Reducing teachers' use of violence toward students: A cluster-randomized controlled trial in secondary schools in Southwestern Uganda

Background: Few interventions addressing teacher violence have been evaluated in low-income countries. Objectives: In Southwestern Uganda we tested the feasibility and effectiveness of the preventative intervention Interaction Competencies with Children - for Teachers (ICC-T) in reducing teachers' use of violence against students. Trial design: Two-arm cluster-randomized controlled trial Participants: The sample comprised of 343 teachers (74.6% male) and 548 students (59.7% boys) from ten public secondary schools in five districts. Outcome: Teachers' use of emotional and physical violence and student's experience of emotional and physical violence were collected before and 3-months after the intervention. Randomization: The selected schools were randomly allocated (using randomly generated numbers) to the intervention condition (that received ICC-T intervention) and the control group (that received no intervention).Blinding: Teachers were not blinded to the intervention assignment. Students and research assistants conducting the follow-up assessment were blinded. Results: 5 of 10 schools were randomly assigned to the intervention group and the other 5 to the control group. All clusters of each group were analyzed. ICC-T was found to be feasible as indicated by high demand, good applicability, and acceptability. Accounting for nested cluster effects in random coefficient models for repeated measures, the effectiveness was supported by significant time x intervention effects regarding teachers' self-reported use of emotional and physical violence (ps < 0.05). These findings were replicated for students' re-ported exposure to emotional violence and were strengthened by positive effects on teachers' attitudes towards the use of physical violence (ps < 0.05). Conclusions: Our study shows that it is possible to reduce teacher violence through school-based interventions. Further studies are needed that replicate these findings beyond this specific context.