A Comparison of the Three Bayes Point Estimates in Cross-Classified Multiple Membership Random Effects Modeling

• In Event: 66.052 - Estimation Accuracy and Conditions for Complex Multilevel Models

Abstract

This simulation study was designed to evaluate the parameter recovery in Bayesian cross-classified multiple membership random effects modeling. To test the accuracy of parameter estimation, three Bayes point estimates were compared: the posterior mean, the posterior median, and the posterior mode. The posterior mean estimates of the level two variance components were substantially over-estimated with 50 units at level two. The posterior mode estimates of the level two variance components, however, were substantially under-estimated with 100 units or less at level two. In general, the use of posterior median estimates resulted in less bias as compared to the use of posterior mean or mode estimates. The coverage rates of 95% credible intervals were close to nominal coverage.

Authors

• Hyewon Chung, Chungnam National University

• Jiseon Kim

• Ryoungsun Park, Wayne State University

• Hyeonjeong Jeon, Chungnam National University