Evaluating the effectiveness of school-bullying prevention programs: An updated meta-analytical review

Abstract

Abstract A comprehensive and extensive systematic review and meta-analysis of the effectiveness of bullying prevention programs is presented. This report updates earlier research conducted by Farrington and Ttofi (2009). Systematic searches of online databases (i.e., Web of Science, PsychARTICLES, PsychINFO, EMBASE, DARE, ERIC, Google scholar, and Scopus) were conducted for primary studies published from 2009 to December 2016. Searches were also conducted for unpublished reports. To be included in the systematic review, primary studies must: (1) describe an evaluation of a school-based anti-bullying program; (2) utilize an appropriate operational definition of school-bullying (e.g., Centers for Disease Control and Prevention, 2014; Farrington, 1993; Olweus, 1992); (3) measure school-bullying perpetration and/or victimization behaviors using quantitative measures; and (4) use an experimental or quasi-experimental design with adequate control group. Following systematic screening of over 20,000 search results, a total of 100 evaluations (with 103 independent effect sizes) were eligible for inclusion in our meta-analysis. Most of the effect sizes are estimated from studies that used RCT designs (n = 45) or quasi-experiments (n = 44 effect sizes), with only 14 effect sizes from age cohort designs. Anti-bullying programs significantly reduce bullying perpetration (random effects OR = 1.309) and bullying victimization (random effects OR = 1.244). These results suggest that anti-bullying programs reduce school-bullying perpetration by approximately 19–20% ad school-bullying victimization by approximately 15–16%. Effect sizes vary greatly across studies, with a significant heterogeneity between studies for both bullying perpetration bullying victimization outcomes. This is anticipated given the variability in a range of moderators, for example, methodological designs, type of program used, or place of implementation. Analyses suggest no publication bias for either meta-analysis. Variability in effect sizes across different methodological designs is investigated. Primary studies employing age cohort designs (n = 14) provide the largest effects in reducing both bullying perpetration (OR = 1.474) and victimization (OR = 1.302). In relation to bullying victimization outcomes, before-after/experimental-control designs provide similar effects (OR = 1.225) to randomized controlled trials (OR = 1.21). Randomized controlled trials (OR = 1.244) are more effective in reducing bullying perpetration than before-after/experimental-control designs (OR = 1.187). In future, we aim to further explain differences across programs by correlating individual effect sizes with varying program components and varying methodological elements available across these 100 evaluations.