Understanding Cyber Delinquency Among Multicultural Adolescents: A Random Forest Analysis

• In Event: Beyond Theory Testing: New Applications of Strain, Learning, and Control Mechanisms

Abstract

As South Korea transitions from an ethnically homogeneous society to one with a growing multicultural population, understanding cyber delinquency among multicultural adolescents is crucial for effective intervention and policy development. Unlike their native Korean peers, multicultural adolescents may face unique stressors, such as family functioning, assimilation challenges, and identity conflicts, which can shape their online behaviors and delinquency patterns. This study applies a criminological framework and machine learning techniques to examine the factors influencing cyber delinquency within this population. Specifically, it seeks to determine the most significant predictors of cyber delinquency and examine how multicultural adolescents’ distinct experiences—such as the influence of family, peer group dynamics, and broader societal attitudes—contribute to this phenomenon, drawing on criminological theories that explain delinquent behavior in the digital age. By employing Random Forest analysis, this study systematically identifies key determinants of cyber delinquency, offering a data-driven approach to understanding these behaviors. Integrating machine learning into criminological research, the findings aim to enhance theoretical perspectives and inform evidence-based policies tailored to the needs of multicultural adolescents, ultimately contributing to early intervention strategies and risk reduction efforts in South Korea’s evolving social landscape.

Authors

• Hyeseon Noh, University of South Carolina

• Juyoung Song, Penn State University

• Hyoungah Park, Saint Peter's University