Bayesian Analysis of Treatment Effects in Experiments with Imperfect Compliance

Grant number: DP0880069 | Funding period: 2008 - 2013



This research has a potential to benefit society by providing methods to evaluate the effectiveness of medical treatments and socio-economic and health related interventions and policies based on experiments with imperfect compliance. The proposed methods are directly relevant to the second national research priority, as they can be used to test potential measures for preventative health care and to evaluate measures for strengthening Australia's social and economic fabric. The usefulness of our methods is demonstrated in the analysis of the effectiveness of a training program for the unemployed in alleviating negative mental health effects from job loss.