Dr. Keil's research is focused on improving the link between epidemiologic studies and policy choices through the application of state-of-the art statistical methods and theory. His primary interest is in providing a rational background for policies on occupational and environmental exposures that take into account exposure across the life course and the complex interactions between exposures in bringing about changes in health. Dr. Keil has worked on studies of exposure to arsenic, ionizing radiation, silica, environmental tobacco smoke, smoking and household pesticides. As part of his long-term goal to improve causal inference in the face of imperfect data (missing data, measurement error, small samples), he and his close colleagues have developed a Bayesian approach to making causal inference from observational data. This approach leverages the strengths of Bayesian methods in augmenting data with prior knowledge, with the inferential power of causal inference methods.