Bayesian model selection for generalized linear mixed models

We propose a Bayesian model selection approach for generalized linear mixed models (GLMMs). We consider covariance structures for the random effects that are widely used in areas such as longitudinal studies, genome-wide association studies, and spatial statistics. Since the random effects cannot be integrated out of GLMMs analytically, we approximate the integrated likelihood function using a pseudo-likelihood approach. Our Bayesian approach assumes a flat prior for the fixed effects and includes both approximate reference prior and half-Cauchy prior choices for the variances of random effects. Since the flat prior on the fixed effects is improper, we develop a fractional Bayes factor approach to obtain posterior probabilities of the several competing models. Simulation studies with Poisson GLMMs with spatial random effects and overdispersion random effects show that our approach performs favorably when compared to widely used competing Bayesian methods including deviance information criterion and Watanabe-Akaike information criterion. We illustrate the usefulness and flexibility of our approach with three case studies including a Poisson longitudinal model, a Poisson spatial model, and a logistic mixed model. Our proposed approach is implemented in the R package GLMMselect that is available on CRAN.

Automated summary

We propose a Bayesian model selection approach for generalized linear mixed models (GLMMs) that uses a pseudo-likelihood method to approximate the integrated likelihood function. Our approach includes flat priors for fixed effects and offers choices of approximate reference priors and half-Cauchy priors for random effects variances. We introduce a fractional Bayes factor approach to obtain posterior probabilities of competing models. Simulation studies demonstrate the favorable performance of our approach compared to other commonly used Bayesian methods.