research article about bullying

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Published: 13 February 2024

Bullying fosters interpersonal distrust and degrades adolescent mental health as predicted by Social Safety Theory

• Dimitris I. Tsomokos   ORCID: orcid.org/0000-0002-9613-7823 1 &
• George M. Slavich   ORCID: orcid.org/0000-0001-5710-3818 2  

Nature Mental Health volume  2 ,  pages 328–336 ( 2024 ) Cite this article

467 Accesses

376 Altmetric

Metrics details

Social Safety Theory predicts that socially threatening experiences such as bullying degrade mental health partly by fostering the belief that others cannot be trusted. Here we tested this prediction by examining how peer bullying in childhood impacted adolescent mental health, and whether this effect was mediated by interpersonal distrust and several other commonly studied mediators—namely diet, sleep and physical activity—in 10,000 youth drawn from the UK’s Millennium Cohort Study. Youth bullied in childhood developed more internalizing, externalizing and total mental health problems in late adolescence, and this effect was partially mediated by interpersonal distrust during middle adolescence. Indeed, adolescents who developed greater distrust were approximately 3.5 times more likely to subsequently experience clinically significant mental health problems than those who developed less distrust. Individual and school-based interventions aimed at reducing the negative impact of bullying on mental health may thus benefit from bolstering youths’ sense of trust in others.

This is a preview of subscription content, access via your institution

Access options

Subscribe to this journal

Receive 12 digital issues and online access to articles

55,14 € per year

only 4,60 € per issue

Buy this article

• Purchase on SpringerLink
• Instant access to full article PDF

Prices may be subject to local taxes which are calculated during checkout

Similar content being viewed by others

The impact of resilience as a protective factor on Health-Related Quality of Life’s psychological dimensions among adolescents who experience peer victimization

The relationship between harsh parenting and adolescent depression

The structural relations of self-control, empathy, interpersonal trust, friendship quality, and mental well-being among adolescents: a cross-national comparative study in China and Canada

Data availability.

The data that support the findings of the present study are publicly available from the Millennium Cohort Study (UK Data Service) by application, under license. For further information on how to obtain the dataset, visit the UK Data Service website ( https://ukdataservice.ac.uk/ ) or the relevant website of the Centre for Longitudinal Studies ( https://cls.ucl.ac.uk/cls-studies/millennium-cohort-study/ ).

Code availability

Details of all the variable names, their processing and the full output of the R code are available on the Open Science Framework website ( https://osf.io/zjq9a ; ref. 5 in the Supplementary Information ). D.I.T. accessed the data and wrote the code.

Bitsko, R. H. et al. Mental health surveillance among children—United States, 2013–2019. MMWR Suppl. 71 , 1–42 (2022).

Article   ADS   PubMed   PubMed Central   Google Scholar  

Ravens-Sieberer, U. et al. Impact of the COVID-19 pandemic on quality of life and mental health in children and adolescents in Germany. Eur. Child Adolesc. Psychiatry 31 , 879–889 (2022).

Article   PubMed   Google Scholar  

Newlove-Delgado, T. et al. Child mental health in England before and during the COVID-19 lockdown. Lancet Psychiatry 8 , 353–354 (2021).

Article   PubMed   PubMed Central   Google Scholar  

Auerbach, R. P. et al. WHO World Mental Health Surveys International College Student Project: prevalence and distribution of mental disorders. J. Abnorm. Psychol. 127 , 623–638 (2018).

Racine, N. et al. Global prevalence of depressive and anxiety symptoms in children and adolescents during COVID-19: a meta-analysis. JAMA Pediatr. 175 , 1142–1150 (2021).

Youth Risk Behavior Survey: Data Summary and Trends Report (Center for Disease Control and Prevention, 2023).

Jones, S. et al. Mental health, suicidality, and connectedness among high school students during the COVID-19 pandemic—Adolescent Behaviors and Experiences Survey, United States, January–June 2021. MMWR Suppl. 71 , 16–21 (2022).

Slavich, G. M. Social safety theory: a biologically based evolutionary perspective on life stress, health, and behavior. Annu. Rev. Clin. Psychol. 16 , 265–295 (2020).

Slavich, G. M. Social Safety Theory: understanding social stress, disease risk, resilience, and behavior during the COVID-19 pandemic and beyond. Curr. Opin. Psychol. 45 , 101299 (2022).

Slavich, G. M. et al. Social Safety Theory: conceptual foundation, underlying mechanisms, and future directions. Health Psychol. Rev. 17 , 5–59 (2023).

Gilbert, P. et al. Feeling safe and content: a specific affect regulation system? Relationship to depression, anxiety, stress, and self-criticism. J. Posit. Psychol. 3 , 182–191 (2008).

Article   Google Scholar  

Hostinar, C. E., Sullivan, R. M. & Gunnar, M. R. Psychobiological mechanisms underlying the social buffering of the hypothalamic–pituitary–adrenocortical axis: a review of animal models and human studies across development. Psychol. Bull. 140 , 256–282 (2014).

Slavich, G. M. & Irwin, M. R. From stress to inflammation and major depressive disorder: a social signal transduction theory of depression. Psychol. Bull. 140 , 774–815 (2014).

Chen, G. Y. & Nuñez, G. Sterile inflammation: sensing and reacting to damage. Nat. Rev. Immunol. 10 , 826–837 (2010).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Blascovich, J. & Mendes, W. B. in Handbook of Social Psychology Vol. 1, 5th edn 194–227 (John Wiley & Sons, 2010).

Flouri, E. et al. Prenatal and childhood adversity and inflammation in children: a population-based longitudinal study. Brain Behav. Immun. 87 , 524–530 (2020).

Article   CAS   PubMed   Google Scholar  

Iob, E. et al. Adverse childhood experiences and severity levels of inflammation and depression from childhood to young adulthood: a longitudinal cohort study. Mol. Psychiatry 27 , 2255–2263 (2022).

Furman, D. et al. Chronic inflammation in the etiology of disease across the life span. Nat. Med. 25 , 1822–1832 (2019).

Arseneault, L. Annual Research Review: the persistent and pervasive impact of being bullied in childhood and adolescence: implications for policy and practice. J. Child Psychol. Psychiatry 59 , 405–421 (2018).

Stewart, J. G. et al. Peer victimization and suicidal thoughts and behaviors in depressed adolescents. J. Abnorm. Child Psychol. 46 , 581–596 (2018).

Sinclair, H. C., Wilson, K. J. & Stubbs-Richardson, M. Advances in youth bullying research. Front. Psychol. 13 , 860887 (2022).

Biswas, T. et al. Global variation in the prevalence of bullying victimisation amongst adolescents: role of peer and parental supports. EClinicalMedicine 20 , 100276 (2020).

Wolke, D. & Lereya, S. T. Long-term effects of bullying. Arch. Dis. Child. 100 , 879–885 (2015).

Gini, G. & Pozzoli, T. Bullied children and psychosomatic problems: a meta-analysis. Pediatrics 132 , 720–729 (2013).

Sweeting, H. et al. Peer victimization and depression in early–mid adolescence: a longitudinal study. Br. J. Educ. Psychol. 76 , 577–594 (2006).

Lereya, S. T. et al. Being bullied during childhood and the prospective pathways to self-harm in late adolescence. J. Am. Acad. Child Adolesc. Psychiatry 52 , 608–618.e2 (2013).

Copeland, W. E. et al. Adult psychiatric outcomes of bullying and being bullied by peers in childhood and adolescence. JAMA Psychiatry 70 , 419–426 (2013).

Sharpe, H. et al. Changes in peer and sibling victimization in early adolescence: longitudinal associations with multiple indices of mental health in a prospective birth cohort study. Eur. Child Adolesc. Psychiatry 31 , 737–746 (2022).

Moore, S. E. et al. Consequences of bullying victimization in childhood and adolescence: a systematic review and meta-analysis. World J. Psychiatry 7 , 60–76 (2017).

Vaillancourt, T. et al. Longitudinal links between childhood peer victimization, internalizing and externalizing problems, and academic functioning: developmental cascades. J. Abnorm. Child Psychol. 41 , 1203–1215 (2013).

Schoeler, T. et al. Quasi-experimental evidence on short- and long-term consequences of bullying victimization: a meta-analysis. Psychol. Bull. 144 , 1229–1246 (2018).

van Geel, M. et al. Does peer victimization predict low self-esteem, or does low self-esteem predict peer victimization? Meta-analyses on longitudinal studies. Dev. Rev. 49 , 31–40 (2018).

Young-Jones, A. et al. Bullying affects more than feelings: the long-term implications of victimization on academic motivation in higher education. Social Psychology of Education 18 , 185–200 (2015).

Gaffney, H., Ttofi, M. M. & Farrington, D. P. Evaluating the effectiveness of school-bullying prevention programs: an updated meta-analytical review. Aggress. Violent Behav. 45 , 111–133 (2019).

Menesini, E. & Salmivalli, C. Bullying in schools: the state of knowledge and effective interventions. Psychol. Health Med. 22 , 240–253 (2017).

Gaffney, H., Ttofi, M. M. & Farrington, D. P. What works in anti-bullying programs? Analysis of effective intervention components. J. School Psychol. 85 , 37–56 (2021).

Marsh, H. W. et al. Peer victimization: an integrative review and cross-national test of a tripartite model. Educ. Psychol. Rev. 35 , 46 (2023).

Bonell, C. et al. Effects of the Learning Together intervention on bullying and aggression in English secondary schools (INCLUSIVE): a cluster randomised controlled trial. Lancet 392 , 2452–2464 (2018).

Herkama, S. et al. Sleeping problems partly mediate the association between victimization and depression among youth. J. Child Family Stud. 28 , 2477–2486 (2019).

Sosnowski, D. W., Kliewer, W. & Lepore, S. J. The role of sleep in the relationship between victimization and externalizing problems in adolescents. J. Youth Adolesc. 45 , 1744–1754 (2016).

Lee, K. S. & Vaillancourt, T. Developmental pathways between peer victimization, psychological functioning, disordered eating behavior, and body mass index: a review and theoretical model. Aggress. Violent Behav. 39 , 15–24 (2018).

Storch, E. A. et al. Peer victimization, psychosocial adjustment, and physical activity in overweight and at-risk-for-overweight youth. J. Pediatr. Psychol. 32 , 80–89 (2007).

Eisenberg, M. & Neumark-Sztainer, D. Peer harassment and disordered eating. Int. J. Adolesc. Med. Health 20 , 155–164 (2008).

Liang, K. et al. Food insecurity and bullying victimization among 170,618 adolescents in 59 countries. Front. Psychiatry 12 , 766804 (2021).

Uslaner, E. M. The Moral Foundations of Trust (Cambridge Univ. Press, 2002).

Rotenberg, K. J. in International Encyclopedia of the Social & Behavioral Sciences (eds Smelser, N. J. & Baltes, P. B.) 7866–7868 (Pergamon, 2001).

Neil, L. et al. Trust and childhood maltreatment: evidence of bias in appraisal of unfamiliar faces. J. Child Psychol. Psychiatry 63 , 655–662 (2022).

Turagabeci, A. R., Nakamura, K. & Takano, T. Healthy lifestyle behaviour decreasing risks of being bullied, violence and injury. PLoS ONE 3 , e1585 (2008).

Williford, A. et al. Effects of the KiVa anti-bullying program on adolescents’ depression, anxiety, and perception of peers. J. Abnorm. Child Psychol. 40 , 289–300 (2012).

Albaladejo-Blázquez, N. et al. Poor dietary habits in bullied adolescents: the moderating effects of diet on depression. Int. J. Environ. Res. Public Health 15 , 1569 (2018).

Vugteveen, J., de Bildt, A. & Timmerman, M. E. Normative data for the self-reported and parent-reported Strengths and Difficulties Questionnaire (SDQ) for ages 12–17. Child Adolesc. Psychiatry Ment. Health 16 , 5 (2022).

Kim, S. S. et al. Association between interpersonal trust, reciprocity, and depression in South Korea: a prospective analysis. PLoS ONE 7 , e30602 (2012).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Fermin, A. S. R. et al. The neuroanatomy of social trust predicts depression vulnerability. Sci. Rep. 12 , 16724 (2022).

Blakemore, S.-J. The social brain in adolescence. Nat. Rev. Neurosci. 9 , 267–277 (2008).

Frith, C. D. The social brain? Philos. Trans. R. Soc. B 362 , 671–678 (2007).

Krueger, F. & Meyer-Lindenberg, A. Toward a model of interpersonal trust drawn from neuroscience, psychology, and economics. Trends Neurosci. 42 , 92–101 (2019).

Donaldson, Z. R. & Young, L. J. Oxytocin, vasopressin, and the neurogenetics of sociality. Science 322 , 900–904 (2008).

Article   ADS   CAS   PubMed   Google Scholar  

Marsh, N. et al. Oxytocin and the neurobiology of prosocial behavior. Neuroscientist 27 , 604–619 (2021).

Yan, Z. & Kirsch, P. in The Neurobiology of Trust (ed. Krueger, F.) 315–337 (Cambridge Univ. Press, 2021).

Kosfeld, M. et al. Oxytocin increases trust in humans. Nature 435 , 673–676 (2005).

Declerck, C. H. et al. A registered replication study on oxytocin and trust. Nat. Hum. Behav. 4 , 646–655 (2020).

Neumann, I. D. & Landgraf, R. Balance of brain oxytocin and vasopressin: implications for anxiety, depression, and social behaviors. Trends Neurosci. 35 , 649–659 (2012).

Xie, S. et al. The association of oxytocin with major depressive disorder: role of confounding effects of antidepressants. Rev. Neurosci. 33 , 59–77 (2022).

Veiga, L. et al. Depressive symptomatology, temperament and oxytocin serum levels in a sample of healthy female university students. BMC Psychol. 10 , 36 (2022).

De Cagna, F. et al. The role of intranasal oxytocin in anxiety and depressive disorders: a systematic review of randomized controlled trials. Clin. Psychopharmacol. Neurosci. 17 , 1–11 (2019).

McQuaid, R. J. et al. Making room for oxytocin in understanding depression. Neurosci. Biobehav. Rev. 45 , 305–322 (2014).

Heinrichs, M. et al. Social support and oxytocin interact to suppress cortisol and subjective responses to psychosocial stress. Biol. Psychiatry 54 , 1389–1398 (2003).

Guzman-Holst, C. et al. Research review: do antibullying interventions reduce internalizing symptoms? A systematic review, meta-analysis, and meta-regression exploring intervention components, moderators, and mechanisms. J. Child Psychol. Psychiatry 63 , 1454–1465 (2022).

Joshi, H. & Fitzsimons, E. The Millennium Cohort Study: the making of a multi-purpose resource for social science and policy. Longitud. Life Course Stud. 7 , 409–430 (2016).

Plewis, I. et al. MCS: Technical Report on Sampling (Centre for Longitudinal Studies, 2004).

Goodman, R. The Strengths and Difficulties Questionnaire: a research note. J. Child Psychol. Psychiatry 38 , 581–586 (1997).

Lundh, L. G., Wångby-Lundh, M. & Bjärehed, J. Self-reported emotional and behavioral problems in Swedish 14 to 15-year-old adolescents: a study with the self-report version of the Strengths and Difficulties Questionnaire. Scand. J. Psychol. 49 , 523–532 (2008).

Hoare, E. et al. Association of child and adolescent mental health with adolescent health behaviors in the UK Millennium Cohort. JAMA Network Open 3 , e2011381 (2020).

Biddle, S. J. H. & Asare, M. Physical activity and mental health in children and adolescents: a review of reviews. Br. J. Sports Med. 45 , 886–895 (2011).

Haraden, D. A., Mullin, B. C. & Hankin, B. L. The relationship between depression and chronotype: a longitudinal assessment during childhood and adolescence. Depress. Anxiety 34 , 967–976 (2017).

Patalay, P. & Fitzsimons, E. Development and predictors of mental ill-health and wellbeing from childhood to adolescence. Soc. Psychiatry Psychiatric Epidemiol. 53 , 1311–1323 (2018).

Patalay, P. & Hardman, C. A. Comorbidity, codevelopment, and temporal associations between body mass index and internalizing symptoms from early childhood to adolescence. JAMA Psychiatry 76 , 721–729 (2019).

Mueller, M. A. E. & Flouri, E. Urban adolescence: the role of neighbourhood greenspace in mental well-being. Front. Psychol. 12 , 712065 (2021).

Blakemore, S.-J. Adolescence and mental health. Lancet 393 , 2030–2031 (2019).

Fitzsimons, E. et al. Poverty dynamics and parental mental health: determinants of childhood mental health in the UK. Soc. Sci. Med. 175 , 43–51 (2017).

Church, D. & Midouhas, E. Data Note: MEDIX Air Pollution Data at Ward Level, Linked to MCS1 and MCS2 (UCL Institute of Education, 2016).

Raghunathan, T. E. et al. A multivariate technique for multiply imputing missing values using a sequence of regression models. Survey Methodol. 27 , 85–96 (2001).

Google Scholar  

Rubin, D. B. Multiple Imputation for Nonresponse in Surveys (John Wiley, 1987).

R Core Team. R: a language and environment for statistical computing. R Foundation for Statistical Computing https://www.R-project.org/ (2018).

Tsomokos, D. I. & Dunbar, R. I. M. The role of religion in adolescent mental health: faith as a moderator of the relationship between distrust and depression. Religion Brain Behav. https://doi.org/10.1080/2153599X.2023.2248230 (2023).

Download references

Acknowledgements

D.I.T. was partially supported by Alphablocks Nursery School Ltd. G.M.S. was supported by grant #OPR21101 from the California Governor’s Office of Planning and Research/California Initiative to Advance Precision Medicine.

Author information

Authors and affiliations.

School of Psychology and Neuroscience, University of Glasgow, Glasgow, UK

Dimitris I. Tsomokos

Department of Psychiatry and Biobehavioral Sciences, University of California, Los Angeles, CA, USA

George M. Slavich

You can also search for this author in PubMed   Google Scholar

Contributions

This article is the work of both authors (D.I.T. and G.M.S.).

Corresponding author

Correspondence to George M. Slavich .

Ethics declarations

Competing interests.

The authors have no competing interests. The funders had no role in designing or planning this study; in collecting, analyzing or interpreting the data; in writing the article; or in deciding to submit this article for publication.

Peer review

Peer review information.

Nature Mental Health thanks Carolina Guzman Holst and the other, anonymous, reviewers for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information.

Supplementary text, Tables 1–6, Fig. 1 and References.

Reporting Summary

Rights and permissions.

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Cite this article.

Tsomokos, D.I., Slavich, G.M. Bullying fosters interpersonal distrust and degrades adolescent mental health as predicted by Social Safety Theory. Nat. Mental Health 2 , 328–336 (2024). https://doi.org/10.1038/s44220-024-00203-7

Download citation

Received : 07 July 2023

Accepted : 08 January 2024

Published : 13 February 2024

Issue Date : March 2024

DOI : https://doi.org/10.1038/s44220-024-00203-7

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

• Research article
• Open access
• Published: 14 December 2021

Bullying at school and mental health problems among adolescents: a repeated cross-sectional study

• Håkan Källmén 1 &
• Mats Hallgren   ORCID: orcid.org/0000-0002-0599-2403 2  

Child and Adolescent Psychiatry and Mental Health volume  15 , Article number:  74 ( 2021 ) Cite this article

110k Accesses

18 Citations

37 Altmetric

Metrics details

To examine recent trends in bullying and mental health problems among adolescents and the association between them.

A questionnaire measuring mental health problems, bullying at school, socio-economic status, and the school environment was distributed to all secondary school students aged 15 (school-year 9) and 18 (school-year 11) in Stockholm during 2014, 2018, and 2020 (n = 32,722). Associations between bullying and mental health problems were assessed using logistic regression analyses adjusting for relevant demographic, socio-economic, and school-related factors.

The prevalence of bullying remained stable and was highest among girls in year 9; range = 4.9% to 16.9%. Mental health problems increased; range = + 1.2% (year 9 boys) to + 4.6% (year 11 girls) and were consistently higher among girls (17.2% in year 11, 2020). In adjusted models, having been bullied was detrimentally associated with mental health (OR = 2.57 [2.24–2.96]). Reports of mental health problems were four times higher among boys who had been bullied compared to those not bullied. The corresponding figure for girls was 2.4 times higher.

Conclusions

Exposure to bullying at school was associated with higher odds of mental health problems. Boys appear to be more vulnerable to the deleterious effects of bullying than girls.

Introduction

Bullying involves repeated hurtful actions between peers where an imbalance of power exists [ 1 ]. Arseneault et al. [ 2 ] conducted a review of the mental health consequences of bullying for children and adolescents and found that bullying is associated with severe symptoms of mental health problems, including self-harm and suicidality. Bullying was shown to have detrimental effects that persist into late adolescence and contribute independently to mental health problems. Updated reviews have presented evidence indicating that bullying is causative of mental illness in many adolescents [ 3 , 4 ].

There are indications that mental health problems are increasing among adolescents in some Nordic countries. Hagquist et al. [ 5 ] examined trends in mental health among Scandinavian adolescents (n = 116, 531) aged 11–15 years between 1993 and 2014. Mental health problems were operationalized as difficulty concentrating, sleep disorders, headache, stomach pain, feeling tense, sad and/or dizzy. The study revealed increasing rates of adolescent mental health problems in all four counties (Finland, Sweden, Norway, and Denmark), with Sweden experiencing the sharpest increase among older adolescents, particularly girls. Worsening adolescent mental health has also been reported in the United Kingdom. A study of 28,100 school-aged adolescents in England found that two out of five young people scored above thresholds for emotional problems, conduct problems or hyperactivity [ 6 ]. Female gender, deprivation, high needs status (educational/social), ethnic background, and older age were all associated with higher odds of experiencing mental health difficulties.

Bullying is shown to increase the risk of poor mental health and may partly explain these detrimental changes. Le et al. [ 7 ] reported an inverse association between bullying and mental health among 11–16-year-olds in Vietnam. They also found that poor mental health can make some children and adolescents more vulnerable to bullying at school. Bayer et al. [ 8 ] examined links between bullying at school and mental health among 8–9-year-old children in Australia. Those who experienced bullying more than once a week had poorer mental health than children who experienced bullying less frequently. Friendships moderated this association, such that children with more friends experienced fewer mental health problems (protective effect). Hysing et al. [ 9 ] investigated the association between experiences of bullying (as a victim or perpetrator) and mental health, sleep disorders, and school performance among 16–19 year olds from Norway (n = 10,200). Participants were categorized as victims, bullies, or bully-victims (that is, victims who also bullied others). All three categories were associated with worse mental health, school performance, and sleeping difficulties. Those who had been bullied also reported more emotional problems, while those who bullied others reported more conduct disorders [ 9 ].

As most adolescents spend a considerable amount of time at school, the school environment has been a major focus of mental health research [ 10 , 11 ]. In a recent review, Saminathen et al. [ 12 ] concluded that school is a potential protective factor against mental health problems, as it provides a socially supportive context and prepares students for higher education and employment. However, it may also be the primary setting for protracted bullying and stress [ 13 ]. Another factor associated with adolescent mental health is parental socio-economic status (SES) [ 14 ]. A systematic review indicated that lower parental SES is associated with poorer adolescent mental health [ 15 ]. However, no previous studies have examined whether SES modifies or attenuates the association between bullying and mental health. Similarly, it remains unclear whether school related factors, such as school grades and the school environment, influence the relationship between bullying and mental health. This information could help to identify those adolescents most at risk of harm from bullying.

To address these issues, we investigated the prevalence of bullying at school and mental health problems among Swedish adolescents aged 15–18 years between 2014 and 2020 using a population-based school survey. We also examined associations between bullying at school and mental health problems adjusting for relevant demographic, socioeconomic, and school-related factors. We hypothesized that: (1) bullying and adolescent mental health problems have increased over time; (2) There is an association between bullying victimization and mental health, so that mental health problems are more prevalent among those who have been victims of bullying; and (3) that school-related factors would attenuate the association between bullying and mental health.

Participants

The Stockholm school survey is completed every other year by students in lower secondary school (year 9—compulsory) and upper secondary school (year 11). The survey is mandatory for public schools, but voluntary for private schools. The purpose of the survey is to help inform decision making by local authorities that will ultimately improve students’ wellbeing. The questions relate to life circumstances, including SES, schoolwork, bullying, drug use, health, and crime. Non-completers are those who were absent from school when the survey was completed (< 5%). Response rates vary from year to year but are typically around 75%. For the current study data were available for 2014, 2018 and 2020. In 2014; 5235 boys and 5761 girls responded, in 2018; 5017 boys and 5211 girls responded, and in 2020; 5633 boys and 5865 girls responded (total n = 32,722). Data for the exposure variable, bullied at school, were missing for 4159 students, leaving 28,563 participants in the crude model. The fully adjusted model (described below) included 15,985 participants. The mean age in grade 9 was 15.3 years (SD = 0.51) and in grade 11, 17.3 years (SD = 0.61). As the data are completely anonymous, the study was exempt from ethical approval according to an earlier decision from the Ethical Review Board in Stockholm (2010-241 31-5). Details of the survey are available via a website [ 16 ], and are described in a previous paper [ 17 ].

Students completed the questionnaire during a school lesson, placed it in a sealed envelope and handed it to their teacher. Student were permitted the entire lesson (about 40 min) to complete the questionnaire and were informed that participation was voluntary (and that they were free to cancel their participation at any time without consequences). Students were also informed that the Origo Group was responsible for collection of the data on behalf of the City of Stockholm.

Study outcome

Mental health problems were assessed by using a modified version of the Psychosomatic Problem Scale [ 18 ] shown to be appropriate for children and adolescents and invariant across gender and years. The scale was later modified [ 19 ]. In the modified version, items about difficulty concentrating and feeling giddy were deleted and an item about ‘life being great to live’ was added. Seven different symptoms or problems, such as headaches, depression, feeling fear, stomach problems, difficulty sleeping, believing it’s great to live (coded negatively as seldom or rarely) and poor appetite were used. Students who responded (on a 5-point scale) that any of these problems typically occurs ‘at least once a week’ were considered as having indicators of a mental health problem. Cronbach alpha was 0.69 across the whole sample. Adding these problem areas, a total index was created from 0 to 7 mental health symptoms. Those who scored between 0 and 4 points on the total symptoms index were considered to have a low indication of mental health problems (coded as 0); those who scored between 5 and 7 symptoms were considered as likely having mental health problems (coded as 1).

Primary exposure

Experiences of bullying were measured by the following two questions: Have you felt bullied or harassed during the past school year? Have you been involved in bullying or harassing other students during this school year? Alternatives for the first question were: yes or no with several options describing how the bullying had taken place (if yes). Alternatives indicating emotional bullying were feelings of being mocked, ridiculed, socially excluded, or teased. Alternatives indicating physical bullying were being beaten, kicked, forced to do something against their will, robbed, or locked away somewhere. The response alternatives for the second question gave an estimation of how often the respondent had participated in bullying others (from once to several times a week). Combining the answers to these two questions, five different categories of bullying were identified: (1) never been bullied and never bully others; (2) victims of emotional (verbal) bullying who have never bullied others; (3) victims of physical bullying who have never bullied others; (4) victims of bullying who have also bullied others; and (5) perpetrators of bullying, but not victims. As the number of positive cases in the last three categories was low (range = 3–15 cases) bully categories 2–4 were combined into one primary exposure variable: ‘bullied at school’.

Assessment year was operationalized as the year when data was collected: 2014, 2018, and 2020. Age was operationalized as school grade 9 (15–16 years) or 11 (17–18 years). Gender was self-reported (boy or girl). The school situation To assess experiences of the school situation, students responded to 18 statements about well-being in school, participation in important school matters, perceptions of their teachers, and teaching quality. Responses were given on a four-point Likert scale ranging from ‘do not agree at all’ to ‘fully agree’. To reduce the 18-items down to their essential factors, we performed a principal axis factor analysis. Results showed that the 18 statements formed five factors which, according to the Kaiser criterion (eigen values > 1) explained 56% of the covariance in the student’s experience of the school situation. The five factors identified were: (1) Participation in school; (2) Interesting and meaningful work; (3) Feeling well at school; (4) Structured school lessons; and (5) Praise for achievements. For each factor, an index was created that was dichotomised (poor versus good circumstance) using the median-split and dummy coded with ‘good circumstance’ as reference. A description of the items included in each factor is available as Additional file 1 . Socio-economic status (SES) was assessed with three questions about the education level of the student’s mother and father (dichotomized as university degree versus not), and the amount of spending money the student typically received for entertainment each month (> SEK 1000 [approximately $120] versus less). Higher parental education and more spending money were used as reference categories. School grades in Swedish, English, and mathematics were measured separately on a 7-point scale and dichotomized as high (grades A, B, and C) versus low (grades D, E, and F). High school grades were used as the reference category.

Statistical analyses

The prevalence of mental health problems and bullying at school are presented using descriptive statistics, stratified by survey year (2014, 2018, 2020), gender, and school year (9 versus 11). As noted, we reduced the 18-item questionnaire assessing school function down to five essential factors by conducting a principal axis factor analysis (see Additional file 1 ). We then calculated the association between bullying at school (defined above) and mental health problems using multivariable logistic regression. Results are presented as odds ratios (OR) with 95% confidence intervals (Cis). To assess the contribution of SES and school-related factors to this association, three models are presented: Crude, Model 1 adjusted for demographic factors: age, gender, and assessment year; Model 2 adjusted for Model 1 plus SES (parental education and student spending money), and Model 3 adjusted for Model 2 plus school-related factors (school grades and the five factors identified in the principal factor analysis). These covariates were entered into the regression models in three blocks, where the final model represents the fully adjusted analyses. In all models, the category ‘not bullied at school’ was used as the reference. Pseudo R-square was calculated to estimate what proportion of the variance in mental health problems was explained by each model. Unlike the R-square statistic derived from linear regression, the Pseudo R-square statistic derived from logistic regression gives an indicator of the explained variance, as opposed to an exact estimate, and is considered informative in identifying the relative contribution of each model to the outcome [ 20 ]. All analyses were performed using SPSS v. 26.0.

Prevalence of bullying at school and mental health problems

Estimates of the prevalence of bullying at school and mental health problems across the 12 strata of data (3 years × 2 school grades × 2 genders) are shown in Table 1 . The prevalence of bullying at school increased minimally (< 1%) between 2014 and 2020, except among girls in grade 11 (2.5% increase). Mental health problems increased between 2014 and 2020 (range = 1.2% [boys in year 11] to 4.6% [girls in year 11]); were three to four times more prevalent among girls (range = 11.6% to 17.2%) compared to boys (range = 2.6% to 4.9%); and were more prevalent among older adolescents compared to younger adolescents (range = 1% to 3.1% higher). Pooling all data, reports of mental health problems were four times more prevalent among boys who had been victims of bullying compared to those who reported no experiences with bullying. The corresponding figure for girls was two and a half times as prevalent.

Associations between bullying at school and mental health problems

Table 2 shows the association between bullying at school and mental health problems after adjustment for relevant covariates. Demographic factors, including female gender (OR = 3.87; CI 3.48–4.29), older age (OR = 1.38, CI 1.26–1.50), and more recent assessment year (OR = 1.18, CI 1.13–1.25) were associated with higher odds of mental health problems. In Model 2, none of the included SES variables (parental education and student spending money) were associated with mental health problems. In Model 3 (fully adjusted), the following school-related factors were associated with higher odds of mental health problems: lower grades in Swedish (OR = 1.42, CI 1.22–1.67); uninteresting or meaningless schoolwork (OR = 2.44, CI 2.13–2.78); feeling unwell at school (OR = 1.64, CI 1.34–1.85); unstructured school lessons (OR = 1.31, CI = 1.16–1.47); and no praise for achievements (OR = 1.19, CI 1.06–1.34). After adjustment for all covariates, being bullied at school remained associated with higher odds of mental health problems (OR = 2.57; CI 2.24–2.96). Demographic and school-related factors explained 12% and 6% of the variance in mental health problems, respectively (Pseudo R-Square). The inclusion of socioeconomic factors did not alter the variance explained.

Our findings indicate that mental health problems increased among Swedish adolescents between 2014 and 2020, while the prevalence of bullying at school remained stable (< 1% increase), except among girls in year 11, where the prevalence increased by 2.5%. As previously reported [ 5 , 6 ], mental health problems were more common among girls and older adolescents. These findings align with previous studies showing that adolescents who are bullied at school are more likely to experience mental health problems compared to those who are not bullied [ 3 , 4 , 9 ]. This detrimental relationship was observed after adjustment for school-related factors shown to be associated with adolescent mental health [ 10 ].

A novel finding was that boys who had been bullied at school reported a four-times higher prevalence of mental health problems compared to non-bullied boys. The corresponding figure for girls was 2.5 times higher for those who were bullied compared to non-bullied girls, which could indicate that boys are more vulnerable to the deleterious effects of bullying than girls. Alternatively, it may indicate that boys are (on average) bullied more frequently or more intensely than girls, leading to worse mental health. Social support could also play a role; adolescent girls often have stronger social networks than boys and could be more inclined to voice concerns about bullying to significant others, who in turn may offer supports which are protective [ 21 ]. Related studies partly confirm this speculative explanation. An Estonian study involving 2048 children and adolescents aged 10–16 years found that, compared to girls, boys who had been bullied were more likely to report severe distress, measured by poor mental health and feelings of hopelessness [ 22 ].

Other studies suggest that heritable traits, such as the tendency to internalize problems and having low self-esteem are associated with being a bully-victim [ 23 ]. Genetics are understood to explain a large proportion of bullying-related behaviors among adolescents. A study from the Netherlands involving 8215 primary school children found that genetics explained approximately 65% of the risk of being a bully-victim [ 24 ]. This proportion was similar for boys and girls. Higher than average body mass index (BMI) is another recognized risk factor [ 25 ]. A recent Australian trial involving 13 schools and 1087 students (mean age = 13 years) targeted adolescents with high-risk personality traits (hopelessness, anxiety sensitivity, impulsivity, sensation seeking) to reduce bullying at school; both as victims and perpetrators [ 26 ]. There was no significant intervention effect for bullying victimization or perpetration in the total sample. In a secondary analysis, compared to the control schools, intervention school students showed greater reductions in victimization, suicidal ideation, and emotional symptoms. These findings potentially support targeting high-risk personality traits in bullying prevention [ 26 ].

The relative stability of bullying at school between 2014 and 2020 suggests that other factors may better explain the increase in mental health problems seen here. Many factors could be contributing to these changes, including the increasingly competitive labour market, higher demands for education, and the rapid expansion of social media [ 19 , 27 , 28 ]. A recent Swedish study involving 29,199 students aged between 11 and 16 years found that the effects of school stress on psychosomatic symptoms have become stronger over time (1993–2017) and have increased more among girls than among boys [ 10 ]. Research is needed examining possible gender differences in perceived school stress and how these differences moderate associations between bullying and mental health.

Strengths and limitations

Strengths of the current study include the large participant sample from diverse schools; public and private, theoretical and practical orientations. The survey included items measuring diverse aspects of the school environment; factors previously linked to adolescent mental health but rarely included as covariates in studies of bullying and mental health. Some limitations are also acknowledged. These data are cross-sectional which means that the direction of the associations cannot be determined. Moreover, all the variables measured were self-reported. Previous studies indicate that students tend to under-report bullying and mental health problems [ 29 ]; thus, our results may underestimate the prevalence of these behaviors.

In conclusion, consistent with our stated hypotheses, we observed an increase in self-reported mental health problems among Swedish adolescents, and a detrimental association between bullying at school and mental health problems. Although bullying at school does not appear to be the primary explanation for these changes, bullying was detrimentally associated with mental health after adjustment for relevant demographic, socio-economic, and school-related factors, confirming our third hypothesis. The finding that boys are potentially more vulnerable than girls to the deleterious effects of bullying should be replicated in future studies, and the mechanisms investigated. Future studies should examine the longitudinal association between bullying and mental health, including which factors mediate/moderate this relationship. Epigenetic studies are also required to better understand the complex interaction between environmental and biological risk factors for adolescent mental health [ 24 ].

Availability of data and materials

Data requests will be considered on a case-by-case basis; please email the corresponding author.

Code availability

Not applicable.

Olweus D. School bullying: development and some important challenges. Ann Rev Clin Psychol. 2013;9(9):751–80. https://doi.org/10.1146/annurev-clinpsy-050212-185516 .

Article   Google Scholar  

Arseneault L, Bowes L, Shakoor S. Bullying victimization in youths and mental health problems: “Much ado about nothing”? Psychol Med. 2010;40(5):717–29. https://doi.org/10.1017/S0033291709991383 .

Article   CAS   PubMed   Google Scholar  

Arseneault L. The long-term impact of bullying victimization on mental health. World Psychiatry. 2017;16(1):27–8. https://doi.org/10.1002/wps.20399 .

Article   PubMed   PubMed Central   Google Scholar  

Moore SE, Norman RE, Suetani S, Thomas HJ, Sly PD, Scott JG. Consequences of bullying victimization in childhood and adolescence: a systematic review and meta-analysis. World J Psychiatry. 2017;7(1):60–76. https://doi.org/10.5498/wjp.v7.i1.60 .

Hagquist C, Due P, Torsheim T, Valimaa R. Cross-country comparisons of trends in adolescent psychosomatic symptoms—a Rasch analysis of HBSC data from four Nordic countries. Health Qual Life Outcomes. 2019;17(1):27. https://doi.org/10.1186/s12955-019-1097-x .

Deighton J, Lereya ST, Casey P, Patalay P, Humphrey N, Wolpert M. Prevalence of mental health problems in schools: poverty and other risk factors among 28 000 adolescents in England. Br J Psychiatry. 2019;215(3):565–7. https://doi.org/10.1192/bjp.2019.19 .

Article   PubMed Central   Google Scholar  

Le HTH, Tran N, Campbell MA, Gatton ML, Nguyen HT, Dunne MP. Mental health problems both precede and follow bullying among adolescents and the effects differ by gender: a cross-lagged panel analysis of school-based longitudinal data in Vietnam. Int J Ment Health Syst. 2019. https://doi.org/10.1186/s13033-019-0291-x .

Bayer JK, Mundy L, Stokes I, Hearps S, Allen N, Patton G. Bullying, mental health and friendship in Australian primary school children. Child Adolesc Ment Health. 2018;23(4):334–40. https://doi.org/10.1111/camh.12261 .

Article   PubMed   Google Scholar  

Hysing M, Askeland KG, La Greca AM, Solberg ME, Breivik K, Sivertsen B. Bullying involvement in adolescence: implications for sleep, mental health, and academic outcomes. J Interpers Violence. 2019. https://doi.org/10.1177/0886260519853409 .

Hogberg B, Strandh M, Hagquist C. Gender and secular trends in adolescent mental health over 24 years—the role of school-related stress. Soc Sci Med. 2020. https://doi.org/10.1016/j.socscimed.2020.112890 .

Kidger J, Araya R, Donovan J, Gunnell D. The effect of the school environment on the emotional health of adolescents: a systematic review. Pediatrics. 2012;129(5):925–49. https://doi.org/10.1542/peds.2011-2248 .

Saminathen MG, Låftman SB, Modin B. En fungerande skola för alla: skolmiljön som skyddsfaktor för ungas psykiska välbefinnande. [A functioning school for all: the school environment as a protective factor for young people’s mental well-being]. Socialmedicinsk tidskrift [Soc Med]. 2020;97(5–6):804–16.

Google Scholar  

Bibou-Nakou I, Tsiantis J, Assimopoulos H, Chatzilambou P, Giannakopoulou D. School factors related to bullying: a qualitative study of early adolescent students. Soc Psychol Educ. 2012;15(2):125–45. https://doi.org/10.1007/s11218-012-9179-1 .

Vukojevic M, Zovko A, Talic I, Tanovic M, Resic B, Vrdoljak I, Splavski B. Parental socioeconomic status as a predictor of physical and mental health outcomes in children—literature review. Acta Clin Croat. 2017;56(4):742–8. https://doi.org/10.20471/acc.2017.56.04.23 .

Reiss F. Socioeconomic inequalities and mental health problems in children and adolescents: a systematic review. Soc Sci Med. 2013;90:24–31. https://doi.org/10.1016/j.socscimed.2013.04.026 .

Stockholm City. Stockholmsenkät (The Stockholm Student Survey). 2021. https://start.stockholm/aktuellt/nyheter/2020/09/presstraff-stockholmsenkaten-2020/ . Accessed 19 Nov 2021.

Zeebari Z, Lundin A, Dickman PW, Hallgren M. Are changes in alcohol consumption among swedish youth really occurring “in concert”? A new perspective using quantile regression. Alc Alcohol. 2017;52(4):487–95. https://doi.org/10.1093/alcalc/agx020 .

Hagquist C. Psychometric properties of the PsychoSomatic Problems Scale: a Rasch analysis on adolescent data. Social Indicat Res. 2008;86(3):511–23. https://doi.org/10.1007/s11205-007-9186-3 .

