🔢 Mann-Whitney U Test Calculator
Paste two groups to run a Mann-Whitney U test — the rank-based alternative to the t-test that doesn't assume normal data.
U statistic
0
z (normal approx.)
-2.802
p-value
0.0051
U₁ = 0 · U₂ = 36 · n₁ = 6, n₂ = 6
Distributions differ significantly at α = 0.05
p = 0.0051 < 0.05, so you reject the null hypothesis that the two groups come from the same distribution.
The Mann-Whitney U test (Wilcoxon rank-sum) is the non-parametric alternative to the two-sample t-test: it compares two independent groups by ranking all values together, so it doesn't assume the data are normally distributed — useful for ordinal data, skewed data or outliers. This tool uses the normal approximation with a tie correction, which is accurate for moderate-to-large samples; for very small samples an exact table is more precise. 🔒 Computed in your browser.
How the mann-whitney u test calculator works
The Mann-Whitney U test ranks all values from both groups together, sums the ranks of each group, and computes the U statistic from those rank sums. The tool uses the normal approximation with a tie correction to get a z score and p-value, and handles tied values with average ranks. Enter each group's numbers separated by commas or spaces.
This is the go-to test when a t-test's normality assumption is doubtful — small samples, skewed distributions, ordinal ratings, or data with outliers — because it works on ranks rather than raw values. Strictly it tests whether one group tends to have larger values than the other (stochastic dominance); it's often described as comparing medians, which holds when the two distributions have similar shapes. The normal approximation is accurate for moderate-to-large samples; for very small groups an exact U table is more precise. It all runs locally in your browser.
Frequently asked questions
When should I use a Mann-Whitney U test?
When comparing two independent groups but the data isn't normally distributed — for example skewed measurements, ordinal ratings, small samples, or data with outliers. It's the non-parametric alternative to the two-sample t-test.
Is Mann-Whitney the same as the Wilcoxon rank-sum test?
Yes — they are equivalent tests with different formulations and names. (Don't confuse it with the Wilcoxon signed-rank test, which is for paired data.)
Does it compare medians?
It tests whether values in one group tend to be larger than in the other. When the two distributions have a similar shape, that difference in "typical" value can be read as a difference in medians; when shapes differ, interpret it as stochastic dominance rather than strictly medians.
How are ties handled?
Tied values receive the average of the ranks they span, and the variance used for the z score includes a tie correction, so repeated values are handled correctly.
Is my data uploaded?
No — the test runs entirely in your browser and works offline, so your data stays on your device.