Mann-Whitney U Test Calculator

Compare two independent samples with ranks, exact or asymptotic p-values, and effect sizes.

At a glance

Computes
Mann-Whitney U, an exact or asymptotic p-value, and rank-based effect sizes.
You supply
Two independent groups of values, an alternative, and a significance level.
Use when
Two independent groups, ordinal or skewed data, and you cannot assume normal means.
Not for
Paired or repeated measurements on the same subjects.
Alternative
P-value method

Mann-Whitney U

Enter samples

Auto chooses exact when ties allow.

Test details
MetricValue
Rank table
ValueGroupRank

Density by Mann-Whitney U; 5 bins of width 1 (fd rule); P-value tail; Observed U 0

0.000.050.100.150.200.250.3001234
Export

U1 = R1 - n1(n1 + 1) / 2 How?

How this is calculated

The calculator ranks all observations together, assigns average ranks to ties, and computes U for group 1. Exact mode uses the no-tie U distribution. Asymptotic mode uses the normal approximation with tie correction and a continuity correction.

Formula: U1 = R1 - n1(n1 + 1) / 2

Sources

  1. SciPy mannwhitneyu. SciPy. Retrieved .
  2. Mann and Whitney 1947 paper. Project Euclid. Retrieved .

Method last reviewed