Structural After Measurement Estimation for Structural Equation Models with Complex Sampling

• In Event: The Nuance of Patterns: Methods for Uncovering Intricate Quantitative Relationships

Abstract

This study extends Structural After Measurement (SAM) estimators to accommodate complex sampling designs, specifically sampling weights, in Structural Equation Models (SEM). While large-scale educational survey data offers rich insights, traditional SEM estimation struggles with model complexity. SAM estimators, which split SEM estimation into two phases, have shown strong performance with complex SEMs and latent interactions, but their application with sampling weights in educational survey data has been unexplored. Through Monte Carlo simulations, we investigated SAM-Croon's performance with simple latent mediation and latent moderated mediation models under complex sampling. Our findings demonstrate that SAM-Croon effectively handles sampling weights and complex SEMs, including those with latent interactions, providing educational researchers with a crucial tool for analyzing complex theories using large-scale survey data.

Authors

• Kyle T. Cox, University of North Carolina - Charlotte

• Richard G. Lambert, University of North Carolina - Charlotte