BAYESIAN-ANALYSIS OF LINEAR AND NONLINEAR POPULATION-MODELS BY USING THE GIBBS SAMPLER

A fully Bayesian analysis of linear and non-linear population models h as previously been unavailable, as a consequence of the seeming imposs ibility of performing the necessary numerical integrations in the comp lex multiparameter structures that typically arise in such models. It is demonstrated that, for a variety of linear and non-linear populatio n models, a fully Bayesian analysis can be implemented in a straightfo rward manner by using the Gibbs sampler. The approach is illustrated w ith examples involving challenging problems of outliers and mean-varia nce relationships in population modelling.