An Additive Model for Spatio-Temporal Smoothing of Cancer Mortality Rates

In this paper, Bayesian hierarchical model is used to anaylze the female breast cancer mortality rates for the State of missouri from 1969 through 2001. The logit transformations of the mortality rates are assumed to be linear over the time with additive spatial and age effect as intercepts and slopes. Objective priors of the hierarchical model are explored. The Bayesian srimates are quite robustness in terms change of the hyper parameters. The spatial correlation are appeared in both intercepts and slopes.