Effects of Modeling Residual Noninvariance Using Growth Mixture Modeling: A Case Study

• In Event: Modeling Latent Variable Interactions and Multiple-Group Models

Abstract

Current best practices for growth mixture modeling emphasize the importance of the proper specification of parameterization, number of classes and invariance assumptions. Although noninvariance in intercepts, slopes and residuals is addressed in methodological literature, the impact of these assumptions on the parameters and latent class composition has not been thoroughly addressed in applied research in multilevel growth mixture models. As a case study, changes in student classroom behavior using Direct Behavior Rating Single Item Scales over one year were analyzed using multilevel growth mixture modeling, to reflect students nested within classrooms. This research provides a description of decisions and results from testing for noninvariance, and how these affect the resulting subgroups and model parameters.

Authors

• Janice Kooken, University of Connecticut

• D. Betsy Mccoach, University of Connecticut

• Sandra M. Chafouleas, University of Connecticut

• Megan E. Welsh, University of California - Davis