PreprintCC BY-NC-SA 4.0arXiv
Response Propensity Estimation and Cross-Fitting
Abstract
This paper investigates whether five fold cross fitting improves nonresponse adjustment in survey estimation when flexible machine learning methods are used to estimate response propensities. We conduct a finite population Monte Carlo simulation with 90 experimental configurations and 2,000 replications per configuration, varying sample size, response rate, and the structure of the response mechanism. Logistic regression is used as a conventional parametric benchmark, while Random Forest and Gradient Boosting Machine are considered as flexible nonparametric models.
Related
- Orthogonal Inference for Conditional Z-Estimation without Cross-Fitting: A Distributional Nearest-Neighbor Approach
- Cross-Fitting for Neural Posterior Estimation
- Unbiased Treatment Effect Estimation under Network Interference via Neighborhood-Excluded Cross-Fitting
- Differentially Private Average Treatment Effect Estimation by Propensity Score Blocking
- A new quantile regression model for bounded responses: mathematical derivation, estimation, and fitting bounded datasets
BibTeX
@misc{rocca2026response,
title = {{Response Propensity Estimation and Cross-Fitting}},
author = {Alessandro La Rocca},
howpublished = {arXiv},
year = {2026},
month = aug,
doi = {10.48550/arxiv.2608.28324},
eprint = {2608.28324},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2608.28324},
note = {CC BY-NC-SA 4.0},
}
RIS
TY - UNPB
TI - Response Propensity Estimation and Cross-Fitting
AU - Alessandro La Rocca
PY - 2026
DA - 2026/08/28
PB - arXiv
DO - 10.48550/arxiv.2608.28324
UR - https://arxiv.org/abs/2608.28324
AB - This paper investigates whether five fold cross fitting improves nonresponse adjustment in survey estimation when flexible machine learning methods are used to estimate response propensities. We conduct a finite population Monte Carlo simulation with 90 experimental configurations and 2,000 replications per configuration, varying sample size, response rate, and the structure of the response mechanism. Logistic regression is used as a conventional parametric benchmark, while Random Forest and Gradient Boosting Machine are considered as flexible nonparametric models.
ER -