When perpetrators are also victims: latent profiles of traditional and cyberbullying among adolescents who bully in 42 countries

Background Adolescents who bully others are not a homogeneous group: some confine their behavior to face-to-face settings, whereas others also bully online or are themselves victimized. This study aimed to identify profiles of bullying perpetrators based on their patterns of traditional and cyberbullying perpetration and victimization, and to compare these profiles on individual well-being, health-related risk behaviors, and family, peer, and school characteristics. Methods Data from the 2017/2018 Health Behaviour in School-aged Children (HBSC) study were used. Latent class analysis was used to identify profiles based on indicators of traditional and cyberbullying perpetration and victimization among adolescents who reported bullying others (n = 48,465; 42 countries). Differences between profiles were explored separately for boys and girls using the Bolck–Croon–Hagenaars approach that takes into account classification error. Sensitivity analyses tested the robustness of the solution to alternative specifications of the sample, weighting, measurement, age-group categories, and frequency of bullying. Results Three profiles emerged: a traditional-only profile, a cyber-involved profile characterized by cyberbullying perpetration, and a multiple bully-victim profile of adolescents who both bullied others and were bullied in traditional and cyber contexts. For both genders, the multiple bully-victim profile was the least favorable, with lower levels of perceived health and life satisfaction, increased occurrence of psychosomatic complaints, greater risk behaviors, lower family and peer support, less ease of communication with their mothers, and more negative school experiences. Problematic use of social media increased from the traditional-only to the cyber-involved profile and was greatest in the multiple bully-victim profile. On most other variables, the cyber-involved profile fell between the other two. Sensitivity analyses mostly supported the three-profile solution, although the cyber-involved profile was not recovered using a stricter criterion related to the frequency of bullying behavior. Conclusions The adolescents under study who engage in bullying behaviors do not constitute a homogeneous group; rather they show distinctive involvement patterns. Those who both bully and are bullied across settings fare poorly on several dimensions of their lives and thus interventions and prevention programs should be designed to deal with the different behavioral and psychosocial profiles.