A new systematic review is challenging a widely held assumption about school violence prevention: that making classrooms safer will automatically make children better learners. The evidence, drawn from randomised controlled trials conducted across different countries and education systems, suggests a more nuanced reality. School-based interventions can improve certain educational outcomes, but their effects are neither universal nor guaranteed. While most of the programmes examined successfully reduced violence, only a subset produced measurable gains in learning or school participation. The findings indicate that safety may be an essential foundation for education, yet it is not always sufficient on its own to raise literacy, numeracy or test performance.
Published in BMC Public Health, the review examined whether interventions designed primarily to prevent violence in and around schools also influenced children’s educational outcomes. The researchers began with a larger systematic review of randomised controlled trials evaluating school-based violence-prevention programmes. From 188 papers screened using searches conducted through June 2025, they identified 16 papers representing 15 distinct intervention trials that reported educational measures alongside violence-related outcomes. Rather than combining all results into a single statistical estimate, the team summarised the evidence narratively, assessed the risk of bias in the studies and used a forest plot to visualise the direction and size of reported educational effects.
The interventions varied considerably, reflecting the many forms that school violence prevention can take. Some programmes focus on changing children’s knowledge, attitudes and behaviour through structured lessons, while others train teachers, establish school-wide rules, involve parents or strengthen systems for identifying and responding to abuse. Although the review’s abstract does not provide a full catalogue of every intervention, the programmes broadly shared the goal of reducing physical, sexual or emotional violence and improving the social environment in which learning takes place. This distinction matters scientifically: an intervention can reduce aggression or victimisation without directly teaching reading, mathematics or other academic skills.
Overall, 13 of the 15 interventions reduced violence, confirming that many school-based programmes can influence the behaviours and relationships they are designed to target. Educational outcomes, however, followed a less predictable pattern. Six interventions reported improvements in at least some educational measures, four produced mixed results and five found no educational effect. The outcomes considered across the trials included measures such as school attendance, progression, classroom engagement, concentration, participation or academic achievement. Because the studies used different designs, populations, follow-up periods and measurement tools, the researchers could not treat their findings as directly interchangeable. The result is not a simple “yes” or “no” answer, but a map of where benefits appear most likely.
The strongest caution concerns conventional academic achievement. None of the trials showed a clear and consistent effect of a violence-prevention programme on literacy or numeracy test scores. This does not prove that safer schools cannot improve learning over the long term. Standardised test scores are relatively distal outcomes: they may depend on many factors beyond daily safety, including teacher quality, instructional time, nutrition, language exposure, household resources and access to learning materials. A programme might improve attendance or reduce fear in the classroom, for example, without producing an immediate change in reading or mathematics scores. Academic gains may also require additional teaching support before they become visible in assessments.
The review found an important geographical pattern. Five of the six interventions that improved educational outcomes were tested in high-income countries. The authors argue that in less well-resourced settings, preventing violence may not be enough to overcome the other barriers that limit educational progress. A child who feels safer at school may still face overcrowded classrooms, shortages of trained teachers, inadequate textbooks, long travel distances, hunger or repeated interruptions to schooling. In such environments, violence prevention could be beneficial in its own right while failing to generate measurable academic improvement unless it is combined with broader investments in education and child wellbeing.
This finding also highlights why the word “intervention” should not be treated as a single scientific category. Programmes differ in intensity, duration, delivery and mechanism. A short classroom curriculum may alter peer behaviour, whereas a whole-school strategy may change supervision, discipline, teacher practice and referral systems. Their educational consequences could therefore emerge through different pathways. Reducing violence may increase attendance by making school feel safer; it may improve concentration by lowering stress; or it may create more teaching time by reducing classroom disruption. These mechanisms are plausible, but the trials did not consistently measure every step between reduced violence and improved learning, making it difficult to identify which components drive educational benefits.
The study’s design strengthens the relevance of its conclusions but also imposes limits. Randomised controlled trials are generally considered the most reliable way to estimate whether an intervention causes an observed change, because assignment to intervention and comparison groups can reduce the influence of pre-existing differences. Yet the review included only 15 trials with educational data, a small evidence base for a question of global importance. Differences in outcome definitions and study quality also complicate comparisons. The authors therefore avoid claiming that violence prevention consistently improves education. Instead, they conclude that it can improve some educational outcomes, particularly in high-income settings, while finding no evidence that such programmes harm children’s educational progress.
The implications extend beyond schools and into public policy. Education is recognised as a major social determinant of health and is central to Sustainable Development Goal 4, which calls for inclusive, equitable and quality education. Violence undermines that goal not only through direct injury, but also through fear, absenteeism, psychological distress and damaged relationships with teachers and peers. The review suggests that governments should continue investing in evidence-based violence prevention, but should not expect those programmes to substitute for policies addressing poverty, instructional quality, infrastructure or basic services. The most effective strategy may be layered: create safer school environments while simultaneously providing academic support and the material conditions children need to learn.
The researchers ultimately deliver a cautiously optimistic message. School-based violence prevention is capable of producing benefits that extend beyond safety, but educational improvement is context-dependent and uneven. The evidence supports the idea that safer schools can help children engage with education, especially where schools have sufficient resources to convert improved security into better learning conditions. It does not support the stronger claim that reducing violence alone will reliably raise literacy or numeracy scores. Future trials will need to measure educational outcomes more consistently, follow students for longer periods and examine how prevention programmes interact with teaching quality and other support services. For now, the science points to a clear priority: protect children from violence, then make sure they have the tools, time and support required to learn.
Subject of Research: The effects of school-based violence-prevention interventions on educational outcomes in children.
Article Title: Do school-based violence interventions improve educational outcomes? A systematic review
Article References: Devries, K., Eldred, E., Pearlman, J. et al. “Do school-based violence interventions improve educational outcomes? A systematic review.” BMC Public Health (2026). https://doi.org/10.1186/s12889-026-28351-8
Image Credits: AI Generated
DOI: 10.1186/s12889-026-28351-8
Keywords: Systematic review, randomised controlled trials, education, violence prevention, school safety, educational outcomes
