School-based interventions for reducing disciplinary school exclusion: a systematic review

37 randomised controlled trials (33 US, 3 UK, 1 unclear), nine intervention types

Mainstream schools, children aged 4-18.

Campbell systematic review restricted to RANDOMISED CONTROLLED TRIALS only — the strictest inclusion rule in this topic, and the reason it is graded B while most behaviour reviews sit at C. It answers the operational question directly: can a school do something that reduces its own use of exclusion?

School-based interventions cause a small and significant drop in exclusion rates in the first six months after intervention, and the effect is NOT SUSTAINED beyond that. Four intervention types had significant desirable effects — enhancement of academic skills, counselling, mentoring/monitoring, and skills training for teachers — but each rests on few studies. There is NO impact on antisocial behaviour: the interventions change how schools respond, not how children behave. Effect-size variation is not explained by participant characteristics, intervention theory, or study quality. The finding this archive weights hardest: INDEPENDENT EVALUATOR TEAMS REPORTED LOWER EFFECT SIZES than research teams also involved in designing or delivering the intervention.

Low — randomised trials only.

The independence gradient documented here is the same pattern the archive records for SEL (Wigelsworth et al. 2016) and for curriculum evaluation generally, now reproduced inside the discipline literature on an RCT-only sample.