By
Sara Valdebenito, Hannah Gaffney, Maria Jose Arosemena-Burbano, Sydney Hitchcock, Darrick Jolliffe, Alex Sutherland
School exclusion—commonly referred to as suspension—is a disciplinary response employed by school authorities to address student misbehaviour. Typically, it involves temporary removal from regular teaching or, in more serious cases, complete removal from the school premises. A substantial body of research has associated exclusion with adverse developmental outcomes. In response, various school-based interventions have been developed to reduce exclusion rates. While some programmes have shown promising effects, the evidence on their effectiveness remains inconclusive. This mixed-methods systematic review and multi-level meta-analysis updates the previous review by Valdebenito et al. (2018), which included literature published between 1980 and 2015. The present update extends the evidence base by including studies until 2022. The primary aim of this review was to assess the effectiveness of school-based interventions in reducing disciplinary exclusions, with secondary aims focused on related behavioural outcomes including conduct problems, delinquency, and substance use. Systematic searches conducted between November and December 2022 yielded over 11,000 references for quantitative studies. Following title and abstract screening, 777 records were reviewed at full text by two independent coders. Thirty-two studies met the inclusion criteria for meta-analysis, comprising 2765 effect sizes from 67 primary evaluations (1980–2022) and representing approximately 394,242 students. Meta-analysis was conducted using a multilevel random-effects model with robust variance estimation to account for the nested structure of the data. Quantitative impact evaluations were eligible if they used a randomised controlled or quasi-experimental design, included both a control group and pre/post-test data, and used statistical methods to minimise selection bias (e.g., propensity score matching or matched cohort design). Studies were excluded if they exhibited substantial baseline differences between treatment and control groups. The qualitative synthesis explored implementation barriers and facilitators based on nine UK-based process evaluations, identified through searches completed in September 2023. Process evaluations were included if they focused on the perceptions of stakeholders—teachers, students, or school leadership—within UK schools. Data collection followed two stages: initial selection based on titles, abstracts, and keywords, followed by full-text review. Two independent coders applied inclusion criteria, extracted data, and resolved discrepancies with the principal investigators. All steps were documented to inform the PRISMA flow chart. To evaluate interventions reducing school exclusions, we conducted a multilevel meta-analysis using robust variance estimation. We explored heterogeneity via meta-regression (e.g., gender, intervention type), conducted sensitivity analyses for outliers and correlation structures, and assessed quality data using the EPOC, ROBIN-I and CASP checklist for methodological quality. Findings indicated that school-based interventions were associated with a small but statistically significant reduction in school exclusion (standardised mean difference [SMD] = 0.104; 95% CI: 0.04 to 0.17; p < 0.001). Compared with the original 2018 review, which reported a slightly larger effect size, this update benefits from a broader evidence base and more advanced statistical modelling. However, the results for secondary behavioural outcomes were more limited: effects on conduct problems and delinquency were negligible or non-significant, and the impact on substance use was small and not statistically significant. Risk of bias was assessed using the Cochrane EPOC 2 tool (Higgins and Green 2011) for randomised controlled trials and ROBINS-I (Sterne et al. 2016) for quasi-experimental designs. Randomised studies generally exhibited lower risk of bias, while quasi-experimental studies showed greater variability in quality. Four major themes emerged from the analysis. First, intervention format mattered: flexible, collaborative, and well-structured interventions facilitated implementation, while outdated materials or content misaligned with local context impeded delivery. Second, consistency in school policies and practice enabled smoother implementation, whereas inconsistency acted as a barrier. Third, staff buy-in—particularly among senior leaders—was essential for successful implementation, although resistance from more experienced staff was noted. Finally, perceived effectiveness played a motivational role: visible improvements in pupil behaviour supported continued engagement with the intervention. In summary, the updated review finds that school-based interventions can modestly but significantly reduce school exclusions. While more serious disciplinary sanctions such as permanent exclusions and out-of-school suspensions appear less responsive, in-school exclusion shows greater potential for reduction. Impacts on other behavioural outcomes remain limited. These findings suggest that targeted, context-sensitive interventions supported by strong implementation strategies and whole-school engagement are most likely to achieve sustained reductions in school exclusion.
