Evaluating Model Fit and Structural Coefficient Bias: A Bayesian Approach to Multilevel Bifactor Model Misspecification

• In Event: Power and Design, Model Misspecification, and Multilevel Modeling

Abstract

Appropriate model specification is fundamental to unbiased parameter estimates and accurate model interpretations in the context of multilevel structural equation modeling. This study evaluates the performance of the Bayesian approach in detecting multilevel bifactor model misspecification and the structural coefficient bias at both between and within levels. Misspecification occurs when group factors at the within level are ignored. Preliminary results showed that with misspecification could be detected with a high level of data dependency. Structural coefficients at both levels would be severely and negatively biased with misspecified models.

Authors

• Yan Wang, University of South Florida

• Eun Sook Kim, University of South Florida