Research Interests

As a statistician, my research focuses on the design of empirical studies, including surveys and experiments, and the analysis of data from such studies. I regard myself as both theoretical and as applied, in the sense that essentially all the statistical work that I have done is stimulated by real problems, but when addressing such problems, I try to find solutions that solve the class of problems represented by the ones being faced. Thus, I have made contributions to the handling of data sets with missing values, including establishing terminology (e.g., missing at random, ignorability), creating (with coauthors) algorithms for parameter estimation with incomplete data (e.g., the EM algorithm and its extensions--ECM, ECME, PXEM), and multiple imputation. These research activities were inspired by real projects on which I consulted, including ones for the US federal government. I have also made contributions to causal inference in randomized experiments and in observational studies. Here, I also established terminology, including potential outcomes, the stability assumption (SUTVA--the stable unit treatment value assumption), and the assignment mechanism. I have also, with others, created propensity score methods and principal stratification. The general framework is often referenced as the "Rubin Causal Model," especially when the mode of inference is Bayesian, which I prefer. I continue to conduct research in these and other areas.

Membership Type


Election Year


Primary Section

Section 32: Applied Mathematical Sciences

Secondary Section

Section 54: Economic Sciences