Emotional Problem Trajectories and Predictors in Adolescence: Using Latent Class Growth Analysis and SEM Forests

• In Event: AERA Roundtable Session Thursday 7:45 am - Gold 4
  In Roundtable Session: Measurement Complexity in SEM: Cross-Loading, Invariance, and Heterogeneity

Abstract

This study aimed to classify trajectories of adolescent emotional problems and explore key predictors influencing these changes. Using waves 1-5 of the Korean Children and Youth Panel Survey 2018 (n=2,071), latent class growth analysis was conducted to classify trajectories, SEM forests were used to explore important predictors, and the three-step approach was applied to analyze their effects on class membership. Three latent classes were classified: Low-Stable (23.9%), Moderate-Increase-Decrease (61.1%), and High-Increase-Decrease (15.0%). SEM forests revealed ten key predictors―six previously studied and four newly derived. The three-step approach showed that happiness, self-esteem, cooperation (group facilitation) and grit (consistency of interest) decreased the likelihood of belonging to classes characterized by higher levels of emotional problems, whereas peer relationships (negative) increased this likelihood.

Authors

• Eunah Jang, Chungnam National University

• Hyewon Chung, Chungnam National University