Title: Propensity Score Matching for causal inference when multiply
imputing missing covariate data 

Abstract: Propensity Scores are a tool used in Observational Studies
to balance the covariate distributions of treated and control
units. This enables inference about the treatment effect to be based
on comparable groups. However when units have only partially observed
covariates, propensity scores cannot be directly estimated. Imputing
the missing values will allow propensity scores to be estimated but we
may be sensitive to the accuracy of our imputation models. In this
talk I will briefly review the ideas behind propensity scores and
multiple imputation and evaluate through some simulations various
methods to perform propensity score matching with multiply imputed
data.