Using SAS PROC CALIS to fit Level-1 error covariance structures of latent growth models

Area of Science:

• Statistics
• Quantitative Psychology
• Longitudinal Data Analysis

Background:

• Latent growth models (LGM) are crucial for analyzing change over time.
• Specifying Level-1 error covariance structures in LGMs is complex.
• Existing methods may lack flexibility in modeling complex growth processes.

Purpose of the Study:

• To demonstrate the application of SAS PROC CALIS for fitting various Level-1 error covariance structures in LGMs.
• To provide practical SAS syntax for modeling growth in both manifest and latent variables.
• To highlight the advantages of the Structural Equation Modeling (SEM) approach for LGMs.

Main Methods:

• Utilizing SAS PROC CALIS, a SEM-based procedure.
• Modeling growth trajectories for manifest and latent constructs.
• Focusing on the specification and identification of Level-1 error covariance structures.

Main Results:

• Successfully demonstrated the fitting of diverse Level-1 error covariance structures using PROC CALIS.
• Provided illustrative SAS code for modeling manifest and latent variable growth.
• Highlighted the flexibility and model-fitting capabilities of the SEM approach within PROC CALIS.

Conclusions:

• SAS PROC CALIS offers a powerful and flexible tool for advanced LGM analysis, particularly for complex error structures.
• The provided tutorial and syntax facilitate the implementation of sophisticated growth models.
• Researchers can leverage this approach for robust analysis of longitudinal data when the growth model is well-defined.