One Step Closer Towards Reliable Causal Inference

Assistant Professor Karen Yan
About This Project

Assistant Professor Karen YanBy: Aselia Urmanbetova

In the paper recently accepted by the Journal Econometric Reviews, Professor Yan and her co-authors, Yu Sun and Qi Li, develop a semi-parametric propensity score model to more accurately identify the effects of a policy or treatment, technically referred to as average treatment effects, or ATE.

In the paper, the authors demonstrate that the proposed method would work well in most of the empirical data sets (finite samples), especially if the data sets have a large number of covariates or variables that can influence the outcome of a treatment.

The authors test their model against right-heart catheterization data and find that those who have received right-heart catheterization face a 5.3% increase in risk of death. The findings of the study suggest that propensity score models play an important role in causal inference – the aspired keystone of all scientific inquiry.

Explore more research by faculty in the School of Economics on our featured project page.