Hagquist C. Ungas psykiska hälsa i Sverige–komplexa trender och stora kunskapsluckor [Young people’s mental health in Sweden—complex trends and large knowledge gaps]. Socialmedicinsk tidskrift [Soc Med]. 2013;90(5):671–83.

Wu W, West SG. Detecting misspecification in mean structures for growth curve models: performance of pseudo R(2)s and concordance correlation coefficients. Struct Equ Model. 2013;20(3):455–78. https://doi.org/10.1080/10705511.2013.797829 .

Holt MK, Espelage DL. Perceived social support among bullies, victims, and bully-victims. J Youth Adolscence. 2007;36(8):984–94. https://doi.org/10.1007/s10964-006-9153-3 .

Mark L, Varnik A, Sisask M. Who suffers most from being involved in bullying-bully, victim, or bully-victim? J Sch Health. 2019;89(2):136–44. https://doi.org/10.1111/josh.12720 .

Tsaousis I. The relationship of self-esteem to bullying perpetration and peer victimization among schoolchildren and adolescents: a meta-analytic review. Aggress Violent Behav. 2016;31:186–99. https://doi.org/10.1016/j.avb.2016.09.005 .

Veldkamp SAM, Boomsma DI, de Zeeuw EL, van Beijsterveldt CEM, Bartels M, Dolan CV, van Bergen E. Genetic and environmental influences on different forms of bullying perpetration, bullying victimization, and their co-occurrence. Behav Genet. 2019;49(5):432–43. https://doi.org/10.1007/s10519-019-09968-5 .

Janssen I, Craig WM, Boyce WF, Pickett W. Associations between overweight and obesity with bullying behaviors in school-aged children. Pediatrics. 2004;113(5):1187–94. https://doi.org/10.1542/peds.113.5.1187 .

Kelly EV, Newton NC, Stapinski LA, Conrod PJ, Barrett EL, Champion KE, Teesson M. A novel approach to tackling bullying in schools: personality-targeted intervention for adolescent victims and bullies in Australia. J Am Acad Child Adolesc Psychiatry. 2020;59(4):508. https://doi.org/10.1016/j.jaac.2019.04.010 .

Gunnell D, Kidger J, Elvidge H. Adolescent mental health in crisis. BMJ. 2018. https://doi.org/10.1136/bmj.k2608 .

O’Reilly M, Dogra N, Whiteman N, Hughes J, Eruyar S, Reilly P. Is social media bad for mental health and wellbeing? Exploring the perspectives of adolescents. Clin Child Psychol Psychiatry. 2018;23:601–13.

Unnever JD, Cornell DG. Middle school victims of bullying: who reports being bullied? Aggr Behav. 2004;30(5):373–88. https://doi.org/10.1002/ab.20030 .

Download references

Acknowledgements

Authors are grateful to the Department for Social Affairs, Stockholm, for permission to use data from the Stockholm School Survey.

Open access funding provided by Karolinska Institute. None to declare.

Author information

Authors and affiliations.

Stockholm Prevents Alcohol and Drug Problems (STAD), Center for Addiction Research and Department of Clinical Neuroscience, Karolinska Institutet, Solna, Sweden

Håkan Källmén

Epidemiology of Psychiatric Conditions, Substance Use and Social Environment (EPiCSS), Department of Global Public Health, Karolinska Institutet, Level 6, Solnavägen 1e, Solna, Sweden

Mats Hallgren

You can also search for this author in PubMed   Google Scholar

Contributions

HK conceived the study and analyzed the data (with input from MH). HK and MH interpreted the data and jointly wrote the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Mats Hallgren .

Ethics declarations

Ethics approval and consent to participate.

As the data are completely anonymous, the study was exempt from ethical approval according to an earlier decision from the Ethical Review Board in Stockholm (2010-241 31-5).

Consent for publication

Competing interests.

The authors declare that they have no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1..

Principal factor analysis description.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Källmén, H., Hallgren, M. Bullying at school and mental health problems among adolescents: a repeated cross-sectional study. Child Adolesc Psychiatry Ment Health 15 , 74 (2021). https://doi.org/10.1186/s13034-021-00425-y

Download citation

Received : 05 October 2021

Accepted : 23 November 2021

Published : 14 December 2021

DOI : https://doi.org/10.1186/s13034-021-00425-y

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Mental health
• Adolescents
• School-related factors
• Gender differences

Child and Adolescent Psychiatry and Mental Health

ISSN: 1753-2000

• Submission enquiries: [email protected]

Advertisement

Understanding Bullying and Cyberbullying Through an Ecological Systems Framework: the Value of Qualitative Interviewing in a Mixed Methods Approach

• Original Article
• Open access
• Published: 10 May 2022
• Volume 4 , pages 220–229, ( 2022 )

Cite this article

You have full access to this open access article

• Faye Mishna   ORCID: orcid.org/0000-0003-2538-826X 1 ,
• Arija Birze   ORCID: orcid.org/0000-0002-1988-8383 1 &
• Andrea Greenblatt   ORCID: orcid.org/0000-0002-6964-8193 1  

5195 Accesses

8 Citations

4 Altmetric

Explore all metrics

Recognized as complex and relational, researchers endorse a systems/social-ecological framework in examining bullying and cyberbullying. According to this framework, bullying and cyberbullying are examined across the nested social contexts in which youth live—encompassing individual features; relationships including family, peers, and educators; and ecological conditions such as digital technology. Qualitative inquiry of bullying and cyberbullying provides a research methodology capable of bringing to the fore salient discourses such as dominant social norms and otherwise invisible nuances such as motivations and dilemmas, which might not be accessed through quantitative studies. Through use of a longitudinal and multi-perspective mixed methods study, the purpose of the current paper is to demonstrate the ways qualitative interviews contextualize quantitative findings and to present novel discussion of how qualitative interviews explain and enrich the quantitative findings. The following thematic areas emerged and are discussed: augmenting quantitative findings through qualitative interviews, contextualizing new or rapidly evolving areas of research, capturing nuances and complexity of perspectives, and providing moments for self-reflection and opportunities for learning.

Similar content being viewed by others

A Comparison of Traditional Victims, Cyber Victims, Traditional-Cyber Victims, and Uninvolved Adolescents: A Social-Ecological Framework

A qualitative meta-study of youth voice and co-participatory research practices: informing cyber/bullying research methodologies, a qualitative exploration of college students’ perceptions of cyberbullying, explore related subjects.

• Artificial Intelligence

Avoid common mistakes on your manuscript.

Introduction

Bullying and cyberbullying are increasingly recognized as complex phenomena that are considered relationship problems (Mishna et al., 2021a ; Pepler et al., 2010 ; Pepler, 2006 ; Spears et al., 2009 ). Appreciating that individuals are embedded in and both shape and are shaped by systems of relationships (Bronfenbrenner & Morris, 2007 ), researchers often endorse an ecological systems framework as paramount and comprehensive in examining bullying and cyberbullying phenomena Footnote 1 (Espelage, 2014 ; Newman et al., 2018 ; Thornberg, 2015 , 2018 ). According to this approach, individuals are embedded in and affected by interconnected and layered systems (Bronfenbrenner, 1979 , 1992 ). Children’s social-emotional development at school is consequently shaped not only by children’s relationships with their teachers and peers, but also by the interconnections between these relationships and the other layers of social ecology, all of which are considered to contribute to social behavioral patterns (O'Moore & Minton, 2005 ). Bullying and cyberbullying are examined across the nested social contexts in which youth live—encompassing individual features, peer relationships, school, family, and ecological climate such as societal norms and conditions as well as online technology (Cross et al., 2015 ; Johnson, 2010 ; Nesi et al., 2018 ). An ecological systems framework is considered an overarching approach that many theories complement and within which they fit (Bauman & Yoon, 2014 ).

The purpose of the current paper is to demonstrate the contributions of qualitative research in understanding the phenomena of bullying and cyberbullying and enriching and complementing the findings of quantitative methodology (Creswell & Creswell, 2018 ). Qualitative inquiry of bullying and cyberbullying provides a research methodology capable of bringing to the fore salient discourses and otherwise invisible nuances that might not be accessed through quantitative studies (Dennehy et al., 2020 ).

There are advantages to utilizing mixed methods in conducting research on various topics including cyberbullying (Creswell & Creswell, 2018 ). When engaging with complex phenomena such as cyberbullying, conceptual and methodological multiplicity offers distinct insights into research questions (McKim, 2017 ; Thornberg, 2011 ). When quantitative and qualitative research are used in combination, it is possible to obtain deeper as well as more comprehensive and accurate understanding of young people’s experiences, which increases the likelihood of informing strategies and responses that can effectively address the needs of children and adolescents (Crivello et al., 2009 ; Darbyshire et al., 2005 ; Fevre et al., 2010 ). The quality of findings may be strengthened when researchers use mixed methods because the data are triangulated (Crivello et al., 2009 ). Data generated through diverse research methods can both complement and contradict each other, which offers an opportunity to better understand the complexities of cyberbullying (Hemming, 2008 ). While quantitative approaches strive for objectivity by examining general concepts, such as cyberbullying, and parceling those concepts into specific, concrete, and understandable behaviors (Fevre et al., 2010 ), qualitative interviews give voice to children and youth, enabling them to express their thoughts and feelings about themselves, their relationships, environments, and the world in which they live (Mishna et al., 2004 ; Chaumba, 2013 ; Dennehy et al., 2020 ; Patton et al., 2017 ).

Through qualitative interviewing, we can step outside the bounds of adult thinking, gaining insights and discovering unanticipated differences in the perceptions of adults and children (Dennehy et al., 2020 ; O’Farrelly, 2021 ). To understand the phenomena of bullying and cyberbullying and inform effective prevention and intervention strategies, it is argued, children’s own views, “are at the heart of these efforts” (O’Farrelly, 2021 , p. 43). Thus, we present findings from the qualitative component of our Canadian federally funded mixed methods longitudinal study on cyberbullying from the perspectives of school-aged youth and their parents and teachers, entitled Motivations for Cyber Bullying: A Longitudinal and Multi-Perspective Inquiry Footnote 2 (Mishna et al., 2016 ).

Background Study Description

The objectives of our longitudinal mixed methods study were to (1) explore youth experiences and perspectives and their parents’ and teachers’ conceptions of cyberbullying; (2) explore how youth and adults view the underlying motivations for cyberbullying; (3) document the prevalence rates of cyberbullying victimization, witnessing, and perpetration; (4) identify risk and protective factors for cyberbullying involvement; and (5) explore social, mental health, and health consequences of cyberbullying among children and youth aged 9 to 18 (grades 4, 7, and 10) over 3 years.

In addressing the objectives, we use an explanatory sequential mixed methods design (Creswell & Creswell, 2018 ). The study comprised a 2-phase data collection approach in which we first collected the quantitative data and then used findings from the first phase to design and plan the qualitative data phase. The quantitative findings informed both our selection of interview participants and the focus of questions we wanted to explore further in the interviews. The overall intent of the qualitative interviews was to enrich and expand upon the quantitative findings and perhaps generate and explore similarities and contradictions (Creswell & Creswell, 2018 ). In the current paper, we briefly review key quantitative findings. We then discuss the qualitative findings and how they provide more depth and insight and demonstrate the complexities of bullying and cyberbullying motivations, behaviors, and attitudes. In so doing, we present novel discussions of how the qualitative interviews augment the quantitative findings.

Participants

Three participant groups were included in the baseline study sample: (1) students in 4th ( n  = 160), 7th ( n  = 243), and 10th ( n  = 267) grades; (2) their teachers ( n  = 103); and (3) their parents ( n  = 246). A stratified random sampling strategy was utilized to select participants. First, a random sample of 19 schools was drawn from one of the largest school boards in North America. Schools were stratified into three categories of need (low, medium, and high) based on an index developed by the school board that ranked schools on external challenges to student achievement (Toronto District School Board, 2014 ). This stratification ensured representation of ethno-cultural and socioeconomic diversity—factors that potentially impact access to Information and Communication Technologies (ICTs), experiences of cyberbullying, and the manifestation of negative outcomes (Lenhart et al., 2015 ; Steeves & Marx, 2014 ). In year 3 of the study, 10 additional schools were recruited for participation to follow those students transitioning from elementary/middle school to middle/secondary school. A total of 29 schools participated in the study. All students in the selected grades at the original participating schools were invited to participate, as were their parents and teachers.

Participating students and their parents provided data in all 3 years of the study, while matching teachers provided data in year 1 only (as student participants’ teachers changed each year). All three participant groups completed quantitative questionnaire packages, and a sub-sample of each group participated in individual interviews. Quantitative data were collected from students and parents in each year of the study, while qualitative data were collected during years 1 and 3, to allow for enough time to elapse for changes in perceptions of cyberbullying to emerge.

Quantitative Measures and Analysis

In year 1, students completed a 45–60-min quantitative questionnaire package in the school setting, while parents completed a questionnaire package by mail. Questionnaires for teachers, which took approximately 45–60 min to complete, were administered in the participating schools. This study utilized several quantitative measures, including standardized measures and measures developed specifically for the study. Student, parent, and teacher surveys obtained information related to experiences with bullying/cyberbullying (Mishna et al., 2012 ; Unpublished Survey), socio-demographics, and Information and Communication Technology (ICT) use. Standardized measures assessing student mental health, health, social, and behavioral issues included Child Behavior Check List (Achenbach, 2001a ), Teacher Report Form (Achenbach, 2001b ), Youth Self Report Form (Achenbach, 2001c ), Self-Perception Profile for Children (Harter, 1985b ), Self-Perception Profile for Adolescents (Harter, 2012 ), Social Support Scale for Children (Harter, 1985a ), and Social Support Behaviors Scale (Vaux et al., 1987 ).

Descriptive analyses were conducted to calculate frequencies for categorical variables and means and standard deviations for continuous variables. We summarized socio-demographic variables among participants in each grade level (4, 7, 10). Items for each outcome scale (e.g., Social Support Scale for Children) were summed to calculate total or subscale scores for each measure.

Findings on Prevalence and Reporting

The quantitative findings in the larger study (Mishna et al., 2015 ) show that rates of cyber witnessing were higher than cyberbullying and victimization at each assessment. In year 1, 24.2 percent reported cyber witnessing, 10.7 percent cyber victimization, and 2.9 percent cyberbullying. In year 2, 21.5 percent reported cyber witnessing, 7.6 percent cyber victimization, and 1.6 percent cyberbullying. In year 3, 25.1 percent reported cyber witnessing, 10.8 percent cyber victimization, and 2.5 percent cyberbullying. Similarly, rates of witnessing traditional bullying were higher than perpetration and victimization at each assessment. In year 1, 53.0% reported witnessing traditional bullying, 23.5% victimization, and 7.8% perpetration. In year 2, 42.6% reported witnessing traditional bullying, 17.3% victimization, and 4.3% perpetration. In year 3, 35.7% reported witnessing traditional bullying, 19.2% victimization, and 5.4% perpetration (Mishna et al., 2015 ). Of note, nearly half of all students (48.3%), who reported cyberbullying involvement in our survey, reported that they had not told an adult about what was happening online (Mishna et al., 2015 ). Moreover, 69.5% of students reported that cyberbullying and physical bullying are equally serious, and 64.5% believed that cyberbullying and “real” life verbal bullying are also equally serious (Mishna et al., 2015 ). These quantitative results serve as a springboard for the following discussion of qualitative findings, demonstrating that qualitative interviews reveal nuanced similarities and differences in the views of adults and youth, elucidating important interconnections among the levels of the ecological system (Mishna et al., 2004 , 2009 ; Dennehy et al., 2020 ).

Qualitative Interview Data Collection and Analysis

Student participants in 4th grade ( n  = 20), 7th grade ( n  = 21), and 10th grade ( n  = 16) in the qualitative sub-sample were purposively selected for interviews from the larger quantitative sample, based on gender, grade, school need level, and whether they reported bullying/cyberbullying victimization, perpetration, or witnessing. After selecting student participants, their teachers ( n  = 30) and parents ( n  = 50) were invited to participate in interviews. Interviews lasted approximately 1 h, ranging in length from 30 to 90 min. All year 1 interviews (with students, parents, and teachers) took place in the school setting and utilized a semi-structured interview guide. Following preliminary analysis, this interview guide was refined for use in the year 3 follow-up phone interviews with the students and parents. Areas explored with students comprised understanding of cyberbullying and how it compares with traditional bullying, experiences of online aggression, and others’ attitudes and responses. Questions were informed by existing literature and the research team’s considerable experience. Parent and teacher interviews included questions on their awareness and understanding of cyberbullying, their child or student’s involvement in cyberbullying, links between cyber and traditional bullying, support, and their responses to cyberbullying.

Using a grounded theory inquiry, data were concurrently analyzed and theorized through constant comparison (Birks & Mills, 2015 ; Corbin & Strauss, 2008 ). Through this iterative process, the team used initial interview data and theoretical categories to inform and refine subsequent interview guides and data collection (Charmaz, 2014 ). The team members individually coded a portion of interviews to establish preliminary analytic focuses and inductively identify preliminary themes. Consistent with a grounded theory approach, no hypotheses guided data analysis and coders sought to bracket their biases through reflexive journaling and team discussions of assumptions (Corbin & Strauss, 2008 ). During team meetings, each interview was collectively coded, building upon, revising, and/or removing codes proposed by the initial coder. Emerging categories were developed and expanded. Axial coding promoted connections within and between categories and subcategories and enabled synthesis and explanation (Birks & Mills, 2015 ; Charmaz, 2014 ; Corbin & Strauss, 2008 ). Numerous preliminary codes were identified based on emerging themes that were generated and discussed. A holistic “middle-order” approach to coding resulted in a condensed number of initial codes (Saldaña, 2015 ). Axial coding was then used to identify connections within and between themes and subthemes (Birks & Mills, 2015 ; Charmaz, 2006 , 2014 ; Corbin & Strauss, 2008 ). Through this iterative process of open, holistic, and focused coding, key themes emerged related to the understanding of traditional and cyberbullying according to the perspectives of the students, parents, and teachers. Measures were employed to ensure trustworthiness and authenticity. Prolonged engagement over the 3 years of the study ensured thick descriptions of the youth and adult narratives (Lietz & Zayas, 2010 ). Rigor was established through documentation for auditing purposes (Padgett, 2008 ). Trustworthiness and transferability were further ensured through reflexive journaling, bracketing, and dense descriptions (Corbin & Strauss, 2008 ).

While we use examples from our published manuscripts derived from our study entitled, “Motivations of Cyberbullying,” in the current manuscript, we identify new thematic areas and demonstrate how our qualitative interviews complement our quantitative findings. In analyzing findings across the study publications and datasets, we have not previously drawn the conclusions. The unique contribution of the current manuscript is the use of findings of previous publications to generate broader conclusions about the benefits of a mixed-methods approach (qualitative interviews and quantitative survey data) that makes visible the connections across ecological systems levels.

In discussing how qualitative research contributes to understanding bullying and cyberbullying and complements quantitative findings, the following new thematic areas are discussed: augmenting quantitative findings through qualitative interviews, contextualizing new or rapidly evolving areas of research, capturing nuances and complexity of perspectives, and providing moments for self-reflection and opportunities for learning.

Augmenting Quantitative Findings Through Qualitative Interviews

By examining process, context, and meaning for participants, qualitative methodology can augment quantitative findings. Quantitative methodology establishes outcomes and causal relationships and puts forth generalization and predictions (Yilmaz, 2013 ). Our background study which was a longitudinal multi-informant mixed methods study (Tashakkori et al., 1998 ) used grounded theory (Strauss & Corbin, 1998 ) and a longitudinal quantitative design to aid understanding of nuances related to cyberbullying (Mishna et al., 2009 ). In creating opportunities for the voices of young people to be heard (Carroll & Twomey, 2020 ; Gilgun & Abrams, 2002 ), qualitative methodology is especially useful for phenomena that are largely unstudied and/or rapidly evolving, such as cyberbullying, by explicating process and a holistic understanding and directions for future research (Mishna & Van Wert, 2013 ; Gilgun & Abrams, 2002 ).

In our paper, “Benchmarks and bellwethers in cyberbullying: The relational process of telling” Footnote 3 (Mishna et al., 2020 ), the qualitative analysis revealed relational processes among students that occurred when they considered whether to tell adults about their bullying and cyberbullying experiences. As noted above, almost half of the students who reported cyberbullying involvement relayed that they had not told an adult. Qualitative findings, however, exposed complex interactions that informed their decision-making processes. Reticent about speaking with adults, students turned to friends. It emerged that in addition to sharing, telling friends often served as a bellwether to gauge whether to proceed and report the situation to an adult. Often minimizing the severity of their ordeal, many students had decided against informing adults, frequently mentioning their concern about making a “big deal.” Participant interviews further revealed that media reports of high-profile cases involving cyberbullying can serve as benchmarks through which to assess the severity of their own personal experiences. The qualitative findings in our study helped to contextualize the quantitative data by unpacking and making visible the reasoning and contributing factors, thus increasing understanding of what informs youth’s decisions regarding whether and who to tell about cyberbullying involvement. By augmenting the quantitative data detailing the proportion of youth who do not tell adults, particulars attained through qualitative interview data help to inform and direct prevention and intervention strategies that are concrete and actionable for addressing the more challenging aspects of cyberbullying involvement and disclosure. In offering insights on the relational dynamics among peers and between youth and adults with respect to cyberbullying, the qualitative analysis gave voice to these interconnected layers of the youths’ ecological environment.

Contextualizing New or Rapidly Evolving Areas of Research

While cyberbullying is no longer considered a new phenomenon, the rapid development of technology is continually altering the cyber landscape, creating a need for perpetual knowledge generation (Odgers & Jensen, 2020 ; Rosa et al., 2019 ) and for evolving definitions, measurements, and responses (Spears et al., 2009 ). Moreover, rapid and ongoing technological advances create unique challenges for practitioners, policy makers, and researchers, in remaining current and responding to cyberbullying (George & Odgers, 2015 ; Jäger et al., 2010 ). With youth at the forefront of technological advances in many ways, qualitative methodology is well suited to elicit the experiences and perspectives of young people in promoting in-depth understanding of youth cultures, dynamics, and processes (Thornberg & Knutsen, 2011 ).

The data collection for our background study occurred between 2012 and 2014, during the early stages of attention to and research on sexting (sending and receiving sexually explicit images, videos, and text among youth). In the quantitative questionnaires, we included one question related to sexting for students in grades 7 and 10 and their parents and teachers. Our quantitative survey found that 15.6% of students in grades 7 and 10 had seen nude or sexual photos of friends, family, boyfriend, girlfriend, or other romantic partner online or over a cell phone. Furthermore, 27.8% of teachers had witnessed or were aware of their students viewing sexually explicit images, video, or text on cell phones at school. The data indicated that digital sending and receiving of sexually explicit images, video, or text was a new phenomenon among youth participants in grades 7 and 10 in a rapidly changing digital environment.

We did not explicitly inquire about sexting in the interviews with students, parents, and teachers. Rather, we asked participants about the students’ negative experiences with cyber technology. During analysis of the interview data, however, sexting emerged as a new and pertinent phenomenon among youth, which generated knowledge about rapidly evolving cyber dynamics that warranted further attention and inspired a paper entitled, “Gendered and sexualized bullying and cyberbullying: Spotlighting girls and making boys invisible” (Mishna et al., 2021b ). The qualitative interview data in this instance confirmed our quantitative findings on sexting among youth and allowed us to delve into the complex and nuanced ways participants articulated sexting behaviors along gender lines that both reinforced and were reinforced by gendered sociocultural norms and pressures. In student accounts, boys’ presence and participation in cyberbullying were frequently invisible, such as the non-consensual sharing of sexual images. Blamed for their poor choices, girls were spotlighted and their behavior problematized through negative characterizations. The participants’ focus on girls as responsible for the gendered cyberbullying and non-consensual sharing of images corresponds with how youth are typically educated about digital technologies through an “online safety model” with the focus on youth protecting themselves and avoiding “risky” activities (Johnson, 2015 ). As such, our findings provided context for this rapidly evolving environment that then allowed us to draw links between individual cyberbullying behaviors, understanding and articulation of these behaviors, and the broader influence of patriarchal structures (Mishna et al., 2021b ). The qualitative findings underscored the need to consider key factors that go beyond individual characteristics and behaviors and to develop education and prevention and intervention strategies that address sociocultural norms and values. The qualitative findings stimulated new research endeavors and collaborations with community organizations and academics.

Capturing Nuances and Complexity of Perspectives

Bullying and cyberbullying are exceedingly complex and must be studied within the contexts of the involved youth as well as within the larger social context of youth (Cross et al., 2015 ; Dennehy et al., 2020 ; Johnson & Puplampu, 2008 ; Sainju, 2020 ; Thornberg, 2011 ). An ecological systems framework is appropriate as it provides insight into the interconnected relationships among varying aspects and social layers of an individual’s world (Bronfenbrenner, 1979 ). While quantitative research considers and articulates context, qualitative interviews provide an occasion to engage with the richness of students’ perspectives, thoughts, and feelings about themselves and their social worlds (Mishna et al., 2004 ) and allow for a deeper understanding of youth culture and social processes from the vantage point of young people (Chaumba, 2013 ; Dennehy et al., 2020 ; Spears et al., 2009 ; Thornberg & Knutsen, 2011 ). Although qualitative studies are generally bound by a particular timeframe, participants bring their life histories and cumulative experiences to the research engagement (Phoenix et al., 2003 ), which can generate a fulsome and holistic understanding of cyberbullying, taking into consideration individual, family, peer, school, cyber, and sociocultural conditions over time.

Qualitative interview data allow for an interpretive approach that draws upon patterns of understanding, similarity, and contradiction, thereby teasing out underlying assumptions that shape how people define and assess experiences and phenomena such as bullying and cyberbullying (Mishna et al., 2020 , 2021a ). In our paper entitled “Looking Beyond Assumptions to Understand Relationship Dynamics in Bullying” Footnote 4 (Mishna et al., 2021a ), analysis of the qualitative interview data exposed persistent and pervasive assumptions about bullying linked to sociocultural norms and understanding of gender. These assumptions shaped participants’ understanding and conclusions of bullying and cyberbullying experiences, behavior, and motivations. Focusing on the visible hurt and injuries associated with physical bullying, participants tended to make comments such as “you’ll heal in a few days,” whereas they noted that with verbal bullying, the mental anguish “might stay for a long term.” This viewpoint that physical bullying was not a relationship problem appeared to be linked to gender stereotypes and social norms regarding the “natural” behavior of girls and boys. These gendered assumptions led participants to suggest that addressing bullying among girls was “complicated” and ongoing, whereas addressing physical bullying among boys was “simpler” and faster, a finding similar to that of Eriksen and Lyng ( 2018 ) who described participants’ descriptions of bullying among boys as “undramatic.” These assumptions appeared to preclude participants from discussing physical bullying among boys in a manner that acknowledged the physical bullying involvement as entrenched in relationship dynamics.

Qualitative interviewing provides an opportunity for participants to express their views and ideas when discussing the topic of interest which can elicit novel conclusions and nuances. As an example, at times, youth who claimed not to have involvement with cyberbullying may go on to describe situations that actually seemed to fit the definition of cyberbullying. In our Spotlighting Girls paper, many participant reports aligned with stereotypes regarding differences in how boys and girls bully others. These stereotypes were shared, however, even when they contradicted participants’ own experiences. For instance, similar to other research findings (Eriksen & Lyng, 2018 ), one participant described a boy as using “guilt trips” as a bullying tactic, yet described boys as only bullying physically. Consequently, relational aggression among boys often goes unnoticed and remains invisible. Similarly, the same behavior displayed by both girls and boys was discounted in boys and highlighted in girls. Boys’ behaviors were often not considered to be bullying because they were positioned as within the bounds of masculine gender norms. For example, one girl reported that “mostly girls, not boys,” bully “because boys would just go over and do some physical things... [Girls would] post embarrassing stuff about the person and do that kind of stuff” (p. 410). It is possible therefore that such actions by boys were not identified as bullying and thus underreported in the quantitative surveys while captured in the interviews. Discrepancies emerged in how cyberbullying had been reported in quantitative measures and how it was described in the interviews. This indicates that qualitative interviews can complement quantitative findings by revealing the complexities and ramifications of social experiences which are not reported in quantitative surveys.

The critical role of witnessing in bullying and cyberbullying is well documented (Salmivalli, 2010 , 2014 ; Spadafora et al., 2020 ; Volk et al., 2014 ). Social experiences related to witnessing are also complex, and bystander decision-making and responses impact both the process and outcomes of bullying incidents (Salmivalli et al., 2011 ). Qualitative research can offer youth the opportunity to explore and explain the motivations and factors they consider in determining whether to intervene, specifically the social costs and benefits of intervening (Spadafora et al., 2020 ). Our qualitative interviews similarly added youth voices concerning the dilemmas they faced in considering whether and how to respond based on emotional and contextual factors (Mishna et al., 2021b ), thus providing nuanced perspectives that serve to augment the quantitative findings related to bystander responses.

Providing Moments for Self-reflection and Opportunities for Learning

Qualitative methodologies are recognized as providing participants opportunities to self-reflect in the context of being listened to empathically (Birch & Miller, 2000 ; Wolgemuth et al., 2015 ). According to a systematic review of quantitative, qualitative, and mixed-methods studies conducted with children and adolescents, participation was mainly considered to be beneficial (Crane & Broome, 2017 ). Negative responses to participating in the research included feeling anxious and upset (Crane & Broome, 2017 ). Research indicates that despite describing negative effects of participating, children and youth reported that overall it was more positive to participate in the research (Crane & Broome, 2017 ) or described the emotional pain they experienced as beneficial in various ways, for example, as “emotionally cleansing” (Wolgemuth et al., 2015 , p. 366). The qualitative research process offers participants the opportunity to come to new understandings and can reveal evolving thoughts within participant narratives (Birch & Miller, 2000 ; Wolgemuth et al., 2015 ). Qualitative processes are iterative and involve probing questions that can prompt dynamic reflection by participants (Wolgemuth et al., 2015 ). Birch and Miller ( 2000 ) explain that they “use the term therapeutic to represent a process by which an individual reflects on, and comes to understand previous experiences in different—sometimes more positive—ways that promote a changed sense of self” (p. 190).

Recognizing the potential risks in research with children and youth (Mishna et al., 2004 ; Crane & Broome, 2017 ), we informed the students in our study of the possible risks should they decide to participate, such as the possibility that they would become upset as we were asking them about hurtful matters, and the limits to confidentiality. Anticipating that some of the questions could lead to a participant becoming distressed or disclosing potentially sensitive or upsetting information, we put in place a protocol (approved by the university and school board research ethics boards) to identify and offer support for students in distress (Mishna et al., 2016 ).

Corresponding with previous research, the reflexivity of sharing their narratives and views seemed to contribute to some participants coming to a different understanding of their experiences. Such reflection was evident in our interviews with students and their parents and teachers. When asked whether he had witnessed cyberbullying, for example, a boy reflected that only in being asked about cyberbullying in the interview did he recognize the behavior as cyberbullying: “When I think about it now, I actually did a few times. I didn’t feel that it’s cyber bullying, I wasn’t thinking that it’s a huge deal. It’s basically a few arguments between people on Facebook, like writing things about each other in public, not in private, chats.”

In another example, a parent reconsidered her views during the interview. This parent first commented that girls and women are “more vindictive” than boys and men, who, she explained, have “your spat, you get over it, and you move on.” After reflecting on her assumptions, she wondered how much of this widely held view of the behavior “is just media driven because I guess the victims that we see on the news, at least in Canada, have been girls, right?… but that doesn’t say that boys aren’t also being bullied.” Similarly, a girl contemplated her assumptions after first casting boys in a favorable light in contrast to girls. In commenting that girls bully each other because of appearance, she praised boys, “because usually they don’t tend to worry about those things...They’re proud of themselves, and they don’t pick on other people. They’re good with what they have.” After pondering these stated differences between boys and girls, this girl surmised, “I think it’s from when we were little because those Barbie dolls are super skinny. We wanted to have blonde hair, blue eyes, and be like Barbie. I think it’s just how maybe we were raised.” Another girl, who asserted that while cyberbullying occurred with equal frequency among boys and girls, added that it was not “a big thing” for boys, in contrast to girls who, “would show it off more, be like oh yah, blah, blah, blah.” Rather than concluding that this difference indicated that cyberbullying was not a big deal for boys, however, this girl attributed the difference between boys and girls to dominant masculinity norms. She asserted that “guys kind of hide it in more” and explained that “they don’t want to show that they’re weak because guys tend to be, they think that they’re very strong, kind of thing.” The evolving perspectives throughout this and the previous exchanges demonstrate the process of deepened understanding that can occur because of qualitative interviewing.

Such new understanding can inspire a desire to act and make change through community engagement. A girl explained that the research was the first time she had spoken with anyone about cyberbullying. This girl’s appraisal of her participation is consistent with findings in which participants may be motivated to take part in research for the opportunity to effect and advocate for change and help others (Cutcliffe & Ramcharan, 2002 ; Wolgemuth et al., 2015 ). She remarked that participating had been a helpful process which led her to,

think of different ways that I could help someone else if I see it happening… Just talking about it makes you think about what could cause it, what could make someone bully someone else. It makes you realize how it could make someone feel. Also, talking about how there isn’t really a support system at school. It makes me want to go and talk to someone to organize it, because it does happen a lot and I know it affects a lot of people

The inclusion of qualitative interviews in mixed methods research brings forth new information about content, process, and meaning that is otherwise not visible. By engaging youth voices as well as adult perspectives through both quantitative measures and qualitative interviews in the mixed methods study discussed in this manuscript, entitled Motivations for Cyberbullying, understanding of bullying and cyberbullying was advanced, thus enriching the quantitative methodology. The findings of the interviews extended knowledge related to bullying and cyberbullying in the following ways, which can inform “bottom-up research and intervention efforts” (Dennehy et al., 2020 , p. 10): augmenting quantitative findings, contextualizing new or rapidly evolving areas of research, capturing nuances and complexity of perspectives, and providing moments for self-reflection and opportunities for learning.

Qualitative research constitutes a significant venue through which to amplify the voices of children and youth (Dennehy et al., 2020 ) and ensures that children and youth’s experiences of the world are represented in understanding social phenomena (Mishna et al., 2004 ; Carroll & Twomey, 2020 ; Chaumba, 2013 ; Dennehy et al., 2020 ; Patton et al., 2017 ). According to Dennehy and colleagues ( 2020 ), engaging youth as co-researchers in cyberbullying research may enhance efforts to ethically and earnestly amplify youth voices. A synthesis by Elsaesser et al. ( 2017 ) supports the view that focusing on collaboratively working with youth to understand and safely navigate the cyber world through education and empowerment is more effective than interventions aimed at restricting ICT use without involving youth. Through quantitative measures and qualitative interviews, our mixed methods study examined participant perspectives regarding bullying and cyberbullying on the various ecological systems levels across the students’ lives. The use of mixed methods facilitated a dialogue between the participant responses to both methodologies, thus highlighting the salience of the overlapping influence and interactions among the systems levels. Such complex and nuanced understanding is necessary to inform meaningful prevention and intervention strategies to address bullying and cyberbullying.

According to the United Nations Convention on the Rights of the Child (Assembly UG, 1989 ), children and youth have the right to discuss their views and experiences. The Convention states that all children have the right to protections, provisions, participation, and non-discrimination (Assembly UG, 1989 ). Participation entails the right for children to express themselves and have a voice in situations that have to do with and affect them. The importance of listening to children’s voices underscores the limits of adult proxies in representing children’s emotional and social worlds (O’Farrelly, 2021 ). Bullying and cyberbullying fundamentally violate these protections, silence children’s voices, and compromise their healthy development (Greene, 2006 ). Our mixed methods study through quantitative measures and qualitative interviews facilitated a dialogue between the participant responses in both methodologies. This interaction of data types maximizes the voices of and collaboration with participants as well as knowledge generation.

Data Availability

Not applicable.

Code Availability

Different terms are used to describe the same approach (e.g., social-ecological framework, ecological systems framework, ecological theory, ecological perspectives). For the purposes of this paper, the term ecological systems framework is used.

All additional references to this research study will be shortened to “Motivations for Cyberbullying.”

All additional references to this paper will be shortened to “Benchmarks and Bellwethers paper.”

All additional references to the paper will be shortened to “Relationship Dynamics paper.”

Achenbach, T. (2001a). Child behavior checklist . ASEBA, University of Vermont. https://store.aseba.org/CHILD-BEHAVIOR-CHECKLIST_6-18/productinfo/201/

Achenbach, T. (2001b). Teacher report form . ASEBA, University of Vermont. https://store.aseba.org/TEACHERS-REPORT-FORM_6-18/productinfo/301/

Achenbach, T. (2001c). Youth self report form . ASEBA, University of Vermont. https://store.aseba.org/YOUTH-SELF-REPORT_11-18/productinfo/501/

Assembly, U. G. (1989). Convention on the Rights of the Child. (1989). United Nations, Treaty Series, 1577 (3), pp. 1–23.

Bauman, S. A., & Yoon, J. (2014). This issue: Theories of bullying and cyberbullying. Theory into Practice, 53 (4), 253–256.  https://doi.org/10.1080/00405841.2014.947215

Article   Google Scholar  

Birch, M., & Miller, T. (2000). Inviting intimacy: The interview as therapeutic opportunity. International Journal of Social Research Methodology, 3 (3), 189–202.

Birks, M., & Mills, J. (2015). Grounded theory: A practical guide . Sage.

Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and design.

Bronfenbrenner, U. (1992). Contexts of child rearing: Problems and prospects. Child and Youth Care Administrator, 5 (1), 59–64.

Google Scholar  

Bronfenbrenner, U., & Morris, P. A. (2007). The bioecological model of human development. In R. M. Lerner (Ed.), Handbook of child psychology (6 ed., Vol. 1, pp. 793–828). Wiley.

Carroll, C., & Twomey, M. (2020). Voices of children with neurodevelopmental disorders in qualitative research: A scoping review. Journal of Developmental and Physical Disabilities , 1–16.

Charmaz, K. (2006). Constructing grounded theory: A practical guide through qualitative research . Sage Publications Ltd.

Charmaz, K. (2014). Constructing grounded theory (2nd ed.). Sage.

Chaumba, J. (2013). The use and value of mixed methods research in social work. Advances in Social Work, 14 (2), 307–333.

Corbin, J., & Strauss, A. (2008). Basics of qualitative research: Techniques and procedures for developing grounded theory . Sage Publications Inc.

Crane, S., & Broome, M. E. (2017). Understanding ethical issues of research participation from the perspective of participating children and adolescents: A systematic review. Worldviews on Evidence-Based Nursing, 14 (3), 200–209.

Article   PubMed   PubMed Central   Google Scholar  

Creswell, J. W., & Creswell, J. D. (2018). Research design: Qualitative, quantitative, and mixed methods approaches (5 ed.). Sage Publications.

Crivello, G., Camfield, L., & Woodhead, M. (2009). How can children tell us about their wellbeing? Exploring the potential of participatory research approaches within young lives. Social Indicators Research, 90 (1), 51–72.

Cross, D., Barnes, A., Papageorgiou, A., Hadwen, K., Hearn, L., & Lester, L. (2015). A social–ecological framework for understanding and reducing cyberbullying behaviours. Aggression and Violent Behavior, 23 , 109–117.

Cutcliffe, J. R., & Ramcharan, P. (2002). Leveling the playing field? Exploring the merits of the ethics-as-process approach for judging qualitative research proposals. Qualitative Health Research, 12 (7), 1000–1010.

Article   PubMed   Google Scholar  

Darbyshire, P., MacDougall, C., & Schiller, W. (2005). Multiple methods in qualitative research with children: More insight or just more? Qualitative Research, 5 (4), 417–436.

Dennehy, R., Meaney, S., Walsh, K. A., Sinnott, C., Cronin, M., & Arensman, E. (2020). Young people’s conceptualizations of the nature of cyberbullying: A systematic review and synthesis of qualitative research. Aggression and Violent Behavior, 51 , 101379.

Elsaesser, C., Russell, B., Ohannessian, C. M., & Patton, D. (2017). Parentng in a digital age: A review of parents’ role in preventing adolescent cyberbullying. Aggression and Violent Behavior, 35 , 62–72.

Eriksen, I. M., & Lyng, S. T. (2018). Relational aggression among boys: Blind spots and hidden dramas. Gender and Education, 30 (3), 396–409.

Espelage, D. L. (2014). Ecological theory: Preventing youth bullying, aggression, and victimization. Theory into Practice, 53 (4), 257–264.

Fevre, R., Robinson, A., Jones, T., & Lewis, D. (2010). Researching workplace bullying: The benefits of taking an integrated approach. International Journal of Social Research Methodology, 13 (1), 71–85.

George, M. J., & Odgers, C. L. (2015). Seven fears and the science of how mobile technologies may be influencing adolescents in the digital age. Perspectives on Psychological Science, 10 (6), 832–851.

