Replication Folder for "Revisiting Bootstrap Inference in Propensity Score Matching"
Published: 18 September 2026| Version 1 | DOI: 10.17632/hpnm327z8g.1
Contributors:
, Bingyue Su, Description
This replication package reproduces the simulation results reported in the paper. The package studies the performance of a bootstrap subsample propensity score matching estimator of the average treatment effect on the treated (ATT) under a threshold assignment rule, relative to a naive difference-in-means baseline and full-sample propensity score matching. Performance is evaluated across a grid of two parameters, U-noise (σ_U) and epsilon noise (σ_ε).
Files
Steps to reproduce
See the README file.
Categories
Bootstrapping, Propensity Score, Applied Statistics, Causal Inference