Bayesian Model Averaging with Applications to the Risk Assessment for Arsenic in Drinking Water

摘要

Model selection often presents a challenge in the risk assessment process, especially when biologically based models are not apparent or fully developed. In a recent dose-response assessment for arsenic in drinking water, risk estimates were found to highly depend on the choice of the model. Two models could fit the data equally well, based on a standard model fit criterion, yet yield quite different risk estimates. Bayesian model averaging takes into account model uncertainty, more appropriately explains overall uncertainty about risk estimates. The current analysis uses lung cancer mortality data from the southwest region of Taiwan where high concentrations of inorganic arsenic were found in the drinking water.