Key Social Norms Take-aways The Good School Toolkit intervention effectively shifted social norms surrounding physical discipline both in schools and in surrounding communities. The study found strong reductions in caregivers’ acceptance of physical violence at school and at home, indicating that school-based interventions can catalyze wider normative change. However, changes in actual violence practices at home were not significant, suggesting that while normative change is possible, behavioral change beyond the school setting may require deeper or more sustained engagement with caregivers.

Objectives To assess the impact of the Good School Toolkit on school operational culture and normative beliefs about physical violence among students, staff, and caregivers, and to evaluate whether the intervention influenced violence against children within homes surrounding participating schools.

Methods This secondary analysis of a cluster-randomised controlled trial was conducted in 42 primary schools in Luwero District, Uganda. Cross-sectional surveys were administered at baseline (2012) and follow-up (2014) to students and staff, and at follow-up to caregivers. Outcomes included changes in school culture across relational, psychological, and structural domains, normative beliefs about violence, and reported experiences of violence at home. Mixed-effects regression models accounted for clustering and adjusted for baseline characteristics where applicable.

Intervention Details The Good School Toolkit is an 18-month, school-wide program developed by Raising Voices to reduce violence and improve school climate. It includes over 60 activities focusing on positive discipline, power dynamics, and student-teacher relationships, facilitated by trained “protagonists” (students and staff) and supported through governance structures like student and parent committees and peer courts.

Results The intervention led to statistically significant improvements in multiple domains of school culture, including emotional support from teachers and peers, identification with school, and reduced acceptance of physical discipline. Caregivers in intervention communities reported significantly lower acceptance of physical violence at school (AMD: −0.77) and at home (AMD: −0.67). However, no significant differences were observed in actual reports of violence against children in the home.

Conclusions The Good School Toolkit is effective at improving school culture and shifting social norms around violence within schools and in the broader community. To achieve measurable reductions in violence at home, future iterations may require more direct caregiver engagement and a longer intervention period. These findings support the potential for school-based interventions to influence community-level norms.