The Impact of Informative Priors for Parameters in Bayesian Multiple-Membership Multilevel Modeling

• In Event: 65.060 - Multilevel Modeling: Nonnested Multilevel Data, Measurement Invariance, and Bayesian Approaches

Abstract

This simulation study examines how the selection of priors within a Bayesian framework affects parameter estimates in multiple membership multilevel modeling. Two priors were used to assess the impact on parameter estimates with Markov chain Monte Carlo (MCMC): non-informative and informative. Substantial over-estimation of the level two variance component was identified for both non-informative and informative priors. With a larger number of level two units, however, the differences of parameter estimates between non-informative and informative priors were diminished. Increasing level one sample size has relatively little impact on the level two variance component parameter estimates for both priors. The results suggest that researchers should be cautious in selecting accurate priors especially with a relatively small number of level two units.

Authors

• Hyewon Chung, Chungnam National University

• Jiseon Kim, University of Washington - Seattle

• Ryoungsun Park, Wayne State University

• Hyeonjeong Jeon, Chungnam National University