Global and Local Structural-After-Measurement Estimators for n-level Structural Equation Models

• In Event: Methodological Innovations in Structural Equation Modeling

Abstract

A core limitation many researchers encounter is how to estimate SEMs for structures with multiple unique levels of nesting or n-level SEMs (i.e., SEMs with an arbitrary structure and number of levels). In this study, we develop a new class of estimators for n-level SEMs. The estimators are well-suited to the types of n-level, cross-classified and multilevel structures because the framework and estimation method easily extends to accommodate arbitrary structures and levels of nesting (e.g., any number of levels with any nesting structure) including interactions between levels of nesting (e.g., interactions between school and neighborhood membership that operationalize intersectionality). An initial evaluation of the estimators suggests they often outperform conventional estimators in convergence, bias, and variance in typical sample sizes.

Authors

• Benjamin Kelcey, University of Cincinnati

• Fangxing Bai, Montana State University