We examined the connections between attitudes, group norms, and students’ behaviour in bullying situations (bullying others, assisting the bully, reinforcing the bully, defending the victim, or staying outside bullying situations). The participants were 1220 elementary school children (600 girls and 620 boys) from 48 school classes from Grades four, five, and six, i.e., 9–10, 10–11, and 11–12 years of age. Whereas attitudes did predict behaviour at the student level in most cases (although the effects were moderate after controlling for gender), the group norms could be used in explaining variance at the classroom level, especially in the upper grades. The class context (even if not classroom norms specifically) had more effect on girls’ than on boys’ bullying-related behaviours.

This study demonstrates the effectiveness of the KiVa antibullying program using a large sample of 8,237 youth from Grades 4-6 (10-12 years). Altogether, 78 schools were randomly assigned to intervention (39 schools, 4,207 students) and control conditions (39 schools, 4,030 students). Multilevel regression analyses revealed that after 9 months of implementation, the intervention had consistent beneficial effects on 7 of the 11 dependent variables, including self- and peer-reported victimization and self-reported bullying. The results indicate that the KiVa program is effective in reducing school bullying and victimization in Grades 4-6. Despite some evidence against school-based interventions, the results suggest that well-conceived school-based programs can reduce victimization.

This study investigated whether the bystanders' behaviors (reinforcing the bully vs. defending the victim) in bullying situations are related to the frequency of bullying in a classroom. The sample consisted of 6,764 primary school children from Grades 3 to 5 (9-11 years of age), who were nested within 385 classrooms in 77 schools. The students filled out Internet-based questionnaires in their schools' computer labs. The results from multilevel models showed that defending the victim was negatively associated with the frequency of bullying in a classroom, whereas the effect of reinforcing the bully was positive and strong. The results suggest that bystander responses influence the frequency of bullying, which makes them suitable targets for antibullying interventions.

This study investigated the effectiveness of the KiVa Antibullying Program in two samples of students, one from Grades 1-3 (7-9 years old, N ϭ 6,927) and the other from Grades 7-9 (13-15 years old, N ϭ 16, 503). The Grades 1-3 students were located in 74 schools and Grades 7-9 students in 73 schools that were randomly assigned to intervention and control conditions. Multilevel regression analyses revealed that after 9 months of implementation, the intervention had beneficial effects in Grades 1-3 on self-reported victimization and bullying (odds ratios Ϸ 1.5), with some differential effects by gender. In Grades 7-9, statistically significant positive results were obtained on 5 of 7 criterion variables, but results often depended on gender and sometimes age. The effects were largest for boys' peer reports: bullying, assisting the bully, and reinforcing the bully (Cohen's ds 0.11-0.19). Overall, the findings from the present study and from a previous study for Grades 4 -6 (Kärnä, Voeten, Little, Poskiparta, Kaljonen, et al., 2011) indicate that the KiVa program is effective in reducing bullying and victimization in Grades 1-6, but the results are more mixed in Grades 7-9.

A positive impact of the intervention programme was found on several outcome variables (e.g. frequencies of bullies and victims, observed and experienced bullying, attitudes and efficacy beliefs, and to some extent, participant role behaviours). The intervention effects were found more often in Grade 4 than in Grade 5, and often only in schools with a high degree of implementation of the programme.