Gilgun, J. F., & Abrams, L. S. (2002). The nature and usefulness of qualitative social work research: Some thoughts and an invitation to dialogue. Qualitative Social Work, 1 (1), 39–55.

Greene, M. (2006). Bullying in Schools: A Plea for a Measure of Human Rights. Journal of Social Issues, 62 (1), 63–79. https://doi.org/10.1111/j.1540-4560.2006.00439.x

Harter, S. (1985a). Manual for the social support scale for children . University of Denver.

Harter, S. (1985b). The self-perception profile for children (manual) . University of Denver.

Harter, S. (2012). Self-perception profile for adolescents: Manual and questionnaires . Univeristy of Denver, Department of Psychology.

Hemming, P. J. (2008). Mixing qualitative research methods in children’s geographies. Area, 40 (2), 152–162.

Jäger, T., Amado, J., Matos, A., & Pessoa, T. (2010). Analysis of experts’ and trainers’ views on cyberbullying. Journal of Psychologists and Counsellors in Schools, 20 (2), 169–181.

Johnson, G. (2010). Internet use and child development: The techno-microsystem. Australian Journal of Educational and Developmental Psychology (AJEDP), 10 , 32–43.

Johnson, G., & Puplampu, K. (2008). A conceptual framework for understanding the effect of the Internet on child development: The ecological techno-subsystem. Canadian Journal of Learning and Technology, 34 , 19–28.

Johnson, M. (2015). Digital literacy and digital citizenship: Approaches to girls’ online experiences. In J. Bailey & V. Steeves (Eds.), eGirls, eCitizens (pp. 339–360). University of Ottawa Press.

Lenhart, A., Duggan, M., Perrin, A., Stepler, R., Rainie, H., & Parker, K. (2015). Teens, social media & technology overview 2015 . Pew Research Center [Internet & American Life Project].

Lietz, C. A., & Zayas, L. E. (2010). Evaluating qualitative research for social work practitioners. Advances in Social Work, 11 (2), 188–202.

McKim, C. A. (2017). The value of mixed methods research: A mixed methods study. Journal of Mixed Methods Research, 11 (2), 202–222.

Mishna, F., Antle, B. J., & Regehr, C. (2004). Tapping the perspectives of children: Emerging ethical issues in qualitative research. Qualitative Social Work: Research and Practice, 3 (4), 449–468.

Mishna, F., Craig, W., Pepler, D., & Daciuk, J. (2012). The Bullying and Cyberbullying: Perpetrators . Victims and Witnesses Survey: Unpublished survey.

Mishna, F., McInroy, L., Lacombe-Duncan, A., & Daciuk, J. (2015). Motivations for cyberbullying study: A longitudinal and multi-perspective inquiry . Toronto, ON: The Authors.

Mishna, F., McInroy, L. B., Lacombe-Duncan, A., Bhole, P., VanWert, M., Schwan, K., et al. (2016). Prevalence, motivations, and social, mental health and health consequences of cyberbullying among school-aged children and youth: Protocol of a longitudinal and multi-perspective mixed method study. JMIR Research Protocols, 5 (2), e83.

Mishna, F., *Saini, M., & *Solomon, S. (2009). Ongoing and online: Children and youth’s perceptions of cyber bullying. Children and Youth Services Review, 31 (12), 1222–1228.

Mishna, F., Schwan, A., *Birze, A., Van Wert, M., McInroy, L., *Lacombe-Duncan, A., Attar-Schwartz, S., & Daciuk, J. (2020). Gendered and sexualized bullying and cyber bullying: Spotlighting girls and making boys invisible. Youth & Society, 52 (3), 403–426. https://doi.org/10.1177/0044118X18757150

Mishna, F., & Van Wert, M. (2013). Qualitative studies. In S. Bauman, D. Cross, & J. Walker (Eds.), Principles of cyberbullying research: Definitions, measures, and methodology (pp. 238–257). New York & London: Routledge.

Mishna, F., Birze, A., Greenblatt, A., & Pepler, D. (2021a). Looking beyond assumptions to understand relationship dynamics in bullying. Frontiers in Psychology, 12, 1–11. https://doi.org/10.3389/fpsyg.2021.661724

Mishna, F., Birze, A., Greenblatt, A., & Khoury-Kassabri, M. (2021b). Benchmarks and bellwethers in cyberbullying: The relational process of telling. International Journal of Bullying Prevention, 3 (4), 241–252. https://doi.org/10.1007/s42380-020-00082-3

Nesi, J., Choukas-Bradley, S., & Prinstein, M. J. (2018). Transformation of adolescent peer relations in the social media context: Part 1—A theoretical framework and application to dyadic peer relationships. Clinical Child and Family Psychology Review, 21 (3), 267–294.

Newman, P. A., Fantus, S., Woodford, M. R., & Rwigema, M.-J. (2018). “Pray that god will change you”: The religious social ecology of bias-based bullying targeting sexual and gender minority youth—A qualitative study of service providers and educators. Journal of Adolescent Research, 33 (5), 523–548.

O’Moore, A. M., & Minton, S. J. (2005). Evaluation of the effectiveness of an anti-bullying programme in primary schools. Aggressive Behavior: Official Journal of the International Society for Research on Aggression, 31 (6), 609–622.

O’Farrelly, C. (2021). Bringing young children’s voices into programme development, randomized controlled trials and other unlikely places. Children & Society, 35 (1), 34–47.

Odgers, C. L., & Jensen, M. R. (2020). Annual research review: Adolescent mental health in the digital age: Facts, fears, and future directions. Journal of Child Psychology and Psychiatry, 61 (3), 336–348.

Padgett, D. (2008). Qualitative Methods in Social Work Research . Sage.

Patton, D. U., Hong, J. S., Patel, S., & Kral, M. J. (2017). A systematic review of research strategies used in qualitative studies on school bullying and victimization. Trauma, Violence, & Abuse, 18 (1), 3–16.

Pepler, D., Craig, W., & O'Connell, P. (2010). Peer processes in bullying: Informing prevention and intervention strategies.

Pepler, D. J. (2006). Bullying interventions: A binocular perspective. Journal of the Canadian Academy of Child and Adolescent Psychiatry, 15 (1), 16.

PubMed   PubMed Central   Google Scholar  

Phoenix, A., Frosh, S., & Pattman, R. (2003). Producing contradictory masculine subject positions: Narratives of threat, homophobia and bullying in 11–14 year old boys. Journal of Social Issues, 59 (1), 179–195.

Rosa, H., Pereira, N., Ribeiro, R., Ferreira, P. C., Carvalho, J. P., Oliveira, S., Coheur, L., Paulino, P., Simão, A. V., & Trancoso, I. (2019). Automatic cyberbullying detection: A systematic review. Computers in Human Behavior, 93 , 333–345.

Sainju, K. D. (2020). Beyond the schoolyard: A multilevel examination of individual, school and school district variables associated with traditional and cyber peer aggression. Child & Youth Care Forum.

Saldaña, J. (2015). The coding manual for qualitative researchers . Sage.

Salmivalli, C. (2010). Bullying and the peer group: A review. Aggression and Violent Behavior, 15 (2), 112–120.

Salmivalli, C. (2014). Participant roles in bullying: How can peer bystanders be utilized in interventions? Theory into Practice, 53 (4), 286–292.

Salmivalli, C., Voeten, M., & Poskiparta, E. (2011). Bystanders matter: Associations between reinforcing, defending, and the frequency of bullying behavior in classrooms. Journal of Clinical Child & Adolescent Psychology, 40 (5), 668–676.

Spadafora, N., Marini, Z. A., & Volk, A. A. (2020). Should I defend or should I go? An adaptive, qualitative examination of the personal costs and benefits associated with bullying intervention. Canadian Journal of School Psychology, 35 (1), 23–40.

Spears, B., Slee, P., Owens, L., & Johnson, B. (2009). Behind the scenes and screens: Insights into the human dimension of covert and cyberbullying. Zeitschrift Für Psychologie/journal of Psychology, 217 (4), 189–196.

Steeves, V., & Marx, G. T. (2014). Safe schools initiatives and the shifting climate of trust. Responding to school violence: Confronting the Columbine effect , 105–124.

Strauss, A., & Corbin, J. (1998). Basics of qualitative research: Techniques and procedures for developing grounded theory (2nd ed.). Sage Publications Inc.

Tashakkori, A., Teddlie, C., & Teddlie, C. B. (1998). Mixed methodology: Combining qualitative and quantitative approaches (Vol. 46). Sage.

Thornberg, R. (2011). ‘She’s weird!’—The social construction of bullying in school: A review of qualitative research. Children & Society, 25 (4), 258–267.

Thornberg, R. (2015). The social dynamics of school bullying: The necessary dialogue between the blind men around the elephant and the possible meeting point at the social-ecological square. Confero: Essays on Education, Philosophy and Politics, 3 (2), 161–203.

Thornberg, R. (2018). School bullying and fitting into the peer landscape: A grounded theory field study. British Journal of Sociology of Education, 39 (1), 144–158.

Thornberg, R., & Knutsen, S. (2011). Teenagers’ explanations of bullying. Child & Youth Care Forum,

Toronto District School Board. (2014). The 2014 Learning Opportunities Index: Questions and answers .

Vaux, A., Riedel, S., & Stewart, D. (1987). Modes of social support: The social support behaviors (SS-B) scale. American Journal of Community Psychology, 15 (2), 209–232.

Volk, A. A., Dane, A. V., & Marini, Z. A. (2014). What is bullying? A Theoretical Redefinition. Developmental Review, 34 (4), 327–343.

Wolgemuth, J. R., Erdil-Moody, Z., Opsal, T., Cross, J. E., Kaanta, T., Dickmann, E. M., & Colomer, S. (2015). Participants’ experiences of the qualitative interview: Considering the importance of research paradigms. Qualitative Research, 15 (3), 351–372.

Yilmaz, K. (2013). Comparison of quantitative and qualitative research traditions: Epistemological, theoretical, and methodological differences. European Journal of Education, 48 (2), 311–325.

Download references

Acknowledgements

We would like to acknowledge first and foremost the Toronto District School Board for their utmost commitment to participating in the study, as well as each school for their dedication to both data collection and ensuring that the mental health needs of students that were identified through the study were addressed. We would like to thank the students, parents, and teachers for sharing their experiences with us. We would like to thank the research assistants, without whom we could not have completed this study.

This research was supported by a grant from the Social Sciences and Humanities Research Council of Canada: Grant Account Number: 410–2011-1001.

Author information

Authors and affiliations.

Factor-Inwentash Faculty of Social Work, University of Toronto, 246 Bloor Street West, Toronto, ON, M5S 1V4, Canada

Faye Mishna, Arija Birze & Andrea Greenblatt

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Faye Mishna .

Ethics declarations

Conflict of interest.

The authors declare no competing interests.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Mishna, F., Birze, A. & Greenblatt, A. Understanding Bullying and Cyberbullying Through an Ecological Systems Framework: the Value of Qualitative Interviewing in a Mixed Methods Approach. Int Journal of Bullying Prevention 4 , 220–229 (2022). https://doi.org/10.1007/s42380-022-00126-w

Download citation

Accepted : 25 April 2022

Published : 10 May 2022

Issue Date : September 2022

DOI : https://doi.org/10.1007/s42380-022-00126-w

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Qualitative research benefits
• Mixed methods research
• Bullying and cyberbullying
• Voices of children and youth
• Find a journal
• Publish with us
• Track your research

ORIGINAL RESEARCH article

Understanding alternative bullying perspectives through research engagement with young people.

• School of Education and Social Care, Faculty of Health, Education, Medicine and Social Care, Anglia Ruskin University, Chelmsford, United Kingdom

Bullying research has traditionally been dominated by largescale cohort studies focusing on the personality traits of bullies and victims. These studies focus on bullying prevalence, risk and protective factors, and negative outcomes. A limitation of this approach is that it does not explain why bullying happens. Qualitative research can help shed light on these factors. This paper discusses the findings from four mainly qualitative research projects including a systematic review and three empirical studies involving young people to various degrees within the research process as respondents, co-researchers and commissioners of research. Much quantitative research suggests that young people are a homogenous group and through the use of surveys and other large scale methods, generalizations can be drawn about how bullying is understood and how it can be dealt with. Findings from the studies presented in this paper, add to our understanding that young people appear particularly concerned about the role of wider contextual and relational factors in deciding if bullying has happened. These studies underscore the relational aspects of definitions of bullying and, how the dynamics of young people’s friendships can shift what is understood as bullying or not. Moreover, to appreciate the relational and social contexts underpinning bullying behaviors, adults and young people need to work together on bullying agendas and engage with multiple definitions, effects and forms of support. Qualitative methodologies, in particular participatory research opens up the complexities of young lives and enables these insights to come to the fore. Through this approach, effective supports can be designed based on what young people want and need rather than those interpreted as supportive through adult understanding.

Introduction

Research on school bullying has developed rapidly since the 1970s. Originating in social and psychological research in Norway, Sweden, and Finland, this body of research largely focusses on individualized personality traits of perpetrators and victims ( Olweus, 1995 ). Global interest in this phenomenon subsequently spread and bullying research began in the United Kingdom, Australia, and the United States ( Griffin and Gross, 2004 ). Usually quantitative in nature, many studies examine bullying prevalence, risk and protective factors, and negative outcomes ( Patton et al., 2017 ). Whilst quantitative research collates key demographic information to show variations in bullying behaviors and tendencies, this dominant bullying literature fails to explain why bullying happens. Nor does it attempt to understand the wider social contexts in which bullying occurs. Qualitative research on the other hand, in particular participatory research, can help shed light on these factors by highlighting the complexities of the contextual and relational aspects of bullying and the particular challenges associated with addressing it. Patton et al. (2017) in their systematic review of qualitative methods used in bullying research, found that the use of such methods can enhance academic and practitioner understanding of bullying.

In this paper, I draw on four bullying studies; one systematic review of both quantitative and qualitative research ( O’Brien, 2009 ) and three empirical qualitative studies ( O’Brien and Moules, 2010 ; O’Brien, 2016 , 2017 ) (see Table 1 below). I discuss how participatory research methodologies, to varying degrees, were used to facilitate bullying knowledge production among teams of young people and adults. Young people in these presented studies were consequently involved in the research process along a continuum of involvement ( Bragg and Fielding, 2005 ). To the far left of the continuum, young people involved in research are referred to as “active respondents” and their data informs teacher practice. To the middle of the continuum sit “students as co-researchers” who work with teachers to explore an issue which has been identified by that teacher. Finally to the right, sit “students as researchers” who conduct their own research with support from teachers. Moving from left to right of the continuum shows a shift in power dynamics between young people and adults where a partnership develops. Young people are therefore recognized as equal to adults in terms of what they can bring to the project from their own unique perspective, that of being a young person now.

Table 1. The studies.

In this paper, I advocate for the active involvement of young people in the research process in order to enhance bullying knowledge. Traditional quantitative studies have a tendency to homogenize young people by suggesting similarity in thinking about what constitutes bullying. However, qualitative studies have demonstrated that regardless of variables, young people understand bullying in different ways so there is a need for further research that starts from these perspectives and focusses on issues that young people deem important. Consequently, participatory research allows for the stories of the collective to emerge without losing the stories of the individual, a task not enabled through quantitative approaches.

What Is Bullying?

Researching school bullying has been problematic and is partly related to the difficulty in defining it ( Espelage, 2018 ). Broadly speaking, bullying is recognized as aggressive, repeated, intentional behavior involving an imbalance of power aimed toward an individual or group of individuals who cannot easily defend themselves ( Vaillancourt et al., 2008 ). In more recent times, “traditional” bullying behaviors have been extended to include cyber-bullying, involving the use of the internet and mobile-phones ( Espelage, 2018 ). Disagreements have been noted in the literature about how bullying is defined by researchers linked to subject discipline and culture. Some researchers for example, disagree about the inclusion or not of repetition in definitions ( Griffin and Gross, 2004 ) and these disagreements have had an impact on interpreting findings and prevalence rates. However, evidence further suggests that young people also view bullying in different ways ( Guerin and Hennessy, 2002 ; Cuadrado-Gordillo, 2012 ; Eriksen, 2018 ). Vaillancourt et al. (2008) explored differences between researchers and young people’s definitions of bullying, and found that children’s definitions were usually spontaneous, and did not always encompass the elements of repetition, power imbalance and intent. They concluded, that children need to be provided with a bullying definition so similarities and comparisons can be drawn. In contrast, Huang and Cornell (2015) found no evidence that the inclusion of a definition effected prevalence rates. Their findings, they suggest, indicate that young people use their own perceptions of bullying when answering self-report questionnaires and they are not influenced by an imposed definition.

Nevertheless, differences in children and young people’s bullying definitions are evident in the research literature and have been explained by recourse to age and stage of development ( Smith et al., 2002 ) and their assumed lack of understanding about what constitutes bullying ( Boulton and Flemington, 1996 ). Naylor et al. (2001) for example, found that younger children think similarly in their definitions of bullying, while Smith et al. (2002) found that 8 year olds did not distinguish as clearly between different forms of behavioral aggression as 14 year olds. Methodological limitations associated with understanding bullying have been identified by Forsberg et al. (2018) and Maunder and Crafter (2018) . These authors postulate that quantitative approaches, although providing crucial insights in understanding bullying, are reliant on pre-defined variables, which can shield some of the complexities that qualitative designs can unravel, as individual experiences of bullying are brought to the fore. Indeed, La Fontaine (1991) suggests that unlike standard self-report questionnaires and other quantitative methods used to collect bullying data, analyzing qualitative data such as those collected from a helpline, enables the voice of young people to be heard and consequently empowers adults to understand bullying on their terms rather than relying solely on interpretations and perceptions of adults. Moore and Maclean (2012) collected survey, as well as interview and focus group data, on victimization occurring on the journey to and from school. They found that what young people determined as victimization varied and was influenced by a multifaceted array of circumstances, some of which adults were unaware of. Context for example, played an important role where certain behaviors in one situation could be regarded as victimization while in another they were not. Specific behaviors including ignoring an individual was particularly hurtful and supporting a friend who was the subject of victimization could lead to their own victimization.

Lee (2006) suggests that some bullying research does not reflect individual experiences, and are thus difficult for participants to relate to. Canty et al. (2016) reiterates this and suggests that when researchers provide young people with bullying definitions in which to position their own experiences, this can mask some of the complexities that the research intends to uncover. Such approaches result in an oversight into the socially constructed and individual experiences of bullying ( Eriksen, 2018 ). Griffin and Gross (2004) further argue that when researchers use vague or ambiguous definitions an “overclassification of children as bullies or victims” (p. 381) ensues. Consequently, quantitative research does not consider children as reliable in interpreting their own lived experiences and therefore some of the interactions they consider as bullying, that do not fit within the conventional definitions, are concealed. This approach favors the adult definition of bullying regarding it as “more reliable” than the definitions of children and young people Canty et al. (2016) . The perceived “seriousness” of bullying has also been explored. Overall, young people and adults are more likely to consider direct bullying (face-to-face actions including hitting, threatening and calling names) as “more serious” than indirect bullying (rumor spreading, social exclusion, forcing others to do something they do not want to do) ( Maunder et al., 2010 ; Skrzypiec et al., 2011 ). This perception of “seriousness,” alongside ambiguous definitions of bullying, has further implications for reporting it. Despite the advice given to young people to report incidents of school bullying ( Moore and Maclean, 2012 ), the literature suggests that many are reluctant to do so ( deLara, 2012 ; Moore and Maclean, 2012 ).

Several factors have been highlighted as to why young people are reluctant to report bullying ( Black et al., 2010 ). deLara (2012) , found apprehension in reporting bullying to teachers due to the fear that they will either not do enough or too much and inadvertently make the situation worse, or fear that teachers will not believe young people. Research also shows that young people are reluctant to tell their parents about bullying due to perceived over-reaction and fear that the bullying will be reported to their school ( deLara, 2012 ; Moore and Maclean, 2012 ). Oliver and Candappa (2007) suggest that young people are more likely to confide in their friends than adults (see also Moore and Maclean, 2012 ; Allen, 2014 ). However, if young people believe they are being bullied, but are unable to recognize their experiences within a predefined definition of bullying, this is likely to impact on their ability to report it.

Research from psychology, sociology, education and other disciplines, utilizing both quantitative and qualitative approaches, have enabled the generation of bullying knowledge to date. However, in order to understand why bullying happens and how it is influenced by wider social constructs there is a need for further qualitative studies, which hear directly from children and young people themselves. The next section of this paper discusses the theoretical underpinnings of this paper, which recognizes that young people are active agents in generating new bullying knowledge alongside adults.

Theoretical Underpinnings – Hearing From Children and Young People

The sociology of childhood ( James, 2007 ; Tisdall and Punch, 2012 ) and children’s rights agenda more broadly ( United Nations Convention on the Rights of the Child, 1989 ) have offered new understandings and methods for research which recognize children and young people as active agents and experts on their own lives. From this perspective, research is conducted with rather than on children and young people ( Kellett, 2010 ).

Participatory methodologies have proven particularly useful for involving young people in research as co-researchers (see for example O’Brien and Moules, 2007 ; Stoudt, 2009 ; Kellett, 2010 ; Spears et al., 2016 ). This process of enquiry actively involves those normally being studied in research activities. Previously, “traditional” researchers devalued the experiences of research participants arguing that due to their distance from them, they themselves are better equipped to interpret these experiences ( Beresford, 2006 ). However, Beresford (2006) suggests that the shorter the distance between direct experience and interpretation, the less distorted and inaccurate the resulting knowledge is likely to be. Jones (2004) further advocates that when young people’s voices are absent from research about them the research is incomplete. Certainly Spears et al. (2016) , adopted this approach in their study with the Young and Well Cooperative Research Centre (CRC) in Australia. Young people played an active role within a multidisciplinary team alongside researchers, practitioners and policymakers to co-create and co-evaluate the learning from four marketing campaigns for youth wellbeing through participatory research. Through this methodological approach, findings show that young people were able to reconceptualize mental health and wellbeing from their own perspectives as well as share their lived experiences with others ( Spears et al., 2016 ). Bland and Atweh (2007) , Ozer and Wright (2012) , highlight the benefits afforded to young people through this process, including participating in dialog with decision-makers and bringing aspects of teaching and learning to their attention.

Against this background, data presented for this paper represents findings from four studies underpinned by the ethos that bullying is socially constructed and is best understood by exploring the context to which it occurs ( Schott and Sondergaard, 2014 ; Eriksen, 2018 ). This socially constructed view focusses on the evolving positions within young people’s groups, and argues that within a bullying situation sometimes a young person is the bully, sometimes the victim and sometimes the bystander/witness, which contrasts the traditional view of bullying ( Schott and Sondergaard, 2014 ). The focus therefore is on group relationships and dynamics. For that reason, Horton (2011) proposes that if bullying is an extensive problem including many young people, then focusing entirely on personality traits will not generate new bullying knowledge and will be problematic in terms of interventions. It is important to acknowledge that this change in focus and view of bullying and how it is manifested in groups, does not negate the individual experiences of bullying rather the focus shifts to the process of being accepted, or not, by the group ( Schott and Sondergaard, 2014 ).

The Studies

This section provides a broad overview of the four included studies underpinned by participatory methodologies. Table 1 presents the details of each study. Young people were involved in the research process as respondents, co-researchers and commissioners of research, along a continuum as identified by Bragg and Fielding (2005) . This ranged from “active respondents” to the left of the continuum, “students as co-researchers” in the middle and “students as researchers” to the right of the continuum. Young people were therefore recognized as equal to adults in terms of what they can bring to the project from their own unique perspectives ( Bradbury-Jones et al., 2018 ).

A key finding from study one ( O’Brien, 2009 ) was the lack of voice afforded to young people through the research process and can be seen to reflect the far left of Bragg and Fielding (2005) continuum, as young people were not directly involved as “active respondents” but their views were included in secondary data analysis and informed the studies that followed. For example, the quantitative studies used an agreed academic definition of bullying which may or may not have influenced how young participants defined bullying within the studies. On the other hand, the qualitative study involved a group of students in deciding which questions to ask of the research participants and in interpreting the findings.

In contrast, study two ( O’Brien and Moules, 2010 ) was commissioned and led by a group of young people called PEAR (Public health, Education, Awareness, Researchers), who were established to advise on public health research in England. PEAR members were based in two large English cities and comprised 20 young people aged between 13 and 20 years. The premise of the study was that PEAR members wanted to commission research into cyber bullying and the effects this has on mental health from the perspectives of young people rather than adult perspectives. This project was innovative as young people commissioned the research and participated as researchers ( Davey, 2011 ) and can be seen to reflect the middle “students as co-researchers” as well as moving toward to right “students as researchers” of Bragg and Fielding (2005) continuum. Although the young people did not carry out the day-to-day work on the project, they were responsible for leading and shaping it. More importantly, the research topic and focus were decided with young people and adults together.

Study three ( O’Brien, 2016 ) involved five self-selecting students from an independent day and boarding school who worked with me to answer this question: What do young people in this independent day and boarding school view as the core issue of bullying in the school and how do they want to address this? These students called themselves R4U (Research for You) with the slogan researching for life without fear . Three cycles of Participatory Action Research (PAR) ensued, where decision making about direction of the research, including methods, analysis and dissemination of findings were made by the research team. As current students of the school, R4U had a unique “insider knowledge” that complemented my position as the “academic researcher.” By working together to generate understanding about bullying at the school, the findings thus reflected this diversity in knowledge. As the project evolved so too did the involvement of the young researchers and my knowledge as the “outsider” (see O’Brien et al., 2018a for further details). Similar to study two, this project is situated between the middle: “students as co-researchers” and the right: “students as researchers” of Bragg and Fielding (2005) continuum.

Study four ( O’Brien, 2017 ) was small-scale and involved interviewing four young people who were receiving support from a charity providing therapeutic and educational support to young people who self-exclude from school due to anxiety, as a result of bullying. Self-exclusion, for the purposes of this study, means that a young person has made a decision not to go to school. It is different from “being excluded” or “truanting” because these young people do not feel safe at school and are therefore too anxious to attend. Little is known about the experiences of young people who self-exclude due to bullying and this study helped to unravel some of these issues. This study reflects the left of Bragg and Fielding (2005) continuum where the young people were involved as “active respondents” in informing adult understanding of the issue.

A variety of research methods were used across the four studies including questionnaires, interviews and focus groups (see Table 1 for more details). In studies two and three, young researchers were fundamental in deciding the types of questions to be asked, where they were asked and who we asked. In study three the young researchers conducted their own peer-led interviews. The diversity of methods used across the studies are a strength for this paper. An over-reliance on one method is not portrayed and the methods used reflected the requirements of the individual studies.

Informed Consent

Voluntary positive agreement to participate in research is referred to as “consent” while “assent,” refers to a person’s compliance to participate ( Coyne, 2010 ). The difference in these terms are normally used to distinguish the “legal competency of children over and under 16 years in relation to research.” ( Coyne, 2010 , 228). In England, children have a legal right to consent so therefore assent is non-applicable ( Coyne, 2010 ). However, there are still tensions surrounding the ability of children and young people under the age of 18 years to consent in research which are related to their vulnerability, age and stage of development ( Lambert and Glacken, 2011 ). The research in the three empirical studies (two, three and four) started from the premise that all young participants were competent to consent to participate and took the approach of Coyne (2010) who argues that parental/carer consent is not always necessary in social research. University Research Ethics Committees (RECs) are nonetheless usually unfamiliar with the theoretical underpinnings that children are viewed as social actors and generally able to consent for themselves ( Lambert and Glacken, 2011 ; Fox, 2013 ; Parsons et al., 2015 ).

In order to ensure the young people in these reported studies were fully informed of the intentions of each project and to adhere to ethical principles, age appropriate participant information sheets were provided to all participants detailing each study’s requirements. Young people were then asked to provide their own consent by signing a consent form, any questions they had about the studies were discussed. Information sheets were made available to parents in studies three and four. In study two, the parents of young people participating in the focus groups were informed of the study through the organizations used to recruit the young people. My full contact details were provided on these sheets so parents/carers could address any queries they had about the project if they wished. When young people participated in the online questionnaire (study two) we did not know who they were so could not provide separate information to parents. Consequently, all participants were given the opportunity to participate in the research without the consent of their parents/carers unless they were deemed incompetent to consent. In this case the onus was on the adult (parent or carer for example) to prove incompetency ( Alderson, 2007 ). Favorable ethical approval, including approval for the above consent procedures, was granted by the Faculty Research Ethics Committee at Anglia Ruskin University.

In the next section I provide a synthesis of the findings across the four studies before discussing how participatory research with young people can offer new understandings of bullying and its impacts on young people.

Although each study was designed to answer specific bullying research questions, the following key themes cut across all four studies 1 :

• Bullying definitions

◦ Behaviors

• Impact of bullying on victim

• Reporting bullying

Bullying Definitions

Young people had various understandings about what they considered bullying to be. Overall, participants agreed that aggressive direct behaviors, mainly focusing on physical aggression, constituted bullying:

“…if someone is physically hurt then that is bullying straight away.” (Female, study 3).

“I think [cyber-bullying is] not as bad because with verbal or physical, you are more likely to come in contact with your attacker regularly, and that can be disturbing. However, with cyber-bullying it is virtual so you can find ways to avoid the person.” (Female, study 2).

Name-calling was an ambiguous concept, young people generally believed that in isolation name-calling might not be bullying behavior or it could be interpreted as “joking” or “banter”:

“I never really see any, a bit of name calling and taking the mick but nothing ever serious.” (Male, study 3).

The concept of “banter” or “joking” was explored in study three as a result of the participatory design. Young people suggested “banter” involves:

“…a personal joke or group banter has no intention to harm another, it is merely playful jokes.” (Female, study 3).

However, underpinning this understanding of “banter” was the importance of intentionality:

“Banter saying things bad as a joke and everyone knows it is a joke.” (Male, study 3).

“Banter” was thus contentious when perception and reception were ambiguous. In some cases, “banter” was considered “normal behavior”:

“…we’ve just been joking about, but it’s never been anything harsh it’s just been like having a joke…” (Male, study 3).

The same view was evident in relation to cyber-bullying. Some participants were rather dismissive of this approach suggesting that it did not exist:

“I don’t really think it exists. If you’re being cyber-“bullied” then there is something wrong with you- it is insanely easy to avoid, by blocking people and so on. Perhaps it consists of people insulting you online?” (Male, study 2).

When young people considered additional factors added to name calling such as the type of name-calling, or aspects of repetition or intention, then a different view was apparent.

“…but it has to be constant it can’t be a single time because that always happens.” (Male, study 3).

Likewise with words used on social media, young people considered intentionality in their consideration of whether particular behaviors were bullying, highlighting important nuances in how bullying is conceptualized:

“Some people they don’t want to sound cruel but because maybe if you don’t put a smiley face on it, it might seem cruel when sometimes you don’t mean it.” (Female, study 2).

Study one also found that young people were more likely to discuss sexist or racist bullying in interviews or focus groups but this information was scarce in the questionnaire data. This is possibly as a result of how the questions were framed and the researchers’ perspectives informing the questions.

Evident across the four studies was the understanding young people had about the effects of continuous name-calling on victims:

“…you can take one comment, you can just like almost brush it off, but if you keep on being bullied and bullied and bullied then you might kind of think, hang on a minute, they’ve taken it a step too far, like it’s actually become more personal, whereas just like a cheeky comment between friends it’s become something that’s more serious and more personal and more annoying or hurtful to someone.” (Female, study 3).

“Cyber-bullying is basically still verbal bullying and is definitely psychological bullying. Any bullying is psychological though, really. And any bullying is going to be harmful.” (Female, study 2).

Aspects of indirect bullying (social exclusion) were features of studies one and three. For the most part, the research reviewed in study one found that as young people got older they were less likely to consider characteristics of social exclusion in their definitions of bullying. In study three, when discussing the school’s anti-bullying policy, study participants raised questions about “ isolating a student from a friendship group .” Some contested this statement as a form of bullying:

“…. there is avoiding, as in, not actively playing a role in trying to be friends which I don’t really see as bullying I see this as just not getting someone to join your friendship group. Whereas if you were actually leaving him out and rejecting him if he tries to be friends then I think I would see that as malicious and bullying.” (Male, study 3).

“Isolating a student from a friendship group – I believe there are various reasons for which a student can be isolated from a group – including by choice.” (Female, study 3).

Cyber-bullying was explored in detail in study two but less so in the other three studies. Most study two participants considered that cyber-bullying was just as harmful, or in some cases worse than, ‘traditional’ bullying due to the use of similar forms of “harassment,” “antagonizing,” “tormenting,” and ‘threatening’ through online platforms. Some young people believed that the physical distance between the victim and the bully is an important aspect of cyber-bullying:

“I think it’s worse because people find it easier to abuse someone when not face to face.” (Male, study 2).

“I think it could be worse, because lots of other people can get involved, whereas when it’s physical bullying it’s normally just between one or two or a smaller group, things could escalate too because especially Facebook, they’ve got potential to escalate.” (Female, study 2).

Other participants in study two spoke about bullying at school which transfers to an online platform highlighting no “escape” for some. In addition, it was made clearer that some young people considered distancing in relation to bullying and how this influences perceptions of severity:

“…when there’s an argument it can continue when you’re not at school or whatever and they can continue it over Facebook and everyone can see it then other people get involved.” (Female, study 2).

“I was cyber-bullied on Facebook, because someone put several hurtful comments in response to my status updates and profile pictures. This actually was extended into school by the bully…” (Male, study 2).

Impact of Bullying on Victim

Although bullying behaviors were a primary consideration of young people’s understanding of bullying, many considered the consequences associated with bullying and in particular, the impact on mental health. In these examples, the specifics of the bullying event were irrelevant to young people and the focus was on how the behavior was received by the recipient.

In study two, young people divulged how cyber-bullying had adversely affected their ability to go to school and to socialize outside school. Indeed some young people reported the affects it had on their confidence and self-esteem:

“I developed anorexia nervosa. Although not the single cause of my illness, bullying greatly contributed to my low self-esteem which led to becoming ill.” (Female, study 2).

“It hurts people’s feelings and can even lead to committing suicide….” (Female, study 2).

Across the studies, young people who had been bullied themselves shared their individual experiences:

“….you feel insecure and it just builds up and builds up and then in the end you have no self-confidence.” (Female, study 2).

“…it was an everyday thing I just couldn’t take it and it was causing me a lot of anxiety.” (Male, study 4).

“I am different to everyone in my class …. I couldn’t take it no more I was upset all the time and it made me feel anxious and I wasn’t sleeping but spent all my time in bed being sad and unhappy.” (Male, study 4).

Young people who had not experienced bullying themselves agreed that the impact it had on a person was a large determiner of whether bullying had happened:

“When your self-confidence is severely affected and you become shy. Also when you start believing what the bullies are saying about you and start to doubt yourself.” (Female, study 3).

“…it makes the victim feel bad about themselves which mostly leads to depression and sadness.” (Male, study 2).

Further evidence around the impact of bullying was apparent in the data in terms of how relational aspects can affect perceived severity. In the case of cyber-bullying, young people suggested a sense of detachment because the bullying takes place online. Consequently, as the relational element is removed bullying becomes easier to execute:

“…because people don’t have to face them over a computer so it’s so much easier. It’s so much quicker as well cos on something like Facebook it’s not just you, you can get everyone on Facebook to help you bully that person.” (Female, study 2).

“Due to technology being cheaper, it is easier for young people to bully people in this way because they don’t believe they can be tracked.” (Male, study 2).

“The effects are the same and often the bullying can be worse as the perpetrator is unknown or can disguise their identity. Away from the eyes of teachers etc., more can be done without anyone knowing.” (Female, study 2).

Relational aspects of bullying were further highlighted with regards to how “banter” was understood, particularly with in-group bullying and how the same example can either be seen as “banter” or bullying depending on the nature of the relationship:

“…we’ve just been joking about, but it’s never been anything harsh it’s just been like having a joke. well, I haven’t done it but I’ve been in a crowd where people do it, so I don’t want to get involved just in case it started an argument.” (Female, study 3).

“But it also depends…who your groups with, for example, if I spoke to my friends from [School]… I wouldn’t like use taboo language with them because to them it may seem inappropriate and probably a bit shocked, but if I was with my friends outside of school we use taboo language, we’ll be ourselves and we’ll be comfortable with it, and if a stranger walked past and heard us obviously they’d be thinking that we’re being bullied ourselves.” (Female, study 3).

Furthermore, how individuals are perceived by others tended to influence whether they were believed or not. In study four for example, participants suggested that who the bullies were within the school might have impacted how complaints were acted upon by school officials:

“When I went to the school about it, the students said I had attacked them – all eight of them! I just realized that no one believes me….” (Female, study 4).

While in study three, a characteristic of bullying was the influence the aggressor has over the victim:

“When the victim starts to feel in danger or start to fear the other person. Consequently he or she tries to avoid the bad guy (or girl!)” (Male, study 3).

These relational and contextual issues also influenced a young person’s ability to report bullying.

Reporting Bullying

Young people were more likely to report bullying when they considered it was ‘serious’ enough. Just under half of participants in study two sought emotional/practical support if they worried about, or were affected by cyber-bullying, with most talking to their parents. In study three, young people were less likely to seek support but when they did, most went to their teachers. In study four, all participants reported bullying in school where they did not feel supported.

Fear of making the bullying worse was captured across the studies as a reason for not reporting it:

“I’m scared that if I tell then the bullying will still go on and they will do more.” (Female, study 3).

“The bully might bully you if he finds out.” (Male, study 3).

Being able to deal with the incident themselves was also a reason for non-reporting:

“…it’s embarrassing and not necessary, my friends help me through it, adults never seem to understand.” (Female, study 2).

“I don’t tend to talk to anyone about it, I just keep it to myself and obviously that’s the worst thing you should ever do, you should never keep it to yourself, because I regret keeping it to myself to be honest….” (Female, study 3).

“…but I think I’d deal with it myself ‘cos. I was quite insecure but now I’m quite secure with myself, so I’ll sort it out myself. I think it’s just over time I’ve just sort of hardened to it.” (Male, study 3).

Most young people seeking support for bullying said they spoke to an adult but the helpfulness of this support varied. This finding is important for understanding relationships between young people and adults. Those who felt supported by their teachers for example, suggested that they took the time to listen and understood what they were telling them. They also reassured young people who in turn believed that the adult they confided in would know what to do:

“So I think the best teacher to talk to is [Miss A] and even though people are scared of her I would recommend it, because she’s a good listener and she can sense when you don’t want to talk about something, whereas the other teachers force it out of you.” (Female, study 3).

“My school has had assemblies about cyber-bullying and ways you can stop it or you can report it anonymously…. you can write your name or you can’t, it’s all up to YOU.” (Male, study 2).

Others however had a negative experience of reporting bullying and a number of reasons were provided as to why. Firstly, young people stated that adults did not believe them which made the bullying worse on some level:

“I went to the teachers a couple of times but, no, I don’t think they could do anything. I did sort of go three times and it still kept on going, so I just had to sort of deal with it and I sort of took it on the cheek….” (Male, study 3).

Secondly, young people suggested that adults did not always listen to their concerns, or in some cases did not take their concerns seriously enough:

“…I had had a really bad day with the girls so I came out and I explained all this to my head of year and how it was affecting me but instead of supporting me he put me straight into isolation.” (Male, study 4).

“I could understand them thinking I maybe got the wrong end of the stick with one incident but this was 18 months of me constantly reporting different incidents.” (Female, study 4).

“If cyber-bullying is brought to our school’s attention, usually, they expect printed proof of the situation and will take it into their own hand depending on its seriousness. However this is usually a couple of detentions. And it’s just not enough.” (Female, study 2).

Finally, some young people suggested that teachers did not always know what to do when bullying concerns were raised and consequently punished those making the complaint:

“I think I would have offered support instead of punishment to someone who was suffering with anxiety. I wouldn’t have seen anxiety as bad behavior I think that’s quite ignorant but they saw it as bad behavior.” (Male, study 4).

It is worth reiterating, that the majority of young people across the studies did not report bullying to anybody , which further underscores the contextual issues underpinning bullying and its role in enabling or disabling bullying behaviors. Some considered it was “pointless” reporting the bullying and others feared the situation would be made worse if they did:

“My school hide and say that bullying doesn’t go on cos they don’t wanna look bad for Ofsted.” (Male, study 2).

“My school is oblivious to anything that happens, many things against school rules happen beneath their eyes but they either refuse to acknowledge it or are just not paying attention so we must suffer.” (Female, study 2).

“That’s why I find that when you get bullied you’re scared of telling because either, in most cases the teacher will – oh yeah, yeah, don’t worry, we’ll sort it out and then they don’t tend to, and then they get bullied more for it.” (Female, study 3).

Young people were concerned that reporting bullying would have a negative impact on their friendship groups. Some were anxious about disrupting the status quo within:

“I think everyone would talk about me behind my back and say I was mean and everyone would hate me.” (Female, study 3).

Others expressed concern about the potential vulnerability they were likely to experience if they raised concerns of bullying:

“I was worried it might affect my other friendships.”(Boy, study 2).

