Interventions That Failed: Factors Associated with the Continuation of Bullying After a Targeted Intervention

We examined how often teachers' targeted interventions fail in stopping bullying and to what extent this varies between schools vs. between students involved. In addition, we investigated which student-level factors were associated with intervention failure. Data were collected annually in 2011-2016 via online questionnaires and included responses from students in 2107 Finnish primary and secondary schools implementing the KiVa antibullying program. During the years of the study, 27% of the 57,835 students who were victims in the cases of bullying addressed by adults reported no improvement in their situation. Among the 44,918 bullying perpetrators who were targeted by an intervention, 21% said they did not bully less as a result. Intervention failures were mostly due to differences between individuals: only 3-12% of the total variance in continued victimization and bullying was due to between-school differences. According to two-level logistic regression results, victim-perceived failure was more likely when the victimized student was in higher grades, had been victimized more frequently and, for a longer time, had been victimized also online, had bullied others, and had fewer friends in the class. Bully-perceived failure was more likely when the bullying student was in higher grades, bullied more frequently, and was victimized. Finally, the bullying students' antibullying attitudes and their perception of teacher's and parents' antibullying attitudes were negatively associated with failure of the intervention.