Are school-based interventions to prevent dating and relationship violence and gender-based violence equally effective for all students? Systematic review and equity analysis of moderation analyses in randomised trials

Abstract

School-based interventions for the prevention of dating and relationship violence (DRV) and gender-based violence (GBV) take advantage of universal opportunities for intervention. Information on differential effectiveness of interventions is important to assess if they ameliorate or worsen social gradients in specific outcomes. This is especially important in DRV and GBV prevention given the gendered context of these behaviours and their common aetiologies in patriarchal gender norms, and social acceptance in school contexts of sexual harassment, such as catcalling or unwanted groping. We undertook a systematic review of moderation analyses in randomised trials of school-based interventions for DRV and GBV prevention. We searched 21 databases and used supplementary search methods without regard to publication type, language or year of publication, and synthesised moderation tests relating to equity-relevant characteristics (principally sex and prior history of the outcome) for DRV and GBV perpetration and victimisation. Across 23 included outcome evaluations, programme effects on DRV victimisation were not moderated by gender or prior experience of DRV victimisation, but DRV perpetration outcomes were greater for boys, particularly for emotional and physical DRV perpetration. Findings for GBV outcomes were counterintuitive. Our findings suggest that practitioners should carefully monitor local intervention effectiveness and equity to ensure that interventions are working as intended. However, one of the most surprising findings from our analysis-with clear relevance for uncertainties in practice-was that differential impacts by sexuality or sexual minority status were not frequently evaluated.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 15 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.