Longitudinal network associations between risk and protective factors for bullying, victimization, and well-being: Effects of an antibullying intervention

Objective: The present study explored the differences in networks of bullying, victimization, and well-being, along with related protective and risk factors. Network differences were investigated in adolescent samples following an antibullying intervention vs. regular school year. Methods: A sample of 671 students (52 % girls; Mage = 10.3 years) from 24 elementary schools participated across three time points. A multigroup lag-1 graphical vector autoregressive cross-lagged panel network model was employed and yielded an acceptable model fit. The final model was pruned and bootstrapped (N = 100 resamplings). Results: Several variables demonstrated stability across time points. Cross-lagged associations differed between intervention and control groups. In the absence of intervention, defending increased only when empathy for victims increased, while social self-efficacy reduced both bullying and moral disengagement. In the intervention group, defending became more interconnected with well-being, and an adverse feedback loop emerged between victimization and bullying. Contemporaneous effects were similar across intervention and control schools and showed expected associations, such as positive interrelatedness between moral disengagement and bullying perpetration. Conclusions: The antibullying intervention changed the dynamic interplay of factors related to bullying, victimization, and well-being in schools. While the outcomes were largely positive, some adverse effects emerged, which could guide future research.