Previous research has shown that many school-based anti-bullying programs are effective. A prior meta-analysis (Gaffney, Ttofi, & Farrington, 2019) found that intervention programs are effective in reducing school-bullying perpetration by approximately 19–20% and school-bullying victimization by approximately 15–16%. Using data from this prior meta-analysis, the aim of the current study was to examine the relationship between effectiveness estimates and specific elements of anti-bullying programs. Specific intervention components in line with a socio-ecological framework were coded as present or absent. Components were coded on the following levels: school, classroom, teacher, parent, peer, individual, and intervention. Meta-analytical subgroup comparisons analogous to ANOVA were computed to examine the relationship between the presence of specific components and the effectiveness in reducing bullying perpetration (n = 82) and victimization (n = 86). Results indicated that the presence of a number of intervention components (e.g., whole-school approach, anti-bullying policies, classroom rules, information for parents, informal peer involvement, and work with victims) were significantly associated with larger effect sizes for school-bullying perpetration outcomes. The presence of informal peer involvement and information for parents were associated with larger effect sizes for school-bullying victimization outcomes. Meta-regression analyses showed no significant relationship between effectiveness and the number of intervention components included in a program. The present report contributes to the understanding of 'what works' in reducing school-bullying perpetration and victimization. The impact of these findings on future anti-bullying research is discussed.

先前的研究表明，许多以学校为基础的反欺凌计划是有效的。之前的一项荟萃​​分析（Gaffney、Ttofi 和 Farrington，2019 年）发现，干预计划可有效减少约 19-20% 的校园欺凌行为和约 15-16% 的校园欺凌行为。使用先前荟萃分析中的数据，当前研究的目的是检查有效性估计与反欺凌计划的特定要素之间的关系。符合社会生态框架的特定干预成分被编码为存在或不存在。组件按以下级别编码：学校、课堂、教师、家长、同伴、个人和干预。计算类似于方差分析的元分析亚组比较，以检查特定成分的存在与减少欺凌行为（n = 82）和受害（n = 86）的有效性之间的关系。结果表明，许多干预成分（例如，全校方法、反欺凌政策、课堂规则、家长信息、非正式同伴参与和与受害者合作）的存在与学校-欺凌行为的结果。非正式同伴参与和家长信息的存在与学校欺凌受害结果的更大影响大小有关。元回归分析表明，有效性与计划中包含的干预成分数量之间没有显着关系。本报告有助于理解在减少校园欺凌行为和受害方面“什么有效”。讨论了这些发现对未来反欺凌研究的影响。