Bayesian Analysis Using the BGLIMM Procedure

This workshop provides an introduction to the BGLIMM procedure in SAS and demonstrates how to use the software for Bayesian estimation and inference. PROC BGLIMM is a high-performance, sampling-based procedure that fits Bayesian generalized linear mixed models (GLMMs). The procedure adopts familiar SAS syntax that provides coding convenience and accessibility for users to work with a wide range of hierarchical models, without requiring significant effort in coding, a key improvement over the existing MCMC procedure. The workshop will use extensive examples to illustrate how to use the procedure to handle nested or nonnested multilevel models, repeated-measures data, missing data, prediction, post fitting inference, model selection, and power prior analysis.