Abstract
Cyberbullying represents a severe public health issue, endangering the mental and physical health of adolescents. Externalizing problem behaviors are significant factors for cyberbullying perpetration. However, the spreading effects between them remain debated. This study applied network analyses to further examine the spreading effects. We focused on two subtypes of cyberbullying perpetration (cyber verbal aggression; cyber relational aggression) and five subtypes of externalizing problem behaviors (academic misbehaviors; vandalism behaviors; aggressive behaviors; interpersonal dishonesty behaviors; risk-taking behaviors). The valid sample consisted of 2075 Chinese adolescents who completed surveys at three time points, spaced 6 months apart. The contemporaneous and cross-lagged panel networks were constructed. Results suggest that the spreading effect is bidirectional over the 6-month interval, but unidirectional in the 12-month interval, and behavior subtypes play different roles. Among the five subtypes of externalizing problem behaviors, interpersonal dishonesty behaviors especially warrant consideration as a risk signal for cyberbullying perpetration, potentially serving as both a precursor and an outcome. These findings advance understandings of developmental cascades at the behavioral level by shifting focus from broad behavior categories to specific subtypes. Identifying such detailed behavioral risk signals enables earlier detection of and more targeted interventions against cyberbullying perpetration, ultimately contributing to protecting adolescents.