Building Latent Growth Models Using PROC CALIS: A Structural Equation Modeling Approach
Teck Kiang Tan, Trivina Kang, David Hogan · 2010
This paper illustrates the structural equation modeling approach of building latent growth models (LGMs) using PROC CALIS. In the past decade, LGM has become one of the commonly used statistical models for analyzing longitudinal data analysis. Although recent years have seen the increase use of LGM to carry out research work in longitudinal analysis, there is limited work that has spelled out a systematic procedure of using PROC CALIS in modeling LGM. This paper serves to fill the gap for data analysts and researchers by showing the practical aspects and theoretical concerns of applying this modeling technique. This paper also includes a new conceptual idea of combining simplex approach and classical LGM to include autoregressive terms and moving average terms in order to improve the way we can conduct longitudinal analysis. The syntaxes of PROC CALIS are illustrated throughout the paper. Using data from a 4-wave longitudinal dataset of secondary school students, various LGMs are illustrated from the simplest of an unconditional LGM to conditional multivariate cross-Lag autoregressive LGM.