Violence has severe and long-lasting negative consequences for children's and adolescents' well-being and psychosocial functioning, thereby also hampering communities' and societies' economic growth. Studies show high prevalence of violence by teachers against children in Sub-Saharan Africa, both in countries where violence is lawful as disciplinary measure at school and in countries where it has been officially banned. In addition to legal and structural factors (e. g. stressful working conditions for teachers), attitudes favoring violence against children as an effective and acceptable discipline method and the lack of access to alternative non-violent strategies are likely to contribute to teachers' ongoing use of violence against children.

Notwithstanding, there are currently very few school-level interventions to reduce violence by teachers that a) have been scientifically evaluated and b) that focus both on changing attitudes towards violence and on equipping teachers with non-violent discipline strategies.

Thus, the present study tests the effectiveness of the preventative intervention Interaction Competencies with Children - for Teachers (ICC-T) in primary and secondary schools in Tanzania, Uganda and Ghana. Previous studies have provided initial evidence on the feasibility and effectiveness of ICC-T to reduce teacher violence in primary and secondary schools in Tanzania and secondary schools in Uganda. This study aims to provide further evidence for the effectiveness of ICC-T to reduce violence and to improve children's functioning (i.e. mental health, well-being, academic performance) across educational settings, societies and cultures.

The study is a multi-site cluster randomized controlled trial with schools (clusters) as level of randomization. Multi-stage random sampling will be applied to select three zones per country (Tanzania, Uganda and Ghana) and one district in each zone. In each district, eight public schools will be randomly selected, resulting in a total number of 72 schools (24 per country). Schools will be stratified based on school type (primary or secondary school) and location (rural or urban), leading to 36 sites or cluster pairs (3 countries x 3 zones x 2 school types x 2 locations). In each site, one school will be randomly allocated to the intervention group (that will receive the ICC-T intervention) and one school to the control group (that will receive no intervention).

At each school, 40 students (stratified by gender) in the third year of primary school or in the first year of secondary/junior high school and all teachers (expected average number: 20) will be recruited. Thus the final sample will comprise 2880 students and at least 1440 teachers.

The study will have three data assessment points: baseline assessment prior to the intervention, the first follow-up assessment six months after the intervention and the second follow-up assessment 18 months after the intervention. In addition, feasibility data will be assessed in the intervention group at the beginning and the end of the intervention.

Primary outcome measures are student- and teacher-reported physical and emotional violence by teachers in the past week. Secondary outcome measures include teachers' attitudes towards violence against students, children's emotional and behavioral problems, quality of life, and cognitive functioning.

General:

Inclusion Criteria:

• All: Written informed consent (if underaged by parents (written) & minors themselves (oral or written)
• Students: Enrollment in class 3 of primary school or in class 1 of secondary school
• Teachers: All teachers employed at the school

Schools:

• Public schools
• Mixed-gender schools
• Day schools
• At least 40 students in selected class/stream

Exclusion criteria:

• Acute drug or alcohol intoxication
• Acute psychotic disorder