RAND Statistics Seminar Series
Bayesian Modeling and Analysis of Treatment-Response Data
Presented by Siddhartha Chib
Harry C. Hartkopf Professor of Econometrics and Statistics
John M. Olin School of Business, Washington University, St. Louis
Thursday, April 17, 2003 4:00 pm
Main Conference Room
In this paper we present a set a models and Bayesian inferential techniques for analyzing observational treatment-response data. In the model formulations we assume that a binary instrumental variable is available, say from the design of the problem. Under this maintained assumption, parametric and semiparametric (potential outcomes based) modeling strategies are outlined for continuous and binary responses, and binary and ordinal treatments. Estimation of the models is by Markov chain Monte Carlo methods, after rewriting the models in line with framework of Albert and Chib (1993), and the comparison of the various models is by marginal likelihoods and Bayes factors, estimated by the method of Chib (1995). We discuss inferences for the treatment effects, outlining ways in which one can calculate the average treatment effect, and the effect of the instrument on the outcome, both from the intrinsic structure of the model and the output of the MCMC simulations. The ideas and methods are illustrated with both simulated and real data.