Permutation Test Calculator
Exact two-sided permutation test on the difference of two group means; all C(nA+nB, nA) group assignments of the pooled sample (each sample capped at 8 values) are enumerated deterministically.
Description
Run an exact two-sided permutation test for the difference between two sample means.
A permutation test evaluates a group difference by comparing the observed statistic with every reassignment of pooled observations to groups of the original sizes. This tool is exact and deterministic for small samples.
When to use Exact Permutation Test Calculator
- Test a difference in means without a normal-theory reference distribution
- Verify a small permutation-test example
- Inspect the number of possible label assignments
How the calculation works
Every combination of nA positions from the pooled data defines one reassignment. The p value is the fraction whose absolute mean difference is at least as large as the observed absolute difference.
Interpreting the result
`observedDifference` retains the sign mean(A) − mean(B), while the p value is two-sided. A p value describes extremeness under exchangeable labels; it does not measure effect size or the probability that the null is true.
Important limitations
- Each sample is capped at eight values because assignment counts grow combinatorially.
- The exact exchangeability assumption must be defensible.
- The test targets means and can be sensitive to outliers.