The Mann-Whitney U test is the non-parametric alternative to the independent-samples t-test: it compares two independent groups when the data are ordinal or the normality assumption fails. Because it works on ranks, the APA 7 write-up looks different from a t-test, and people often report the wrong summary statistics. This guide gives you the exact format, a worked example you can copy, and the mistakes reviewers catch.
What you need before you write a single word
An APA 7 write-up of a Mann-Whitney U test needs these values from your output:
- The median (Mdn) for each group. Report medians, not means, because the test is about ranks.
- The U statistic.
- For larger samples, the z approximation, which lets you compute an effect size.
- The exact p value.
- An effect size: the rank-biserial correlation, or r = z divided by the square root of N.
Report the group sizes (n for each group) as well.
The APA 7 format template
Report the result in running text using this pattern:
A Mann-Whitney U test showed that [dependent variable] was significantly higher/lower in [group 1] (Mdn = XX) than in [group 2] (Mdn = XX), U = XXX, z = X.XX, p = .XXX, r = .XX.
All 53 templates, ready to paste with the italics set: the Rescue Vault, $27 →
Formatting rules reviewers actually check:
- Italicize Mdn, U, z, p, and r.
- Report medians for each group, not means and standard deviations.
- No leading zero on p or on r.
- Round U to a whole number when it is a count of comparisons; round z and r to two decimals.
- If it is below .001, write p < .001.
A worked example
Suppose you compared pain ratings (0 to 10, ordinal) between a treatment group and a control group.
- Treatment (n = 20): Mdn = 3
- Control (n = 22): Mdn = 5
- Result: U = 128.50, z = -2.61, p = .009, r = .40
Written up in APA 7, that becomes:
A Mann-Whitney U test showed that pain ratings were significantly lower in the treatment group (Mdn = 3) than in the control group (Mdn = 5), U = 128.50, z = -2.61, p = .009, r = .40.
The sentence carries the medians, the U statistic, the z approximation, the significance, and the effect size.
Now do it with your numbers
Paste the table straight out of your SPSS output window. You’ll get the APA 7 sentence, the effect size with its confidence interval, and a check that your p-value actually matches your test statistic.
No account. Your dataset never leaves your computer — only the summary numbers you paste are sent. If we can’t read your table we keep a copy of it so we can add support for it; nothing we keep identifies you.
The APA 7 table (optional)
For a results chapter, a small table keeps the summary out of the prose. APA tables use horizontal rules only:
| Group | n | Mdn |
|---|---|---|
| Treatment | 20 | 3 |
| Control | 22 | 5 |
Note. Pain rated on a 0 to 10 scale. The difference was significant, U = 128.50, z = -2.61, p = .009, r = .40.
Reporting a non-significant result
Use the same statistics; change the wording, and do not call it a trend:
A Mann-Whitney U test found no significant difference in pain ratings between the treatment group (Mdn = 4) and the control group (Mdn = 5), U = 175.50, z = −1.12, p = .263, r = .17.
Where to find these numbers in SPSS
- Ranks: N and mean rank per group (report medians from Explore or Descriptives instead of mean ranks).
- Test Statistics: Mann-Whitney U, Z and Asymp. Sig. (2-tailed) (or the Exact Sig. for small samples). Compute r as z/√N yourself.
Mistakes reviewers catch
- Reporting means and SDs. The Mann-Whitney test ranks the data, so report medians, not means.
- No effect size. Report r (z divided by the square root of the total N), or the rank-biserial correlation. Roughly, .10 is small, .30 is medium, .50 is large.
- Calling it a test of medians. Strictly, it tests whether one group tends to have higher ranks; it only becomes a clean test of medians when the two distributions have the same shape. Describe it as a difference in ranks or distributions.
- Using it on paired data. For two related measurements use the Wilcoxon signed-rank test, not Mann-Whitney.
- Writing p = .000. Report p < .001.
- A leading zero on p or r. APA drops it.
When to use it instead of a t-test
Use the Mann-Whitney U test when:
- The dependent variable is ordinal (ranks, Likert-type ratings), or
- The dependent variable is continuous but badly non-normal, especially in small samples, or
- There are extreme outliers that would distort a mean.
If the data meet the assumptions of an independent t-test, that test is more powerful and is usually preferred. See our guide on reporting an independent t-test.
Let KyroStat do the write-up for you
Formatting a rank-based test, with medians and the right effect size, is where errors creep in. KyroStat runs the Mann-Whitney U test on your data, reports the medians and the z approximation, computes the effect size, and hands you the finished APA 7 sentence and the underlying Python or R code. Upload your spreadsheet, and the report is done in seconds.
Frequently asked questions
Should I report means or medians for a Mann-Whitney test? Medians. The test works on ranks, so medians describe the groups appropriately.
What effect size goes with a Mann-Whitney U test? The rank-biserial correlation, or r = z divided by the square root of the total sample size. Report it without a leading zero.
Is the Mann-Whitney test the same as the Wilcoxon rank-sum test? Yes, they are equivalent. Wilcoxon rank-sum and Mann-Whitney U give the same p value. The Wilcoxon signed-rank test is a different, paired test.
My p value shows as .000. What do I write? Report p < .001. A p value is never exactly zero.
Which related guides might I need? Often these: Welch's t-test, descriptive statistics and p values.