Convention 2012

Intensive Learning
Advanced Methodology and Statistics Seminars
Open All      Hide All  

The AMASS program is a special series of offerings for applied researchers, presented by nationally renowned research scientists.

Thursday, 1:00 p.m. - 5:00 p.m.


Longitudinal Data Analysis Using Structural Equation Modeling: Latent Growth Models

James M. Henson, Old Dominion University

Matthew R. Pearson, Old Dominion University

The analysis of change over time is critical for clinical research. Longitudinal data are uniquely suited to study the development of psychopathology over time and the effectiveness of therapeutic interventions. Therefore, longitudinal data are ubiquitous in clinical psychology, which highlights the importance of using and understanding the most modern and appropriate statistical methods for analyzing change over time. Although longitudinal data provide the ability to test unique research hypotheses, many hypotheses can also be difficult to analyze with traditional statistical techniques like repeated measures ANOVA, which also imposes strict statistical assumptions on the data.

The present AMASS will give an applied introduction to latent growth curve modeling (LGCM) for continuous outcomes using clinical research examples. We will demonstrate how LGCM is uniquely suited to answer some of the most relevant and interesting longitudinal research questions, and will compare and contrast LGCM to alternative methods (e.g., repeated measures ANOVA, hierarchical linear modeling). We will review how LGCM can be used to model linear or nonlinear growth (i.e., change), how LGCM can be used to simultaneously model multiple growth processes, how LGCM can be used to compare growth across different groups, and how LGCM can be used to test mediation hypotheses. We will provide clinical examples that will be made available to attendees with instructions for how to use three structural equation modeling packages (Mplus, AMOS, and EQS) to analyze the data. Attendees should have a basic understanding of structural equation modeling and multiple regression; experience with LGCM is not required.

You will learn:

  • How to describe advantages of LGCM over alternative methods for analyzing longitudinal data
  • How to correctly interpret LGCM parameters for both linear and nonlinear change over time
  • How to conduct LGCM using one of the following statistical packages: Mplus, AMOS, or EQS
Recommended Readings: Duncan, T. E., & Duncan, S. C. (2004). An introduction to latent growth curve modeling. Behavior Therapy, 35, 333-363. Duncan, T. E., Duncan, S. C., & Strycker, L. A. (2006). An introduction to latent variable growth curve modeling: Concepts, issues, and applications (2nd ed.). Mahwah, NJ: Lawrence Erlbaum.



Association for Behavioral and Cognitive Therapies
305 7th Avenue, 16th Fl., New York, NY 10001 | Phone (212) 647-1890 | Fax: (212) 647-1865
Copyright 2003 - 2017 ABCT. All rights reserved. Terms of Use | Site Map