“I’m scared that if I tell, then the bullying will still go on and they will do more.” (Female, study 3).

“….because they might tell off the bullies and then the bullies will like get back at you.” (Female, study 3).

These findings underscore the importance of contextual and relational factors in understanding bullying from the perspectives of young people and how these factors influence a young person’s ability or willingness to report bullying.

Finally one young person who had self-excluded from school due to severe bullying suggested that schools:

“…need to be looking out for their students’ mental wellbeing – not only be there to teach them but to support and mentor them. Keep them safe really… I missed out on about three years of socializing outside of school because I just couldn’t do it. I think it’s important that students are encouraged to stand up for each other.” (Female, study 4).

The studies presented in this paper illustrate the multitude of perceptions underpinning young people’s understandings of what constitutes bullying, both in terms of the behavior and also the impact that this behavior has on an individual. In turn, the ambiguity of what constitutes bullying had an impact on a young person’s ability to seek support. Discrepancies in bullying perceptions within and between young people’s groups are shown, highlighting the fluid and changing roles that occur within a bullying situation. Findings from quantitative studies have demonstrated the differing perceptions of bullying by adults and young people (see for example Smith et al., 2002 ; Vaillancourt et al., 2008 ; Maunder et al., 2010 ; Cuadrado-Gordillo, 2012 ). However, by combining findings from participatory research, new understandings of the relational and contextual factors important to young people come to the fore.

Young people participating in these four studies had unique knowledge and experiences of bullying and the social interactions of other young people in their schools and wider friendship groups. The underpinning participatory design enabled me to work alongside young people to analyze and understand their unique perspectives of bullying in more detail. The research teams were therefore able to construct meaning together, based not entirely on our own assumptions and ideologies, but including the viewpoint of the wider research participant group ( Thomson and Gunter, 2008 ). Together, through the process of co-constructing bullying knowledge, we were able to build on what is already known in this field and contribute to the view that bullying is socially constructed through the experiences of young people and the groups they occupy ( Schott and Sondergaard, 2014 ).

With regards to understanding what bullying is, the findings from these studies corroborate those of the wider literature from both paradigms of inquiry (for example Naylor et al., 2001 ; Canty et al., 2016 ); that being the discrepancies in definitions between adults and young people and also between young people themselves. Yet, findings here suggest that young people’s bullying definitions are contextually and relationally contingent. With the exception of physical bullying, young people did not differentiate between direct or indirect behaviors, instead they tended to agree that other contextual and relational factors played a role in deciding if particular behaviors were bullying (or not). The participatory research design enabled reflection and further investigation of the ideas that were particularly important to young people such as repetition and intentionality. Repetition was generally seen as being indicative of bullying being “serious,” and therefore more likely to be reported, and without repetition, a level of normality was perceived. This finding contradicts some work on bullying definitions, Cuadrado-Gordillo (2012) for example found that regardless of the role played by young people in a bullying episode (victim, aggressor or witness), the criteria of ‘repetition’ was not important in how they defined bullying.

Relational factors underpinning young people’s perception of bullying and indeed it’s “seriousness” were further reflected in their willingness or otherwise to report it. Fear of disrupting the status quo of the wider friendship group, potentially leading to their own exclusion from the group, was raised as a concern by young people. Some were concerned their friends would not support them if they reported bullying, while others feared further retaliation as a result. Friendship groups have been identified as a source of support for those who have experienced bullying and as a protective factor against further bullying ( Allen, 2014 ). Although participants did not suggest their friendship groups are unsupportive it is possible that group dynamics underscore seeking (or not) support for bullying. Other literature has described such practices as evidence of a power imbalance ( Olweus, 1995 ; Cuadrado-Gordillo, 2012 ) but young people in these studies did not describe these unequal relationships in this way and instead focused on the outcomes and impacts of bullying. Indeed Cuadrado-Gordillo (2012) also found that young people in their quantitative study did not consider “power imbalance” in their understanding of bullying and were more likely to consider intention. This paper, however, underscores the relational aspects of definitions of bullying and, how the dynamics of young people’s friendships can shift what is understood as bullying or not. Without such nuances, some behaviors may be overlooked as bullying, whereas other more obvious behaviors draw further attention. This paper also shows that contextual issues such as support structures can shift how young people see bullying. Contextual factors were evident across the four studies through the recognition of bullying being enabled or disabled by institutional factors, including a school’s ability to respond appropriately to bullying concerns. Young people suggested that schools could be influenced by bullies, perceiving them as non-threatening and consequently not dealing appropriately with the situation. Indeed some young people reported that their schools placed the onus on them as victims to change, consequently placing the “blame” on victims instead. These findings raise questions about who young people feel able to confide in about bullying as well as issues around training and teacher preparedness to deal with bullying in schools. Evidenced in these four studies, is that young people feel somewhat disconnected from adults when they have bullying concerns. Those who did report bullying, identified particular individuals they trusted and knew would support them. Novick and Isaacs (2010) identified teachers who young people felt comfortable in approaching to report bullying and described them as “most active, engaged and responsive.” (p. 291). The bullying literature suggests that as young people get older they are more likely to confide in friends than adults ( Moore and Maclean, 2012 ; Allen, 2014 ). However, findings from this paper indicate that although fewer young people reported bullying, those who did confided in an adult. Young people have identified that a variety of supports are required to tackle bullying and that adults need to listen and work with them so nuanced bullying behaviors are not recognized as “normal” behaviors. Within the data presented in this paper, “banter” was portrayed as “normal” behavior. Young people did not specify what behaviors they regarded as “banter,” but suggested that when banter is repeated and intentional the lines are blurred about what is bullying and what is banter.

Exploring bullying nuances in this paper, was enhanced by the involvement of young people in the research process who had a unique “insider” perspective about what it is like to be a young person now and how bullying is currently affecting young people. In studies one and four, young people were “active respondents” ( Bragg and Fielding, 2005 ) and provided adults with their own unique perspectives on bullying. It could be argued that study one did not involve the participation of young people. However, this study informed the basis of the subsequent studies due to the discrepancies noted in the literature about how bullying is understood between adults and young people, as well as the lack of young people’s voice and opportunity to participate in the reviewed research. Accordingly, young people’s data as “active respondents” informed adult understanding and led to future work involving more active research engagement from other young people. Participation in study four provided an opportunity for young people to contribute to future participatory research based on lived experiences as well as informing policy makers of the effects bullying has on the lives of young people ( O’Brien, 2017 ). In studies two and three, young people were involved further along Bragg and Fielding (2005) continuum as “co-researchers” and “students as researchers” with these roles shifting and moving dependent on the context of the project at the time ( O’Brien et al., 2018a ). These young researchers brought unique knowledge to the projects ( Bradbury-Jones et al., 2018 ) that could not be accessed elsewhere. Perspectives offered by the young researchers supported adults in understanding more about traditional and cyber-bullying from their perspectives. Furthermore, this knowledge can be added to other, quantitative studies to further understand why bullying happens alongside bullying prevalence, risk and protective factors, and negative outcomes.

Findings from the four studies offer an alternative perspective to how bullying is understood by young people. Complexities in defining bullying have been further uncovered as understanding is informed by individual factors, as well as wider social and relational contexts ( Horton, 2011 ; Schott and Sondergaard, 2014 ). This has implications for the type of support young people require. This paper highlights how definitions of bullying shift in response to relational and contextual aspects deemed important to young people. Because of this, further nuances were uncovered through the research process itself as the respective studies showed discrepancies in bullying perceptions within and between young people’s groups.

These understandings can act as a starting point for young people and adults to collaborate in research which seeks to understand bullying and the context to which it occurs. Furthermore, such collaborations enable adults to theorize and understand the complexities associated with bullying from the perspective of those at the center. There is a need for additional participatory research projects involving such collaborations where adults and young people can learn from each other as well as combining findings from different methodologies to enable a more comprehensive picture of the issues for young people to emerge. Further research is needed to unravel the complexities of bullying among and between young people, specifically in relation to the contextual and relational factors underscoring perceptions of bullying.

Data Availability

The raw data supporting the conclusions of this manuscript will be made available by the authors, without undue reservation, to any qualified researcher.

Ethics Statement

Ethical approval was granted for all four studies from the Faculty of Health, Education, Medicine and Social Care at the Anglia Ruskin University. The research was conducted on the premise of Gillick competency meaning that young people (in these studies over the age of 12 years) could consent for themselves to participate. Parents/carers were aware the study was happening and received information sheets explaining the process.

Author Contributions

The author confirms being the sole contributor of this work and has approved it for publication.

These four studies were conducted at the Anglia Ruskin University. Study one was part of a wider masters degree funded by the Anglia Ruskin University, Study two was funded by a group of young people convened by the National Children’s Bureau with funding from the Wellcome Trust (United Kingdom). Study three was a wider Doctoral study funded by the Anglia Ruskin University and Study four was also funded by the Anglia Ruskin University.

Conflict of Interest Statement

The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Acknowledgments

I would like to thank Dr. Grace Spencer, Ruskin Fellow at the Anglia Ruskin University for providing the critical read of this manuscript and offering constructive feedback. I would also like to thank the two independent reviewers for their feedback on the drafts of this manuscript.

• ^ These findings focus on perceptions and data from the young people in the four studies. For a full discussion on adult perceptions please refer to the individual studies.

Alderson, P. (2007). Competent children? Minors’ consent to health care treatment and research. Soc. Sci. Med. 65, 2272–2283. doi: 10.1016/j.socscimed.2007.08.005

PubMed Abstract | CrossRef Full Text | Google Scholar

Allen, M. (2014). Local Action on Health Inequalities: Building Children and Young People’s Resilience in Schools , London: Public Health England.

Google Scholar

Beresford, P. (2006). Making the connections with direct experience: from the western front to user-controlled research. Educ. Action Res. 14, 161–170.

Black, S., Weinles, D., and Washington, E. (2010). Victim strategies to stop bullying. Youth Violence Juv. Justice 8, 138–147. doi: 10.1177/1541204009349401

CrossRef Full Text | Google Scholar

Bland, D., and Atweh, B. (2007). Students as researchers: engaging students’ voices in PAR. Educ. Action Res. 15, 337–349. doi: 10.1080/09650790701514259

Boulton, M. J., and Flemington, I. (1996). The effects of a short video intervention on secondary school Pupils’ involvement in definitions of and attitudes towards bullying. Sch. Psychol. Int. 17, 331–345. doi: 10.1177/0143034396174003

Bradbury-Jones, C., Isham, L., and Taylor, J. (2018). The complexities and contradictions in participatory research with vulnerable children and young people: a qualitative systematic review. Soc. Sci. Med. 215, 80–91. doi: 10.1016/j.socscimed.2018.08.038

Bragg, S., and Fielding, M. (2005). “It’s an equal thing. It’s about achieving together: student voices and the possibility of a radical collegiality,” in Improving Schools Through Collaborative Enquiry , eds H. Street, and J. Temperley, (London: Continuum), 105–135.

Braun, V., and Clarke, V. (2006). Using thematic analysis in Psychology. Qual. Res. Psychol. 3, 77–101.

Canty, J., Stubbe, M., Steers, D., and Collings, S. (2016). The trouble with bullying–deconstructing the conventional definition of bullying for a child-centred investigation into Children’s use of social media. Child. Soc. 30, 48–58. doi: 10.1111/chso.12103

Coyne, I. (2010). Research with children and young people: the issue of parental (proxy) consent. Child. Soc. 24, 227–237.

Cuadrado-Gordillo, I. (2012). Repetition, power imbalance, and intentionality: do these criteria conform to teenagers’ perception of bullying? A role-based analysis. J. Interpers. Violence 27, 1889–1910. doi: 10.1177/0886260511431436

Davey, C. (2011). Evaluation of the PEAR Project. London: National Children’s Bureau.

deLara, E. W. (2012). Why adolescents Don’t disclose incidents of bullying and harassment. J. Sch. Violence 11, 288–305. doi: 10.1080/15388220.2012.705931

Eriksen, I. M. (2018). The power of the word: students’ and school staff’s use of the established bullying definition. Educ. Res. 60, 157–170. doi: 10.1080/00131881.2018.1454263

Espelage, D. L. (2018). Understanding the complexity of school bully involvement. Chautauqua J. 2:20.

Forsberg, C., Wood, L., Smith, J., Varjas, K., Meyers, J., Jungert, T., et al. (2018). Students’ views of factors affecting their bystander behaviors in response to school bullying: a cross-collaborative conceptual qualitative analysis. Res. Pap. Educ. 33, 127–142. doi: 10.1080/02671522.2016.1271001

Fox, R. (2013). Resisting participation: critiquing participatory research methodologies with young people. J. Youth Stud. 16, 986–999. doi: 10.1080/13676261.2013.815698

Griffin, R. S., and Gross, A. M. (2004). Childhood bullying: current empirical findings and future directions for research. Aggr. Violent Behav. 9, 379–400. doi: 10.1016/s1359-1789(03)00033-8

Guerin, S., and Hennessy, E. (2002). Pupils’ definitions of bullying. Eur. J. Psychol. Educ. 17, 249–261. doi: 10.1007/bf03173535

Horton, P. (2011). School bullying and social and moral orders. Child. Soc. 25, 268–277. doi: 10.1111/j.1099-0860.2011.00377.x

Huang, F. L., and Cornell, D. G. (2015). The impact of definition and question order on the prevalence of bullying victimization using student self-reports. Psychol. Assess. 27:1484. doi: 10.1037/pas0000149

James, A. (2007). Giving voice to children’s voices: practices and problems, pitfalls and potentials. Am. Anthropol. 109, 261–272. doi: 10.1525/aa.2007.109.2.261

Jones, A. (2004). “Involving children and yong people as researchers,” in Doing Research with Children and Young People , eds S. Fraser, V. Lewis, S. Ding, M. Kellett, and C. Robinson, (London: Sage Publications), 113–130.

Kellett, M. (2010). Small shoes, Big Steps! Empowering children as active researchers. Am. J. Commun. Psychol. 46, 195–203. doi: 10.1007/s10464-010-9324-y

La Fontaine, J. (1991). Bullying: The Child’s View – an Analysis of Telephone Calls to ChildLIne about Bullying. London: Calouste Gulbenkian Foundation.

Lambert, V., and Glacken, M. (2011). Engaging with children in research: theoretical and practical implications of negotiating informed consent/assent. Nurs. Ethics 18, 781–801. doi: 10.1177/0969733011401122

Lee, C. (2006). Exploring teachers’ definitions of bullying. Emot. Behav. Diffic. 11, 61–75. doi: 10.1080/13632750500393342

Maunder, R. E., and Crafter, S. (2018). School bullying from a sociocultural perspective. Aggr. Violent Behav. 38, 13–20. doi: 10.1016/j.avb.2017.10.010

Maunder, R. E., Harrop, A., and Tattersall, A. J. (2010). Pupil and staff perceptions of bullying in secondary schools: comparing behavioural definitions and their perceived seriousness. Educ. Res. 52, 263–282. doi: 10.1080/00131881.2010.504062

Moore, S., and Maclean, R. (2012). Victimization, friendship and resilience: crossing the land in-between. Pastor. Care Educ. 30, 147–163. doi: 10.1080/02643944.2012.679956

Naylor, P., Cowie, H., and del Rey, R. (2001). Coping strategies of secondary school children in response to being bullied. Child Psychol. Psychiatry Rev. 6, 114–120. doi: 10.1111/j.1469-7610.2009.02137.x

Novick, R. M., and Isaacs, J. (2010). Telling is compelling: the impact of students reports of bullying on teacher intervention. Educ. Psychol. 30, 283–296. doi: 10.1080/01443410903573123

O’Brien, N. (2009). Secondary school teachers’ and pupils’ definitions of bullying in the UK: a systematic review. Evid. Policy 5, 399–426.

PubMed Abstract | Google Scholar

O’Brien, N. (2014). “I Didn’t Want to be Known as a Snitch”: Using PAR to Explore Bullying in a Private day and Boarding School. Childhood Remixed. Conference Edition. Suffolk: University Campus Suffolk, 86–96.

O’Brien, N. (2016). To ‘Snitch’ or Not to ‘Snitch’? Using PAR to Explore Bullying in a Private Day and Boarding School. Available at: http://arro.anglia.ac.uk/700970/ (accessed September 20, 2018).

O’Brien, N. (2017). An Exploratory Study of Bullied Young People’s Self-Exclusion from School. Evidence: Presented at Meetings of the All Party Parliamentary Group on Bullying 2011-2016. Project Report. All Party Parliamentary Group on Bullying. Available at: http://arro.anglia.ac.uk/id/eprint/702024 (accessed September 20, 2018).

O’Brien, N., and Moules, T. (2007). So round the spiral again: a reflective participatory research project with children and young people. Educ. Action Res. J. 15, 385–402. doi: 10.1080/09650790701514382

O’Brien, N., and Moules, T. (2010). The Impact of Cyber-Bullying on Young People’s Mental Health. Project Report. Chelmsford: Anglia Ruskin University.

O’Brien, N., and Moules, T. (2013). Not sticks and stones but tweets and texts: findings from a national cyberbullying project. Pastor. Care Educ. 31, 53–65. doi: 10.1080/02643944.2012.747553

O’Brien, N., Moules, T., and Munn-Giddings, C. (2018a). “Negotiating the research space between young people and adults in a PAR study exploring school bullying,” in Reciprocal Relationships and Well-Being: Implications for Social Work and Social Policy , eds M. Torronen, C. Munn-Giddings, and L. Tarkiainen, (Oxon: Routledge), 160–175. doi: 10.4324/9781315628363-11

O’Brien, N., Munn-Giddings, C., and Moules, T. (2018b). The repercussions of reporting bullying: some experiences of students at an independent secondary school. Pastor. Care Educ. 36, 29–43. doi: 10.1080/02643944.2017.1422004

O’Brien, N., Munn-Giddings, C., and Moules, T. (2018c). The Ethics of Involving Young People Directly in the Research Process. Childhood Remixed. Conference Edition , 115–128. Available at: www.uos.ac.uk/content/centre-for-study-children-childhood (accessed May 2018).

Oliver, C., and Candappa, M. (2007). Bullying and the politics of ‘telling’. Oxford Rev. Educ. 33, 71–86. doi: 10.1080/03054980601094594

Olweus, D. (1995). Bullying or peer abuse at school: facts and intervention. Curr. Dir. Psychol. Sci. 4, 196–200. doi: 10.1111/1467-8721.ep10772640

Ozer, E. J., and Wright, D. (2012). Beyond school spirit: the effects of youth-led participatory action research in two urban high schools. J. Res. Adolesc. 22, 267–283. doi: 10.1111/j.1532-7795.2012.00780.x

Parsons, S., Abbott, C., McKnight, L., and Davies, C. (2015). High risk yet invisible: conflicting narratives on social research involving children and young people, and the role of research ethics committees. Br. Educ. Res. J. 41, 709–729. doi: 10.1002/berj.3160

Patton, D. U., Hong, J. S., Patel, S., and Kral, M. J. (2017). A systematic review of research strategies used in qualitative studies on school bullying and victimization. Trauma Violence Abuse 18, 3–16. doi: 10.1177/1524838015588502

Popay, J., Roberts, H., Sowden, A., Petticrew, M., Arai, L., Rodgers, M., et al. (2006). Guidance on the conduct of narrative synthesis in systematic reviews. Eur. Soc. Res. Council Methods Program. doi: 10.13140/2.1.1018.4643

Schott, R. M., and Sondergaard, D. M. (2014). “Introduction: new approaches to school bullying,” in School Bullying: New Theories in Context , eds R. M. Schott, and D. M. Sondergaard, (Massachusetts, MA: Cambridge University Press), 1–17.

Skrzypiec, G., Slee, P., Murray-Harvey, R., and Pereira, B. (2011). School bullying by one or more ways: does it matter and how do students cope? Sch. Psychol. Int. 32, 288–311. doi: 10.1177/0143034311402308

Smith, P. K., Cowie, H., Olafsson, R. F., and Liefooghe, A. P. D. (2002). Definitions of bullying: a comparison of terms used, and age and gender differences, in a fourteen-country international comparison. Child Dev. 73, 1119–1133. doi: 10.1111/1467-8624.00461

Spears, B., Taddeo, C., Collin, P., Swist, T., Razzell, M., Borbone, V., et al. (2016). Safe and Well Online: Learnings from Four Social Marketing Campaigns for Youth Wellbeing. Available at: https://researchdirect.westernsydney.edu.au/islandora/object/uws:36405/datastream/PDF/view (accessed July 1, 2019).

Stoudt, B. G. (2009). The role of language & discourse in the investigation of privilege: using participatory action research to discuss theory. Dev. Methodol. Interrupt. Power Urban Rev. 41, 7–28.

Thomson, P., and Gunter, H. (2008). Researching Bullying with students: a lens on everyday life in an ‘innovative school’. Int. J. Inclusive Educ. 12, 185–200. doi: 10.1080/13603110600855713

Tisdall, E. K. M., and Punch, S. (2012). Not so ‘new’? Looking critically at childhood studies. Child. Geogr. 10, 249–264. doi: 10.1080/14733285.2012.693376

United Nations Convention on the Rights of the Child (1989). Available at: http://www.unicef.org.uk/Documents/Publication-pdfs/UNCRC_PRESS2009 10web.pdf (accessed January 19, 2014).

Vaillancourt, T., McDougall, P., Hymel, S., Krygsman, A., Miller, J., Stiver, K., et al. (2008). Bullying: are researchers and children/youth talking about the same thing? Int. J. Behav. Dev. 32, 486–495. doi: 10.1177/0165025408095553

Keywords : bullying, young people, participatory research, social constructionism, young people as researchers, collaboration, bullying supports

Citation: O’Brien N (2019) Understanding Alternative Bullying Perspectives Through Research Engagement With Young People. Front. Psychol. 10:1984. doi: 10.3389/fpsyg.2019.01984

Received: 28 February 2019; Accepted: 13 August 2019; Published: 28 August 2019.

Reviewed by:

Copyright © 2019 O’Brien. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Niamh O’Brien, [email protected]

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Numbers, Facts and Trends Shaping Your World

Read our research on:

Full Topic List

Regions & Countries

• Publications
• Our Methods
• Short Reads
• Tools & Resources

Read Our Research On:

• Teens and Cyberbullying 2022

Nearly half of U.S. teens have been bullied or harassed online, with physical appearance being seen as a relatively common reason why. Older teen girls are especially likely to report being targeted by online abuse overall and because of their appearance

Table of contents.

• Acknowledgments
• Methodology

Pew Research Center conducted this study to better understand teens’ experiences with and views on bullying and harassment online. For this analysis, we surveyed 1,316 U.S. teens. The survey was conducted online by Ipsos from April 14 to May 4, 2022.

This research was reviewed and approved by an external institutional review board (IRB), Advarra, which is an independent committee of experts that specializes in helping to protect the rights of research participants.

Ipsos recruited the teens via their parents who were a part of its  KnowledgePanel , a probability-based web panel recruited primarily through national, random sampling of residential addresses. The survey is weighted to be representative of U.S. teens ages 13 to 17 who live with parents by age, gender, race, ethnicity, household income and other categories.

Here are the  questions used for this report , along with responses, and  its methodology .

While bullying existed long before the internet, the rise of smartphones and social media has brought a new and more public arena into play for this aggressive behavior.

Nearly half of U.S. teens ages 13 to 17 (46%) report ever experiencing at least one of six cyberbullying behaviors asked about in a Pew Research Center survey conducted April 14-May 4, 2022. 1

The most commonly reported behavior in this survey is name-calling, with 32% of teens saying they have been called an offensive name online or on their cellphone. Smaller shares say they have had false rumors spread about them online (22%) or have been sent explicit images they didn’t ask for (17%).

Some 15% of teens say they have experienced someone other than a parent constantly asking them where they are, what they’re doing or who they’re with, while 10% say they have been physically threatened and 7% of teens say they have had explicit images of them shared without their consent.

In total, 28% of teens have experienced multiple types of cyberbullying.

Defining cyberbullying in this report

This report measures cyberbullying of teens using six distinct behaviors:

• Offensive name-calling
• Spreading of false rumors about them
• Receiving explicit images they didn’t ask for
• Physical threats
• Constantly being asked where they are, what they’re doing, or who they’re with by someone other than a parent
• Having explicit images of them shared without their consent

Teens who indicate they have personally experienced any of these behaviors online or while using their cellphone are considered targets of cyberbullying in this report. The terms “cyberbullying” and “online harassment” are used interchangeably throughout this report.

Age and gender are related to teens’ cyberbullying experiences, with older teen girls being especially likely to face this abuse

Teens’ experiences with online harassment vary by age. Some 49% of 15- to 17-year-olds have experienced at least one of the six online behaviors, compared with 42% of those ages 13 to 14. While similar shares of older and younger teens report being the target of name-calling or rumor spreading, older teens are more likely than their younger counterparts (22% vs. 11%) to say someone has sent them explicit images they didn’t ask for, an act sometimes referred to as cyberflashing ; had someone share explicit images of them without their consent, in what is also known as revenge porn (8% vs. 4%); or been the target of persistent questioning about their whereabouts and activities (17% vs. 12%).

While there is no gender difference in having ever experienced online abuse, teen girls are more likely than teen boys to say false rumors have been spread about them. But further differences are seen when looking at age and gender together: 15- to 17-year-old girls stand out for being particularly likely to have faced any cyberbullying, compared with younger teen girls and teen boys of any age. Some 54% of girls ages 15 to 17 have experienced at least one of the six cyberbullying behaviors, while 44% of 15- to 17-year-old boys and 41% of boys and girls ages 13 to 14 say the same. These older teen girls are also more likely than younger teen girls and teen boys of any age to report being the target of false rumors and constant monitoring by someone other than a parent.

White, Black and Hispanic teens do not statistically differ in having ever been harassed online, but specific types of online attacks are more prevalent among certain groups. 2 For example, White teens are more likely to report being targeted by false rumors than Black teens. Hispanic teens are more likely than White or Black teens to say they have been asked constantly where they are, what they’re doing or who they’re with by someone other than a parent.

There are also differences by household income when it comes to physical threats. Teens who are from households making less than $30,000 annually are twice as likely as teens living in households making $75,000 or more a year to say they have been physically threatened online (16% vs. 8%).

Beyond those differences related to specific harassing behaviors, older teen girls are particularly likely to say they experience multiple types of online harassment. Some 32% of teen girls have experienced two or more types of online harassment asked about in this survey, while 24% of teen boys say the same. And 15- to 17-year-olds are more likely than 13- to 14-year-olds to have been the target of multiple types of cyberbullying (32% vs. 22%).

These differences are largely driven by older teen girls: 38% of teen girls ages 15 to 17 have experienced at least two of the harassing behaviors asked about in this survey, while roughly a quarter of younger teen girls and teen boys of any age say the same.

Beyond demographic differences, being the target of these behaviors and facing multiple types of these behaviors also vary by the amount of time youth spend online. Teens who say they are online almost constantly are not only more likely to have ever been harassed online than those who report being online less often (53% vs 40%), but are also more likely to have faced multiple forms of online abuse (37% vs. 21%).

These are some of the findings from a Pew Research Center online survey of 1,316 U.S. teens conducted from April 14 to May 4, 2022.

Black teens are about twice as likely as Hispanic or White teens to say they think their race or ethnicity made them a target of online abuse

There are numerous reasons why a teen may be targeted with online abuse. This survey asked youth if they believed their physical appearance, gender, race or ethnicity, sexual orientation or political views were a factor in them being the target of abusive behavior online.

Teens are most likely to say their physical appearance made them the target of cyberbullying. Some 15% of all teens think they were cyberbullied because of their appearance.

About one-in-ten teens say they were targeted because of their gender (10%) or their race or ethnicity (9%). Teens less commonly report being harassed for their sexual orientation or their political views – just 5% each.

Looking at these numbers in a different way, 31% of teens who have personally experienced online harassment or bullying think they were targeted because of their physical appearance. About one-in-five cyberbullied teens say they were targeted due to their gender (22%) or their racial or ethnic background (20%). And roughly one-in-ten affected teens point to their sexual orientation (12%) or their political views (11%) as a reason why they were targeted with harassment or bullying online.

The reasons teens cite for why they were targeted for cyberbullying are largely similar across major demographic groups, but there are a few key differences. For example, teen girls overall are more likely than teen boys to say they have been cyberbullied because of their physical appearance (17% vs. 11%) or their gender (14% vs. 6%). Older teens are also more likely to say they have been harassed online because of their appearance: 17% of 15- to 17-year-olds have experienced cyberbullying because of their physical appearance, compared with 11% of teens ages 13 to 14.

Older teen girls are particularly likely to think they have been harassed online because of their physical appearance: 21% of all 15- to 17-year-old girls think they have been targeted for this reason. This compares with about one-in-ten younger teen girls or teen boys, regardless of age, who think they have been cyberbullied because of their appearance.

A teen’s racial or ethnic background relates to whether they report having been targeted for cyberbullying because of race or ethnicity. Some 21% of Black teens report being made a target because of their race or ethnicity, compared with 11% of Hispanic teens and an even smaller share of White teens (4%).

There are no partisan differences in teens being targeted for their political views, with 5% of those who identify as either Democratic or Republican – including those who lean toward each party – saying they think their political views contributed to them being cyberbullied.

Black or Hispanic teens are more likely than White teens to say cyberbullying is a major problem for people their age

In addition to measuring teens’ own personal experiences with cyberbullying, the survey also sought to understand young people’s views about online harassment more generally.

The vast majority of teens say online harassment and online bullying are a problem for people their age, with 53% saying they are a major problem. Just 6% of teens think they are not a problem.

Certain demographic groups stand out for how much of a problem they say cyberbullying is. Seven-in-ten Black teens and 62% of Hispanic teens say online harassment and bullying are a major problem for people their age, compared with 46% of White teens. Teens from households making under $75,000 a year are similarly inclined to call this type of harassment a major problem, with 62% making this claim, compared with 47% of teens from more affluent homes. Teen girls are also more likely than boys to view cyberbullying as a major problem.

Views also vary by community type. Some 65% of teens living in urban areas say online harassment and bullying are a major problem for people their age, compared with about half of suburban and rural teens.

Partisan differences appear as well: Six-in-ten Democratic teens say this is a major problem for people their age, compared with 44% of Republican teens saying this.

Roughly three-quarters of teens or more think elected officials and social media sites aren’t adequately addressing online abuse

In recent years, there have been several initiatives and programs aimed at curtailing bad behavior online, but teens by and large view some of those behind these efforts – including social media companies and politicians – in a decidedly negative light.

According to teens, parents are doing the best of the five groups asked about in terms of addressing online harassment and online bullying, with 66% of teens saying parents are doing at least a good job, including one-in-five saying it is an excellent job. Roughly four-in-ten teens report thinking teachers (40%) or law enforcement (37%) are doing a good or excellent job addressing online abuse. A quarter of teens say social media sites are doing at least a good job addressing online harassment and cyberbullying, and just 18% say the same of elected officials. In fact, 44% of teens say elected officials have done a poor job addressing online harassment and online bullying.

Teens who have been cyberbullied are more critical of how various groups have addressed online bullying than those who haven’t

Teens who have experienced harassment or bullying online have a very different perspective on how various groups have been handling cyberbullying compared with those who have not faced this type of abuse. Some 53% of teens who have been cyberbullied say elected officials have done a poor job when it comes to addressing online harassment and online bullying, while 38% who have not undergone these experiences say the same (a 15 percentage point gap). Double-digit differences also appear between teens who have and have not been cyberbullied in their views on how law enforcement, social media sites and teachers have addressed online abuse, with teens who have been harassed or bullied online being more critical of each of these three groups. These harassed teens are also twice as likely as their peers who report no abuse to say parents have done a poor job of combatting online harassment and bullying.

Aside from these differences based on personal experience with cyberbullying, only a few differences are seen across major demographic groups. For example, Black teens express greater cynicism than White teens about how law enforcement has fared in this space: 33% of Black teens say law enforcement is doing a poor job when it comes to addressing online harassment and online bullying; 21% of White teens say the same. Hispanic teens (25%) do not differ from either group on this question.

Large majorities of teens believe permanent bans from social media and criminal charges can help reduce harassment on the platforms

Teens have varying views about possible actions that could help to curb the amount of online harassment youth encounter on social media.

While a majority of teens say each of five possible solutions asked about in the survey would at least help a little, certain measures are viewed as being more effective than others.

Teens see the most benefit in criminal charges for users who bully or harass on social media or permanently locking these users out of their account. Half of teens say each of these options would help a lot in reducing the amount of harassment and bullying teens may face on social media sites.

About four-in-ten teens think that if social media companies looked for and deleted posts they think are bullying or harassing (42%) or if users of these platforms were required to use their real names and pictures (37%) it would help a lot in addressing these issues. The idea of forcing people to use their real name while online has long existed and been heavily debated: Proponents see it as a way to hold bad actors accountable and keep online conversations more civil , while detractors believe it would do little to solve harassment and could even  worsen it .

Three-in-ten teens say school districts monitoring students’ social media activity for bullying or harassment would help a lot. Some school districts already use digital monitoring software to help them identify worrying student behavior on school-owned devices , social media and other online platforms . However, these programs have been met with criticism regarding privacy issues , mixed results and whether they do more harm than good .

Having personally experienced online harassment is unrelated to a teen’s view on whether these potential measures would help a lot in reducing these types of adverse experiences on social media. Views do vary widely by a teen’s racial or ethnic background, however.

Black or Hispanic teens are consistently more optimistic than White teens about the effectiveness of each of these measures.

Majorities of both Black and Hispanic teens say permanently locking users out of their account if they bully or harass others or criminal charges for users who bully or harass on social media would help a lot, while about four-in-ten White teens express each view.

In the case of permanent bans, Black teens further stand out from their Hispanic peers: Seven-in-ten say this would help a lot, followed by 59% of Hispanic teens and 42% of White teens.

• It is important to note that there are various ways researchers measure youths’ experiences with cyberbullying and online harassment. As a result, there may be a range of estimates for how many teens report having these experiences. In addition, since the Center last polled on this topic in 2018, there have been changes in how the surveys were conducted and how the questions were asked. For instance, the 2018 survey asked about bullying by listing a number of possible behaviors and asking respondents to “check all that apply.” This survey asked teens to answer “yes” or “no” to each item individually. Due to these changes, direct comparisons cannot be made across the two surveys. ↩
• There were not enough Asian American teen respondents in the sample to be broken out into a separate analysis. As always, their responses are incorporated into the general population figures throughout the report. ↩

Sign up for our weekly newsletter

Fresh data delivery Saturday mornings

Sign up for The Briefing

Weekly updates on the world of news & information

• Online Harassment & Bullying
• Teens & Tech
• Teens & Youth

Teens and Video Games Today

How teens and parents approach screen time, teens, social media and technology 2023, teens and social media: key findings from pew research center surveys, gun deaths among u.s. children and teens rose 50% in two years, most popular, report materials.

901 E St. NW, Suite 300 Washington, DC 20004 USA (+1) 202-419-4300 | Main (+1) 202-857-8562 | Fax (+1) 202-419-4372 |  Media Inquiries

Research Topics

• Email Newsletters

ABOUT PEW RESEARCH CENTER  Pew Research Center is a nonpartisan fact tank that informs the public about the issues, attitudes and trends shaping the world. It conducts public opinion polling, demographic research, media content analysis and other empirical social science research. Pew Research Center does not take policy positions. It is a subsidiary of  The Pew Charitable Trusts .

© 2024 Pew Research Center

• Religious Studies
• Abrahamic Religions

Model Konseling Islam Dalam Menangani Korban Bullying Pada Remaja

• Diversity Jurnal Ilmiah Pascasarjana 4(1):01-15
• This person is not on ResearchGate, or hasn't claimed this research yet.

Discover the world's research

• 25+ million members
• 160+ million publication pages
• 2.3+ billion citations

No full-text available

To read the full-text of this research, you can request a copy directly from the authors.

• Ibnu Mahmudi
• Y S J Amini
• Djauharah Bawazier
• N Dan Warsiah
• Metode Penelitian Kuantitatif
• Abdul Hayat
• D Deharnita
• Aunur Rahim Faqih
• Syahrum Salim Dan
• Recruit researchers
• Join for free
• Login Email Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google Welcome back! Please log in. Email · Hint Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google No account? Sign up

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Campbell Syst Rev
• v.17(2); 2021 Jun

Effectiveness of school‐based programs to reduce bullying perpetration and victimization: An updated systematic review and meta‐analysis

Hannah gaffney.

1 Institute of Criminology, University of Cambridge, Cambridge UK

Maria M. Ttofi

David p. farrington, executive summary/abstract.

Bullying first emerged as an important topic of research in the 1980s in Norway (Olweus), and a recent meta‐analysis shows that these forms of aggression remain prevalent among young people globally (Modecki et al.). Prominent researchers in the field have defined bullying as any aggressive behavior that incorporates three key elements, namely: (1) an intention to harm, (2) repetitive in nature, and (3) a clear power imbalance between perpetrator and victim (Centers for Disease Control and Prevention; Farrington). There are many negative outcomes associated with bullying perpetration, such as: suicidal ideation (Holt et al.), weapon carrying (Valdebenito et al.), drug use (Ttofi et al.), and violence and offending in later life (Ttofi et al.). Bullying victimization too is associated with negative outcomes such as: suicidal ideation (Holt et al.), anxiety, low self‐esteem and loneliness (Hawker& Boulton). Therefore, school bullying is an important target for effective intervention, and should be considered a matter of public health concern.

The objective of this review is to establish whether or not existing school‐based antibullying programs are effective in reducing school‐bullyng behaviors. This report also updates a previous meta‐analysis conducted by Farrington and Ttofi. This earlier review found that antibullying programs are effective in reducing bullying perpetration and victimization and a primary objective of the current report is to update the earlier analysis of 53 evaluations by conducting new searches for evaluations conducted and published since 2009.

Search Methods

Systematic searches were conducted using Boolean combinations of the following keywords: bully*; victim*; bully‐victim; school; intervention; prevention; program*; evaluation; effect*; and anti‐bullying . Searches were conducted on several online databases including, Web of Science, PscyhINFO, EMBASE, EMBASE, DARE, ERIC, Google Scholar, and Scopus. Databases of unpublished reports, such as masters' and doctoral theses (e.g., Proquest) were also searched.

Selection Criteria

Results from systematic searches were screened thoroughly against the following inclusion criteria. To be included in this review, a study must have: (1) described an evaluation of a school‐based antibullying program implemented with school‐age participants; (2) utilized an operational definition of school‐bullying that coincides with existing definitions; (3) measured school‐bullying perpetration and/or victimization using quantitative measures, such as, self‐, peer‐, or teacher‐report questionnaires; and (4) used an experimental or quasi‐experimental design, with one group receiving the intervention and another not receiving the intervention.

Data Collection and Analysis

Of the 19,877 search results, 474 were retained for further screening. The majority of these were excluded, and after multiple waves of screening, 100 evaluations were included in our meta‐analysis. A total of 103 independent effect sizes were estimated and each effect size was corrected for the impact of including clusters in evaluation designs. Included evaluations were conducted using both randomized ( n  = 45; i.e., randomized controlled trials/RCTs) and nonrandomized ( n  = 44; i.e., quasi‐experimental designs with before/after measures; BA/EC) methodologies. All of these studies included measures of bullying outcomes before and after implementation of an intervention. The remaining 14 effect sizes were estimated from evaluations that used age cohort designs. Two models of meta‐analysis are used to report results in our report. All mean effects computed are presented using both the multivariance adjustment model (MVA) and random effects model (RE). The MVA model assigns weights to primary studies in direct proportion to study level sampling error as with the fixed effects model but adjusts the meta‐analytic standard error and confidence intervals for study heterogeneity. The RE model incorporates between‐study heterogeneity into the formula for assigning weights to primary studies. The differences and strengths/limitations of both approaches are discussed in the context of the present data.

Our meta‐analysis identified that bullying programs significantly reduce bullying perpetration (RE: odds ratio [OR] = 1.309; 95% confidence interval [CI]: 1.24–1.38; z  = 9.88; p  < .001) and bullying victimization (RE: OR = 1.244; 95% CI: 1.19–1.31; z  = 8.92; p  < .001), under a random effects model of meta‐analysis. Mean effects were similar across both models of meta‐analysis for bullying perpetration (i.e., MVA: OR = 1,324; 95% CI: 1.27–1.38; z  = 13.4; p  < .001) and bullying victimization (i.e., MVA: OR = 1.248; 95% CI: 1.21–1.29; z  = 12.06; p  < .001). Under both computational models, primary studies were more effective in reducing bullying perpetration than victimization overall. Effect sizes varied across studies, with significant heterogeneity between studies for both bullying perpetration ( Q  = 323.392; df  = 85; p  < .001; I 2  = 73.716) and bullying victimization ( Q  = 387.255; df  = 87; p  < .001; I 2  = 77.534) outcomes. Analyses suggest that publication bias is unlikely. Between‐study heterogeneity was expected, given the large number of studies included, and thus, the number of different programs, methods, measures and samples used.

