Beta 6 months of Pro free for the first 500. Join the waitlist
Contact sales Join the waitlist

Guide

How to report a one-way ANOVA in APA 7

What to include, how to format it, and example sentences you can adapt, for a standard ANOVA and for Welch's test when variances differ.

Join the waitlist All guides

What to report

A one-way ANOVA tests whether the means of three or more groups differ. In APA 7 style, report:

  • The descriptive statistics for each group: the mean (M) and standard deviation (SD), and the group sizes, in the text or a table.
  • The test statistic: F with its two degrees of freedom in parentheses, between groups first, then within groups.
  • The exact p value to two or three decimal places, or p < .001 when it is smaller than that.
  • An effect size, usually eta squared (η²) or omega squared (ω²), ideally with a confidence interval.
  • Post hoc comparisons when the overall test is significant, naming the test you used.
  • How you checked the assumptions: normality of residuals and equality of variances, and what you did if one failed.

Formatting rules

  • Italicise statistical symbols: F, p, M, SD, n. Greek letters such as η are not italicised.
  • Leave out the leading zero for values that can't be greater than 1, such as p = .03 and η² = .14, but keep it for others, such as M = 0.85.
  • Put spaces around the equals sign: F(3, 140) = 125.81, not F(3,140)=125.81.
  • Round to two decimal places for most statistics; give p to two or three.

Example: a standard one-way ANOVA

Kinase activity was compared across four treatment groups (vehicle, 1 µM, 5 µM and 10 µM), with 36 samples in each.

Example write-up: A one-way ANOVA showed that kinase activity differed across treatments, F(3, 140) = 125.81, p < .001, η² = .73. Tukey's HSD post hoc tests showed that [name the groups that differed, with their p values].

When variances are unequal: Welch's ANOVA

If Levene's test or a similar check shows the group variances differ, report Welch's F instead. Its denominator degrees of freedom are adjusted and usually not a whole number, so report them to two decimals, and use the Games-Howell test for post hoc comparisons.

Example write-up: Because variances differed across groups, Welch's ANOVA was used. Kinase activity differed across treatments, Welch's F(3, 73.12) = 181.62, p < .001. Games-Howell comparisons showed that [name the groups that differed].

Doing this in OrcStat

OrcStat runs the assumption checks for you and switches to Welch's ANOVA and Games-Howell when variances differ. With the APA 7 table look, its tables are ready for Word, and the report writer drafts a results paragraph like the examples above, with every number checked against the output. See APA 7 tables and results.