A1328
Title: Where to weight: Estimating population causal effects with weighted double score matching in complex surveys
Authors: Guangyu Tong - Yale University (United States) [presenting]
Abstract: Double score matching is an emerging method for causal inference that combines propensity and prognostic scores to achieve dimension reduction and model robustness. However, when observational data are drawn from a complex survey, the existing literature does not provide theoretical guidance on the proper use of survey weights in matching for population-level causal inference. Weighted double score matching (WDSM) provides a unified framework for incorporating survey weights throughout the score estimation, matching, and outcome analysis under both retrospective (informative) and prospective (non-informative) sampling designs. Asymptotic normality is established under both known and estimated scores, and a linearization-based wild bootstrap procedure for variance estimation that preserves the survey-weighted matching structure is developed. An extensive series of simulation studies examines the performance of the estimators and compares the results to those of alternative methods. Two case studies illustrate the application of the proposed methods with complex survey data. The findings include practical guidelines for weighted double score matching analysis of observational data from complex surveys.
