Abstract
In this paper, we propose to use sufficient dimension reduction (SDR) in conjunction with nonparametric techniques to estimate the average treatment effect on the treated (ATT), a parameter of common interest in causal inference. The proposed method is applicable under a general low-dimensional structure in the data and avoids both the risk of model misspecification and the "curse of dimensionality," for which it often outperforms the existing parametric and nonparametric methods. We develop the theoretical properties of the proposed method, including its asymptotic normality, its asymptotic super-efficiency, and its equivalent form as an augmented inverse probability weighting estimator. We also consider the impact of SDR estimation in the asymptotic studies. These theoretical results are further illustrated by the simulation studies at the end.
Recommended Citation
L. Li et al., "Estimating Average Treatment Effect on the Treated Via Sufficient Dimension Reduction," Stat, vol. 10, no. 1, article no. e367, Wiley, Dec 2021.
The definitive version is available at https://doi.org/10.1002/sta4.367
Department(s)
Mathematics and Statistics
Publication Status
Full Access
International Standard Serial Number (ISSN)
2049-1573
Document Type
Article - Journal
Document Version
Citation
File Type
text
Language(s)
English
Rights
© 2024 Wiley, All rights reserved.
Publication Date
01 Dec 2021
Comments
National Natural Science Foundation of China, Grant 11971170