Authors' Conclusions

We conclude that overall, school‐based antibullying programs are effective in reducing bullying perpetration and bullying victimization, although effect sizes are modest. The impact of evaluation methodology on effect size appears to be weak and does not adequately explain the significant heterogeneity between primary studies. Moreover, the issue of the under‐/over‐estimation of the true treatment effect by different experimental designs and use of self‐reported measures is reviewed. The potential explanations for this are discussed, along with recommendations for future primary evaluations. Avenues for future research are discussed, including the need further explain differences across programs by correlating individual effect sizes with varying program components and varying methodological elements available across these 100 evaluations. Initial findings in the variability of effect sizes across different methodological moderators provide some understanding on the issue of heterogeneity, but future analyses based on further moderator variables are needed.

1. PLAIN LANGUAGE SUMMARY

1.1. interventions to reduce school bullying perpetration and victimization are effective.

Bullying is a ubiquitous form of aggression in schools worldwide. Intervention and prevention programs targeting school bullying perpetration and victimization are effective, yet more research is needed to understand variability in effectiveness.

The main findings of our review are that bullying programs were effective in reducing bullying perpetration outcomes by roughly 18–19% and bullying victimization by roughly 15–16%. There are substantial variations in effects, and the reasons for these variations require further research.

1.2. What is this review about?

Bullying is defined as aggressive behaviors that occur repeatedly over time between two or more individuals. Typically, there is a clear power imbalance between victims and bullies, either socially or physically. Furthermore, bullying behaviors are those that are committed intentionally to harm the victim.

What is the aim of this review?

The aim of this review is to summarise findings from studies of the effectiveness of school‐based antibullying programs in reducing both bullying perpetration and victimization will be reported. The review summarizes 100 studies, with the largest number being from the United States.

1.3. What studies are included?

To be included in this review, primary studies must have evaluated a specific intervention program that targeted bullying perpetration and/or victimization outcomes in school‐aged children, that is, typically between four and 18 years old. Studies must have used two experimental groups of children, one that received the intervention, and one that did not, and applied quantitative measures of bullying behavior (perpetration and/or victimization) that coincided with our operational definition of bullying.

Our final meta‐analytic review includes 100 studies of the effectiveness of antibullying programs. The largest number of studies came from the United States, with most other studies from Canada and Europe.

1.4. What are the findings of this review?

Antibullying programs are effective in reducing bullying perpetration outcomes by roughly 18–19% and bullying victimization by roughly 15–16%.

Variability in the effectiveness of antibullying programs was associated with differences in methodological designs, types of programs and geographical regions. Interventions evaluated using age cohort designs collectively gave the largest overall effect for both bullying perpetration and bullying victimization.

Limitations of the results are similar to those of previous reviews; for example, the reliance of self‐reported measurements of bullying may suggest the change is in reports of bullying perpetration/victimization and not behavioral change.

1.5. What do the findings of this review mean?

The findings indicate that school‐based bullying intervention and prevention programs can be effective in reducing both bullying perpetration and victimization, although the effect is, overall, modest.

The effectiveness of antibullying programs is an important finding with implications for public health and educational policy. However, our review did identify that there are variations in the effectiveness of intervention programs. Future research is needed to explore the reasons for these variations.

1.6. How up‐to‐date is this review?

This report forms an update of an earlier review (Farrington & Ttofi,  2009 ). The review authors searched for studies published up to December 2016.

2. BACKGROUND

Bullying first emerged as an important topic of research in the 1980s, following the tragic suicides of young boys in Norway, the reason for which was attributed to bullying victimization (Olweus,  1993 ). Today, this form of aggressive behavior remains a prevalent problem among young people globally. For example, a recent meta‐analysis of 80 international studies discovered prevalence levels of 34.5% and 36% for bullying perpetration and bullying victimization respectively (Modecki et al.,  2014 ).

Notably, bullying is a matter of public health, impacting the life outcomes of both bullies and victims, in varying ways (Arseneault et al.,  2010 ; Masiello & Schroeder,  2014 ; Ttofi et al.,  2012 ). Given its long‐term effects, it is imperative that effective intervention efforts are put in place in order to alleviate this troubling school phenomenon (Ttofi,  2015 ).

2.1. Defining school bullying

In order to adequately determine which interventions will effectively reduce bullying behaviors, it is important that researchers and educators start by accurately assessing the prevalence of involvement in school bullying (Swearer et al.,  2010 ). There remains some degree of disagreement in relation to definitive cut‐off points for involvement in bullying (Solberg & Olweus,  2003 ; Swearer et al.,  2010 ) and methods utilized for the assessment of bullying (Smith et al.,  2002 ; Swearer et al.,  2010 ). However, there is better agreement in regard to the defining criteria for school bullying.

Prominent researchers in the field have defined bullying as any aggressive behavior that incorporates three core elements, namely: (1) an intention to harm, (2) repetitive in nature, and (3) a clear power imbalance between perpetration and victim (Centers for Disease Control and Prevention,  2014 ; Farrington,  1993 ; Olweus,  1993 ). In other words, bullies are individuals who intend to cause harm to their victims through their actions, over a long period of time. Furthermore, victims of bullying are typically less powerful than bullies, or groups of bullies, and feel that they cannot easily defend themselves. This may be due to a physical or social power imbalance.

There are many forms of bullying, for example, school‐bullying, workplace bullying, sibling bullying and, most recently, cyberbullying. The present review is concerned only with face‐to‐face school‐bullying, namely, bullying that occurs in schools between individuals, usually aged between 4 and 18 years old. In the school context, bullying is a complex social phenomenon, that often does not happen between the bully and victim in isolation (Salmivalli,  2010 ). For example, individuals can be involved in bullying, not only as bullies, victims, or bully‐victims, but also as bystanders, defenders, or reinforcers (Zych et al.,  2017 ).

Cyberbullying is another form of aggressive behaviors that may occur within a school community, and previous research has found a significant overlap between offline (i.e., school‐bullying or face‐to‐face bullying) and online bullying (Baldry et al.,  2017 ). There is currently very little information about the effectiveness of intervention programs designed to reduce cyberbullying or whether school‐based programs that also target face‐to‐face bullying can impact online bullying concurrently.

2.2. The importance of addressing school bullying

School‐bullying is a strong risk marker for several negative behavioral, health, social, and/or emotional problems. A recent comprehensive review of systematic reviews highlighted that the impact of school‐bullying can occur concurrently with perpetration and/or victimization, but also later in life (Zych et al.,  2015 ). Previous studies have found that bullying victimization is often followed by negative mental health outcomes such as: increased suicidal ideation (e.g., Holt et al.,  2015 ); generalized or social anxiety, low self‐esteem and loneliness (e.g., Hawker & Boulton,  2000 ); psychotic symptoms (e.g., van Dam et al.,  2012 ); depression (e.g., Ttofi et al.,  2011a ,  2011b ); sleeping problems (Geel et al.,  2016 ); and other psychosomatic symptoms (Gini & Pozzoli,  2013 ).

Bullying perpetration, on the other hand, has been linked to several negative outcomes such as: suicidal ideation and suicidal attempts (Holt et al.,  2015 ); weapon carrying (Valdebenito et al.,  2018 ); drug use (Ttofi et al.,  2016 ); and violence and offending in later life (Ttofi et al.,  2011b ,  2012 ). Although involvement in school bullying is not necessarily a causal factor for undesirable life outcomes, research has found that there is an apparent association. It may be the case that the experience of school bullying functions as a stepping stone toward undesirable life outcomes (Arseneault et al.,  2010 ).

Moreover, involvement in school bullying, as either a bully or a victim, has been found to correlate with factors such as low academic achievement (Strøm et al.,  2013 ), truancy from school (Gastic,  2008 ), and drug use (Valdebenito et al.,  2015 ). Such factors are common risk factors for youth offending and delinquency (Farrington & Welsh,  2008 ). Therefore, a bullying prevention program could serve as a crime prevention program, as well as a form of promoting public health.

3. OBJECTIVES

It is clear that school bullying is an important target for effective intervention and prevention. Bullying is an ethical problem as well as a developmental one: targeting school bullying facilitates the process of optimal psychological development but it also addresses the question of human rights, especially the rights of the child (Sercombe & Donnelly,  2013 ). The aim of this paper is to provide an up‐to‐date systematic and meta‐analytical exploration of the effectiveness of school‐based antibullying programs. As such, the present report updates an earlier systematic and meta‐analytic review (Farrington & Ttofi, 2009 ; Ttofi & Farrington,  2011 ), by including evidence from an earlier report, and all available evaluations of antibullying programs since 2009.

It is hoped that this new evidence base will assist policy‐makers and practitioners working in the field of bullying prevention. Farrington and Ttofi's ( 2009 ) review concluded that school‐based antibullying programs are effective in reducing both bullying perpetration (OR = 1.36; 95% CI: 1.26–1.47; z  = 7.86; p  < .0001) and bullying victimization (OR = 1.29; 95% CI: 1.18–1.42; z  = 5.61; p  < .0001). Their review had a major impact on the field of bullying intervention and prevention, and in the 9 years that have passed since its publication there has been a wealth of new research.

Therefore, the aim of the present report is to conduct systematic searches for new evaluations of antibullying programs, and also update earlier analysis by including their 53 evaluations.

The initial stage of any meta‐analysis involves conducting a thorough and systematic search of all the existing and relevant literature (Lipsey & Wilson,  2001 ; Littell et al.,  2008 ). Using predetermined keywords and strict inclusion/exclusion criteria, a systematic review aims to identify, screen, appraise, and synthesize all relevant empirical studies (Zych et al.,  2017 ). In this way, systematic bias is avoided.

4.1. Inclusion and exclusion criteria

To be included in the present systematic review, a set of strict inclusion and exclusion criteria were employed to guide searches. These criteria were identical to those used in the previous meta‐analysis (Farrington & Ttofi,  2009 ). Specifically, to be included, primary studies must:

• (1) Describe an evaluation of a school‐based antibullying program implemented with school‐age participants (depending on the site of evaluation, ages may vary between 4 and 18 years of age);
• (2) Utilize an operational definition of school‐bullying that coincides with existing definitions (e.g., CDC,  2014 ; Farrington,  1993 ; Olweus,  1993 );
• (3) Measure school‐bullying perpetration and/or victimization using quantitative measures, such as, self‐, peer‐, or teacher‐report questionnaires; and
• (4) Use an experimental or quasi‐experimental design, with one group receiving the intervention and another (control group) not receiving the intervention. Nonrandomized studies had to measure outcomes before and after the intervention.

As a result, the present systematic review excludes studies that evaluate the effectiveness of intervention programs targeting alternative forms of bullying, such as cyber‐bullying (e.g., Del Rey et al.,  2015 ), general aggression (e.g., Leff et al.,  2010 ), and school violence (e.g., Giesbrecht et al.,  2011 ). Other studies were excluded because they measured bullying‐related nonbehavioral outcomes, for example, “attitudes towards bullying” (e.g., Earhart,  2011 ), or coping strategies for dealing with victimization (e.g., Watson et al.,  2010 ).

In addition, studies conducted with special needs, delinquent, or psychiatric populations were excluded (e.g., Espelage et al.,  2015 ), so that results could be generalizable to the wider mainstream school population. Studies using qualitative measures of effectiveness, such as participant perceptions of the effectiveness of the program (e.g., Fletcher et al.,  2015 ), were also excluded.

4.2. Searches 1

In order to identify potentially includable studies, Boolean searches were conducted using multiple combinations of the following keywords: bully*; victim*; bully‐victim; school; intervention; prevention; program*; evaluation; effect*; and anti‐bullying . A full description of the syntax used is provided in Appendix A.

Searches were conducted on several online databases, including, but not limited to: Web of Science, 2 PsychINFO, EMBASE, DARE, ERIC, and Scopus. Google scholar ( www.scholar.google.co.uk ) was also searched. A full list of databases searched is provided in Table  1 . EBSCOhost was used as a platform to search multiple databases concurrently and such databases are indicated in Table  1 .

Online platforms and databases manually searched

Note: EBSCOhost was used as a platform to search multiple databases concurrently. Such databases are marked with an *.

Databases of unpublished reports (e.g., ProQuest Dissertations and Theses Solutions) were also searched to include gray literature in our review. This should help to minimize potential publication bias linked to larger or significant effect sizes (Easterbrook et al.,  1991 ; McAuley et al.,  2000 ). In addition, evaluation studies included by previous systematic reviews were scanned, based on the name of each program, for additional‐updated evaluation results (i.e., Cantone et al.,  2015 ; Chalamandaris & Piette,  2015 ; Evans et al.,  2014 ; Jiménez‐Barbero et al.,  2012 ,  2016 ).

Studies included in the previous review (Farrington & Ttofi,  2009 ; Ttofi & Farrington,  2011 ), were also included in the present systematic review. Searches for the present review were conducted up to the end of December 2016, 3 for empirical studies published during and since 2009.

4.3. Screening

Our searches of the literature produced approximately 19,877 reports that were screened for eligibility. Based on the title and abstract, a total of 474 primary studies were identified as relevant, were obtained and subjected to further screening. Studies were allocated to six categories based on their relevance to the current meta‐analysis. A description of each category is provided in Table  2 . Screening was undertaken by the first author (H. G.), under the supervision of the second author (M. T.), in a collaborative format. H. G. reviewed eligible studies, and any queries were settled in discussion with M. T.

Relevance scale categories used in screening

The initial wave of screening excluded 258 of these primary studies. At this stage, studies were excluded because they: (1) did not evaluate a specific antibullying program (Category 1; n  = 107); (2) reviewed several different antibullying programs (Category 2; n  = 108); or (3) did not report empirical quantitative data from an evaluation of a specific antibullying program (Category 3; n  = 43).

A second wave of screening excluded a further 133 studies (Category 4; see Table  3 ). Primary studies were excluded at this stage because they: (1) reported irrelevant outcomes; (2) did not have an adequate control group; or (3) did not meet specified methodological criteria. The screening process is described in detail in Figure  1 . In total, 83 studies published since 2009 were included in our updated systematic review (Category 5).

Descriptions of category four studies

Screening of studies

In addition, five studies were identified during searches conducted for a meta‐analytical review of cyberbullying prevention programs (Gaffney et al.,  2018 ). These studies were missed during systematic searches for the current review (i.e., Kaljee et al.,  2017 ; Ortega‐Ruiz et al.,  2012 ; Ostrov et al.,  2015 ; Silva et al.,  2016 ; Solomontos‐Kountouri et al.,  2016 ). One of these studies (i.e., Kaljee et al.,  2017 ) has a publication date outside of the range of our searches. However, it was included because it was available online in 2016.

To provide the most up‐to‐date analysis of school‐based bullying prevention and intervention programs, therefore, a total of 88 newly identified studies are included in the present systematic review.

5. DATA EXTRACTION

After identifying studies eligible for inclusion in the present systematic and meta‐analytical review detailed information about the antibullying programs, sample involved, and evaluation design were extracted from primary studies. The following chapter outlines the coding framework applied in greater detail.

Table  4 also outlines each piece of information extracted. Information was extracted from primary studies under four main headings: (1) Descriptives, (2) Design, (3) Program, and (4) Outcomes. Additionally, the following section outlines information extracted from primary studies in order to create a risk of bias index. Table  5 outlines the items utilized to assess risk of bias for each of the methodological designs included in the present report. Details of the risk of bias results for each study is provided in Appendix B.

Coding framework

Abbreviations: BA/EC, quasi‐experiments with before and after measures of bullying (nonrandomized); exp, experimental group; OBPP, Olweus Bullying Prevention Program; RCT, randomized controlled trial.

Risk of bias tool

Abbreviations: AC, age cohort design; BA/EC, quasi‐experimental design with before and after measures of bullying; RCT, randomized controlled trial.

This procedure was carried out by the first author in consultation with the second and third authors. 4 There were a number of studies from the previous Campbell Collaboration report (i.e., Farrington & Ttofi,  2009 ) for which full texts were unavailable and thus, were excluded from several of the moderator analyses.

5.1. Descriptive

Various pieces of descriptive information were extracted from each of the 100 evaluations included in the present report. Information specific to the evaluation, such as the location or the start/end date, were recorded along with detailed information concerning the sample.

The total sample size and also the n of the relevant experimental and control groups were recorded. Age was extracted in two ways. First, where studies reported the mean age, or the age range (i.e., 8–10 years old) of participants this was recorded. Second, some studies did not report the age in years of participants, but we were able to record the school grade of included samples (i.e., Grades 4–6). Where reported, the % of females and males included in the sample was extracted.

We also coded descriptive information about the publication of the evaluation. Specifically, the type of publication and the publication year was recorded. The former represents a categorical moderator reflected whether or not the evaluation was published via the following channels, in order of hypothesized negative correlation with bias: (1) peer‐reviewed journal article; (2) chapter in an edited book/book; (3) governmental report or similar; (4) correspondence; and (5) unpublished masters or doctoral theses.

Correspondence was included to reflect data obtained from multiple evaluations of the Olweus Bullying Prevention Program (OBPP) sent to the second (M. M. T.) and third (D. P. F.) authors in preparation of their earlier Campbell review. Where evaluation data had been published in multiple formats, we favored the category associated with the least potential bias. For example, Domino ( 2011 ) reported the results of an evaluation of Take the LEAD program in a doctoral dissertation, but later published these results in a peer‐reviewed journal (i.e., Domino,  2013 ). In this scenario, the included study was coded as “article.”

5.2. Design

Included studies were further categorized according to several aspects of the research design used. We coded information regarding both the measures (i.e., instruments to measure bullying behaviors) and research design.

In relation to measurements of bullying, we recorded the timeframe (i.e., past 3 months or “ever”) in which participants were asked to report on experiences of bullying, the type of report used (i.e., self‐, peer‐, or teacher‐report), and data collection points (i.e., baseline, postintervention, 3‐month follow‐up, etc.). We also noted whether the measure was a continuous scale or a global item and whether bullying perpetration, victimization, or both, outcomes were measured.

As for the research design, we recorded information regarding the unit of allocation (or unit of randomization for RCTs; see below), the number of “clusters” included, whether groups were matched at baseline, and the number of experimental or control groups. For example, Elledge et al. ( 2010 ) included multiple control groups: matched controls and nonmatched controls.

Information about the evaluation methodology was also extracted from primary reports. The types of evaluation methodologies included in the present report are now described in further detail.

5.2.1. Evaluation methodology

In order to optimize the comparability of effect sizes, primary studies included in a meta‐analysis should use the same, or at least conceptually similar, research designs (Wilson,  2010 ). Following Farrington and Ttofi's ( 2009 ) criteria, we searched for evaluations using any of the following four research designs:

• (1) Randomized controlled trials (RCTs);
• (2) Before‐after/quasi‐experimental‐control designs (BA/EC);
• (3) Other quasi‐experimental designs; and
• (4) Age cohort designs.

Each of these methodologies varied on four key elements: as randomization of participants (or clusters of participants); use of experimental and control groups; and administration of quantitative bullying measures before and after intervention.

For example, all studies coded as RCT had to include random assignment to experimental conditions (i.e., intervention and control groups) but did not have to use before and after measures of bullying outcomes. RCTs are considered to be the “gold standard” of experimental evaluations (Weisburd et al.,  2001 ). Random assignment of a large number of units is used as a way in which evaluators can also randomize possible confounding variables between groups. As a result, we can infer that any observed differences result from the experimental manipulation (Farrington,  1983 ). The assumption is that randomization ensures that both observed and unobserved variables that may impact the results of an evaluation are also randomly distributed between groups. However, problems may arise if the unit‐of‐allocation, the unit‐of‐randomization, and the unit‐of‐analysis do not align.

Before‐after/quasi‐experimental‐control (BA/EC) designs, are conceptually similar to RCTs, but they do not involve random assignment to experimental conditions. Instead, participants or clusters of participants may be assigned to the intervention or control group on a self‐selected basis (e.g., Menesini et al.,  2012 ), for convenience (e.g., Sapouna et al.,  2010 ), or based on a greater need for intervention (e.g., Losey,  2009 ). Thus, BA/EC designs may be subject to selection biases (Farrington & Petrosino,  2001 ) that may reduce the validity of the results. These can be controlled if outcomes are measured before and after the intervention. Studies coded as BA/EC in the present report all used experimental and control groups but did not randomly assign participants to conditions. They also had to measure bullying outcomes before and after implementation of the intervention.

In contrast, studies categorized in the current review as using “other quasi‐experimental” designs utilized experimental and control conditions, without random assignment, but did not measure bullying behaviors before the intervention. Bullying outcomes were only measured after the implementation of an intervention in these studies. Therefore, selection bias is may be a threat to the internal validity of the results in such designs, which could have possibly attributed to pre‐existing differences between the groups (Farrington, 2003 ). For this reason, a decision was made to omit these designs from this updated meta‐analysis. Thus, relevant evaluations identified in the earlier Campbell Review and any new evaluations (since 2009) using this methodological design were excluded from the new meta‐analyses (see later).

In an age cohort design, students of a particular age X are initially assessed in the 1st year and serve as the control group for the evaluation of an intervention. Then, all students receive the intervention, and different students of the same age X (in the same school, in the 2nd year) serve as the experimental group (see Kärnä et al.,  2013 ). This design, which is largely used in evaluations of the OBPP, deals with some selection effects, since it ensures that experimental and control children are matched on age and school, and it deals with some threats to internal validity (e.g., ageing and maturation). However, this design may be influenced by period and testing effects, and the experimental and control groups may differ on other uncontrolled variables.

Studies employing RCTs, BA/EC, and age cohort designs were included in the present systematic and meta‐analytic review. Because of the potential threat to internal validity, we excluded studies ( n  = 9) in the other quasi‐experimental design category because they are poorly controlled and vulnerable to selection effects. Additionally, the four studies included in the earlier review that used an “other quasi‐experimental” design were excluded from the present systematic review.

5.3. Program

Using a socio‐ecological systems theory framework (Bronfenbrenner, 1979 ) and the previous meta‐analysis (i.e., Farrington & Ttofi,  2009 ) as guidelines, information about the specific intervention program was recorded. General details about the intervention, such as the name of the program (where relevant) and the aim of the intervention (e.g., Silva et al.,  2016 ) were noted along with more detailed information about the antibullying programs.

Intervention components at multiple levels of the socio‐ecological model (i.e., individual, peer, parent, and teacher, etc.) were recorded, such as work with peers, parental involvement, teacher training and whole‐school‐approach. Therefore, a brief description of each antibullying program based on this information is provided in Table  6 .

Systematic review results

In addition to specific program elements included in interventions, we also coded for possible sources of bias in evaluations and intervention development. Conflict of interest (COI) has previously been reported to impact evaluation results of many interventions and is a growing area of interest (COI; Eisner & Humphreys,  2012 ) with studies identified as having higher COI associated with larger overall effect sizes. Eisner and Humphreys outline many other possible sources of COI, such as financial gain to the evaluator, but this information was difficult to obtain for antibullying programs. Thus, a simple indication of potential COI was utilized.

We primarily focused on the overlap between individuals included as author/coauthor on the evaluation study, is also included on previous evaluations of the same program (e.g., NoTrap!; Menesini et al.,  2012 ; Palladino et al.,  2012 ,  2016 ), or is in fact referenced as the developer of that particular program (e.g., Tsiantis et al.,  2013 ). If no reference to a publication relating to the specific program was included, we concluded that the author had developed the program, and thus, the evaluation was deemed high risk.

Program specificity refers to whether the intervention program was specifically targeting bullying outcomes, or if many other outcomes were also included. Targeted programs are suggested to be more effective than generalized programs that aim to reduce many different behaviors in one intervention. Highly specific programs (i.e., those that only included bullying outcomes and very few others) were coded as “high.” Thus, programs that were less specific and included many other outcomes in addition to bullying measures were considered “low.” A third category was created (i.e., “medium”) to include studies that did multiple other outcomes in addition to bullying outcomes, but these additional variables were bullying‐related.

5.4. Outcomes

We also extracted several pieces of statistical information from primary studies that was required for the estimation of effect sizes. Statistics for bullying behaviors, for example, means and standard deviations or sample sizes and percentage of bullies and/or victims, were extracted for experimental and control groups at baseline and immediately postintervention timepoints.

We also coded bullying data for additional follow‐up timepoints where this information was reported by primary studies. Data was extracted and recorded separately for independent samples (i.e., female and male, Palladino et al.,  2016 ; older and younger, Baldry & Farrington, 2001) and different measures. For example, data for both self‐ and peer‐report measures were extracted from Beery and Hunt (2009) and for different forms of bullying (e.g., Frey et al.,  2005 ).

5.5. Risk of bias

As per the Campbell Collaboration reporting guidelines, a risk of bias index was created for the purpose of the present report. The EPOC tool was utilized to assess the risk category of each study on several items relating to the methodological quality of evaluations. Following earlier Campbell review (e.g., Valdebenito et al.,  2018 ) this tool was also used for nonrandomized studies as other risk of bias measurement instruments were considered inappropriate for nonscientific or medical trials.

The following section describes the procedure for addressing risk of bias in the present meta‐analysis. Each primary evaluation was measured on the following items: (1) allocation sequence (AS); (2) Allocation concealment (AC); (3) Baseline equivalence on outcomes (BE); (4) Baseline equivalence on participant characteristics (BC); (5) Incomplete outcome data (ID); (6) Contamination protection (CP); and (7) Selective outcome reporting (SOR). The applicability of these categories for each of the methodological designs included in the present report is outlined in Table  5 . Each study was categorized as being high, low, or unclear (if insufficient information was available) risk on each of these EPOC items.

6. INCLUDED INTERVENTIONS

In total, 67 different school‐based antibullying programs were evaluated by primary studies included in our updated meta‐analysis. Descriptions of each of these interventions is provided in the following section of this report. These narrative reviews of included antibullying programs are based on the best available information provided by the primary studies. Twenty‐one of the evaluated antibullying programs were included (only) in the previous meta‐analysis (Farrington & Ttofi,  2009 ). A number of popular school‐based antibullying programs (n = 7; i.e., Bully Proofing Your School [BPYS], Friendly Schools, KiVa, OBPP, Steps to Respect, ViSC, and Youth Matters) had been re‐evaluated or additional publications since 2009. Hence, the majority of programs evaluated in our updated meta‐analysis ( n  = 40) are new bullying prevention and intervention programs.

The following sections provides detailed summaries of each antibullying program included in our systematic review. Descriptions marked with an * were taken from the previous review (Farrington & Ttofi,  2009 ). To provide the reader with a detailed overview of existing antibullying programs studies subsequently excluded from the meta‐analysis are also included here.

6.1. *Antibullying intervention in Australian secondary schools

This antibullying intervention consisted of several activities that aimed to increase awareness and identification of bullying, to promote empathy for targets of bullying and to provide students with strategies to cope with bullying (Hunt,  2007 , p. 22). The intervention was based on an educational antibullying program, which was delivered by teachers. There was no specific training for teachers. Information about bullying was provided at parent and teacher meetings. Teacher meetings were held in conjunction with regular staff meetings while parent meetings were held after hours. A summary of the information covered at parent meetings was also published in the school newsletter in an attempt to target the wider parent population. Finally, the program includes a 2‐h classroom‐based discussion of bullying (offered by teachers) using activities from an antibullying workbook written by Murphy and Lewers ( 2000 ).

6.2. Anti‐Bullying Pledge Scheme (ABPS)

The ABPS describes a number of local antibullying schemes implemented in UK schools as a result of government recommendations and guidance (Pryce & Frederickson,  2013 ). Schools adopted a declaration of commitment, and intervention components followed a theoretical framework guided by the Theory of Planned Behavior (Ajzen,  1991 ).

The ABPS is a universal prevention program, that aims to reduce the prevalence of bullying perpetration and victimization in schools and increase students' perceptions of safety and support within the school environment (Pryce & Frederickson,  2013 ). Participating schools were assigned a facilitator, referred to as a “pledge supporter,” and a detailed intervention manual. The manual outlined the stages involved in implementing the ABPS program. The stages are as follows:

• Initial meeting with school management and the pledge supporter
• Intervention planning meeting
• School representatives make a declaration of commitment to the intervention
• Staff, student, and parent surveys are circulated
• Results from the surveys were collated and used to tailor intervention components to the individual schools' needs
• Ongoing visits and support from the pledge supporter throughout implementation.

6.3. *Be‐prox program

The Be‐Prox program was specifically designed to tackle bullying and victimization among kindergarten students. According to Alsaker and Valkanover ( 2001 , pp. 177–178), the somewhat higher adult‐children ratio, the interest of preschool teachers in socialization, the greater flexibility as to scheduling and teaching, and the admiration of many preschoolers for their teachers are ideal conditions for the implementation of preventive programs against bully/victim problems. The basic principle of Be‐Prox was to enhance preschool teachers' capacity to handle bully/victim problems (Alsaker,  2004 , p. 291). The program engaged teachers in an intensive focused supervision for approximately 4 months. Central features of Be‐Prox were the emphasis on group discussions, mutual support and co‐operation between consultants and teachers and between teachers and parents (Alsaker,  2004 , pp. 292–293).

The teacher training was provided in six steps (Alsaker,  2004 ; fig. 15.1, p. 292). Initially, teachers were given information about victimization (step 1) and the implications of this information was discussed (step 2). During the third step, specific implementation tasks were introduced and the teachers worked in groups in preparation for the practical implementation (step 4). After this preparation, teachers implemented specific preventive elements in the classroom (step 5) for a specific period of time. After that, teachers met and discussed their experiences of the implementation of the preventive measures (step 6).

In eight meetings over a 4‐month period, issues related to the prevention of bullying were addressed. The main purpose of the first meeting was sensitization. Teachers were asked to describe any possible bully/victim problems in their schools and were then given information about bullying and other types of aggressive behavior. They were also presented with the main principles of the program. The importance of contact between kindergarten teachers and children's parents was also emphasized and teachers were advised to consider the possibility of organizing a meeting with parents. In the second meeting, the importance of setting limits and rules to preschool children was discussed. Teachers were invited to elaborate some behavior codes in their classroom in collaboration with the children and to be ready to present them during the third meeting. Also, as a second homework task, teachers were asked to organize a parent meeting.

During the third meeting, teachers discussed their experiences of implementing classroom rules against bullying. The main focus of this meeting was the need for consistent teacher behavior, the difference between positive and negative sanctioning and the use of basic learning principles in the classroom. The main focus of the fourth session was on the role and responsibility of children who were not involved in bullying and of bystanders in the prevention of victimization. Teachers were asked to draw some kind of personality profiles of passive and aggressive victims and of bullies and to present them to the rest of the group. After this task, teachers were presented with research findings regarding the characteristics of children who were or were not involved in bullying. As a homework task for the next meeting, teachers were asked to systematically observe noninvolved children and to develop some means of involving them in the prevention of victimization.

During the fifth meeting, research‐based information about motor development and body awareness among preschool children was presented to teachers. A discussion between teachers and program researchers of children's self‐perceptions of strength, of peers' perceptions of strengths of victims of bullies, and other motor characteristics of children, aimed to yield important insights. The overall discussion and exchange of information among teachers aimed to promote teachers' understanding about how to change these perceptions within the classroom setting. Specific goals to be achieved within the classroom were clearly set, such as training in empathy and body awareness among children, participation and involvement of noninvolved children and talks with all the children about the situation in their kindergarten. During the sixth meeting, time was given to reflect on the goals formulated at the beginning of the prevention program. Teachers were also given time to discuss their experiences with implementing the goals of the fifth meeting within the classroom settings. The last two meetings followed a similar format, with time given for reflection on goals achieved, problems dealt with, and an overall evaluation of the program.

6.4. *Befriending intervention

Befriending intervention was an antibullying program that relied mainly on a peer support model. The overall aims of the program were: (a) to reduce bullying episodes through developing in bullies an awareness of their own and others' behavior; (b) to enhance children's capacity to offer support to the victims of bullying; (c) to enhance responsibility and involvement on the part of bystanders; and (d) to improve the quality of interpersonal relationships in the class group (Menesini et al.,  2003 , p. 1).

The antibullying intervention was offered in five steps (Menesini et al.,  2003 , p. 5). During the first phase, which targeted the class level (class intervention), several activities were offered aiming to increase children's awareness of prosocial and helping behaviors and to promote positive attitudes toward others. Through work at the class level, the school authorities sensitized and prepared the whole school population for the new service that the school unit was about to implement. In this way, another goal was achieved, namely developing values and attitudes toward “peer support activities” in the whole school population.

During the second phase of the program, the “peer supporters” were selected. Approximately three to four supporters were allocated in each classroom and were selected based on a combination of techniques, such as self‐ and peer‐nominations. These children were then trained in special full‐day sessions or in regular meetings during school time (phase three) so that they knew how to deal with other children and how to facilitate interactions among other children. Teachers and other professionals (psychologists and social workers) took part in these sessions as well. The overall aim of this phase of the antibullying program was to help peer supporters to enhance their listening and communication skills since they would be the mediators in the interactions among children.

During the fourth phase of the program, peer supporters worked in their classes with the assistance and close monitoring of their teachers. The teachers in each class organized “circle meetings” during which the needs of specific children involved in bullying (target children) were identified. Target children were contacted and, after their consent and cooperation, were offered help by the peer supporters. Peer supporters were not only assigned to specific tasks involving the target children but were also supervised by the teachers so that they were given constant feedback on their on‐going work in the class.

During the final phase of the Befriending Intervention, the leading group of peer supporters were involved in training other children in the class, so that more children could be involved in the program (in the transmission of training and passing on the roles).

6.5. *Behavioral program for bullying boys

This program targeted male youth, from a low socio‐economic area, predominately inhabited by individuals of color, involved in bullying. The program was based on the findings of an in‐depth needs assessment within three schools and targeted a specific number of male students aged sixteen who (based on the results of the questionnaire that had been administered) were “considered to be a serious threat to the harmonious functioning of everyday school life” (Meyer & Lesch,  2000 , p. 59). The theoretical basis of the program could be found in the Social Interactional Model for the development of aggression (Meyer & Lesch,  2000 , p. 61) and involved a behavioral approach for tackling the problem of bullying. The program was implemented by psychology students for ten nonconsecutive weeks, with 20‐h‐long sessions held twice weekly at the school, during school hours.

The components of the 17‐session behavioral program included homework tasks, modeling, self‐observation, role‐plays, and a token economy system for reinforcing positive behaviors. According to the program designers “the chief contingency for behavioral change was the token economy system, using Wonderland Games tokens, chocolates and cinema tickets as reward for non‐bullying behavior” (Meyer & Lesch,  2000 , p. 62). Each participant was monitored by himself and by a “buddy” who was selected in each session prior to the monitoring. Each session included an opportunity for feedback on the students' progress in the week, a discussion of a relevant applied topic, role‐playing, games, and drawing. The program designers pointed out the limitations of the intervention strategy. As they indicate (Meyer & Lesch,  2000 , p. 67) “the program was too short and structured to address the issues that were disclosed in sessions, as the severity of the nature of the aggression in the schools and vast social problems was seriously underestimated.”

6.6. Beyond the Hurt

Sutherland ( 2010 ) implemented the Beyond the Hurt program, a peer‐led school‐based bullying intervention and prevention program, developed by the Red Cross. Beyond the Hurt is a high school program and emphasizes education, prevention and intervention to reduce prevalence of bullying perpetration and victimization. Sutherland ( 2010 , p. 84) describes the four key components of the intervention: (1) education and training of peer facilitators, (2) in‐class presentations given by peer facilitators, (3) teacher workshops, and (4) online training material for teachers and community members.

This peer‐led program trains and educates select peer facilitators, who become the implementers of the intervention program within participating schools. These students are guided by a teacher and Red Cross professional throughout training and implementation of class presentations highlighting several bullying‐related issues. The overarching aim of the Beyond the Hurt program is to create a positive school and class climate in which students are encouraged to develop and maintain healthy prosocial relationships, and bullying perpetration and victimization are not supported. The program aims to promote antibullying attitudes among participants and encourage empathy and prosocial support for victims of bullying.

6.7. *Bulli and Pupe

Bulli and Pupe was an intervention program concerned with bullying and family violence. The program, developed by Baldry (2001), was “directed towards the individual and peer group, and aimed to enhance awareness about violence and its negative effects” (Baldry & Farrington,  2004 , p. 3). The intervention package consisted of three videos and a booklet divided into three parts; each video was linked to one part of the booklet. Each part of the booklet was meant to take the form of an interactive lesson where professionals, experienced in school and juvenile processes, discussed three issues according to the structure of the manual.

The first part of the booklet, entitled “Bullying among peers,” emphasized teen violence among peers. The booklet presented vignettes and graphics that reported research findings on bullying in an attempt to raise students' awareness of this issue. The corresponding video showed teenagers talking about bullying based on their own experiences and judgments. The second part of the booklet, entitled “Children witnessing domestic violence,” analyzed the effects of domestic violence on children and the repercussions for school achievement and peer relations. In the accompanying video, children in a shelter for battered women were presented, talking about their personal experiences and emotions. Finally, the third part of the booklet, entitled “Cycle of violence,” dealt with the long‐term effects of violence on adults who were victims of violence in their childhood. The corresponding video consisted of an interview conducted with a 19‐year old boy who had a violent father.

The program was in the first place delivered in 3 days by experts who, together with teachers, discussed about bullying, read the booklet and analyzed its content. The program was taken over by teachers who once a week created a facilitation group and allowed children to discuss any problems they encountered with their peers. The program was more effective with secondary students because it required its participants to have good interpersonal and cognitive skills (Baldry & Farrington,  2004 , p. 4).

6.8. The Bully Prevention Challenge Course Curriculum (BPCCC)

Battey ( 2009 ) implemented the BPCCC (Haggas,  2006 ) to students over two 45 min classes, on 4 days of one school week. The program was implemented by trained facilitators, whom included the schools' physical education/health teacher. The program commenced by providing participants with name tags and organizing some warm‐up physical activities. Next, the physical education/health teacher provided participants with information about bullying, such as, identifying and addressing bullying, who to talk to and where to seek support. Subsequent group discussions focused on empathy and understanding each other's differences. Audience participation activities also required the students to engage to represent the number of students whom had been a victim or bully.

6.9. Bully Proofing Your School

“Bully‐Proofing Your School” was a comprehensive, school‐based intervention program for the prevention of bullying (Menard & Grotpeter,  2014 ; Menard et al.,  2008 ; Toner,  2010 ). The program involved three major components: (1) heightened awareness of the problem of bullying, involving a questionnaire to measure the extent of bullying and the creation of classroom rules related to zero tolerance for bullying; (2) teaching students protective skills for dealing with bullying, resistance to victimization and providing assistance to potential victims by teaching assertiveness skills; and (3) creation of a positive school climate where students were encouraged to work as positive and supportive bystanders (Menard et al.,  2008 , p. 7).

The primary targets of BPYS were elementary and middle school students. School staff were involved as both secondary targets of intervention (since changes in their behavior was a requirement for the construction of a positive antibullying school environment) and as agents delivering the intervention to students. Teachers were given information and strategies to help them recognize bullying incidents among their students and how to effectively deal with these behaviors (Menard & Grotpeter,  2014 ).

The intervention in the classes consisted of a classroom curriculum, which included seven sessions of approximately 30–40 min. Each session was delivered by a teacher or by mental health staff. After completion of the classroom curriculum materials, teachers were encouraged to hold weekly classroom meetings during which students could be helped to reflect on their behaviors. Parents were offered information through newsletters. Individual parents of students involved in bullying as either perpetrators or victims were given consultation (Menard & Grotpeter,  2014 ).

6.10. Chinese antibullying intervention

Ju et al. ( 2009 ) implemented an antibullying program in a Chinese primary school employing an action research framework. There were two main aims of this intervention program. First, the program aimed to reduce bullying perpetration and victimization both on students' way to, and from, school. Second, the study aimed to investigate practical intervention elements that could be applied nationwide to Chinese primary school children (Ju et al.,  2009 ).

The initial step in this intervention was the training of teachers on the fundamental principles of action research. This training program targeted the following components of educational research: (1) research methodology in education; (2) knowledge of school bullying; (3) components of action research; and (4) intervention skills, such as brainstorming and role‐playing. Second, a 5‐week intervention program was designed and implemented by teachers in classrooms. Components that targeted both victims and bullies specifically were also incorporated into the intervention.

6.11. The Confident Kids program

The Confident Kids program is an antibullying intervention designed for early adolescent males who were experiencing anxiety as a result of being bullied at school (Berry & Hunt,  2009 ). The foundations of the program lie in cognitive‐behavioral therapy, employing both anxiety management techniques and antibullying elements. Based on the “Cool Kids Program” (Lyneham et al., 2003), this intervention program aims to reduce bullying victimization by targeting factors that increase the likelihood of victimization. Therefore, this program focuses primarily on issues such as: self‐esteem, coping strategies; social skills; emotional regulation; and internalizing behaviors.

