Seminar Details

Seminar Details


Jan 13

3:30 pm

Causal Inference via Markov Chain Monte Carlo

David Draper


University of Bath

A major challenge in the analysis of observational studies is the assessment of the likely influence of unmeasured potential confounding factors on the estimated effect of the principal causal factor of interest. Two leading classes of models for such data are selection models (developed by the economist J Heckman) and counterfactual models (dating back to J Neyman, more recently developed by D Rubin). The fully Bayesian analysis of such models with Markov Chain Monte Carlo methods poses several technical problems, including multi-modality of the posterior distribution and extremely high serial correlation of the Monte Carlo draws, and informative prior distributions must be elicited for parameters not well addressed by the data. In this talk I will illustrate the solution of these problems with examples from medicine and psychobiology.