Examining the Effect of Nested Data on Class Enumeration and Model Fit in Latent Profile Analysis

• In Event: Evaluating Model Fit in Structural Equation Models

Abstract

This proposal aims to evaluate the impact of nesting on class enumeration and model fit in latent profile analysis (LPA), a method increasingly popular in educational research for identifying meaningful subgroups within complex datasets. Ignoring the hierarchical structure of educational data, such as students nested within classrooms and schools, can lead to biased parameter estimates and incorrect standard errors, compromising research validity. Through a comprehensive simulation study, this research assesses the effectiveness of Mplus’s "type=complex" option in accurately modeling nested data in LPA. By simulating data from multiple indicators across subpopulations within a two-level model, the study aims to provide critical insights into best practices for conducting LPAs with multilevel data, significantly advancing LPA methodology for better-targeted educational interventions.

Authors

• Angela D. Starrett, University of South Carolina

• Katherine E. Masyn, Georgia State University