Propensity score-based methods for causal inference in observational studies with non-binary treatments

Statistical Methods in Medical Research - Tập 29 Số 3 - Trang 709-727 - 2020
Shandong Zhao1, David A. van Dyk2, Kosuke Imai3
1Department of Statistics, University of California, Irvine, CA, USA
2Department of Mathematics, Imperial College, London, UK
3Department of Government and Department of Statistics, Harvard University, Cambridge, MA, USA

Tóm tắt

Propensity score methods are a part of the standard toolkit for applied researchers who wish to ascertain causal effects from observational data. While they were originally developed for binary treatments, several researchers have proposed generalizations of the propensity score methodology for non-binary treatment regimes. Such extensions have widened the applicability of propensity score methods and are indeed becoming increasingly popular themselves. In this article, we closely examine two methods that generalize propensity scores in this direction, namely, the propensity function (PF), and the generalized propensity score (GPS), along with two extensions of the GPS that aim to improve its robustness. We compare the assumptions, theoretical properties, and empirical performance of these methods. On a theoretical level, the GPS and its extensions are advantageous in that they are designed to estimate the full dose response function rather than the average treatment effect that is estimated with the PF. We compare GPS with a new PF method, both of which estimate the dose response function. We illustrate our findings and proposals through simulation studies, including one based on an empirical study about the effect of smoking on healthcare costs. While our proposed PF-based estimator preforms well, we generally advise caution in that all available methods can be biased by model misspecification and extrapolation.

Từ khóa


Tài liệu tham khảo

10.1093/biomet/70.1.41

10.1198/016214504000001187

10.1093/biomet/87.3.706

10.1097/00001648-200009000-00011

10.1214/17-AOAS1101

10.1515/jci-2014-0022

10.1162/REST_a_00177

10.1093/oxfordjournals.aje.a010011

10.1198/016214501753381896

10.1080/01621459.1987.10478441

10.1002/sim.1144

10.1191/0962280204sm351ra

10.1177/0962280209340213

10.2202/1557-4679.1198

10.1111/j.1467-985X.2007.00527.x

Fan J, 1996, Local polynomial modelling and its applications

10.1016/S0304-4076(02)00157-4

10.1080/01621459.2000.10474263

Duan N, 1983, J Business Economic Stat, 1, 115, 10.1080/07350015.1983.10509330

10.1177/1536867X1401400307

Thông tin
Thông tin xuất bản

Statistical Methods in Medical Research

Tập 29 Số 3

709-727

Thông tin tác giả