How to report Mann-Whitney U in APA format

The basic form

A Mann-Whitney U test indicated that scores were significantly higher for the treatment group (Mdn = 24.0) than the control group (Mdn = 18.5), U = 142.5, z = -2.87, p = .004, r = .41.

What each element is

Medians, not means. Mann-Whitney is a rank-based test, so medians are the matching summary statistic. Report them for each group.

U is the test statistic. Report it to one decimal place.

z appears when a normal approximation is used, which most software does for larger samples. Include it if your output gives it.

p to three decimals, or p < .001 below that.

r is the effect size, calculated as z divided by the square root of the total sample size. Interpreted roughly as .1 small, .3 medium, .5 large. Most write-ups omit this, and reviewers increasingly ask for it.

Also report sample sizes

Give n for each group, either in the sentence or in an accompanying table. Rank-based statistics are hard to interpret without knowing how many observations produced them.

What the test actually tests

Not "a difference in medians", strictly. It tests whether one group's values tend to be larger than the other's. When the two distributions have a similar shape it can be interpreted as a difference in medians, which is the usual case and the usual wording. If the shapes differ markedly, phrase it as a difference in distributions.

The paired equivalent

For paired data use the Wilcoxon signed-rank test, reported the same way with T or W as the statistic.

In Plotara

Mann-Whitney, Wilcoxon signed-rank, Kruskal-Wallis and Friedman, with the Methods panel writing the APA sentence including the effect size.