A0885
Title: Causal inference under interference: Regression adjustment and optimality
Authors: Xinyuan Fan - Tsinghua University (China)
Chenlei Leng - University of Warwick (United Kingdom)
Weichi Wu - Tsinghua University (China) [presenting]
Abstract: In randomized controlled trials without interference, regression adjustment is widely used to enhance the efficiency of treatment effect estimation. The purpose is to extend this efficiency principle to settings with network interference, where a unit's response may depend on the treatments assigned to its neighbors in a network. Three key contributions are made: (1) a central limit theorem is established for a linear regression-adjusted estimator, and its optimality in achieving the smallest asymptotic variance is proven within a class of linear adjustments; (2) a novel, consistent estimator is developed for the asymptotic variance of this linear estimator; and (3) a nonparametric estimator that integrates kernel smoothing and trimming techniques is proposed, demonstrating its asymptotic normality and its optimality in minimizing asymptotic variance within a broader class of nonlinear adjustments. Extensive simulations validate the superior performance of the estimators, and a real-world data application illustrates their practical utility. Findings underscore the power of regression-based methods and reveal the potential of kernel-and-trimming-based approaches for further enhancing efficiency under network interference.
