Sharp Variance Estimator and Causal Bootstrap in Stratified Randomized Experiments
Statistics in Medicine · 2025
DOI: 10.1002/sim.70139 ↗Kurz gefasst
Fachartikel in der Zeitschrift Statistics in Medicine, 2025. Verfasst von Haoyang Yu und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Statistics and Data Science, Tsinghua University, Beijing, China. Hauptthemen laut MEDLINE: Randomized Controlled Trials as Topic, Models, Statistical. Die Arbeit nennt 55 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Beijing Natural Science Foundation, Natural Science Foundation of Beijing Municipality.
Abstract (Originalsprache)
Randomized experiments are the gold standard for estimating treatment effects, and randomization serves as a reasoned basis for inference. In widely used stratified randomized experiments, randomization‐based finite‐population asymptotic theory enables valid inference for the average treatment effect, relying on normal approximation and a Neyman‐type conservative variance estimator. However, when the sample size is small or the outcomes are skewed, the Neyman‐type variance estimator may become overly conservative, and the normal approximation can fail. To address these issues, we propose a sharp variance estimator and two causal bootstrap methods to more accurately approximate the sampling distribution of the weighted difference‐in‐means estimator in stratified randomized experiments. The first causal bootstrap procedure is based on rank‐preserving imputation, and we prove its second‐order refinement over normal approximation. The second causal bootstrap procedure is based on constant‐treatment‐effect imputation and is further applicable in paired experiments. In contrast to traditional bootstrap methods, where randomness originates from hypothetical super‐population sampling, our analysis for the proposed causal bootstrap is randomization‐based, relying solely on the randomness of treatment assignment in randomized experiments. Numerical studies and two real data applications demonstrate the advantages of our proposed methods in finite samples. The R package CausalBootstrap implementing our method is publicly available.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Statistics and Data Science, Tsinghua University, Beijing, China.
- Department of Statistics, North Carolina State University, Raleigh, North Carolina, USA.
- Department of Biostatistics and Bioinformatics, Duke University, Durham, North Carolina, USA.
ORCID-Profile: Ke Zhu, Hanzhong Liu
Förderung
- National Natural Science Foundation of China
- Beijing Natural Science Foundation
- Natural Science Foundation of Beijing Municipality
Zitieren
APA
Yu, H., Zhu, K., & Liu, H. (2025). Sharp Variance Estimator and Causal Bootstrap in Stratified Randomized Experiments. Statistics in Medicine. https://doi.org/10.1002/sim.70139
Vancouver
Yu H, Zhu K, Liu H. Sharp Variance Estimator and Causal Bootstrap in Stratified Randomized Experiments. Statistics in Medicine. 2025. doi:10.1002/sim.70139
BibTeX
@article{Yu2025,
author = {Yu, Haoyang and Zhu, Ke and Liu, Hanzhong},
title = {Sharp Variance Estimator and Causal Bootstrap in Stratified Randomized Experiments},
journal = {Statistics in Medicine},
year = {2025},
doi = {10.1002/sim.70139}
}
EUVIMED