exact p value, effect size with its label, post hoc test named with significant pairs reported, non-significant results reported honestly, and F, p, and eta italicized with no leading zero on p values and effect sizes. If every box is ticked, your ANOVA reporting is APA-ready.