Bayesian inference for Latent Class model via MCMC with application to capture-recapture data

In this paper we propose a Bayesian Latent Class model for capture-recapture data. Through two appliations, the first concerning a sample of snowshoe hares and the second concerning a sample of diabetics in a small Italian town, we show how the proposed approach may be effectively used to obtain point estimates and credibility intervals for the size of a closed-population. To estimate the model we use the Reversible Jump algorithm and the Delayed Rejection strategy to improve its efficiency.