Geostatistical mixed beta regression: a Bayesian approach

Abstract

This paper develops regression techniques for geostatistical data, with an emphasis in proportions mea- sured on a continuous scale. Specifically, it deals with Beta regression models with mixed effects to control the spatial variability from a Bayesian approach. We use a suit- able parametrization of the Beta distribution in terms of its mean and the precision parameter, allowing for both parameters to be modeled through regression structures that may involve fixed and random effects. Specification of prior distributions is discussed, computational implemen- tation via Gibbs sampling is provided, and the methodol- ogy is illustrated using simulated and real data.