Individual Differences in Early Self-Regulatory Mechanisms: Bayesian Regularized Latent Class Analysis of Preschoolers' Learning Trajectories

• In Event: Unpacking Self-Regulated Learning: Cognitive, Motivational, and Methodological Insights Across Developmental and Disciplinary Contexts

Abstract

Self-regulation mediates how preschoolers' competencies translate into academic success, yet most research assumes uniform pathways, thereby overlooking individual differences. We applied Bayesian Regularized Latent Class Structural Equation Modeling to 477 preschoolers to (1) identify developmental profiles, (2) discover subgroups with differential self-regulatory patterns, and (3) examine mediation differences across subgroups. Surprisingly, social-emotional competencies, not cognitive abilities, predicted self-regulatory development, with social skills showing strongest effects (β = 0.264). Analysis revealed three profiles: High-Achieving Regulated (46.3%), High-Potential Dysregulated (9.6%), and Language-Strength Moderate (44.0%). Self-regulatory mediation operated conditionally: critically for High-Achieving Regulated, irrelevantly for High-Potential Dysregulated, and domain-specifically for Language-Strength Moderate children. These findings reveal that 54% of children follow non-traditional pathways, highlighting the need for precision educational interventions.

Authors

• YUNQI WANG, The Education University of Hong Kong

• Chengrui Wang, Fudan University

• DA GONG, The Education University of Hong Kong