Utilizing Moderated Nonlinear Factor Analysis Models for Integrative Data Analysis

• In Event: AERA Poster Session 9
  In Poster Session: Structural Equation Modeling (San Diego)

Abstract

Integrative data analysis (IDA), often referred to as data synthesis, involves analyzing raw data that has been pooled across multiple, independent studies. Moderated nonlinear factor analysis (MNLFA), a novel modeling approach which allows for moderation of multiple covariates simultaneously on factor and item parameters, provides an appealing approach suitable for use within IDA. The aim of this paper is to provide a detailed tutorial on the process of MNLFA model building and implementation. Using empirical data from four school-based randomized trials, we explore the effects of a latent disruptive/aggressive behavior factor on student learning ability. Annotated syntax, as well as the dataset used for analyses, are also provided.

Authors

• Joseph M. Kush, Johns Hopkins University

• Rashelle Musci

• Katherine E. Masyn, Georgia State University