APS
29th APS Annual Convention · 2017
Comparing Exploratory Structural Equation Modeling and Existing Modeling Approaches for Multiple Regression with Latent Variables
- Yujiao Mai
University of Notre Dame - Zhiyong Zhang
University of Notre Dame - Zhonglin Wen
South China Normal University
Abstract
This study compared the performance of structural equation modeling (ESEM) with structural equation modeling (SEM) and regression (REG) in multiple regression with latent variables. The Monte Carlo simulations showed that: (1) ESEM had the least estimation bias with non-zero cross-factor loadings; and (2) higher statistical power than SEM across conditions.
Quantitative
