A tutorial on bayesian multiple-group comparisons of latent growth curve models with count distributed variables.

Area of Science:

• Quantitative Psychology
• Developmental Psychology
• Criminology

Background:

• Latent variable product terms are complex for analyzing moderation in longitudinal structural equation models, especially with many panel waves.
• Categorical moderation variables necessitate simpler, more precise estimation techniques for complex longitudinal models.

Purpose of the Study:

• To demonstrate multiple-group comparisons as a simpler alternative to latent variable product terms for moderation analysis in longitudinal data.
• To model developmental trajectories of juvenile delinquency using latent growth curve analysis and Bayesian estimation.
• To examine group differences (gender, school type) and the moderating effects of these variables on delinquency trajectories.

Main Methods:

• Latent growth curve modeling applied to count data representing juvenile delinquency trajectories.
• Bayesian estimation implemented using Mplus software.
• Data processing and analysis of group differences using the R programming language.

Main Results:

• Significant group differences in unconditional growth trajectories were found for gender and school type.
• A conditional growth model revealed a significant moderating effect of school type on the relationship between legal norm acceptance and growth trajectories.
• Multiple-group comparisons provided a straightforward method for differentiating effects in complex longitudinal models.

Conclusions:

• Multiple-group comparisons are a practical and effective technique for analyzing moderation with categorical variables in complex longitudinal models.
• The study highlights the utility of this method for understanding developmental trajectories, particularly in developmental psychology and criminology research.
• Findings underscore the importance of considering school type as a moderator in understanding factors influencing juvenile delinquency.