The program was implemented over a period of 8 weeks, and included student and parent involvement. Students participated in weekly group sessions led by a team of assistant and qualified clinical psychologists. These sessions incorporated a combination of tasks including: skill demonstration; role‐playing; and group discussion. Homework was allocated after each session and participants were encouraged to apply skills acquired in real‐life settings between each session.

Sessions covered a variety of issues, including both cognitive‐behavioral anxiety management techniques and antibullying information. Seven core sessions focused on the following topics: psycho‐education; cognitive restructuring (2 sessions); graded exposure; adaptive coping strategies; improving social skills; and self‐esteem. A final session targeted relapse prevention and provided a general overview of the skills learned throughout the program. Parents participated in sessions that ran parallel to the student program. Group discussions targeted the strategies being taught to student participants and also possible parent factors that could influence effectiveness of intervention for their children, for example, parental anxiety.

6.12. Cyberprogram 2.0

Cyberprogram 2.0 is a cyberbullying intervention program that also incorporates elements on school bullying (Garaigordobil & Martínez‐Valderrey,  2015 ). The intervention is delivered over 19 sessions, and outlines the following four main goals:

• To outline and conceptualize bullying and cyberbullying, including identifying the different roles involved (e.g., bullies, victims, and bystanders).
• To illustrate the consequences of bullying and cyberbullying for all those involved
• To develop coping strategies in order to reduce bullying and cyberbullying behaviors.
• Developing positive social and emotional skills, such as empathy, active listening, anger management, conflict resolution strategies, and diversity tolerance.

A wide range of activities and techniques are used, such as, role‐playing, brainstorming, case studies, and guided discussion. The Cyberprogram 2.0 intervention followed a specific methodological framework, employing four key components for implementation. They are as follows: (1) inter‐session constancy: intervention was delivered in weekly 1‐h sessions; (2) spatial‐temporal constancy: intervention was delivered in the same place and at the same time each week; (3) constancy of adult facilitator: intervention was implemented by the same adult, who same psycho‐pedagogical training, each week; and (4) constancy in the session structure: sessions being with group instruction and activities. There is then a following reflection phase that is led by the adult.

6.13. Daphne III

Daphne III was an international antibullying initiative implemented and developed in association with numerous organizations. In this study (Papacosta et al., 2014), school antibullying programs were coordinated in Cyprus by the Association for the Psychosocial Health of Children and Adolescents (APHCA). Other influential “partners” included the Cyprus Ministry of Health, mental health services, Department of Child and Adolescent Psychiatry, Ministry of Education and Culture, and Educational Psychology services. Organizations from other European countries included: Child Line [ Vsi Vaiku Linija ], in Lithuania, and Nicolaus Copericus University, in Poland, were also involved.

The overarching aim of this initiative was to educate 5th and 6th grade primary school students about bullying, and the many different forms it can take (Papacosta et al., 2014). Teachers implemented the program in their classrooms, and were trained by psychology and mental health professionals. There were eleven workshops involved in the program that followed a structured curriculum manual. This manual also provided schools with suggestions and recommendations on ways in with they could prevent, and intervene in, bullying situations.

6.14. *Dare to Care: BPYS program

“Dare to Care; Bully Proofing Your School” was a modification of the “Bully Proofing Your School” program (Beran et al.,  2004 , p. 103), which in turn was modeled on the Olweus Program. This antibullying program placed emphasis on clinical support to victims and perpetrators of bullying in the form of individual and group counseling. It also enabled collaboration with community services. The essence of the program was to encourage accountability for creating solutions among all parties involved in the education system (Beran et al.,  2004 , p. 104).

The program included several steps. Program facilitators provided to school personnel information and training on issues related to bullying in schools (in a full‐day professional development workshop). This workshop aimed to ensure that the program principles would be reflected in the overall curriculum and would be sustained over time. Information was also given to parents. Then, students, parents and school staff collaborated in the development of a school antibullying policy. This policy had the aim of identifying caring and aggressive behaviors and consequences of those behaviors, but with a focus on reparation rather than punishment. The antibullying policy was posted throughout the school. Finally, the program involved the implementation, on behalf of the teachers, of a classroom curriculum that educated children about the nature of bullying and strategies to avoid victimization. The curriculum included discussion, role‐plays, artwork, books, videos and skits presented to school staff, parents, and other children.

6.15. Defeat Bullying

The Defeat Bullying program is a curriculum‐based antibullying program that was published by the National Society for prevention of Cruelty to Children (NSPCC, UK) in 2007 (Herrick,  2012 ). The program materials were available to download online, as part of a nationwide campaign to reduce bullying perpetration and victimization in UK schools. The overarching aim of the Defect Bullying program is to raise awareness and improve attitudes toward bullying, educate about bullying‐related feelings and emotions, and to develop effective intervention and conflict resolution strategies (Herrick,  2012 , p. 85). Based on social identity theory (Tajfel & Turner,  1979 ), the program aims to establish an in‐class antibullying norm, so that students will be encouraged to adopt this norm, and thus, reduce levels of bullying perpetration and victimization.

There are five key lessons implemented throughout the program, and each incorporates a range of individual, class and group activities (Herrick,  2012 ). The lessons cover the following five themes: (1) understanding attitudes and values toward bullying; (2) educating about the feelings that occur as a result of bullying; (3) embracing diversity; (4) safety awareness; and (5) encouraging bystanders to get involved in antibullying strategies. The available intervention materials were also reviewed by groups of teachers, and any necessary amendments were incorporated. For example, Herrick ( 2012 ) describes that following teacher discussion groups, homework assignments relating to each lesson were developed and implemented. Parents of participating students were also invited to attend an antibullying workshop led by the researcher.

6.16. *Dutch antibullying program

The antibullying initiative in the Netherlands was inspired by the Olweus program (Fekkes et al.,  2006 , p. 639). The program was specifically designed to tackle bullying behavior by involving teachers, parents and students. It offered a 2‐day training session for teachers in order to inform them about bullying behavior and to instruct them about how to deal with bullying incidents in schools. During the intervention period, teachers had access to the training staff for additional advice. Intervention schools were supported by an external organization named KPC, which specialized in training school staff and in assisting schools in setting up new curricula and guidelines. The core intervention program included: (1) antibullying training for teachers, (2) a bullying survey, (3) antibullying rules and a written antibullying school policy, (4) increased intensity of surveillance, and (5) information meetings or parents.

During the intervention, there was careful dissemination of the antibullying program to intervention schools. Also, the researchers provided information about the number of intervention and control schools, which have used the above‐mentioned elements of intervention. Finally, intervention schools were supplied with the booklet “Bullying in schools: how to deal with it” and with a “Bullying Test,” a computerized questionnaire that children could complete anonymously in the classroom.

6.17. Dutch Skills for Life

The Skills for Life program is a Dutch universal school‐based behavioral and health prevention program for adolescents aged 13–16 years old (Diekstra,  1996 ; Gravesteijn & Diekstra,  2013 ). The program targets prosocial behavior, self‐awareness, social awareness, self‐control, interpersonal skills, and ethical decision making to reduce behavioral and health problems (Fekkes et al.,  2016 ). The program is based on social learning theory and Rational Emotive Behavioral Therapy. As a result, the program aims to reduce bullying by enabling students to learn from each other in a classroom setting through behavioral modeling.

The program is implemented by teachers, who attend two 3‐day training workshops prior to implementation and receive “booster” training sessions throughout the intervention (Fekkes et al.,  2016 ). The intervention is comprised of 25 lessons that are delivered over the course of two academic years. First, four lessons address awareness and handling of thoughts and feelings. Skills such as interpersonal problem solving, emotional regulation, and critical thinking are targeted. There are twelve additional lessons in the 1st year, and nine more lessons in the 2nd year of implementation. These generally focus on skills that are applicable to particular behavioral or health experiences. For example, lessons are aimed at: dealing with bullying; setting and respecting boundaries; substance use; norms and values; friendships; sexuality; suicidal ideation; and conflicts with peers and/or teachers. Various activities are utilized throughout the program, including, active enactment, DVDs, role play, discussion and feedback.

6.18. Dynamic Approach to School Improvement (DASI)

The DASI (Kyriakides, Creemers, Papastylianou, et al.,  2014 ; Kyriakides, Creemers, Muijs, et al.,  2014 ) was a whole‐school approach to bullying prevention implemented in several European countries, such as: Cyprus, Greece, UK, Belgium and the Netherlands. This approach draws factors from the educational effectiveness model (Creemers & Kyriakides,  2008 ,  2012 ). The intervention targets specific school factors, that is, (1) school teaching policy, (2) school learning environment, and (3) school evaluation. This framework was previously found to improve academic achievement (e.g., Kyriakides,  2008 ).

At the beginning of the intervention, the research team held training for participating school staff. The theoretical framework was introduced, and a detailed manual was provided. The aim of the handbook was to facilitate school stakeholders to develop strategies and action plans that were specific to the schools' needs (Kyriakides, Creemers, Papastylianou, et al.,  2014 ). Support was offered to each school by the research team throughout the process.

Teacher surveys were distributed prior to implementation in order to highlight specific areas that needed improvement. The next phase of the intervention involved school stakeholders coming together to form cooperative committees with representatives of parents, students, and teachers. These committees then collaborated to develop action plans and strategies to address specific problems in their schools. Committees formulated plans to implement particular intervention components that best suited their specific needs. Therefore, the schools participating did not necessarily implement the same intervention components or activities. Schools were required to retain log books of activities undertaken.

Kyriakides, Creemers, and Papastylianou, et al. ( 2014 ) provide an outline of the intervention components implemented in one experimental school involved in their trial. For example, the following are identified as essential elements implemented in order to reduce bullying:

• “Student behavior outside the classroom”—involves developing clear and efficient antibullying policy, increased teacher vigilance in bullying “hot spots” and effective supervision of students.
• Improved school learning environment
• “Rewarding good behavior”—enforcing a system that acts as a nonpunitive approach to antibullying, by motivating students to behave in a prosocial manner.
• “Collaboration and interaction between teachers”—encouraging teachers to work together and communicate effectively about bullying issues in their schools.
• Other intervention components, including, encouraging and supporting peer bystanders; identifying and support “at risk” and vulnerable students; and creating student‐made videos about bullying issues.

6.19. *Ecological antibullying program

The Ecological antibullying program examined peer group and school environment processes “utilizing a systemic interactional model with evaluations at each level of intervention” (Rahey & Craig,  2002 , p. 283). The overall aim of the program was the creation of a supportive and safe school environment in which firm limits against bullying were established. The specific goals of the program included raising awareness of the problem of bullying, increasing empathy, encouraging peers to speak against bullying and formulating clear rules against bullying.

The 12‐week program was based on the “Bully Proofing Your School” program which was designed to increase the understanding of bullying and decrease the incidence of bullying (Rahey & Craig,  2002 , p. 285). The program elements included a psycho‐educational component implemented within each classroom, a peer mediation component and specialized groups for children involved in bullying.

At the school‐wide level, the psycho‐educational program was implemented by psychology students who received training sessions and manuals prior to intervention. Prior to the program, at a school assembly the program was introduced to students. The assembly signaled the formal beginning of the intervention. The classroom programs involved interactive educational approaches such as role playing and puppet techniques. The topics addressed were bullying and victimization, conflict resolution, empathy, listening skills and individual differences (Rahey & Craig,  2002 , p. 286).

Individual programs for children involved in bullying were also part of the intervention. The relevant sessions consisted of social skills, listening, empathy training and supportive counseling. Each weekly session lasted 45 min. The program also included intervention at the teacher level. Teacher programs consisted of meetings with teachers to discuss bullying, intervention approaches, and student support for those directly involved in bullying. During the intervention, the program coordinators met with principals and teachers to offer support.

6.20. Emotional Literacy Intervention

Knowler and Frederickson ( 2013 ) evaluated the effectiveness of an emotional literacy intervention targeted on bullying behaviors to reduce bullying victimization in UK schools. Selected schools were previously implementing the Social and Emotional Aspects of Learning (SEAL; Department for Education and Skills,  2005 ) program. One of the themes included in the SEAL program is “Say no to bullying” (Knowler & Frederickson,  2013 ), however the overall program aims to improve students' social relationships, motivation, learning strategies, and holistic school improvement.

The specific emotional literacy intervention implemented and evaluated by Knowler and Frederickson ( 2013 ) involved teaching emotional literacy skills to small groups of students (Faupel,  2003 ). In the current evaluation, the intervention was delivered to groups of “low emotional literacy” and “high emotional literacy” groups distinguished by scores above, or below, median scores on the Emotional Literacy assessment‐pupil form (ELA‐PF; Faupel,  2003 ). The intervention program employed 12 weekly lessons and was implemented by trained teaching aids (Knowler & Frederickson,  2013 ). The program consisted of four main concepts: (1) self‐awareness, (2) self‐regulation, (3) empathy, and (4) social skills. Lessons employed a variation of behavioral and cognitive‐behavioral elements (Faupel,  2003 ).

6.21. Empathy training program

This intervention program was developed for children identified as bullies and aimed to increase their empathetic skills in order to reduce their bullying behaviors (Şahin,  2012 ). The empathy training program was implemented over eleven 75‐min sessions that were based on a curriculum lesson plan developed by the author. Several cognitive techniques were utilized throughout the program, such as: recognizing, evaluating and naming feelings; diadtic, experimental, modeling and role‐playing, in order to improve the students' cognitive abilities in relation to empathy. Each lesson required the students to work together to develop a slogan that emulated the content of the session. The following is an outline of the first 4 weekly lessons, and the associated slogan developed, (for a full outline see: Şahin,  2012 , p. 1327; Table  2 ).

Slogan: Be kind, loving and forgiving to each other to lead a happy life .

Slogan: Living without the awareness of feelings is like driving a car with its brakes on .

Slogan: One who claims to know everything about the universe but nothing about himself, actually knows nothing .

Slogan: We can look at the same thing but view it differently .

6.22. *Expect respect

Expect Respect was a school‐based program that aimed to promote awareness and effective responses to bullying and sexual harassment. The project was developed by Safe Place, the sole provider of comprehensive sexual and domestic violence prevention and intervention services in Austin, Texas (Rosenbluth et al.,  2004 , p. 211). The program targeted the involvement of all members of the school community in recognizing and responding to bullying and sexual harassment. The overall project design was inspired by the work of Olweus (Rosenbluth et al.,  2004 , p. 212). Expect Respect consisted of five core program components, namely a classroom curriculum, staff training, policy development, parent education and support services.

The classroom curriculum was based on 12 weekly sessions adapted from a specific manual called “Bullyproof: a teachers” guide on teasing and bullying for use with fourth and fifth grade students' (Whitaker et al.,  2004 , p. 330). The Bullyproof curriculum was designed to be taught in conjunction with literature typically read by fourth and fifth graders. Although the antibullying curriculum was designed to be implemented by teachers, within the framework of the Expect Respect program, it was jointly led by Safe Place Staff and teachers or school counselors (Whitaker et al.,  2004 , p. 331). The curriculum aimed to increase the ability and willingness of bystanders to intervene in bullying situations, thus reducing the social acceptability of bullying and sexual harassment. The Bullyproof lessons included writing assignments, role‐plays of how to intervene in bullying situations, class discussions and so on.

With regard to the staff training, a 6‐h training was provided to project staff, counselors, and fifth grade teachers. The training was given by the author of the specific manual and aimed to prepare school personnel to respond effectively to bullying incidents. In addition, 3‐h training sessions were provided once per semester for all personnel, including bus drivers, cafeteria workers, hall monitors and office staff. The training presentation included research on bullying and sexual harassment; strategies to enhance mutual respect among students; practice in using lessons from the curriculum; and methods for integrating the lessons into other subject areas including language arts and health.

School administrators were encouraged to develop an antibullying policy (policy development) in their school to ensure consistent responses by all staff members to incidents of bullying and sexual harassment. Principals were expected to present the policy to school staff, students and parents. In order to facilitate the overall procedure of policy development, Expect Respect staff provided an initial policy template to school administrators (Whitaker et al.,  2004 , p. 332) and each school was encouraged to expand this initial policy in accordance with the specific needs of their unit.

The Expect Respect program also included parent training. Educational presentations were offered to parents, twice a year, providing information about the project. The information given to parents through these meetings (as well as through parent newsletters sent home) was aimed at enhancing parents' strategies to help children involved in bullying as bullies, victims, bully‐victims, or bystanders.

Further support services were provided such as continuous assistance of school counselors by Safe Place staff. School counselors were given a specialized session on how to deal with students who were repeatedly involved in bullying as either perpetrators or victims. They were also provided with a comprehensive resource manual containing reading and resource materials on bullying, sexual harassment and domestic violence.

6.23. fairplayer.manual

The fairplayer.manual is a structured, curriculum‐based antibullying program for Grade 7–9 students (Bull et al.,  2009 ; Wölfer & Scheithauer,  2014 ). The overarching aim of the intervention is to reduce bullying and relational aggression by improving students' social and moral competencies. The program focuses on raising awareness, changing attitudes, and encouraging bystander intervention.

The program is implemented over 15‐weekly 90 min lessons, and can be delivered either by trained teachers (Bull et al.,  2009 ), or psychologists (Wölfer & Scheithauer,  2014 ). Intervention lessons employ cognitive‐behavioral techniques and target nine specific topics. The first introductory lesson introduces the program to students, and class antibullying rules are developed. Two following lessons are concerned with raising awareness about bullying‐related issues, such as, the various forms of bullying and the consequences associated with perpetration and victimization. One lesson subsequently focuses on improving students' understanding of their own and peers' feelings. A further two lessons highlight the numerous participant roles involved in bullying, for example, bullies, victims, outsiders (i.e., noninvolved), assistants, and re‐inforcers (Wölfer & Scheithauer,  2014 ). The latter roles describe different forms of bystanders, those who witness bullying and allow it to happen and those who reinforce bullying behaviors. Social dynamics in the classroom is also addressed in one intervention session. By addressing the different dynamics, networks and norms socially in the class, this lesson aims to improve the classroom climate and encourage co‐operation among students. Another intervention lesson models and promotes bystander intervention in order to encourage noninvolved children to become actively engaged with intervening in bullying situations that they may witness.

Following these core awareness‐raising and knowledge‐improving lessons, participating students undertake five social skill‐training session s. These lessons focus on developing social, emotional, and moral skills of participants, in order to combat bullying. Perspective taking, empathy, and moral dilemmas are just some of the issues that are included. Diversity is the topic addressed in one of the following lessons, where students learn to respect and appreciate diversity. Finally, a concluding lesson brings together all of the issues covered by the intervention and demonstrates ways in which participants can utilize skills and knowledge in their everyday lives.

6.24. FearNot!

The FearNot! (Fun with Empathetic Agents to achieve Novel Outcomes in Teaching; Sapouna et al.,  2010 ) was an immersive learning intervention that aimed to reduce bullying victimization. Students from British and German primary schools participated in the virtual learning program for weekly 30‐min sessions over the course of three consecutive weeks. Participating schools were required to have adequate computer facilities in order to be able to run the program.

During intervention sessions bullying scenarios were enacted by male and female 3D animated characters. The content of these scenarios reflected the characters' genders, for example, scenarios involving male characters included more incidents of physical bullying, whereas female characters demonstrated more relational bullying. Following each of the bullying episodes, participants were asked to interact and provide the animated victim of bullying with a suitable coping strategy to prevent future victimization. The program then enabled students to see the outcomes of their suggested strategy. In some circumstances, the animated victim of bullying responded that they did not feel emotionally adequate enough to carry out the suggested coping strategy (e.g., not strong enough to stand up to the bully).

Based on previous research (e.g., Kochenderfer & Ladd,  2000 ), students were then provided with an indication of how successful their proposed coping mechanism would be in real‐world bullying scenarios. For example, students were provided with a score on a scale of zero (never successful) to ten (always successful; Sapouna et al.,  2010 ). In addition to the computerized program, teachers in intervention schools were provided with a detailed intervention manual. However, during the FearNot! program, teachers were instructed only to assist students with issues of comprehension, and not to guide them on suitable responses to the bullying scenarios.

6.25. Fourth R

The Fourth R: Strategies for Healthy Youth Relationships is a dating violence prevention program that targeted bullying perpetration and victimization as secondary outcomes (Cissner & Ayoub,  2014 ). This curriculum‐based intervention program was based on social learning theory (Bandura,  1978 ), and was implemented in classrooms by trained teachers during health and physical education classes. Participating teachers completed an intensive 1‐day training session that provided them with the skills to implement the program effectively. Detailed manuals and lesson outlines/materials were provided, and the Fourth R curriculum was integrated into existing health and physical education curricula.

The Fourth R was designed as a 21‐lesson curriculum that incorporates a variety of activities and lessons. Role‐playing, individual, pair and group work, and detailed examples/scenarios of conflict are examples of Fourth R‐style tasks. Program lessons were categorized into the following 3 units: (1) Personal Safety and Injury Prevention; (2) Healthy Growth and Sexuality; and (3) Substance Use and Abuse. Each unit consisted of seven 45‐min lessons. The Fourth R was also designed to be implemented in either gender‐segregated or co‐ed classrooms.

6.26. *Friendly Schools Project

“Friendly Schools” was a theoretically grounded program. Its educational techniques (e.g., role modeling, drama activities, skills training, etc.) were based on notions derived from Social Cognitive theory, the Health Belief Model and Problem Behavior theory (Cross et al.,  2004 ,  2011 ). An interesting aspect of this program is that it was based on the results of a systematic review (Cross et al.,  2004 , p. 187), which provided a set of key elements to be included in the final intervention strategy. The program targeted bullying at three levels: (a) the whole‐school community, (b) the students' families, and (c) the fourth and fifth grade students and their teachers.

With regard to the whole‐school intervention component, in each school, a Friendly Schools Committee was organized with key individuals (e.g., a parent representative, a school psychologist, a school nurse, teaching staff) who could co‐ordinate and successfully sustain the antibullying initiative. Each committee was provided with a 4‐h training, designed to build members' capacity to address bullying. Each member was provided with a specific strategy manual. The manual was a step‐by‐step guide on how to implement the antibullying initiative. It included among others the Pikas “Method of Shared Concern” and the “No Blame” approach (Cross et al.,  2011 ; Pikas,  2002 ).

With regard to the family intervention component, this included home activities linked to each classroom‐learning activity. Parents were also provided with 16 skills‐based newsletter items (eight for each year of the intervention) that aimed to provide research information on bullying as well as advice to parents on what to do if their child was a perpetrator or a victim of bullying behavior.

Moving on to the Grade 4 and 5 classroom curricula, the Friendly Schools curriculum consisted of nine learning activities per year. The curriculum was offered by trained teachers in three blocks of three 60‐min lessons, over a three‐school‐term period. The learning activities aimed to promote awareness of what was bullying behavior; to help students to become assertive and talk about bullying with teachers and parents; and to promote peer and adult discouragement of bullying behavior.

Finally, the Friendly Schools program offered manuals to teachers. The teacher manuals were designed to be entirely self‐contained so as to maximize the likelihood of teacher implementation. Friendly Schools project staff also provided teacher training (a 6‐h course) for all intervention teachers.

6.27. *Granada antibullying program

This program was a pilot antibullying program with the following aims: (a) to establish children's involvement in bullying within different participant roles/categories; (b) to reduce the number of students involved in the phenomenon as bullies, victims and bully‐victims; (c) to increase the number of students who are categorized as noninvolved in bullying, through the enhancement of prosocial skills; and (d) to identify the threats to fidelity of the program and establish the validity of the pilot program with the possibility of replicating it in future (Martin et al.,  2005 , p. 376). Forty‐nine sixth graders from one Spanish primary school in Granada participated in the program.

The program designers gathered information about the social, educational and economic background of the school, of the students' families and the community in general. That was done during 3 meetings/seminars of 3 h each. Parents, teachers and members of the educational team attended those meetings. Through these meetings, it was established that the program should target interpersonal relationships of the children. It was decided that the program would be curriculum‐based as part of the normal program of the school. It was decided that the program would be implemented by one of the researchers because the teachers did not have enough qualifications to do it and because of lack of time and resources for teacher training. Parents and teachers were provided with information about bullying (a dossier/file) that they could use to discuss the problem of bullying with children. Also, teachers could attend the intervention program so that later they would be able to implement it by themselves. Parents were invited to attend some talks on bullying that would be given by the implementation team so that the program could be continued outside the school. The program was implemented for 5 months at the classroom level (30 sessions; 3 sessions per week with one tutor, i.e., one of the evaluators).

During the first 5 sessions, the tutor informed the children about peer bullying. Topics covered in the first 5 sessions involved issues such as concept of bullying, types of bullying, how to identify it, individual and group differences in bullying, and classroom rules against bullying. From the 6th to the 21st sessions, the program emphasis was on the emotional and social abilities of the children. Several topics were covered such as: identification and expression of emotions during bullying situations; communication abilities; ability to pose questions; ability of children to give and receive complements and complaints; ability to say no in life; ability to ask for a change of behavior; and ability to solve interpersonal problems. From the 17th to the 21st sessions, the program placed emphasis on mediation.

From the 22nd to the 25th sessions, the program emphasis was on human rights. Several topics were covered such as: freedom and equality, respect of private life, respect for other people's belongings, and respect for others' opinions. Similarly, from the 26th to 30th sessions, the emphasis was on moral education. During the whole program (sessions 1–30), there was also an emphasis on the inhibition of impulsivity and enhancement of reflexivity. For the enhancement of reflexivity, the program designers used a specific program called “Programa de Intervencion para Aumentar la Attention y la Reflixividad” [PIAAR] developed by Gargallo (2000) (see Martin et al.,  2005 , p. 378). This focuses on cognitive techniques that aim to inhibit impulsivity and enhance self‐control. The program also included role‐playing, peer mediation, guided discussion, brainstorming, and drawings.

The authors acknowledge several problems with the implementation of the program such as: little involvement by parents and teachers; implementation of the program lessons during recess time or during the physical education program; lack of time to cover all the topics; no second follow‐up because of difficulties of following the children; problems with the size and selection of the sample; the instrument they used; and possible contamination of results because of the way they categorized the children (Martin et al.,  2005 , p. 382). These pitfalls could easily be spotted. For example, the evaluators indicate that they implemented the program with the most aggressive sixth graders who had the worst interpersonal problems (Martin et al.,  2005 , p. 738). This made it difficult to know whether any changes in bullying in the experimental condition were attributable to the effectiveness of the program or to regression to the mean. Also, even though they distributed a self‐report questionnaire, they categorized children based on those questionnaires only after teachers' suggestions.

6.28. *Greek antibullying program (1)

The Greek antibullying initiative was a 4‐week intervention program that aimed to minimize both bullying and victimization. The conceptual framework of the Greek antibullying program was based on the theoretical model proposed by Salmivalli in 1999 (Andreou et al.,  2007 , p. 696), according to which changing an individual's behavior (e.g., the bully's behavior) entailed motivating not only the particular person but also the rest of the group members (participant roles' approach).

The program was embedded within the wider curriculum of the fourth‐, fifth‐, and sixth‐grade classrooms and consisted of eight instructional hours, each hour corresponding to one curricular activity. The curricular activities were presented to students by their classroom teachers who received training beforehand. The teacher training consisted of five 4‐h meetings and aimed to increase awareness of the bullying problem and its seriousness as well as to raise teachers' self‐efficacy in implementing the program (Andreou et al.,  2007 , p. 697).

The Greek antibullying curriculum was divided into three parts in accordance with the three main theoretical axes proposed by Salmivalli in 1999, namely: (1) awareness‐raising; (2) self‐reflection; and (3) commitment to new behaviors (Andreou et al.,  2007 , pp. 697–698).

In line with the first axis (awareness‐raising), small‐group and whole‐class discussions were conducted (over three instructional hours) that aimed to increase students' awareness of the bullying problem. Corresponding materials included a real snap‐shot from the playground, a story entitled “A new friend” and students' own drawings. In line with the second theoretical axis (self‐reflection), two instructional hours involving classroom discussions were conducted. These discussions placed emphasis on the participant roles that students took in the bullying process. Corresponding materials involved each students' completion of open‐ended sentences. Through this activity students were intended to reflect on critical issues around the causes, benefits, feelings, and consequences of adopting different roles. In line with the final axis (commitment to new behaviors), three instructional hours of small‐group and whole‐class discussions were conducted concerning different ways of approaching or solving the peer‐conflict situation and the formulation of class rules. Corresponding materials involved an open‐ended comic‐strip for group completion to find a solution to the bullying situation presented in the relevant story.

6.29. Greek antibullying program (2)

This antibullying program was implemented in Greek elementary schools during the academic year 2011/2012 (Tsiantis et al.,  2013 ). The school‐based program incorporated many elements and was implemented by teachers. Participating teachers attended a 2‐day training seminar before implementation began. A teacher's manual (Tsiantis,  2011 ) was also provided and outlined the detailed and systematic procedures involved in the intervention. Throughout the program teachers were provided with additional support from two mental health professionals whom acted as program co‐ordinators.

The program comprised of 11 weekly workshops that were implemented for two 45‐min class periods (90‐min in total). Class activities included group discussions, games and the formation and signing of class antibullying rules (Tsiantis et al.,  2013 ). Parent meetings were also organized to increase parent participation with the intervention. The first meeting provided parents with information about the intervention program and bullying issues. During the second parent session, students presented the achievements they had made during the intervention.

6.30. Inclusive

The INCLUSIVE program is a whole‐school restorative approach to bullying prevention and intervention (Bonnell et al.,  2015 ). The program involves creating an “action group” within each participating school in order to combat bullying. These groups are comprised of a minimum of six students and six members of staff, with at least one representative from senior management, teaching, support, and pastoral staff. Each action group is appointed an external expert facilitator for the duration of the intervention. It is the facilitators' role to provide ongoing support and training to each member of the action group. Action groups were required to meet regularly throughout the intervention year, approximately once every half term.

The INCLUSIVE intervention was designed to include several core standardized intervention components, including staff training in restorative practices, and a student social and emotional skills curriculum. However, the program also allows for schools to adapt the intervention according to school‐specific needs. These needs were established using a needs assessment survey distributed to year 8 students prior to commencement of the intervention. This survey aimed to establish student views on bullying and aggression in their schools, while providing information regarding school engagement and connectedness, perceptions of safety/risks, social support and social skills, relationships, and teaching in personal, social and health (PSHE) classes. Results of the needs assessment survey were then employed by the action group to tailor the INCLUSIVE intervention to target specific needs. The action groups also utilized this information to review and improve schools' existing policies, procedures and schemes (e.g., peer mediation and “buddying” schemes).

In relation to the core components of the INCLUSIVE intervention, all school staff were provided with introductory training in restorative practices by their affiliated expert facilitator. A minimum of twenty school staff were also required to attend intensive training provided by a specialist training provider. Restorative practices, such as “Circle Time,” were taught to staff to improve school climate and student‐staff communication. This technique involves teachers and staff sitting together in a circle discussing various emotional, social, and curricular issues. Each member of the circle is considered a valued contributor, and all inputs are treated equally. Circle time aims to support student communication and promote positive relationships. Another restorative technique used in the INCLUSIVE program was “formal conferencing,” which aimed to deal with serious bullying and aggressive incidents directly. Formal conferencing involves bringing together teachers, parents and students to establish appropriate punishment and ways in which the harm caused can be repaired. This approach emphasizes a nonjudgmental and inclusive environment so that both victims and perpetrators of bullying and/or aggression are involved.

Year 8 students also completed 5–10 h of social and emotional skills training throughout the process of the INCLUSIVE intervention. These lessons were based on the Gatehouse Project curriculum and could be delivered as either stand‐alone modules or integrated into existing academic curriculums. Modules covered included: (1) Establishing respectful relationships; (2) Emotion management; (3) Understanding and creating trusting relationships; (4) Exploring others' needs and avoiding conflict; and (5) Maintaining and repairing relationships.

6.31. *KiVa

The name of this project is an acronym of the expression “Kiusaamista Vastaan” which means “against bullying.” The word “kiva” in Finnish means “nice” and this is why this acronym was chosen for the specific antibullying initiative in Finland. Regarding the overall perspective of the program, the KiVa project included a universal and an indicated intervention (Kärnä et al.,  2011a ,  2011b ,  2013 ; Nocentini & Menesini, 2016; Salmivalli et al., 2007). The universal intervention referred to efforts made to influence the group norms while the indicated intervention referred to the way in which specific cases were handled in schools through individual and group discussions between the teacher and the students involved (Salmivalli et al., 2007, p. 6).

The KiVa program included a large variety of concrete materials for students, teachers, and parents. It also utilized the Internet and virtual learning environments (e.g., computer games against bullying) aiming in this way to enhance students' attitudes against bullying. Also, students received their own personal user ID, which they could use as a password before the completion of each web‐based questionnaire on bullying. KiVa included 20‐h student lessons, which were carried out by student teachers. The lessons involved discussions, group work, short films about bullying, and role‐playing exercises. After each lesson, a class rule was adopted, based on the central theme of the lesson.

A unique feature of the KiVa program was the use of an antibullying computer game. The game involved five levels and the teacher always activated the next level of the game after the relevant lesson was completed. Students were able to begin using the game after the third lesson; the second level of the program was played after the fifth lesson, and so on until the end of the school year. Each level of the computer game included three components that were named as “I know,” “I can,” and “I do.” In the first component, students were informed about basic facts on bullying. In the second component, the “I can”‐component, students moved around in the virtual school and faced different challenging bullying incidents. Finally, the third component was used to encourage students to make use of their knowledge and skills in real life situations.

Another important element of the KiVa project was the teacher training. Teachers were also provided with vests that they could use during playtime while supervising the school yard. This simple technique aimed to enhance teachers' visibility in the schoolyard and to signal that bullying was taken seriously in the school. Also, all teachers carrying out the KiVa program could seek advice from a web‐based discussion forum, where they could share experiences and ideas about bullying with other colleagues.

Within the school framework, the program also facilitated the use of a peer support group for victims of bullying. The classroom teacher was expected to arrange a group with 2–4 classmates—those who were pro‐social and had high status in the class—who were expected to provide support to victimized students, thus sustaining healthy peer relationships. An interesting element in the KiVa program is that it incorporated both punitive and nonblame approaches when dealing with perpetrators of bullying. Half of the school teams were instructed to use more punitive approaches (e.g., what you have done is wrong and it has to stop right now) while the rest of the school teams were instructed to use no‐blame approaches in their discussions with children (e.g., “your classmate is also having a hard time and this is why he behaves like that; what could we do to help him?”). There was also co‐operative group work among experts when dealing with children involved in bullying.

Finally, the KiVa program involved parents. A parents' guide was sent to the home and provided information about bullying and advice on how parents could be involved to reduce this problem. Information nights for parents were also organized and provided.

6.32. Lead Peace Intervention

The Lead Peace intervention is based on a resiliency conceptual framework (Resnik,  2000 ), thus, aims to reduce youth problem behaviors using an assets‐based approach (Harpin,  2011 ; Sieving & Widome,  2008 ). The intervention was developed as a school‐based “service learning and health education” program to reduce risk of violence and school failure in middle school students (Sieving, 2006). Developed from the Points of Light Youth Leadership curriculum for 9th to 12th grade students (Sieving, 2006), the program was adapted for use with Grade 6–8 students (Harpin,  2011 ).

The core curriculum targets factors on three levels: (1) environmental (e.g., adult resources and supports, family norms and behaviors, peer norms and behaviors, school/community opportunities and social connectedness); (2) personal (e.g., attitudes, beliefs, perceived norms, emotional distress); and (3) behavioral (e.g., social and emotional skills, coping behaviors, school performance). The program aims to reduce risky health and social behaviors (e.g., interpersonal aggression, physical fighting, bullying) in order to promote positive and reduce risky behaviors. The curriculum is implemented for 3 years, and can be delivered in two “doses”: (1) Lead Peace program (basic)—includes 15–20 intervention lessons each year; or (2) Lead Peace plus program—includes 30 intervention lessons, 15–20 additional community service hours, and health education and family outreach activities.

6.33. Lunch Buddy Mentoring program

The Lunch Buddy mentoring program was a school‐based antibullying program that aimed to reduce bullying victimization in elementary school children (Elledge et al.,  2010 ). The program was based on previous research that suggests youth mentoring can be utilized as an effective prevention technique (Dortch, 2000). In comparison to peer‐mentoring antibullying program, the Lunch Buddy program employed college student mentors based on prior success of college student mentoring aggressive children (Cavell & Hughes,  2000 ).

Mentors were provided with training prior to implementation of the program and participated in weekly meetings throughout the program. Children were identified as potential participants using a self‐ and teacher‐report victimization index. The self‐report School Experiences Questionnaire (Kochenderfer & Ladd,  2000 ) and teacher ratings of child victimization due to physical, verbal and relational aggressive were combined to create this index. School principals also collaborated with counselors to identify potentially suitable candidates. Eligible participants were then matched with same‐sex college student mentors, based on the availability of mentors during the mentees scheduled lunchtimes. Mentors visited the mentees twice a week, over the course of 5–6 months. During these visits mentors were required to sit with their mentee and their peers during lunchtime. Each mentor was also required to complete a log sheet after each visit.

6.34. Media Heroes

Chaux et al. ( 2016 ) evaluated the effectiveness of the cyberbullying prevention program “Media Heroes” [ Medienhelden ] on reports of traditional school bullying. The Media Heroes program is based theoretically on the Theory of Planned Behavior (Ajzen,  1991 ) and the social context of participant roles in bullying (Salmivalli,  2010 ). The program aims to reduce cyberbullying perpetration by enhancing empathy, increasing awareness and knowledge about what constitutes cyberbullying, the safety risks associated with Internet activity, and by providing assertive and useful methods in which bystanders can intervene in cyberbullying (Chaux et al.,  2016 ).

There are two versions of Media Heroes: (1) a short version implemented over four 90‐min lessons that take place in one school day; and (2) a long version that is implemented over 15‐weekly 45‐min lessons (Schultze‐Krumbholz et al.,  2012 ). Intervention activities include, role‐playing, class debates, news and film content, group learning and student‐parent presentations (Chaux et al.,  2016 ). Measures of both traditional‐ and cyber‐bullying were implemented in this evaluation, due to the significant overlap in the prevalence of these behaviors.

6.35. NoTrap!

Noncadiamointrappola (Let's Not Fall into a Trap), or NoTrap!, is a web‐based antibullying program that has been developed, implemented and refined over several studies (Menesini et al.,  2012 ; Palladino et al.,  2012 ,  2016 ). Initially implemented in two Italian schools in 2008, the program involves students actively engaging in the development of a website promoting antibullying (Menesini et al.,  2012 ). A selected number of students per school are provided with training and enroll as online peer‐educators. These students acted as online moderators of an antibullying forum, regulating discussion threads and responding to users' questions and concerns (Menesini et al.,  2012 ). In addition, peer‐educators also conducted face‐to‐face awareness raising workshops and meetings with their classmates, to highlight the key issues surrounding traditional‐ and cyber‐bullying (Palladino et al.,  2016 ).

Subsequent editions of the NoTrap! program incorporated additional elements based on findings from previous evaluations. For example, Palladino et al. ( 2012 ) placed more emphasis on: (1) victims' roles and victim support, (2) involving bystanders, (3) greater involvement of teachers in antibullying activities, and (3) creation of a Facebook group to supplement online materials. The third revision of the NoTrap! program incorporated standardization of the face‐to‐face antibullying activities led by peer educators (Palladino et al.,  2016 ). New peer‐led activities involved group work that targeted empathy and problem‐solving skills (Palladino et al.,  2016 ).

6.36. *Olweus Bullying Prevention Program

The OBPP was a multilevel program aiming at targeting the individual, the school, the classroom and the community level. Apart from marked mass‐media publicity, the program started with a 1‐day school conference during which the problem of bullying was addressed between school staff, students, and parents. This signaled the formal commencement of the intervention. Two different types of materials were produced: a handbook or manual for teachers (entitled “Olweus” core program against bullying and antisocial behavior') and a folder with information for parents and families. The program also included: (1) CD‐program that was used for assessing and analyzing the data obtained at the pre‐test period, so that school‐specific interventions could then be implemented; (2) a video on bullying; (3) the Revised Olweus Bully/Victim Questionnaire and (4) the book “Bullying at school: what we know and what we can do.”

The antibullying measures mainly targeted three different levels of intervention: the school, the classroom and the individual. At the school level, the intervention included:

• Meetings among teachers to discuss ways of improving peer‐relations; staff discussion groups.
• Parent/teacher meetings to discuss the issue of bullying.
• Increased supervision during recess and lunchtime.
• Improvement of playground facilities so that children have better places to play during recess time.
• Questionnaire surveys.
• The formation of a coordinating group.

At the classroom level the intervention included:

• Students were given information about the issue of bullying and were actively involved in devising class rules against bullying.
• Classroom activities for students included role‐playing situations that could help students learn how to deal better with bullying.
• Class rules against bullying.
• Class meetings with students.
• Meetings with the parents of the class.

