A risk assessment based on cyberbully experiences towards students performance

The expansion of Information and Communication Technology (ICT) has transformed the ways adolescents connect, learn, and communicate, yet it has simultaneously introduced critical challenges, particularly the rising prevalence of cyberbullying. In Malaysia, secondary school students are among the most vulnerable groups, with increasing social media engagement exposing them to online harassment that threatens emotional well-being, behavioral stability, and academic performance. This research develops and validates a probability-based risk assessment model to evaluate the impact of cyberbullying on students’ academic outcomes. Employing an action research design, the study integrates expert input from school counsellors with survey data from 258 secondary students in Kota Setar, Kedah. six key dimensions victimization experience, social media, guardian, personality, lack of knowledge usage the internet, and peer influence were identified and weighted to construct a Cyberbullying Probability Flowchart (CbPF) and a risk matrix for quantifying severity levels. The model was tested and refined through sensitivity analyses to ensure robustness and practical applicability. Findings reveal that students who spend more than three hours daily on social media face significantly higher cyberbullying risks, leading to declines in academic engagement, time management, and learning motivation. Gender and residence emerged as critical risk factors, with female students and parents’ residents disproportionately affected by emotional distress and academic disruption. Common forms of cyberbullying included online harassment, exclusion from peer groups, rumor dissemination, and non-consensual sharing of personal content. Theoretically, this thesis advances knowledge by integrating digital behavior, psychosocial vulnerabilities, and academic performance into a unified risk assessment framework. Practically, it provides educators, parents, and policymakers with actionable strategies for early intervention, targeted prevention, and student support. Methodologically, it introduces a novel data-driven model that combines expert validation with probabilistic analysis, offering a replicable tool for assessing cyber-related risks in educational contexts. Ultimately, the study underscores the urgent need for collaborative and preventive measures to safeguard adolescents’ well-being and academic success in an increasingly digital society