Associate professor of health care policy Jose R. Zubizarreta, PhD, has been awarded the 2020 Cochran Award by Observational Studies for his paper “Handling Limited Overlap in Observational Studies with Cardinality Matching,”  published in collaboration with assistant professor of political science at Purdue University Giancarlo Visconti, PhD.

In their paper, Visconti and Zubizarreta analyze the issue of limited overlap in covariate distributions across treatment groups in observational studies. Instead of the common practice of restricting analyses to the portion of the treatment groups that overlap in terms of the estimated propensity or discretized versions of the observed covariates, they suggest cardinality matching, a flexible matching method that uses integer programming to find the largest matched sample that is balanced according to criteria specified before matching by the investigator. Cardinality matching can be used to directly balance several features of the covariates, and can be extended to build matched samples that are representative of a target population by design.

The 2020 Cochran Award was awarded to Visconti and Zubizarreta for writing the best paper published in Observational Studies during the journal’s past five years (2015-2019). Papers were judged by a committee of associate editors on originality, importance, clarity of thought and excellence in writing. The Cochran Award is named in honor of William Cochran, a prominent statistician and founding member of Harvard University’s statistics department. An Institute of Mathematical Studies Affiliated Journal, Observational Studies publishes papers on all aspects of causal inference in observational studies.