At the individual level the intervention included:

• Talks with bullies and their parents and enforcement of nonhostile, nonphysical sanctions.
• Talks with victims, providing support and providing assertiveness skills training to help them learn how to successfully deal with bullying; also, talks with the parents of victims.
• Talks with children not involved to make them become effective helpers.

An interesting feature of the OBPP is that it offered guided information about what schools should do at both the intervention and the maintenance period. The Olweus program demands significant commitment from the school during the 'introductory period' which covers a period of about 18 months. Later the methodology acquired by the staff and the routines decided by the school may be maintained using less resources … Yet, even for the maintenance period, the program offers a point by point description of what the school should do to continue its work against bullying in accordance with Olweus methodology (Olweus, 2004c, p. 1). Also, at the school level training was offered to the whole school staff, with additional training provided to the coordinators and key personnel. These were responsible for coordinating the overall antibullying initiative in their school. The program also included cooperation among experts and teachers (e.g., psychologists) who worked with children involved in bullying.

6.37. Positive Action program

The Positive Action Program is a generalized school‐based “well‐being” program (Lewis et al.,  2013 ). The program targets both distal (e.g., school climate and teacher classroom management) and proximal (e.g., students' thoughts, feelings, and self‐efficacy) facets are targeted in order to impact a range of health‐ and behavioral‐related outcomes (Li et al.,  2011 ). The program is based on three core elements.

First, the Positive Action philosophy. Based on the theory of self‐concept (Combs,  1962 ; Purkey,  1970 ; Purkey & Novak,  1996 ) and a Positive Psychology (Frederickson, 2000; Seligman & Csikszentmihalyi,  2000 ) approach, the philosophy emphasizes positive feelings about the self, to encourage positive behaviors toward others (Flay & Allerd,  2010 ). Second, the Thoughts‐Actions‐Feelings Circle concept is used throughout the program to illustrate the reinforcing cycle of thoughts, feelings and actions. This is delivered to outline that positive thoughts lead to positive actions, positive actions in turn lead to positive feelings, which then reinforce positive thoughts. Third, a strict six‐unit curriculum that involves daily lessons, interactive learning and social‐emotional skill development.

The PA curriculum is designed to be adapted for kindergarten to Grade 12 students, and is based on six key concepts: (1) self‐concept; (2) social and emotional positive actions for managing oneself responsibly; (3) positive actions relating to a healthy body and mind; (4) honesty with oneself; (5) getting along with others; and (6) continuous self‐improvement (Lewis et al.,  2013 ). The intervention program also involves teacher, parent/family and community training. Schools implementing the PA program receive support from developers throughout implementation by training, manuals, school‐wide climate development, counselors, family classes, and individual consultations for staff with a PA implementation coordinator.

6.38. Preventure and Adventure CBT

The Preventure and Adventure intervention programs were part of two 2 year longitudinal projects that targeted adolescent alcohol use and bullying behaviors (Topper,  2011 ). Intervention components were primarily personality‐targeted cognitive behavioral therapy (CBT) for “high risk” students. Participants were screened prior to taking part in the intervention for four individual personality domains: (1) hopelessness; (2) anxiety‐sensitivity; (3) sensation seeking; and (4) impulsivity. Students who were classified as being “high risk” on any of the four domains were invited to participate, and assigned to one of four potential intervention workshops. These intervention sessions were CBT‐based and were aimed at each of the four personality domains. Thus, a student who scored highly on the impulsivity measure was assigned to the impulsivity‐focused CBT session. For participants that scored above the mean on multiple measures, they were assigned to the session that corresponded to the personality domain that they deviated the most from standardized scores.

High risk students in each school were randomly assigned to either the intervention or control condition, as were “low risk” students, for comparison. The Preventure study took place between 2005 and 2007, and either a chartered counseling psychology, an experienced special needs teacher, or a master‐level research assistant implemented intervention workshops. In comparison, the Adventure study took place between 2007 and 2009, and although the intervention sessions followed the same procedure, they were implemented by trained teachers in each school.

6.39. *Pro‐ACT+E program

Pro‐ACT+E was a universal, multidimensional program that aimed to prevent bullying in secondary schools (Sprober et al., 2006). It involved a cognitive‐behavioral approach to the problem of bullying and victimization by building up prosocial behavior. The program was universal: it did not involve specific work with perpetrators or victims of bullying. However, it included both teacher and parent training and a 2‐h classroom discussion with students about violence problems. The program offered curriculum materials that aimed to increase awareness in relation to the problem of bullying and placed emphasis on specific issues such as classroom management and classroom rules against bullying.

6.40. *Progetto Pontassieve

The program was delivered in a period of 3 years, and it consisted of two main parts. During the 1st two years it was delivered more at the school level whereas the 3rd year was more at the class and individual level (Ciucci & Smorti,  1998 ). During the 1st year a training course for teachers took place addressing psychosocial risks for children and bully‐victim problems. At the end of the training, a study was conducted to reveal how serious was the problem of bullying and what were its characteristics. The 2nd year of the intervention included a counseling service for each individual who was affected by bullying.

The intervention took place in the 3rd year and was based on the use of two different methods: Quality Circles, where pupils had to cooperate to find practical solutions to their problems, with the use of the Interpersonal Process Recall which consisted of the recording of one Quality Circle and discussion about it. The other method used was Role Playing conducted in small groups with subsequent class discussions, which helped students to examine possible strategies to face and overtake bullying problems. The aims of both of these methods were to make students aware that they could intervene in an efficient way to reduce bullying.

6.41. *Project Ploughshares for Peace

Project Ploughshares Puppets for Peace (P4 program) was an antibullying program that aimed to educate elementary school students about bullying and conflict resolution (Beran & Shapiro,  2005 , p. 703). The P4 program used puppets and a 30‐min script. Using three‐feet, hand‐and‐rod puppets, two puppeteers enacted a story that involved direct and indirect bullying, as well as a successful resolution to this scenario. These behaviors occurred among two female puppets and a male puppet friend.

After watching the play, students were invited to identify the bullying behaviors. During the discussion, four main strategies—presented as “4 Footsteps”—to deal with bullying were suggested to pupils: (1) ignore, (2) say stop, (3) walk away, and (4) get help. The show took approximately 45 min and aimed to increase children's awareness about which behaviors could be categorized as bullying and to show various strategies that children who were bullied and/or who witnessed bullying could use to discourage it (Beran & Shapiro,  2005 , p. 703).

6.42. Rational Emotive Behavioral Education (REBE) and ViSC

Trip et al. ( 2015 ) implemented a dual program consisting of REBE (Trip & Bora,  2010 ) and ViSC social competence (Strohmeier et al.,  2012 ) elements. These components were combined to address both social and emotional factors involved in bullying and positive youth development (PYD). This program approaches bullying from a sociological perspective, including factors on the individual, family, peer, classroom, and school levels (Espelage & Horne,  2008 ; Swearer & Espelage,  2011 ).

ViSC social competence program is a systemic approach to antibullying that targets students, teachers and parents (Strohmeier et al.,  2012 ). Implemented by teachers in the classroom, the program comprises several intervention units that aim to: (1) foster empathy and perspective training, (2) enhance responsibility, and (3) improve students' behavioral responses to bullying (Trip et al.,  2015 , p. 733).

REBE elements employed by Trip et al. ( 2015 ) on the other hand, target specific elements of aggression that are lacking in the ViSC units. Based on the theory of Rational Emotive Behavioral Therapy (Ellis,  1962 ), the REBE elements of the intervention program target the difference between desire and reality (Trip & Bora,  2010 ) and anger. The REBE program activities target specific elements of anger, specifically, anger triggers, personal experiences of anger and the consequences of anger (Trip et al.,  2015 ).

6.43. Restorative Whole‐school Approach (RWsA)

The RWsA (Hopkins,  2004 ; Morrison,  2002 ) was a school‐based antibullying initiative that employs a restorative justice inspired philosophy. Hence, the program focuses on creating a positive school environment to prevent bullying in the long‐term, rather than a short‐term disciplinary and punishment approach (Wong et al.,  2011 ). The program had three core goals: (1) to create a positive and harmonious school learning environment; (2) implement an interactive classroom curriculum; and (3) encourage an effective partnership between teachers, students, parents and relevant professionals.

A whole‐school antibullying nonpunitive ethos and policy is implemented as the core of the intervention (Wong et al.,  2011 ). This policy aims to establish a positive school environment in order to combat bullying‐related risk factors. The curriculum lessons incorporated elements on various issues, including, empathy, assertiveness, coping, problem‐solving, and conflict resolution.

6.44. Resourceful Adolescent Program (RAP)

The RAP is a classroom‐based CBT intervention designed for adolescents aged 12–15 years of age (Stallard et al.,  2013 ). The program is a depression prevention program, however, bullying problems were included as secondary outcomes. The program incorporates a detailed manual and student workbooks, and was implemented over nine sessions, of approximately 50–60 min each. The core components include: psycho‐education, helpful thinking, identifying personal strengths, keeping calm, problem solving, support networks, and keeping the peace. The program was designed to flexible and adaptable to participating schools' varying busy timetables.

6.45. *S.S. Grin

The Social Skills Group Intervention (S.S. GRIN) was a school‐based program that aimed to help children enhance their social skills. S.S. GRIN was designed as a social‐skills training intervention for peer‐rejected, victimized, and socially anxious children. It could be applied to an array of problems that are social in nature (e.g., aggression, low self‐esteem, depression, social anxiety, social withdrawal) not just bullying (DeRosier & Marcus,  2005 , p. 140). The authors argued that the program went beyond the most common social‐skills training (De Rosier & Marcus, 2005, p. 141) by emphasizing the cognitive aspects of relations and emotions. That is, children were not only taught prosocial skills, but they were also taught, on the cognitive level, how to identify negative perceptions and behaviors in an effort to help children to regulate their own emotions as well as enhance their coping skills.

Overall, the program was a combination of social‐learning and cognitive‐behavioral techniques, used to help children build social skills and positive relationships with peers. It was a highly structured, manualized program (DeRosier,  2004 , p. 197) with a number of sessions containing scripts and activities to undertake. Each session included didactic instruction combined with active practice such as role‐playing, modeling and hands‐on activities (De Rosier, 2004, p. 197). The children participated in group sessions for eight consecutive weeks. Each session lasted approximately an hour. The groups were led by each school's counselor and an intern, who were trained and supervised by one of the program instructors (De Rosier & Marcus, 2005, p. 143).

6.46. School‐based Drama program

This school‐based antibullying program was based on drama (Owens & Barber,  1998 ) and social cognitive theories (Bandura,  1978 ). The main aim of this project was to design and implement a drama‐based program to improve social relationships and social/emotional well‐being in children, which in turn may help to reduce bullying (Joronen et al.,  2011 ). Targeted concepts included: empathy; social competence; student‐teacher interaction; child–parent interaction; and recognition of values/emotions.

This program was developed by the combined efforts of researchers, drama experts and teachers. It was implemented in‐class by trained teachers and school nurses over a period of 6 months. Teachers and school nurses attended a 2‐day seminar and received two drama handbooks, however, there was no manual or fixed program outline provided. Support was provided through email communication between teachers and researchers for the duration of program implementation. Teachers conducted one drama session per month with their class. These sessions covered a variety of topics, including, bullying, friendship, loss of a friend, supporting a bullied peer, tolerance, and child abuse.

6.47. School‐wide Positive Behavioral Interventions and Supports (SWPBIS)

SWPBIS was a universal behavioral intervention program that targets school‐level factors in order to improve school climate and promote positive student and staff behaviors (Waasdorp et al.,  2012 ). Instead of following a specific antibullying curriculum, SWPBIS aimed to reduce bullying by targeting schools' discipline and behavioral management strategies. A SWPBIS team in each school organized and facilitated the intervention implementation.

These teams were responsible for developing a set of “positive expectations” for the school. These were a number of statements that outlined what the school expected in relation to student and staff behavior, for example, “be responsible, respectful, and ready to learn” (Waasdorp et al.,  2012 , p. 150). Posters highlighting the expectation statements were then displayed all around participating schools, both in classrooms and outside of classrooms, and are positively reinforced using reward systems. Furthermore, data from student surveys and discipline referrals were employed throughout the intervention to inform teachers of potential bullying “hot spots” that require increased supervision and monitoring. School staff also received training on classroom management and how to respond consistently and effectively to bullying. Additionally, students identified as being “high risk” or vulnerable to bullying behaviors or victimization were provided with selective intensive intervention.

6.48. School bus antibullying intervention

This intervention program was a universal antibullying program designed to reduce the prevalence of bullying behaviors on school buses (Krueger,  2010 ). The program was purposefully developed and utilizes materials and content from the “Take a Stand, Lend a Hand, Stop Bullying Now!” tools that are available free of charge.

The intervention was implemented with elementary school children over five consecutive days, during the final 20‐min of the school day. Lessons were delivered by the school's social worker and principal to two groups (kindergarten to 2nd grade students, and 3rd to 5th grade students) of participants. The program followed this format from days 2–5, however, on day 1, all participants completed the introductory lesson together. The school‐bus antibullying program primarily utilized DVD materials from the “Take a Stand” content. These video clips depicted cartoon characters engaging in different bullying scenarios.

On day 1 (i.e., the introductory lesson) an overview of school bullying and related issues, including bystander intervention, was provided to participants. The associated DVD clip depicted a male character physically bullying another child in the playground while other students watched. Participants then discussed the clip in groups, and were introduced to the “Three Steps to Stop Bullying Chart.” This technique involves three steps, Stop, Help , and Tell , that bystanders can take if they witness bullying.

On each subsequent day, a new DVD clip was shown to participants and the Stop, Help , and Tell concepts were revisited. The school's social worker or principal led discussion groups by posing questions to the students concerning the feelings and emotions experienced by the victim of bullying, potential coping strategies that the victim could use, and possible bystander behaviors. Participants also shared their previous experiences with similar situations. Furthermore, using the Stop, Help , and Tell paradigm, participants brainstormed potential ways to tell a bully to stop behaving in a certain manner, ways to help the victim and appropriate trusted adults that they can tell about the situation.

6.49. Second Step

The Second Step: Student Success Through Prevention is a middle school Social‐Emotional Learning (SEL) program that aims to reduce bullying, peer victimization, physical aggression, homophobic name‐calling and sexual violence (Espelage et al.,  2013 ,  2015 ). The intervention curriculum is taught in‐class by trained teachers. Lessons are interactive and engaging, requiring students to take part in whole‐class, small group and individual work. A take home task is also given after each lesson to reinforce skills learned. DVDs are also used to accompany and enrich lesson content.

The 6th grade Second Step curriculum involves 15 weekly lessons on various social and emotional skills and bullying‐related topics. The following outlines the curriculum: (1) empathy and communication—five lessons; (2) bullying—two lessons; (3) emotion regulation (e.g., coping with stress)—three lessons; (4) problem‐solving—two lessons; and (5) substance abuse prevention—four lessons.

Each lesson has clearly outlined learning objectives to reduce problem behaviors and increase prosocial behaviors. For example, lessons on bullying target the peer context by increasing knowledge, improving attitudes, and encouraging bystander intervention in order to reduce bullying perpetration and victimization. Students are educated about the differences between types of bullying, importance and responsibilities of bystanders in preventing bullying and a number of positive bystander behaviors are modeled. The 7th grade Second Step curriculum involves a similar lesson structure, with some slight changes. The intervention is delivered over 13 weekly lessons, and cyber‐bullying and sexual harassment issues are incorporated into bullying modules.

6.50. Shared Concern

Wurf ( 2012 ) assessed the effectiveness of the whole‐school approach to bullying intervention and prevention, with a particular emphasis on Pikas' ( 2002 ) nonpunitive method of shared concern. The Pikas method of Shared Concern is a teacher, or counselor, implemented intervention, that is divided into five key stages. First, the intervener identifies the students involved in bullying and talks with them individually. These discussions aim to provide nonpunitive and constructive options for both bullies and victims (Wurf,  2012 ). The second and third stages involve providing empathy and ongoing support to the victims of bullying. Finally, the fourth stage incorporates a mediation session between bullies and victim(s). A conflict resolution approach to prevent bullying is agreed upon and implemented by all involved. The fifth and final stage occurs during the follow‐up period, whereby the teacher or counselor monitors the involved students to ensure that the bullying has stopped.

6.51. *Short Intensive Intervention in Czechoslovakia

The antibullying intervention in Czechoslovakia was inspired by the OBPP and borrowed elements from it, such as the Olweus videocassette on bullying (Rican et al.,  1996 , p. 399). The Olweus bullying questionnaire was used to measure several aspects of bullying within the schools. A peer nomination technique was also used to identify bully and victim scores. The relevant results from both measurement scales were presented to teachers in the intervention schools to increase awareness of the problem of bullying. The program researchers discussed with the teachers “possibilities of an individual approach to the bullies as well as to the victims” (Rican et al.,  1996 , p. 399).

As another intervention element, teachers were instructed to introduce relevant ethical aspects into the curriculum where possible: the ideal of knighthood was suggested for history classes and the ideal of consideration for the weak was introduced in sentences used for dictation and analysis (Rican et al.,  1996 , p. 400). Another element of the intervention involved the use of a method called “class charter.” Specifically, children were asked to indicate how they would like their teachers and other classmates to behave toward them as well as how students should behave toward teachers and among themselves. The final aim of this classroom activity was the construction of a set of rules and principles, which was then signed by all pupils in the classroom and placed there in a visible position. Finally, the Olweus video‐cassette on bullying was shown to children and was used as a means of promoting the antibullying idea in the school.

6.52. *Short Video Intervention

This antibullying strategy, involved a single viewing of an antibullying video, entitled Sticks and Stones, and aimed to examine its effects on secondary school students' views of, and involvement in, bullying. The program aimed to examine both attitudes toward bullying and the actual behavior since “it would not be unreasonable to propose that these attitudes will influence actual behavior” (Boulton & Flemington,  1996 , p. 334). The program involved only one school that had no prior antibullying policy.

The video presented pupils (either in groups or on their own) talking about bullying, their views about this phenomenon and their personal experiences of bullying. The video also involved a number of bullying scenes (see Boulton & Flemington,  1996 , p. 337 for examples).

6.53. Social and Emotional Training (SET) intervention

This intervention program was a school‐based SET mental health program for Swedish school children (Kimber et al.,  2008 ). The SET program was primarily focused on mental health, but also targeted other aspects of participants' lives, such as bullying. Both internalizing and externalizing aspects of child mental health are addressed.

Trained teachers delivered the program over the course of two academic years. Intensity of program implementation varied according to the age of students. Junior students (i.e., grades 1–5) received the program in 45‐min sessions twice a week, while senior students (i.e., grades 6–9) completed one 45‐min session per week. Program developers provided each participating teacher with detail manuals for implementing the program with each grade and grade‐specific student workbooks. Role‐playing and modeling tasks covered many themes, including: social problem solving; conflict management; dealing with strong emotions; and resisting peer pressure. Teachers were also supervised once a month during the 1st year of implementation, and students were encouraged to practice skills both at school and at home.

6.54. Social Norms Project

Lishak ( 2011 ) implemented an antibullying program based on social norms theory (Perkins,  2003 ) with middle school students. The program was implemented over a period of 12 weeks and was developed based on student responses to an anonymous web‐based survey and student discipline and suspension reports (Lishak,  2011 ). Student surveys collected information regarding perceptions of bullying in the school and results were then relayed to participants via weekly lessons, assemblies, posters, and media content throughout the school. Data from school discipline, suspension and visitation logs were collated to estimate the prevalence of bullying and school violence.

6.55. *Social Skills Training (STT) program

STT was a program specifically designed to support “chronic victims” of bullying (Fox & Boulton,  2003 , p. 237). The general aim of the program was to help children improve their social skills, therefore reducing a child's individual risk of victimization (Fox & Boulton, 2003 , p. 234). The program involved an 8‐week course during which children learnt how to use both problem‐solving and relaxation skills, how to think positively, how to modify their nonverbal behavior and how to use some verbal strategies such as “fogging” and “mirroring” (Fox & Boulton,  2003 , p. 235).

During the program, victims of bullying were gathered in groups of five to ten and were exposed to the aims of the program for 1 h/week. Two trainers delivered the 1‐h sessions throughout the program. The 1st week was dedicated to children introducing each other and listening each other's problem. The next two sessions dealt with issues of friendship and aimed to help children form strong friendships (e.g., having conversations; asking to join in), while the fourth session dealt with issues of body language: teaching children how to modify their nonverbal behavior in a way that would protect them from being victimized. During the fifth session children learned how to be assertive while in the next two sessions children were taught how to deal with the bully. The eighth session signaled the end of the program.

6.56. *SPC and CAPSLE program

This evaluation compared the effects of two intervention packages with a treatment‐as‐usual condition (Fonagy et al.,  2009 ). Nine schools were randomly allocated to the two experimental and one control (treatment‐as‐usual) conditions after a stratified allocation procedure, which was used to stratify schools based on the percentage of low‐income students (indicated by students' free‐ and reduced‐lunch status). In the experimental conditions, the full intervention was offered for 2 years (the efficacy phase) with a limited 3rd year of intervention (the maintenance phase).

The first experimental condition involved a “School Psychiatric Consultation” (SPC), a manualized protocol that aims to address mental health issues of children with disruptive behavioral problems, internalizing problems, or poor academic performance. SPC was a school‐level intervention focused on individual children. Three child psychiatry residents, supervised biweekly by a senior child psychiatrist, delivered mental health consultation following the SPC manual for 4 h/week. The psychiatric residents attended weekly school resource meetings and consulted directly with teachers, parents and other school personnel, through classroom observations and meetings, providing 140 consultations for 65 students in year 1 and 97 consultations for 45 students in year 2.

The second experimental condition involved the implementation of CAPSLE (“Creating a Peaceful School Learning Environment”), a manualized psychodynamic approach addressing the cocreated relationship between bullies, victims and bystanders. In contrast to SPC, CAPSLE represents a whole‐school intervention approach. It aimed to modify the educational and disciplinary school climate. A CAPSLE team drawn from school staff in the pilot project led implementation in the two intervention years using a training manual. In year 1, teachers received a day of group training, students received nine sessions of self‐defense training, and the CAPSLE team consulted with school staff monthly. Year 2 started with a school‐wide half‐day refresher self‐defense course, and consultation continued with counselors, teachers and adult/peer mentor programs. In year 3 (the maintenance phase), self‐defense training continued as in year 2.

CAPSLE includes several antibullying materials that can be used by teachers such as a Teacher Discipline Manual (used in the teacher training), a Student Workbook, Buttons and Magnets and Patches (used as a way of reinforcing of desirable student behavior), Parent Warning Notes (notifying parents about specific problem behavior of the child) as well as antibullying videos that can be used during the physical education lessons (and videos that can be used by parents). CAPSLE also includes the Gentle Warrior Program, a 12‐week curriculum specifically designed for physical education teachers. For CAPSLE, intervention fidelity was assessed using a teacher self‐report measure that required teachers to state the frequency with which various CAPSLE program components were implemented.

6.57. Standard CBT and CBT plus media program

This intervention program combined elements of standardized CBT and DVD bullying‐related materials in order to reduce bullying perpetration and victimization among elementary school children (McLaughlin,  2009 ). The standardized CBT lessons were delivered by a trained counselor, and focused on bullying and aggression relation issues. Two experimental groups were employed, one of which received only the CBT lessons, and the other completed the CBT lessons and were shown the bullying DVDs.

The program was implemented over 4 weekly lessons that followed a strict outline. In week 1, the lesson focused on defining bullying, identifying bullying roles and different forms of bullying, and exploring the possible characteristics of bullies, victims, and bystanders. Week 2's lesson was concerned with establishing the consequences of bullying for all those involved, including the bully, victim and bystanders. Empathy for victims of bullying was also developed. Activities included creating feeling lists, and participating in role plays. Lesson three aimed to promote bystander intervention by developing awareness and knowledge of appropriate responses to bullying, suitable ways to intervene, and promoting assertiveness. Classes are taught using educational and informative posters. The final lesson, in week 4, aimed to outline the gender differences in bullying, why these occur, and ways to combat gender‐specific forms of bullying. In their classes, students establish class antibullying rules and are taught about the support available in school to stop bullying.

In addition, students in the CBT + media experimental group watched three DVDs that highlighted the issues outlined in the weekly lessons. The DVDs that were shown are as follow: (1) Let's Get Real , which shows young people talking about their personal experiences of bullying; (2) The Deepest Hurt , that depicts girls role‐playing various scenarios of relational aggression; and (3) The Broken Toy , a dramatization of the damage bullying can cause. Following the videos, students engaged in group discussions led by the counselor about the issues illustrated in each DVD.

6.58. *Stare bene a scuola: Progetto di prevensione del bullismo

This intervention was based on the curriculum activities and the whole school approach because it tried to involve all people in a school (Gini et al.,  2003 ). The program was delivered to 6 schools and included several activities. Teachers were first trained in 3 days on “cooperative learning” and in particular on the Jigsaw technique. Teachers then had an on‐going supervision once every 15 days. The intervention in the class lasted 4 months with two meetings a week. The intervention was directed toward the following areas: (1) awareness of the body and what it feels; (2) emotional awareness; and (3) bullying awareness. These areas were dealt with in each of the sessions, starting from the first one. For each thematic area, several activities were conducted and several methods were used.

6.59. Start Strong

“Start Strong: Building Healthy Teen Relationships” was a school‐based curriculum focused teen dating‐violence prevention program (Williams et al.,  2015 ). The program was implemented over 2 years in four experimental schools (that implemented the program) and four comparison schools (that did not implement the program). Schools were matched based on: school size, percentage of students eligible for free school lunches, race/ethnicity, and socioeconomic status. The effectiveness of the program was measured for outcomes that included the perpetration and victimization of teen dating‐violence, bullying and sexual harassment.

6.60. *Steps to Respect

The Step to Respect program aimed to tackle bullying by: (1) increasing staff awareness; (2) fostering socially responsible beliefs; and (3) teaching social‐emotional skills so as to promote healthy relationships (Frey et al.,  2005 , p. 481). The program included staff and family training manuals, a program guide and lesson‐based curricula for third‐ through sixth‐grade classrooms (Hirschstein & Frey,  2007 , p. 7).

Components at a whole school level consisted of an antibullying policy and procedures, staff training and parent meetings, all aiming at sharing understanding of bullying and its consequences and increasing adult awareness, monitoring, and involvement. At the classroom level, the proposed activities consisted of teaching friendship skills, emotion regulation skills, identifying types of bullying, teaching prevention strategies and peer group discussion. The aim was to improve peer relations and reduce the risk of victimization, assess level of safety and recognize, report and refuse bullying. At the individual level, students involved in bullying were approached and coached based on the “Four‐A Responses”: affirm behavior, ask questions, assess immediate safety and act.

The S to R training manual consisted of an instructional session for all school staff and two in‐depth training sessions for counselors, administrators, and teachers. There were also videos accompanying the program. With regard to staff training, there were two levels of training: all school staff received an overview of the program goals and principal aspects of the program (program guide). Teachers, counselors, and administrators received additional training in how to coach students involved in bullying, based on behavioral skills training, cooperative learning and role‐playing.

The student curriculum comprised skills and literature‐based lessons delivered by third‐ through sixth‐grade teachers during a 12–14‐week period. The intervention consisted of 10 semi‐scripted skills lessons with topics such as joining groups, distinguishing reporting from tattling and being a responsible bystander.

Finally, with regard to the parent intervention, administrators informed parents about the program and the school's antibullying policy and procedures. Parents could also benefit from other resources such as letters provided to them and newsletters describing whole‐school antibullying activities undertaken at school.

6.61. Strengths in Motion (SIM)

The SIM (Rawana et al.,  2011 ) program was a strength‐based whole school antibullying intervention. There were several components involved in the program, all of which centered around a strength‐based approach. This technique involves highlighting and enhancing individuals' strengths in order to develop positive mental health (Duckworth et al.,  2005 ). In the context of the present evaluation, Rawana et al. ( 2011 ) requested that each participating school allocated one room as a designated intervention resource room. In the first instance, this room acted as a “Good Start Centre” (p. 287) where new students to the school were provided with two half‐day orientation sessions prior to starting school. Part of these orientation sessions was individualized strength assessments. It was predicted that by providing new students with guidance on how to best use their strengths to integrate successfully into school life the likelihood of future bullying and victimization would be reduced.

The second use of the intervention room was as a “Cool Down & Prevention,” where students experiencing behavioral or emotional problems could go to calm down. Staff were on hand to prevent the behaviors from escalating and offer helpful advice. The room also acted as an alternative to suspension from school, whereby students could be mandated to spend a certain number of days in the “Good Choices Room.” An ambassador's club for students identified as being at high risk for bullying perpetration or victimization was also held in the resource room. Finally, mental health professionals provided student and parent workshops and staff received tailored training on the strength‐based approach to bullying prevention and intervention.

6.62. Take the LEAD (TTL)

The TTL (Domino,  2011 ,  2013 ) program was designed to increase the social competencies of participants in order to reduce bullying behaviors. The intervention is based on SEL and PYD theories.

Various social and emotional skills are targeted during the 16‐weekly lesson curriculum, including: (1) Self‐awareness; (2) Self‐management; (3) Social‐awareness; (4) Relationship skills; (5) Decision making; (6) Problem solving; and (7) Leadership. Trained teachers taught TTL lessons during normal class periods on a weekly basis. Participating teachers were trained on the skill‐based curriculum by the developers of the TTL program. During training, teachers were taught about specific learning objectives and goals of the intervention program, and also about the lesson plans and activities involved in “Take the LEAD.” Information evenings for parent were also held as part of the TTL intervention and aimed to raise parents' awareness of key social‐emotional issues.

Each of the sixteen TTL lessons involved specific learning objectives and goals. Lessons involved a combination of knowledge and skill development and an application component, so that participants were given the opportunity to apply skills in real‐world settings. For example, the “Communication skills” lesson aimed to “explore elements of communication that enhance interpersonal skills and foster positive relationships (Domino,  2013 , p. 432). During this lesson students brainstormed ideas about effective and positive communication techniques and were then required to practice these skills (e.g., eye contact, active listening and showing empathy) in pairs. Finally, participants were required to practice these techniques in an interview with a classmate, and later with a parent.

6.63. *Toronto antibullying program

The Toronto antibullying program was inspired by the OBPP (Pepler et al.,  2004 , p. 125). It was based on the understanding that bullying is a problem that extends far beyond the individual children; it involved the peer group and the teachers, as well as the parents of children (Pepler et al.,  2004 , p. 127). The program included several preventive elements implemented at the school, parent, and classroom levels, as well as additional work with specific students involved in bullying as perpetrators or victims.

The level of implementation of the program varied across the intervention schools. However, in all intervention schools three critical elements were found: staff training, codes of behavior and improved playground supervision. At the school level an emphasis was placed on developing a positive code of behavior among students, engaging teachers, and promoting positive playground interactions. At the parent level, information nights were held during which parents were informed about the problem of bullying in their school. Also, information about the program and its objectives was sent home. At the classroom level, children were involved in developing classroom rules against bullying. Further classroom activities aimed to change students' attitudes and to promote healthy relationships among peers. At the individual level, children involved in bullying as perpetrators or victims received specialized intervention through consultation and though engaging their parents. Follow‐up monitoring of these cases helped school authorities to establish that bullying incidents were terminated or discontinued.

6.64. *Transtheoretical‐based Tailored antibullying program

This antibullying initiative involved “transtheoretical‐based tailored programs that provided individualized and interactive computer interventions to populations of middle and high school students involved in bullying as bullies, victims and/or passive bystanders” (Evers et al.,  2007 , p. 398). The intervention involved only three 30‐min computer sessions during the school year for the students and a 10‐page manual for staff and parents with optional activities. According to the program designers, the transtheoretical model is “a theory of behavior change that applies particular change processes like decision‐making and reinforcement to help individuals progress at particular stages of change” (Evers et al.,  2007 , p. 398).

Intervention materials included the “Build Respect, Stop Bullying” program, which is a multicomponent, internet‐based computer system (Evers et al.,  2007 , p. 402). Students initiated the program by running a multimedia CD which brought them to the program website. Students could use the program by creating a login name based on personal information and a password. Once the students registered for the program, logged in and consented to be involved in the intervention study, they were given instructions on how to proceed. This multi‐media program also included short movies (videos) of students giving testimonials about bullying (Evers et al.,  2007 , p. 403).

Other elements of the program included: (1) a 10‐page family guide, sent to children's homes, which provided brief information about the multi‐media program and its relation to the antibullying initiative; and (2) a 10‐page staff guide, which included general information about bullying and how to support student change, classroom activities and information on how to work with parents. Teachers were not provided with any training.

6.65. Utrecht Healthy Schools

The Utrecht Healthy Schools program was a comprehensive educational program that targeted adolescent health behaviors (Busch et al.,  2013 ). The integrated program aims to improve various different health‐related behaviors exhibited by Dutch secondary school students, such as, nutrition, exercise, sexual health, substance and alcohol use, smoking behaviors, bullying, and excessive use of television, gaming and Internet use. The program was implemented as a whole‐school approach and consisted of five key components.

First, participating schools implemented a “healthy school” policy outlining a zero‐tolerance attitude toward risky or violent behaviors, such as alcohol use, smoking or bullying. Second, the program aimed to create a healthy school environment by offering healthy options in the canteen, removing vending machines, ensuring proper sports facilities, hosting alcohol‐free school parties and implementing a smoke‐free school yard. In the third instance, the program aimed to involve parents in intervention activities by providing parent workshops and/or take‐home activities for students. Finally, curriculum materials focused on personal skill development and the program aimed to incorporate public health services into the intervention program.

6.66. *Viennese Social Competence Training program (ViSC)

The ViSC aimed to provide students “with systematic theoretically‐based guidance in becoming responsible and competent actors in conflict situations” (Atria & Spiel,  2007 ; Yanagida et al.,  2019 ). It was specifically designed for disadvantaged adolescents aged fifteen to nineteen who were considered at risk for future problems (Atria & Spiel,  2007 , p. 179). The theoretical basis of the programs drew its main ideas from social information processing theory and from research that approached the problem of bullying as a group phenomenon (Gollwitzer et al.,  2006 , p. 126).

The ViSC program consisted of thirteen lessons which were divided into three phases: (1) impulses and group dynamics; (2) reflection; and (3) action. The first phase, entitled “impulses and group dynamics,” consisted of six lessons and the main aim was to enhance students' competence in dealing with critical situations by teaching them how to look at social situations from different perspectives using vignette stories, discussions and role‐plays. The second phase, reflection , involved one lesson during which pupils reflected on what had been learned in the first phase of the program.

The last phase, action , consisted of six lessons during which the trainer asked students to define how they wanted to benefit from the remaining lessons. The trainer collected students' individual ideas, evaluated them and—along with the students—put them in practice in alignment with the global goal of the program: enhancing pupils' social competence. The third phase of the program was flexible and it could involve several projects suggested by pupils such as a movie production, a work of art, the organization of a party, and so on. This flexibility was allowed and was, in fact, a main feature of ViSC because organizing such projects “involves a variety of critical situations, in which alternative, nonaggressive response options can be probed, rehearsed, and evaluated for success” (Gollwitzer et al.,  2006 , p. 126).

Based on the design of the program, the training of students was conducted by specialist trainers, not their teachers. The trainers participated in instruction workshops and were also supervised during the training by the ViSC developers' team at the University of Vienna (Gollwitzer et al.,  2006 , p. 127). According to the principles of the program, it was essential for the trainer to avoid receiving any information about individual students offered by teachers; students' assessments should be based on standardized diagnostic measures (Atria & Spiel,  2007 , p. 184). Moreover, the training was conducted during regular class time and teachers were advised to attend the lessons, so that the program was taken seriously by the students. ViSC has been implemented and evaluated three times: by Gollwitzer (2005), by Atria and Spiel ( 2007 ) and by Gollwitzer et al. ( 2006 ).

6.67. Youth‐led program

The Youth‐led program (YLP; Connolly et al.,  2015 ) was a generalized middle school violence prevention program. This program was developed by a community agency, and involved training high school students to lead violence prevention workshops with middle school students in order to increase the latter's knowledge and attitudes of peer aggression and victimization.

Experienced mental health professionals were employed to select and supervise male and female high school students that would become “youth leaders.” These students received training in afterschool sessions on skills and knowledge of peer aggression. Topics covered included bullying perpetration and victimization, but also peer aggression, violence, and harassment.

The final sessions of this training required the youth leaders to create two individualized presentations; one covering bullying and the other discussing general aggression. Mixed gender pairs of youth leaders then conducted these presentations in middle school classrooms under the supervision of a mental health worker. These presentations lasted for approximately 45 min each.

6.68. *Youth Matters

The Youth Matters program used “a curricular and a modified systemic approach to bullying prevention” (Jenson & Dieterich,  2007 , p. 287). The aim of the curriculum was to strengthen peer and school norms against antisocial behaviors by addressing critical issues (issue modules) such as the difference between teasing and bullying, building empathy, risks and norms surrounding aggression and so on. The curriculum also aimed to promote skills (skill modules; structured skills training sessions) that students could use in order to stay safe at school, cope with bullying, enhance their social skills and improve their peer relationships. To address systemic issues associated with bullying, curriculum modules terminated with the development of classroom or school‐wide projects, which placed emphasis on the negative consequences of bullying for students.

The curriculum consisted of 10‐session modules. Each module included a 30–40‐page story, the content of which was directly linked to the structured skills training sessions. When looking at the implementation of the program, all curriculum materials were “language sensitive”: translated into Spanish for use in the three Spanish‐speaking classrooms included in the evaluation. Youth Matters curriculum modules were offered to fourth and fifth graders. According to Jenson and Dieterich ( 2007 , p. 287), grades 4 and 5 were selected “based on an appropriate fit between developmental ability and curricula.”

The Youth Matters program was based on a theoretically grounded curriculum. The curriculum was based on theoretical constructs derived from the Social Development Model. The latter integrated perspectives from three theories (i.e., social control theory, social learning theory and differential association theory) and proposed that four factors inhibit the development of antisocial development in children. These were: (1) bonding or attachment to family, schools and positive peers; (2) belief in the shared values or norms of the above‐mentioned social units; (3) external constraints or consistent standards against antisocial behavior; and (4) social, cognitive and emotional skills that can be seen as protective tools for children to solve problems and perform adequately in social situations. The Youth Matters curriculum addressed each of these four core areas.

6.69. Zero program

The Zero antibullying program is based on the idea that bullying is predominately a version of proactive aggression (Roland et al.,  2010 ). The program aims to create a school environment that prevents these forms of proactive aggression. The intervention places the majority of responsibility for bullying prevention and intervention with the adults within the school environment (Roland et al.,  2010 ). School staff were required to define clear standards of positive prosocial behavior among the students and to ensure that these standards are met. Thus, the adults within the school context adhere to a “zero tolerance” policy toward bullying. Another key feature of the intervention is that students are instructed to treat all school property appropriately and respectfully and the intervention philosophy is carried into classroom activities and standards also.

During the intervention, class teachers engage their respective classes in active discussions about issues relating to bullying in adherence with the intervention guidelines. The preventative function of the Zero program takes both a direct and indirect approach (Roland & Galloway,  2004 ). Teachers are also expected to be vigilant and visible in school corridors and playgrounds during nonclass time and follow intervention procedures when dealing with specific instances of bullying (Roland et al.,  2010 ). When particular instances of bullying are identified, the victim is first approached and takes part in a few sessions with trained staff being comforted and assured. Parental involvement also occurs at this point. Finally, the perpetrators are invited to attend meetings and conflict resolution occurs under a restorative justice model.

6.70. Zippy's Friends

Zippy's Friends is a universal school‐based program for children aged 6–8 years old (Holen et al.,  2013 ; Mishara & Ystgaard,  2006 ). The overarching aim of the program is to develop and improve participants' coping strategies in order to reduce and prevent psychological problems. Zippy' Friends has been funded by the global suicide prevention organization “Befrienders International,” and is now distributed internationally by the nonprofit group “Partnership for Children.”

The intervention is delivered over the course of 24 weekly lessons, that are implemented by classroom teachers. The program is based around six stories of the imaginary character “Zippy,” three children, and their families and friends. A structured curriculum outline for each lesson allows participants to engage and discuss the various themes that emerge in each of the stories. Themes that are incorporated include: emotions; communication; friendships; conflict resolution; loss and change.

Teachers are provided with a detailed manual for the program and are required to guide their classrooms through the intervention while also encouraging active engagement with the content. Typical activities that are involved in the Zippy's friends program include: drawing, role‐playing, performing exercises, play and dialogue.

7. RESULTS OF SYSTEMATIC REVIEW

In addition to the newly identified studies ( n  = 88), primary evaluations ( n  = 53) discovered by Farrington and Ttofi ( 2009 ) are also included in the present systematic review, giving a total of 141 studies. However, this updated systematic review has excluded evaluations that used an “other” experimental‐control design ( n  = 13). Next, a detailed explanation is provided about studies which were excluded from the current review and justifications for this decision.

