Supermartingales for One-Sided Tests: Sufficient Monotone Likelihood Ratios are Sufficient

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Grünwald, Peter D., Koolen, Wouter M.
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917916013756416
author Grünwald, Peter D.
Koolen, Wouter M.
author_facet Grünwald, Peter D.
Koolen, Wouter M.
contents The t-statistic is a widely-used scale-invariant statistic for testing the null hypothesis that the mean is zero. Martingale methods enable sequential testing with the t-statistic at every sample size, while controlling the probability of falsely rejecting the null. For one-sided sequential tests, which reject when the t-statistic is too positive, a natural question is whether they also control false rejection when the true mean is negative. We prove that this is the case using monotone likelihood ratios and sufficient statistics. We develop applications to the scale-invariant t-test, the location-invariant $χ^2$-test and sequential linear regression with nuisance covariates.
format Preprint
id arxiv_https___arxiv_org_abs_2502_04208
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Supermartingales for One-Sided Tests: Sufficient Monotone Likelihood Ratios are Sufficient
Grünwald, Peter D.
Koolen, Wouter M.
Statistics Theory
The t-statistic is a widely-used scale-invariant statistic for testing the null hypothesis that the mean is zero. Martingale methods enable sequential testing with the t-statistic at every sample size, while controlling the probability of falsely rejecting the null. For one-sided sequential tests, which reject when the t-statistic is too positive, a natural question is whether they also control false rejection when the true mean is negative. We prove that this is the case using monotone likelihood ratios and sufficient statistics. We develop applications to the scale-invariant t-test, the location-invariant $χ^2$-test and sequential linear regression with nuisance covariates.
title Supermartingales for One-Sided Tests: Sufficient Monotone Likelihood Ratios are Sufficient
topic Statistics Theory
url https://arxiv.org/abs/2502.04208