A Comparison of Methods for Estimating Relationships in the Change Between Two Time Points for Latent Variables

Area of Science:

• Psychometrics
• Longitudinal Data Analysis
• Statistical Modeling

Background:

• Longitudinal data analysis is crucial for understanding developmental processes.
• Collecting data at more than two time points offers deeper insights but is not always feasible.
• Limited data points (two waves) restrict standard modeling techniques like growth curve modeling.

Purpose of the Study:

• To compare the performance of the two-wave latent change score model and the latent difference factor model.
• To evaluate methods for estimating correlations between changes in latent variables with only two data points.
• To identify factors influencing the accuracy of these estimation methods.

Main Methods:

• A simulation study was conducted to compare two statistical models.
• The models assessed were the two-wave latent change score model and the latent difference factor model.
• The study focused on estimating correlations among changes in latent variables between two time points.

Main Results:

• Both the latent change score model and latent difference factor model produced generally accurate estimates of latent trait change correlations.
• Estimation bias and standard errors were reduced with larger sample sizes, higher factor loadings, and more indicators per factor.
• Both methods demonstrated relatively small standard errors in the simulation.

Conclusions:

• Both the two-wave latent change score model and latent difference factor model are viable for analyzing latent variable change correlations with two-wave data.
• Researchers can confidently use these methods, considering the impact of sample size and model specification.
• The findings provide guidance for selecting appropriate statistical approaches in two-wave longitudinal research.