5 Reasons Reviewers Send Back Your Stats Section (and How to Fix Each One)

Reviewer comment #2 always looks the same: "Please justify your choice of statistical test and report effect sizes / assumption checks." Here are the 5 most common reasons a stats section gets bounced back, and the one-line fix for each:

1. No effect size reported — a p-value alone doesn't tell a reader how big the effect is. Add Cohen's d (t-tests), eta-squared (ANOVA), or r (correlation) next to every significance test.
2. No assumption check — reviewers now expect a one-sentence note on normality/homogeneity even when the test is robust to violations. State it, don't skip it.
3. Multiple comparisons left uncorrected — running 5 pairwise t-tests without Bonferroni/Tukey inflates your Type I error and reviewers know it.
4. Wrong test for the data structure — using independent t-test on paired/repeated data (or vice versa) is the single most common silent error in submitted manuscripts.
5. Results reported inconsistently with APA 7 — df, exact p, and test statistic formatting mismatches across the same paper signal a rushed writeup.

Each of these takes under a minute to check once you have your test statistic and sample size — I built free calculators for all 20 common tests (with the effect size and assumption check included in the output) at statmate.org if you want to double check before you resubmit.