Rerandomization for covariate balance mitigates p-hacking in regression adjustment
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916717212467200 |
|---|---|
| author | Lu, Xin Ding, Peng |
| author_facet | Lu, Xin Ding, Peng |
| contents | Rerandomization enforces covariate balance across treatment groups in the design stage of experiments. Despite its intuitive appeal, its theoretical justification remains unsatisfying because its benefits of improving efficiency for estimating the average treatment effect diminish if we use regression adjustment in the analysis stage. To strengthen the theory of rerandomization, we show that it mitigates false discoveries resulting from $p$-hacking, the practice of strategically selecting covariates to get more significant $p$-values. Moreover, we show that rerandomization with a sufficiently stringent threshold can resolve $p$-hacking. As a byproduct, our theory offers guidance for choosing the threshold in rerandomization in practice. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_01137 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Rerandomization for covariate balance mitigates p-hacking in regression adjustment Lu, Xin Ding, Peng Statistics Theory Rerandomization enforces covariate balance across treatment groups in the design stage of experiments. Despite its intuitive appeal, its theoretical justification remains unsatisfying because its benefits of improving efficiency for estimating the average treatment effect diminish if we use regression adjustment in the analysis stage. To strengthen the theory of rerandomization, we show that it mitigates false discoveries resulting from $p$-hacking, the practice of strategically selecting covariates to get more significant $p$-values. Moreover, we show that rerandomization with a sufficiently stringent threshold can resolve $p$-hacking. As a byproduct, our theory offers guidance for choosing the threshold in rerandomization in practice. |
| title | Rerandomization for covariate balance mitigates p-hacking in regression adjustment |
| topic | Statistics Theory |
| url | https://arxiv.org/abs/2505.01137 |