The Role of Smart Counseling System Development in Mediating the Influence of Parenting Style on Early Detection of Adolescent Bullying Behavior

Abstract

Bullying among adolescents remains a serious global and national issue, with documented impacts on mental health, academic achievement, and social relationships.  Early detection of bullying behavior is therefore crucial to prevent the escalation of psychosocial problems. Building on an existing Indonesian study, this article analyzes the role of a Smart Counseling System (SCS) as a technological mediator between parenting style and early detection of bullying in secondary school students. Using a quantitative explanatory design with Structural Equation Modeling (SEM) and Partial Least Squares (PLS), the model incorporates three latent variables: parenting style (PAQ), bullying detection via Strengths and Difficulties Questionnaire (SDQ), and the implementation of SCS. The measurement model indicates that PAQ is the most robust construct, with high standardized loading factors (SLF), excellent reliability (α = 0.899; ω = 0.902), and strong convergent validity (AVE = 0.753), whereas SDQ and SCS show moderate to weak psychometric properties. The structural model reveals that bullying detection explains 91.7% of the variance in parenting style and, together with parenting, explains 53.2% of the variance in SCS.  Nonetheless, the indirect effect of SDQ on PAQ through SCS is negative, suggesting that, in its current form, SCS does not optimally mediate the relationship. The article elaborates theoretical and practical implications for integrating AI-based counseling systems with family and school-based interventions to enhance early detection and prevention of adolescent bullying.