August 2023 Inference via randomized test statistics
Nikita Puchkin, Vladimir Ulyanov
Author Affiliations +
Ann. Inst. H. Poincaré Probab. Statist. 59(3): 1508-1529 (August 2023). DOI: 10.1214/22-AIHP1299

Abstract

We show that external randomization may enforce the convergence of test statistics to their limiting distributions in particular cases. This results in a sharper inference. Our approach is based on a central limit theorem for weighted sums. We apply our method to a family of rank-based test statistics and a family of phi-divergence test statistics and prove that, with overwhelming probability with respect to the external randomization, the randomized statistics converge at the rate O(1/n) (up to some logarithmic factors) to the limiting chi-square distribution in Kolmogorov metric.

Nous montrons que l’ajout de randomisation externe peut guider la convergence des statistiques de test vers leurs lois limites dans certains cas particuliers. Il en résulte une inférence plus précise. Notre approche est basée sur un théorème central limite pour les sommes pondérées. Nous appliquons notre méthode à une famille de statistiques de test basées sur les rangs et à une famille de statistiques de test de phi-divergence et nous prouvons que, avec grande probabilité par rapport à la randomisation externe, les statistiques randomisées convergent à un taux O(1/n) (en négligeant certains facteurs logarithmiques) vers la distribution limite du chi-deux dans la distance de Kolomogorov.

Acknowledgements

The article was prepared within the framework of the HSE University Basic Research Program. The first author is a Young Russian Mathematics award winner and would like to thank its sponsors and jury. The second author carried out the research within the scope of the Moscow Center for Fundamental and Applied Mathematics at Moscow State University. The authors are grateful to the associate editor and two anonymous referees for the careful reading of the paper and for their valuable and constructive remarks that improved the quality of this work.

Citation

Download Citation

Nikita Puchkin. Vladimir Ulyanov. "Inference via randomized test statistics." Ann. Inst. H. Poincaré Probab. Statist. 59 (3) 1508 - 1529, August 2023. https://doi.org/10.1214/22-AIHP1299

Information

Received: 16 December 2021; Revised: 7 April 2022; Accepted: 3 July 2022; Published: August 2023
First available in Project Euclid: 31 August 2023

MathSciNet: MR4635717
Digital Object Identifier: 10.1214/22-AIHP1299

Subjects:
Primary: 62E20 , 62H10

Keywords: central limit theorem , Phi-divergence test statistics , Power divergence test statistics , Rank-based test statistics , weighted sums

Rights: Copyright © 2023 Association des Publications de l’Institut Henri Poincaré

JOURNAL ARTICLE
22 PAGES

This article is only available to subscribers.
It is not available for individual sale.
+ SAVE TO MY LIBRARY

Vol.59 • No. 3 • August 2023
Back to Top