Dynamic latent class structural equation modeling: A hands-on tutorial for modeling intensive longitudinal data

Area of Science:

• Psychological statistics
• Quantitative psychology
• Computational statistics

Background:

• Complex dynamic latent class structural equation models (DLCSEMs) are powerful tools for analyzing intricate psychological data.
• Applied researchers often require accessible guidelines for implementing advanced statistical models.

Purpose of the Study:

• To provide a hands-on, step-by-step tutorial for implementing complex dynamic latent class structural equation models (DLCSEMs) in JAGS.
• To equip applied researchers with the knowledge to utilize the flexible DLCSEM framework for their own analyses.

Main Methods:

• The tutorial builds foundational blocks starting with confirmatory factor and time-series analysis.
• It extends these to multilevel and dynamic structural equation models, integrating hidden Markov switching models.
• Implementation is demonstrated using a clinical psychology example with anxiety and therapist-patient alliance data.

Main Results:

• The tutorial successfully demonstrates the integration of various statistical components to form DLCSEMs.
• It provides clear model implementations and result interpretations for each stage.
• The example illustrates the practical application of DLCSEMs in analyzing treatment data.

Conclusions:

• This tutorial offers a comprehensive guide for applied researchers to implement complex dynamic latent class structural equation models (DLCSEMs) in JAGS.
• The provided framework enhances the ability of researchers to conduct sophisticated longitudinal and person-specific analyses.
• The step-by-step approach and practical example facilitate the adoption of DLCSEMs in psychological research.