Detecting p-hacking
Econometrics
2022-05-13 v5 General Economics
Economics
Abstract
We theoretically analyze the problem of testing for -hacking based on distributions of -values across multiple studies. We provide general results for when such distributions have testable restrictions (are non-increasing) under the null of no -hacking. We find novel additional testable restrictions for -values based on -tests. Specifically, the shape of the power functions results in both complete monotonicity as well as bounds on the distribution of -values. These testable restrictions result in more powerful tests for the null hypothesis of no -hacking. When there is also publication bias, our tests are joint tests for -hacking and publication bias. A reanalysis of two prominent datasets shows the usefulness of our new tests.
Keywords
Cite
@article{arxiv.1906.06711,
title = {Detecting p-hacking},
author = {Graham Elliott and Nikolay Kudrin and Kaspar Wuthrich},
journal= {arXiv preprint arXiv:1906.06711},
year = {2022}
}