p-value peeking and estimating extrema
Statistics Theory
2020-11-04 v1 Methodology
Machine Learning
Statistics Theory
Abstract
A pervasive issue in statistical hypothesis testing is that the reported -values are biased downward by data "peeking" -- the practice of reporting only progressively extreme values of the test statistic as more data samples are collected. We develop principled mechanisms to estimate such running extrema of test statistics, which directly address the effect of peeking in some general scenarios.
Cite
@article{arxiv.2011.01343,
title = {p-value peeking and estimating extrema},
author = {Akshay Balsubramani},
journal= {arXiv preprint arXiv:2011.01343},
year = {2020}
}