Open Access
2010 Multiple hypothesis testing on composite nulls using constrained p-values
Zhiyi Chi
Electron. J. Statist. 4: 271-299 (2010). DOI: 10.1214/08-EJS318


Multiple hypothesis testing often encounters composite nulls and intractable alternative distributions. In this case, using p-values that are defined as maximum significance levels over all null distributions (“pmax”) often leads to very conservative testing. We propose constructing p-values via maximization under linear constraints imposed by data’s empirical distribution, and show that these p-values allow the false discovery rate (FDR) to be controlled with substantially more power than pmax.


Download Citation

Zhiyi Chi. "Multiple hypothesis testing on composite nulls using constrained p-values." Electron. J. Statist. 4 271 - 299, 2010.


Published: 2010
First available in Project Euclid: 22 February 2010

zbMATH: 1329.62210
MathSciNet: MR2645485
Digital Object Identifier: 10.1214/08-EJS318

Primary: 62G10 , 62H15
Secondary: 62G20

Keywords: composite null , DKW inequality , empirical process , FDR , multiple testing , p-value , stopping time

Rights: Copyright © 2010 The Institute of Mathematical Statistics and the Bernoulli Society

Back to Top