Propensity Score Analysis

Abstract

Propensity score analysis (PSA) is a class of statistical methods developed for estimating treatment effects with nonexperimental data or causality analysis in general. Specifically, PSA offers an approach to program evaluation when randomized trials are infeasible or unethical, or when researchers need to assess treatment effects or causal effects from survey data, census data, administrative data, medical records data, or other types of observational data. In the social and health sciences, researchers often face a fundamental task of drawing conditioned causal inferences from quasi-experimental studies. Analytical challenges in making causal inferences can be addressed by a variety of statistical methods including a range of new approaches emerging in the field of PSA. This entry begins with a discussion of the counterfactual framework and two related assumptions. After providing the definition of propensity score and various methods to estimate the score, it discusses seven methods of applying the estimated propensity score in causal analysis, including greedy matching, optimal matching, propensity score subclassification, propensity score weighting, matching estimators, propensity score analysis with nonparametric regression, and propensity score analysis of categorical or continuous treatments. The entry ends with a discussion about the strengths and limitations of the propensity score approach, including the criticism about the method of nearest neighbor matching within a caliper. Selection bias due to unmeasured covariates remains a problem in PSA. The entry concludes that among various approaches, propensity score subclassification, propensity score weighting, and matching estimators are highly recommended.

locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles