Explorations of Individual Change Processes and Their Determinants: A Novel Approach and Remaining Challenges.

Area of Science:

• Developmental Psychology
• Quantitative Psychology
• Statistical Modeling

Background:

• Traditional conditional growth models dominate the analysis of individual change and between-person differences.
• Existing exploratory models like growth mixture models (GMM) and structural equation modeling (SEM) trees offer increased flexibility.
• A need exists for more adaptable methods to model complex developmental trajectories and their determinants.

Purpose of the Study:

• To introduce and illustrate a novel statistical approach: growth mixture modeling with membership trees (GMM-MT).
• To enhance the flexibility in modeling individual change and its determinants.
• To evaluate how measured covariates predict latent class membership within a recursive partitioning framework.

Main Methods:

• Integration of growth mixture modeling (GMM) with structural equation modeling trees (SEM trees).
• Utilizes a recursive partitioning algorithm to model class membership prediction.
• Application of the MplusTrees package in R for analysis.

Main Results:

• The study demonstrates the practical application of GMM-MT using longitudinal reading data.
• The combined approach allows for a more nuanced understanding of how covariates influence developmental trajectories.
• Successful prediction of class membership based on measured covariates is shown.

Conclusions:

• Growth mixture modeling with membership trees offers a powerful and flexible extension to existing modeling techniques.
• This method advances the analysis of individual change and between-person differences in developmental research.
• The approach provides valuable insights into the predictors of distinct developmental trajectories.