Group permutation tests control Type I error at most 2α (sharp) for linear model coefficients, improve power over i.i.d. permutations via design-adaptive group choice, and extend to non-exchangeable errors with TV-distance degradation bounds.
Now we explain how different distribution ofZas well asn, pinfluences the gap||J 2|(¯b2 − ¯b)(¯c2 −¯c) + |J3|(¯b3 − ¯b)(¯c3 −¯c)|ofλ 2 in (22)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Group Permutation Testing in Linear Model: Sharp Validity, Power Improvement, and Extension Beyond Exchangeability
Group permutation tests control Type I error at most 2α (sharp) for linear model coefficients, improve power over i.i.d. permutations via design-adaptive group choice, and extend to non-exchangeable errors with TV-distance degradation bounds.