CMStatistics 2020: Start Registration
View Submission - CMStatistics
Title: Covariate balancing for robust estimation of causal effects of general treatment regimes Authors:  Xavier de Luna - Umea University (Sweden) [presenting]
Abstract: Novel robust estimators for categorical and continuous treatment regimes are proposed by using a strategy based on covariate balancing propensity score and inverse probability weighting. The resulting estimators are shown to be consistent and asymptotically normal for causal contrasts of interest, either when the covariate dependent treatment assignment model is correctly specified, or when the correct set of bases for the outcome models in the space spanned by the covariates has been chosen, and the assignment model is sufficiently rich. For the categorical treatment regime case, we show that the estimator attains the semiparametric efficiency bound when all both models are correctly specified. For the continuous case, the causal contrasts of interest are functions. The latter are not parametrized and the estimators proposed are shown to have bias and variance of the classical nonparametric rate. Asymptotic results are complemented with simulations illustrating the finite sample properties. Our analysis of a data set suggests a nonlinear causal effect of BMI on the decline in self-reported health.