Quantitative Meta-Analysis of Interventions to Reduce Gender-Based Violence

Abstract

The objective is to determine the effects of quantitative and mixed interventions aimed at reducing gender-based violence, identify the associated variables, and describe the types of interventions carried out. Scientific articles from empirical studies registered in the Dimensions, Scielo, EBSCO, and Google Scholar databases are examined. The eligibility criteria are language (Spanish and English), period (January 1, 2010, to November 13, 2025), and quantitative or mixed interventions. Method of synthesis with effect size metrics: Standardized Mean Difference (SMD), Model: Random Effects, Heterogeneity Estimation: REML, Heterogeneity tests/Indices: Cochrane's Q and I^2, using JASP software. Of the 767 articles identified following the PRISMA 2020 model, 9 were finally included, with pooled effects reported: -0.323 (95% CI [-0.527, -0.118]). The interventions identified focus on gender attitudes and stereotypes, violence reduction, and skills development. Main conclusion: a negative and significant pooled effect (interpreted as a reduction in variables associated with Gender Based Violence). Key limitation: high heterogeneity (reported I^2=88.53). Publication Bias: Trim and Fill suggested 1 study missing, adjusted effect = -0.309 (minimum change). Robustness: reported Fail-Safe N (Rosenthal): 293. These results support that interventions with a gender focus are effective tools to eliminate VG; therefore, it is recommended to replicate them and integrate them into school and community activities.