Restricted Block Permutation for Two-Sample Testing

TL;DR

This paper introduces a structured permutation scheme for two-sample testing that improves statistical power and validity by restricting permutations to block-based cross-swaps, with explicit formulas and variance analysis.

Contribution

It develops a new block-restricted permutation method for two-sample testing, providing exact validity, variance reduction, and explicit critical value formulas, enhancing power over classical methods.

Findings

Block-restricted permutations have lower variance than full relabeling.

The method achieves higher statistical power while maintaining exact validity.

Explicit data-dependent critical values are derived for the tests.

Abstract

We study a structured permutation scheme for two-sample testing that restricts permutations to single cross-swaps between block-selected representatives. Our analysis yields three main results. First, we provide an exact validity construction that applies to any fixed restricted permutation set. Second, for both the difference of sample means and the unbiased $\widehat{\mathrm{MMD}}^{2}$ estimator, we derive closed-form one-swap increment identities whose conditional variances scale as $O(h^{2})$, in contrast to the $\Theta(h)$ increment variability under full relabeling. This increment-level variance contraction sharpens the Bernstein--Freedman variance proxy and leads to substantially smaller permutation critical values. Third, we obtain explicit, data-dependent expressions for the resulting critical values and statistical power. Together, these results show that block-restricted one-swap permutations can achieve strictly higher power than classical full permutation tests while maintaining exact finite-sample validity, without relying on pessimistic worst-case Lipschitz bounds.