Bayesian Structural Equation Modeling: Goodness of Fit and Estimation Precision Across Sample Sizes

• In Event: Evaluating Model Fit in Structural Equation Models

Abstract

The current study compared the goodness of fit and estimation precision of Bayes and maximum likelihood (ML) structural equation modeling across sample sizes. Data consisted of Markov Chain Monte Carlo generated samples of 50, 75, 100, 125, 150, 200, 250, 500, 750, 1000, and 1500 observations. The structural equation model included 40 parameters, including 12 continuous observed indicators, four latent variables, and three structural paths. Analyses were replicated 500 times with each of the 22 conditions (11 samples x 2 estimation methods). As the sample size increased, ML and Bayes model fit and estimation precision improved, and both methods performed well with samples of 200 or larger; however, ML demonstrated a marginally superior fit and estimation precision with smaller samples.

Author

• Diana L. Mindrila, University of West Georgia