7.1. Studies excluded because of missing information

A certain amount of statistical information is needed in order to produce meaningful effect sizes in a meta‐analysis. We estimated an antibullying program's effectiveness as the difference between the experimental and control groups on bullying outcomes, either measured as the percentage of bullies/nonbullies or victims/nonvictims or based on mean scores on measurement instruments before and after implementation of the intervention.

However, 21 studies identified by our systematic review did not present sufficient effect size information, and so the primary authors of these publications were contacted. We were able to obtain relevant information for the majority of these studies, but three authors were unable to provide required statistics and seven did not respond to our email communication.

Thus, 10 studies had to be excluded from our meta‐analysis because of a lack of information regarding quantitative outcomes. These relate to: Gradinger et al. ( 2015 ); Harpin ( 2011 ); Kyriakides et al. ( 2014 ); Lewis et al. ( 2013 ); Lishak ( 2011 ); Low and Van Ryzin ( 2014 ); van der Ploeg et al. ( 2016 ); Sahin (2012); Schroeder et al. ( 2012 ); and Wurf (2010). In the previous review by Farrington and Ttofi ( 2009 ), 44 out of 53 evaluations provided sufficient information on quantitative outcomes.

7.2. Studies excluded because of nonindependent samples

One further stipulation of a meta‐analysis is that the final samples must be independent of one another (Borenstein et al.,  2009 ; Ellis,  2010 ). Overlapping samples are statistically dependent, and thus the variance of the summary effect size produced by the meta‐analysis would be under‐estimated (Wilson,  2010 ). Therefore, before conducting our meta‐analysis we ensured that all samples were independent of one another.

This issue of nonindependent samples was particularly relevant for the multiple evaluations of the KiVa antibullying program. Our thorough systematic searches identified 16 potentially includable studies presenting evaluation data from implementation of the KiVa program (i.e., Ahtola et al.,  2012 ,  2013 ; Garandeau, Lee, et al.,  2014 , Garandeau, Poskiparta, et al.,  2014 ; Haataja et al.,  2014 ; Hutchings & Clarkson,  2015 ; Kärnä et al.,  2011a ,  2011b ,  2013 ; Nocentini & Menesini,  2015 ; Noland,  2011 ; Sainio et al.,  2012 ; Salmivalli et al.,  2012 ; Williford et al.,  2012 ,  2013 ; Yang & Salmivalli,  2015 ). For a description of each of these studies, see Table  7 .

Description of KiVa studies

* Included in meta‐analysis.

However, following further screening, only four of the aforementioned studies were subsequently included in the systematic and meta‐analytic review (i.e., Kärnä et al.,  2011a ,  2011b ,  2013 ; Nocentini & Menesini, 2016). These four studies presented independent results of the KiVa program from the initial nationwide evaluation in Finland. Kärnä et al. ( 2011a ) used an age cohort design with adjacent cohorts and reported the initial results from the nationwide implementation in Finland. Second, Kärnä et al. ( 2011b ) reported the results from the RCT with Finnish students in grades 4–6, and Kärnä et al. ( 2013 ) reported results for students in grades 1–3 and 7–9. In addition, Nocentini and Menesini (2016) reported the results of the implementation and evaluation of KiVa in Italian schools. The remaining 12 publications relating to the KiVa program utilized data from the RCT evaluation in Finland (i.e., Kärnä et al.,  2013 or Kärnä et al.,  2011b ) but explored different facets of the program's effectiveness.

Four studies identified in our systematic searches replaced evaluations included in the earlier review. For example: (1) Menard and Grotpeter ( 2014 ) was a continuation of the Menard et al. ( 2008 ) evaluation; (2) Cross et al. ( 2011 ) was a republication of the Cross et al. ( 2004 ) evaluation included in the previous review; (3) Jenson et al. ( 2013 ) and Jenson et al. ( 2010 ) presented data from additional follow‐up points to the Jenson et al. ( 2007 ) evaluation; and (4) Frey et al. ( 2009 ) used an age cohort design to evaluate follow‐up effects from the earlier Frey et al. ( 2005 ) study. In cases such as these, the most recent publication, or the publication with the most statistical information, was included in the meta‐analysis.

Ten studies (published both before and since 2009) were identified as reporting the effectiveness of an antibullying program from the same sample, or were repeat publications of earlier studies (e.g., DeRosier,  2004 and DeRosier & Marcus,  2005 ; Domino,  2011 and Domino,  2013 ; Espelage et al.,  2013 and Espelage et al.,  2015 ; Jenson et al.,  2013 and Jenson et al.,  2010 ; and Menesini et al.,  2012 ; Study 2 and Palladino et al.,  2012 ). In these instances, the most recent publications were selected, and as a result, five studies were excluded from the meta‐analysis.

7.3. Included studies

Therefore, 128 studies are included. Table  5 summarizes the intervention programs and methodological components of the 79 newly identified studies that are included in the present systematic review. For details of the remaining 49 studies please refer to Farrington and Ttofi ( 2009 ).

7.4. Moderator analysis

The following moderators were selected a priori for further analysis, under the descriptive label (i.e., location of intervention, publication type, publication year), design label (i.e., evaluation method and unit of allocation/randomization), and the program heading (i.e., name of intervention, COI, and program specificity). Results of these moderator analyses analogous to the analysis of variance (ANOVA) are presented in Sections 8.5.1 to 8.5.7 of the present report.

7.4.1. Evaluation method

The primary moderator chosen for further analysis was evaluation method. Specifically, whether the evaluation was conducted using a RCT, quasi‐experimental with before and after measures (BA/EC) or age cohort (AC) design.

Overall, in relation to bullying perpetration outcomes, 36 evaluations used RCT designs, 31 used BA/EC designs and 14 used age cohort designs. However, due to some evaluations reporting data for multiple independent samples, a total of 40 effect sizes were estimated for bullying perpetration outcomes from RCT designs. A further 36 were estimated from BA/EC designs and 14 effect sizes came from evaluations using age cohort designs.

For bullying victimization outcomes, overall, 33 evaluations used RCT designs that gave 37 independent effect sizes for bullying victimization and 37 evaluations used BA/EC designs and gave 42 independent effect sizes. Similar to perpetration outcomes, 14 evaluations used age cohort designs to evaluate the effect of antibullying programs on bullying victimization outcomes.

7.4.2. Location of intervention

Evaluations included in the present analysis were conducted in many different countries around the world. However, there were only a few countries in which multiple evaluations of antibullying programs had been published.

Specifically, in the following countries only one evaluation was included in the present report: Austria (i.e., Yanagida et al.,  2019 ); Brazil (i.e., Silva et al.,  2016 ); China (i.e., Ju et al., 2009); Czechoslovakia (modern day Czech Republic and Solvakia; i.e., Rican et al.,  1996 ); Hong Kong (i.e., Wong et al.,  2011 ); Ireland (O'Moore and Milton,  2004 ); Malaysia (i.e., Yaakub et al.,  2010 ); Romania (i.e., Trip et al.,  2015 ); Sweden (i.e., Kimber et al.,  2008 ); Switzerland (Alsaker & Valkanover,  2001 ); South Africa (Meyer & Lesch,  2000 ); and Zambia (Kaljee et al.,  2017 ).

If these evaluations were to be included in further moderator analysis, we would be examining the differences based on only one sample and effect size. Therefore, moderator analysis was conducted only between locations in which multiple evaluations of antibullying programs had been conducted.

So, of the 100 evaluations included in our meta‐analysis of school‐based antibullying programs, the majority (80 for perpetration, 84 for victimization) were conducted in one of 12 different countries. With respect to bullying perpetration outcomes, these countries were as follows: Australia ( n  = 2); Canada ( n  = 6); Cyprus ( n  = 3); Finland ( n  = 6); Germany ( n  = 5); Greece ( n  = 2); Italy ( n  = 11); Netherlands ( n  = 3); Norway ( n  = 8); Spain ( n  = 3); UK ( n  = 4); and United States ( n  = 26). With respect to bullying victimization outcomes, these countries were as follows: Australia ( n  = 3); Canada ( n  = 7); Cyprus ( n  = 3); Finland ( n  = 6); Germany ( n  = 4); Greece ( n  = 2); Italy ( n  = 10); the Netherlands ( n  = 3); Norway ( n  = 7); Spain ( n  = 3); UK ( n  = 6); and United States ( n  = 28).

7.4.3. Publication type and year

Overall, the majority of evaluations were published in peer‐reviewed journal articles, for both bullying perpetration ( n  = 67) and bullying victimization ( n  = 72) outcomes. Two evaluations were published in chapters of edited books and both reported effects of a program on both bullying victimization and perpetration. No evaluations identified were published as entire books. Moreover, 12 unpublished dissertations were identified that published evaluation data for bullying perpetration and bullying victimization outcomes. Data was also retrieved for both outcomes from three governmental reports. Four of the effect sizes included in the present report were estimated from data emailed to authors (M. M. T. and D. P. F.) in preparation of the previous Campbell report (i.e., Farrington & Ttofi,  2009 ).

We also categorized included evaluations according to whether they were included in the previous report (i.e., “2009” studies), or only included in the present report (i.e., “2016” studies). In relation to bullying perpetration outcomes, 37 studies were coded as 2009 studies and 53 studies were coded as 2016 studies. Similarly, more studies were coded as 2016 ( n  = 54) studies in comparison to 2009 ( n  = 39) studies for bullying victimization outcomes.

7.4.4. Intervention program

We found that very few specific antibullying programs had been implemented and evaluated more than once using independent samples. Sixty‐five different school‐based bullying intervention and prevention programs were included in our meta‐analysis, but only eight were repeatedly evaluated. Moderator analysis with respect to the specific intervention program therefore, focused on programs that had been repeatedly evaluated.

In relation to reducing bullying perpetration outcomes the intervention programs thus included in our moderator analysis were: BPYS ( n  = 3; e.g., Menard & Grotpeter,  2014 ); fairplayer.manual ( n  = 2; e.g., Bull et al.,  2009 ); KiVa ( n  = 6; Kärnä et al.,  2011b ); NoTrap! ( n  = 4; e.g., Menesini et al.,  2012 ); Second Step ( n  = 3; e.g., Espelage et al.,  2015 ); Steps to Respect ( n  = 2; e.g., Frey et al.,  2005 ); ViSC ( n  = 5; e.g., Yanagida et al.,  2019 ).

Similarly, these interventions were included in our moderator analysis in relation to bullying victimization outcomes with the exception of the fairplayer.manual program. This intervention was evaluated twice only in relation to bullying perpetration outcomes.

Additionally, multiple evaluations of the OBPP were included in our meta‐analysis. Overall, 12 independent evaluations of this intervention were included in our analysis in relation to bullying perpetration and victimization outcomes. These are included in our moderator analysis as a collective subgroup and also as further subgroups. Evaluations of the OBPP conducted in the United States (perpetration n  = 6; victimization n  = 7) and those conducted in Norway (perpetration n  = 5; victimization n  = 5) were included in the moderator analysis separately. There was one evaluation of the OBPP conducted in Malaysia is included in the overall category ( n  = 12).

7.4.5. Unit of allocation/randomization

Systematic review findings showed that one consistent issue with included intervention programs was that the unit of allocation of participants, or clusters of participants, was different to the unit of analysis in most evaluations. Age cohort designs were omitted from this moderator analysis as the unit of allocation was largely unclear due to the logistics of this experimental design.

The majority of RCT and BA/EC evaluations assigned schools to experimental conditions (perpetration n  = 44; victimization n  = 47) yet the unit of analysis was individual students. A number of evaluations (perpetration n  = 19; victimization n  = 15) assigned classes to experimental conditions yet the unit of analysis was individual students. Less than 10 evaluations (perpetration n  = 7; victimization n  = 9) included assigned students to experimental and control conditions. One study randomly assigned districts to experimental conditions, and information was not available for five studies in relation to bullying perpetration outcomes and four studies in relation to bullying victimization.

7.4.6. Conflict of interest

In the present report, 40 studies were categorized as high COI. A large number of studies (perpetration n  = 36; victimization n  = 39) were considered low COI, and 14 were categorized as possible COI. Information concerning COI was unavailable for 4 evaluations in relation to bullying perpetration outcomes.

7.4.7. Program specificity

Overall, a small number ( n  = 11) of studies included in our analysis were coded as low on the program specificity variable. The vast majority of evaluations were considered highly specific (i.e., were mostly concerned with only bullying behavioral outcomes; n  = 59). Additionally, 18 studies were categorized as medium in relation to specificity, where extra outcome variables were measured but these variables were related to bullying (e.g., school climate).

7.5. Risk of bias analysis

Figure  2 presents the results of the risk of bias analysis for each of the items on the EPOC tool and the additional items we included. The following section describes each of these categories in more detail, with examples of high‐ and low‐risk studies included. The main limitation in assessing risk of bias was the lack of information reported by primary studies. Thus, while the best effort was made to categorize each primary evaluation as being high or low risk, a large number of studies were recorded as “unclear” risk.

Risk of bias analysis results. AC, allocation concealment; AS, allocation sequence; BC, baseline equivalence on participant characteristics; BE, baseline equivalence on outcomes; BOA, blind outcome assessment; COI, conflict of interest; CP, contamination protection; ID, incomplete outcome data; SOR, selected outcome reporting

As seen in Figure  2 , the fewest studies were considered unclear risk on CP and selected outcome reporting. Furthermore, a large number of studies were considered low risk on these items.

For the purpose of analysis, the categories high, unclear, and low risk were transformed into scores of 3, 2, and 0 respectively. A continuous “risk of bias” variable was then estimated as the sum total of scores on each of the EPOC items. As such, the lowest possible score a study could be given was zero and the maximum score was 24.

Descriptive statistical analysis identified that risk of bias scores ranged from 0 to 17, with a mean score of 9.62. Meta‐regression analysis was conducted to assess the relationship between risk of bias and effect sizes. The result of this analysis is included in Section  7 of this report. The following sections provide more detail about each of the risk categories.

7.5.1. Allocation sequence

AS refers to the way in which participants, or clusters of participants, were assigned to experimental conditions. For example, low‐risk studies were those where a random number generator or another randomization software was used. In total, 30 studies were categorized as high risk on the AS item. Moreover, 29 studies were low risk and 32 were unclear risk.

7.5.2. Allocation concealment

AC item refers to whether the method of allocation was concealed from participants or not. In total, 36 studies were categorized as high risk on the AC item. A further 19 studies were considered low risk, and 34 were unclear risk.

7.5.3. Baseline equivalence: Outcome

Baseline equivalence refers to the comparability of experimental and control participants before the intervention has taken place. This item specifically refers to equivalence on relevant outcomes, in this case, school bullying perpetration and victimization. When experimental and control participants are not statistically significant at baseline then we can be more certain that any changes are a result of the intervention. Overall, 14 studies were categorized as high risk on the baseline equivalence on bullying outcomes item. A total of 54 studies were low risk and 21 were unclear risk.

7.5.4. Baseline equivalence: Characteristics

Similarly, baseline equivalence on participant characteristics increases the chance that any change is a result of the intervention, and not a confounding variable such as differential participant characteristics at baseline. Overall, 15 studies were categorized as high risk on the baseline equivalence in participant characteristics item, 64 studies were low risk, and 11 were unclear risk.

7.5.5. Incomplete outcome data

Included evaluations were required to incorporate pre‐ and post‐intervention measures of bullying (except if randomization was used). However, because of this, it is likely that there will be some attrition in primary studies. The incomplete outcome data item referred to the risk associated with differential attrition between experimental groups and/or ways in which attrition and missing cases were dealt with by primary studies. Twelve studies were categorized as high risk on the incomplete outcome data item. Additionally, 48 studies were low risk and 29 were unclear risk.

7.5.6. Blind outcome assessment

This item assesses the risk associated with any bias which may arise if outcome measurements are not conducted blindly. In other words, if the individual, or individuals, who administer and collect the measurement instruments are aware of the experimental conditions of participants at the time of measurement. Overall, 27 studies were categorized as high risk on the BOA item. Twenty studies were low risk and 43 were unclear risk.

7.5.7. Contamination protection

Risk of contamination occurs when there is a possibility that experimental and control participants may interact or encounter one another during the course of the evaluation. Thus, the effects of the intervention may “spill over” to control students and impact the results of the evaluation. In our analysis, 35 studies were categorized as high risk on the CP item, 47 studies were low risk, and 9 were unclear risk.

7.5.8. Selective outcome reporting

SOR occurs when the outcomes reported in an evaluation study differ from the outcomes of interest proposed originally. For example, if a trial protocol proposed different outcomes than those actually reported in the publication of the trial results. Two studies were categorized as high risk on the SOR item. Eighty‐four studies were low risk, and three were unclear risk.

8. META‐ANALYSIS

After accounting for missing information, studies excluded because of their methodology (i.e., “other experimental‐control” designs), and studies with overlapping samples, a total of 41 studies were excluded from the meta‐analysis. Thus, a total of 100 studies were eligible for inclusion in our meta‐analysis. Table  8 outlines the raw data from these studies used to estimate effect sizes. The Comprehensive Meta‐Analysis (CMA) software was used to estimate all summary effect sizes in the present meta‐analysis.

Raw data from included evaluations

Abbreviations: A, after; B, before; C, control; E, experimental; M , mean; N , sample size; n , group sample size.

8.1. Effect sizes

A meta‐analysis aims to estimate comparable effect sizes from multiple primary studies. The choice of effect size depends on how statistical information is reported by primary studies (Borenstein et al.,  2009 ). In meta‐analyses such as this one, the data is largely presented in continuous (e.g., means, standard deviations, sample sizes) or dichotomous (e.g., prevalence or percentages) forms (Wilson,  2010 ). Thus, primary effect sizes estimated were Cohen's d and Odds Ratios.

As previously mentioned, we aimed to estimate one effect size for each independent sample included in primary studies. Therefore, where studies reported results separately for male and female participants, or primary and secondary school students, one effect size was calculated for each group.

For primary studies that presented results as percentages or frequencies of participants identifying as either bullies or victims, the odds ratio (OR) effect size was estimated. The ORs for before and after intervention time‐points were calculated independently. The CMA™ software that we used to analyze effect sizes in the present report did not allow us to enter raw data for before and after time‐points for primary studies that reported dichotomous outcomes separately. Thus, we were unable to use this software to calculate a pre‐post intervention estimate for these studies. Hence, these calculations were carried out manually, 5 by the first author, using the method outlined by Farrington and Ttofi ( 2009 ).

Cohen's d was estimated for primary studies when results were reported in the form of continuous data. Cohen's d is estimated as the difference between experimental and control means divided by the pooled standard deviation (Wilson,  2010 , p. 184). Effects were assigned a positive direction in cases where bullying was less in the experimental group compared to the control group or where the reduction in bullying outcomes was larger in the experimental group in comparison to the change in the control group. Following this logic, a negative effect was found when there was: (1) a larger reduction in the control group compared to the experimental group; or (2) there was no change or increase in bullying perpetration/victimization in the experimental group but a reduction or smaller increase in the control group.

For comparability, all effect sizes were converted to ODs. Summary mean effects for bullying perpetration, bullying victimization, and for each of the moderator subgroup are thus reported as odds ratios. In the present review, odds ratios greater than one represent a positive, or desirable, intervention effect. Namely, a reduction of bullying in the experimental group, that is comparably larger than the change in bullying in the control group. Therefore, the change is attributed to have occurred because of the intervention program. Similarly, odds ratios less than one represent a negative, or undesirable, intervention effect and odds ratios that equal one represents a null effect.

8.2. Corrections for clustering

As the present review aims to evaluate the effectiveness of school‐based antibullying programs, cluster‐randomized trials were included. Clustering is a common phenomenon in educational evaluations (Donner & Klar,  2002 ), and occurs when “clusters,” not individuals, are randomly assigned to experimental conditions (Higgins et al.,  2011 ). In other words, primary studies sometimes assigned classes or schools to intervention and control conditions, rather than individual students.

Often this approach is utilized in evaluation studies to reduce treatment contamination and increase administrative convenience (Donner et al.,  2001 ). However, one of the main issues with incorporating cluster‐randomized trials in a meta‐analysis is that participants within a cluster are likely to be more homogeneous than participants in another cluster (Higgins et al.,  2011 ). Thus, the variance of estimates of treatment effectiveness will be under‐estimated (Donner & Klar,  2002 , p. 2974). Clustering could occur for several reasons in studies included in the present report. For example: (1) classes of children, not individual children, were e randomized to intervention or control condition; (2) the intervention was implemented at the classroom level (i.e., to a class or group of children at one time); or (3) the intervention was targeted at teachers, who were trained to implement the intervention in their respective classrooms.

Therefore, effect sizes in the present meta‐analysis were corrected for the inclusion of clusters in primary studies. This is achieved by estimating a design effect:

where M represents the mean cluster size in each study (e.g., the mean number of students per classroom 6 ) and the ICC is the intraclass correlation coefficient.

The ICC is rarely reported by primary studies (Higgins et al.,  2011 ; Valdebenito et al.,  2018 ). Based on Murray and Blitse ( 2003 ), and subsequently the strategy followed by Farrington and Ttofi ( 2009 ), an ICC of 0.025 was assumed in the current meta‐analysis. The variances of effect sizes were then multiplied by this design effect estimated for each study. In the present meta‐analysis, there were only four studies where corrections for clustering were not required. Three studies (i.e., Berry & Hunt,  2009 ; Knowler & Frederickson,  2013 ; Meyer & Lesch,  2000 ) randomly assigned participants to experimental conditions, and Elledge et al. ( 2010 ) described an intervention that was not implemented in a classroom (i.e., the intervention occurred in one‐on‐one sessions with victims of bullying).

8.3. Computational models

The results of our meta‐analysis are presented using two different models. First, we will report the results as estimated using a random effects model that weights studies, largely in proportion to the between‐study variance and accounting for sampling error, thus allowing for the natural variation that occurs between primary studies (Borenstein et al.,  2009 ). We also present the results under the MVA model (Jones, 2005; Farrington & Welsh,  2013 ). which uses the same estimation of a mean effect size as the fixed effects model in that it assigns greater weight to larger evaluations, but also accounts for the between‐study heterogeneity. The MVA model takes account of the heterogeneity of effect sizes to fit the data exactly and yields the same mean effect size as a fixed effect model, but with and increased confidence interval. 7

Farrington and Welsh ( 2013 ) have argued that larger evaluations should be given more weight, and that adding to the variance of effect sizes in order to reduce the heterogeneity is not an optimal method of estimating the weighted mean effect size. When there is considerable heterogeneity in effect sizes, all studies tend to be given much the same weighting in a random effects model. Therefore, several effect sizes from independent samples in one study (e.g., a multisite evaluation) will have a greater weight in the random effects model than in the fixed effects model.

Comparing six models of estimating mean effect sizes for the impact on CCTV on crime rate, Farrington and Welsh ( 2013 ) found that five of the six models produced very similar mean odds ratio effect sizes, with the exception of the random effects model. In this case the random effects model estimated a much higher mean odds ratio (Farrington & Welsh,  2013 , p. 11).

The MVA model is suggested as an alternative approach that overcomes the issues of the random effects model. This technique can be seen as an adjustment to the fixed effects model and combines both the strengths of the fixed effects model (i.e., larger studies = larger weights) and the random effects model (i.e., adjusting for highly probable between‐study variance), and has been used in several meta‐analyses from both the behavioral sciences (e.g., Portnoy & Farrington,  2015 ; Ttofi et al.,  2016 ; Zych, Baldry, et al.,  2019 ; Zych, Viejo, et al.,  2019 ) and medical sciences, where this is known as the “Shore adjustment” (e.g., Ayieko et al.,  2014 ; Carlos‐Wallace et al.,  2016 ; Erren et al.,  2009 ; Steinmaus et al.,  2008 ).

A full review of the strengths and limitations of this model is beyond the scope of the current review. Therefore, in our current meta‐analysis we report mean effect sizes for the impact of antibullying programs on bullying perpetration and bullying victimization using both the random effects model and the MVA model. In later sections, we discuss the differences in the weighted mean effect sizes according to the model chosen.

8.4. Moderator analysis

In traditional empirical research when one wishes to compare two mean values to evaluate the difference between two participants, or two groups of participants, a t test is the standard statistical test. In meta‐analysis, we want to compare subgroups of studies rather than sub‐groups of individuals, so the analysis is slightly different. We followed guidelines provided by noted meta‐analysts for this type of analysis (Borenstein et al.,  2009 ; Lipsey & Wilson,  2001 ).

Our approach involved two steps: (1) computing the mean effect and variance for each subgroup; and (2) comparing the mean effects between subgroups (Borenstein et al.,  2009 , p. 152). This approach has been used previously by researchers to conduct similar analyses (e.g., Kaminski et al.,  2008 ; Ttofi & Farrington,  2011 ).

Comparing the mean effect sizes for subgroups involves a method that is analogous to a one‐way ANOVA in primary research (Hedges,  1982 ; Lipsey & Wilson,  2001 ; Wilson, 2002). The meta‐analyst creates mutually exclusive categories of primary studies and then compares the between‐studies ( Q B ) and the within‐studies ( Q W ) variance.

The between‐studies heterogeneity is the value used to evaluate whether the difference between subgroups is statistically significant (i.e., whether the difference in weighted mean effect sizes for subgroups is, at least partially, explained by the relevant intervention component). Similar to a one‐way analysis of variance, this approach partitions the variance and compares the variability between‐groups. The following formula is used to estimate the Q B :

The degrees of freedom for the between‐studies heterogeneity is estimated as j  − 1 and the statistical significance is determined using a χ 2 distribution. As Q B is estimated using the weights assigned to observed effect sizes, the value will vary between the fixed effects model and the random effects model. Q B is not reported for comparisons of subgroups with very unequal numbers of studies (e.g., location of the evaluation). Under the MVA model, the heterogeneity between groups is estimated by dividing the fixed effects Q B by Q/df . The present report presents results from moderator analysis under both the random effects and MVA models.

8.5. Meta‐regression analysis

CMA™ version 3 software was used to conduct meta‐regression analysis to explore the relationship between continuous moderator variables and perpetration and victimization outcomes. Weighted regression analysis (Lipsey & Wilson,  2001 ) were used to explore which moderators were independently related to school bullying perpetration and victimization. Meta‐regression analyses were only conducted for continuous moderator variables.

Meta‐regression analyses were computed under a fixed effects model, and the standard error of regression coefficients were adjusted using the MVA model. The Q and df of Q for the mean summary effect sizes for subgroups were used to adjust the standard error to reflect between‐study variance.

9. RESULTS OF META‐ANALYSIS

In total, 100 studies were included in our meta‐analysis of the effectiveness of school‐based antibullying programs. From these evaluations, we were able to estimate 103 independent effect sizes. These are presented for bullying perpetration and bullying victimization outcomes in Tables  8 and  9 , respectively. The majority of these effect sizes were estimated from studies that used RCT designs ( n  = 45 effect sizes) or BA/EC designs ( n  = 44 effect sizes). We estimated the remaining 14 effect sizes from age cohort designs.

Meta‐analysis results: School‐bullying perpetration outcomes

Abbreviations: BA/EC, before‐after/experimental control designs; CI, confidence intervals; MVA, multiplicative variance adjustment; OR, odds ratio; RCT, randomized controlled trial; Sig, statistically significant.

9.1. School‐bullying perpetration outcomes

Overall, we found that antibullying programs significantly reduced bullying perpetration under both computational models of meta‐analysis. The effect sizes for each evaluation are presented in Table  9 . The mean summary effect sizes were similar under both the multivariance adjustment model (MVA: OR = 1.324; 95% CI 1.27–1.38; z  = 13.4; p  < .001; I 2  = 81.42) and the random effects model (RE: OR = 1.309; 95% CI: 1.24–1.38; z  = 9.88; p  < 0.001; τ 2  = 0.044).

This result indicates that participants in primary studies who received an antibullying intervention were less likely to report engaging in bullying others after completing the program in comparison to control students who did not partake in the program.

Analysis of the funnel plot (Figure  3 ) suggests that publication bias is not present, as studies are symmetrically distributed around the mean effect size. In addition, point estimates did not vary using Duval and Tweedie's trim and fill procedure under a random effects model (in both cases: OR = 1.308; 95% CI 1.240–1.380). Based on these results, it was reasonable to assume that publication bias was not likely.

Publication bias analysis: school‐bullying perpetration

9.2. School‐bullying victimization outcomes

Overall, we found that antibullying programs significantly reduced bullying victimization under both computational models of meta‐analysis. The effect sizes for each evaluation are presented in Table  10 . The mean summary effect sizes were very similar under both the multivariance adjustment model (MVA: OR = 1.248; 95% CI 1.21–1.29; z  = 12.06; p  < .001; I 2  = 78.327) and the random effects model (RE: OR = 1.244; 95% CI: 1.19–1.31; z  = 8.92; p  < 0.001; τ 2  = 0.032).

Meta‐analysis results: School‐bullying victimization outcomes

This result suggests that students who participated in an antibullying program were significantly less likely to report being bullied by others after receiving the intervention in comparison to students who did not receive the intervention.

The funnel plot in Figure  4 indicates that no publication bias is present in analysis of bullying victimization effect sizes, as the studies fall symmetrically around the mean effect size. Duval and Tweedie's trim and fill procedure highlighted some minor differences between observed effect sizes (OR = 1.245; 95% CI 1.186–1.306; Q  = 460.97) and adjusted effect sizes (OR = 1.241; 95% CI 1.182–1.303; Q  = 473.43). However, this difference is negligible. Based on these results, it was reasonable to assume that publication bias was not likely.

Publication bias analysis: school‐bullying victimization

9.3. Analysis of heterogeneity

In a meta‐analysis, heterogeneity ( Q ) is the between‐study spurious variance that occurs partly because of true variation in effect sizes, but also as a result of random error (Borenstein et al.,  2009 ). Heterogeneity is estimated as the excess variation that exists when we compare the total amount of between‐study variance and within‐study random error.

In the present meta‐analysis, there was significant heterogeneity between studies for both bullying perpetration ( Q  = 323.392; df  = 85; p  < 0.001; I 2  = 73.716) and bullying victimization ( Q  = 387.255; df  = 87; p  < 0.001; I 2  = 77.534) outcomes. Multiple moderator analyses were conducted to explore possible explanations for this heterogeneity.

9.4. Risk of bias analysis

Scores on each of the risk of bias items were summed to estimate a total risk of bias score. This continuous variable was then used to examine the relationship between effectiveness and risk of bias in meta‐regression models.

For perpetration outcomes, risk of bias was not associated with effect size under a random effects model of meta‐regression ( b  = 0.003; SE  = 0.006; z  = 0.50; p  = .621) or under the MVA model ( b  = 0.014; SE  = 0.014; z  = 1.01; p  = .156). Similarly, risk of bias scores did not significantly predict bullying victimization effect sizes under a random effects meta‐regression ( b  = 0.007; SE  = 0.005; z  = 1.30; p  = .195) or the MVA model ( b  = 0.012; SE  = 0.012; z  = 1.006; p  = .157).

9.5. Moderator analyses 8

9.5.1. evaluation method.

Our meta‐analysis further investigated the effectiveness of antibullying programs in relation to the methodological designs used by evaluation studies. The breakdown of results by methodological design is also shown in Tables  9 and  10 for bullying perpetration and victimization outcomes respectively.

Primary studies employing age cohort designs associated with the largest effect sizes for both bullying perpetration (OR = 1.474; 95% CI, 1.39–1.56; p  < .001) and bullying victimization (OR = 1.302; 95% CI, 1.230–1.378; p  < .001) under a random effects model. Similarly, AC studies were associated with the largest effect sizes under the MVA model also (perpetration OR = 1.422; 95% CI, 1.36–1.46; p  < .001) and victimization OR = 1.289; 95% CI, 1.29–1.35; p  < .001).

Under the MVA model of meta‐analysis, mean effect sizes were the same for RCT evaluations (OR = 1.171; 95% CI, 1.08–1.27; p  < .001) and BA/EC evaluations (OR = 1.170; 95% CI, 1.05–1.31; p  = .005) for bullying perpetration outcomes. Moreover, the differences between RCT evaluations (OR = 1.117; 95% CI, 1.03–1.22; p  = .01) and BA/EC evaluations (OR = 1.188; 95% CI, 1.07–1.33; p  = .002) were marginal for bullying victimization outcomes under the MVA model.

In relation to bullying victimization outcomes, before‐after/experimental‐control designs gave the second largest mean effect size (OR = 1.225; 95% CI, 1.085–1.383; p  = 0.001), followed by RCTs (OR = 1.210; 95% CI, 1.091–1.342; p  < .001) under a random effects model. However, the result was the opposite for bullying perpetration outcomes under a random effects model (RCT: OR = 1.244; 95% CI, 1.123–1.379; p  < .001; BA/EC: OR = 1.187; 95% CI, 1.044–1.350; p  = 0.009).

Due to the marginal differences and lack of clear pattern in which method was associated with the largest effect sizes (between RCT and BA/EC) further moderator analysis was not conducted.

9.5.2. Location of intervention

Mean effects for bullying perpetration and bullying victimization outcomes are presented graphically in Figures  5 and  6 , respectively. Table  11 outlines the mean effects for each of the 12 countries for both bullying perpetration and victimization outcomes under both the MVA model and the random effects model.

Forest plot of effect size by location: school‐bullying perpetration

Forest plot of effect sizes by location: school‐bullying victimization

Moderator analyses results: Location of evaluation

Evaluations conducted in Greece were associated with the largest effect sizes for bullying perpetration outcomes, followed by Norway, Italy, United States, and Finland under the MVA model of meta‐analysis. Evaluations conducted in Italy were associated with the largest mean effect sizes in relation to bullying victimization, followed by Spain, Norway, United States, and Finland under the MVA model of meta‐analysis. Additionally, evaluations conducted in Germany and the UK gave significant mean effects when computed using the MVA model.

Under the random effects model, Greek evaluations were similarly associated with the largest effect sizes for bullying perpetration, followed by Spanish and Norwegian evaluations. Evaluations conducted in Italy and the United States were also associated with significant mean effects for reductions in bullying perpetration. In relation to bullying victimization, evaluations conducted in Spain and Italy were associated with very similar mean effect sizes and were the largest of the 12 effect sizes, followed by evaluations conducted in Norway. Evaluations conducted in Australia were also associated with significant mean effects in reducing bullying victimization ( p  < .05) and evaluations conducted in Finland and the United States were nearly statistically significant ( p  = .05 and p  = .06, respectively) under the random effects model.

Due to the large number of different countries and the unequal number of studies in each location, further subgroup analyses were not conducted.

9.5.3. Publication type and year

Table  12 outlines the mean summary effect sizes for each of the publication type moderators for bullying perpetration and victimization outcomes. Evaluations for which data was received via email correspondence from evaluators gave the largest mean effect sizes for both bullying perpetration and bullying victimization. Differences in the mean effect sizes for evaluations reported via unpublished dissertations, either masters or doctoral theses, gave the smallest mean effect sizes for both bullying perpetration and victimization outcomes. Subgroup analysis was not conducted further using these categorizations due to the imbalance in numbers of evaluations in each category (i.e., evaluations were overwhelmingly published in peer‐reviewed journal article format).

Moderator analyses results: Publication type

However, additional analysis was conducted to examine any potential differences between peer reviewed and nonpeer reviewed evaluations. Therefore, the above categories were collapsed, and evaluations reported by dissertation, chapter, correspondence and governmental reports (perpetration n  = 23; victimization n  = 21) were compared to evaluations published via peer‐reviewed journal article.

Under the MVA model, non‐peer‐reviewed evaluations gave a larger (OR = 1.493; 95% CI, 1.266–1.761; p  < .001) mean effect size than peer‐reviewed evaluations (see Table  11 ). Moreover, moderator analysis indicated that the difference was statistically significant ( Q B  = 12.861; df  = 1; p  < .001). However, under the random effects model, both groups gave similar effect sizes for bullying perpetration outcomes, and the difference between peer‐reviewed (see Table  11 ) and non‐peer‐reviewed (OR = 1.309; 95% CI, 1.137–1.508; p  < .001) was not statistically significant ( Q B  = 0.595; df  = 1; p  = .441).

For bullying victimization outcomes, similar results were obtained. Under the MVA model, non‐peer‐reviewed evaluations gave statistically significant larger mean effect sizes (OR = 1.403; 95% CI, 1.262 1.560; p  < .001) than peer‐reviewed evaluations (see Table  11 ; Q B  = 27.197; df  = 1; p  < .001). Yet, there was a marginal difference under the random effects model between peer‐reviewed (see Table  11 ) and non‐peer‐reviewed (OR = 1.231; 95% CI, 1.059–1.431; p  = .007) and the difference was not statistically significant ( Q B  = 0.048; df  = 1; p  = .827).

The mean summary effect size for “2009” studies on the year of publication moderator was OR = 1.487 (95% CI, 1.430–1.546; p  < .001) under the MVA model and OR = 1.411 (95% CI, 1.315–1.513; p  < .001) under the random effects model for bullying perpetration outcomes. Across both computational models these summary effects were larger than those for studies labeled “2016” on bullying perpetration for the MVA model (OR = 1.243; 95% CI, 1.667–1.324; p  < .001) and the RE model (OR = 1.184; 95% CI, 1.087–1.289; p  < .001). Moderator analysis analogous to the ANOVA showed that this difference was statistically significant ( Q B  = 76.412; df  = 1; p  < .001) under fixed effects and mixed effects analysis ( Q B  = 9.676; df  = 1; p  = .002).

In relation to bullying victimization, the mean summary effect size for studies labeled “2009” was larger (OR = 1.322; 95% CI, 1.220–1.432; p  < .001) under the MVA model than the mean summary effect size for studies labeled “2016” (OR = 1.229; 95% CI, 1.175–1.285; p  < .001). Moderator analysis analogous to the ANOVA found that this difference was statistically significant ( Q B  = 10.115; df  = 1; p  = .001) but the difference between odds ratios was marginal. However, under the random effects model the minimal difference between the “2009” studies (OR = 1.215; 95% CI, 1.094–1.350; p  < .001) was not statistically different to the mean summary effect size for “2019” studies (OR = 1.223; 95% CI, 1.139–1.313; p  < .001; Q B  = 0.010; df  = 1; p  = .920).

9.5.4. Intervention program

The mean summary effect sizes for 10 different intervention programs in relation to reducing bullying perpetration behaviors and 9 different intervention programs in relation to reducing bullying victimization behaviors. Table  13 outlines the effectiveness of specific antibullying programs in reducing both school‐bullying perpetration and victimization. The effectiveness of these programs varied greatly.

Moderator analyses results: Intervention program

In relation to school‐bullying perpetration outcomes, the OBPP was associated with the largest mean effect sizes. In addition, evaluations of the OBPP in Norway were associated with larger summary effect sizes than evaluations of OBPP conducted in the United States. However, the difference was not statistically significant for school‐bullying perpetration outcomes when moderator analysis analogous to the ANOVA was conducted ( Q b  = 3.65; df  = 1; p  = 0.06).

Other programs were significantly effective in reducing school‐bullying perpetration behaviors, for example KiVa, Second Step, and Steps to Respect. Positive effect sizes (i.e., OR > 1) were also observed for the BPYS and NoTrap! programs but these effects were not statistically significant in relation to reduction in bullying perpetration outcomes. Negative effects were found for two antibullying programs, the fairplayer manual and ViSC, although these effects were not statistically significant.

In relation to school‐bullying victimization outcomes, NoTrap! was associated with the largest mean effect size, followed by the BPYS Program, and then the OBPP. Our analysis identified that other antibullying programs were also significantly effective in reducing school‐bullying victimization, for example, Steps to Respect and KiVa.

Again, effect sizes for the OBPP varied between evaluations conducted in Norway and evaluations conducted in the United States for bullying victimization outcomes. Moreover, our analysis found that the difference in the magnitude of these effect sizes was statistically significant ( Q b   =  74.95; df  = 1; p  < 0.001). Our analysis also identified negative effects of the Second Step program in relation to bullying victimization outcomes. Evaluations of the ViSC program also had a negative effect on bullying victimization, although this effect was not statistically significant.

9.5.5. Unit of allocation/randomization

Table  14 outlines the mean effects for subgroups of studies according to how participants were allocated to experimental or control groups. Results are presented for bullying perpetration and victimization outcomes for all studies that allocated studies in classes, schools, or individual students. The mean effects for RCT and BAEC for each allocation unit are also presented separately.

Moderator analyses results: Unit of allocation/randomization

In relation to bullying perpetration outcomes, under the MVA model, studies that assigned participants in classes were associated with the largest effect sizes. However, the difference between the mean effect for all evaluations that used classes or schools as the unit of allocation were verging on statistically significance ( Q b   =  3.705, df  = 1, p  = .054). Under the random effects model, evaluations that assigned students to experimental conditions were associated with the largest effect size for bullying perpetration outcomes when all designs were included, and for RCT evaluations and BA/EC evaluations individually. However, the mean effect size for many of the subgroups were not collectively statistically significant overall under the random effects model.

Similarly, under the MVA model, evaluations conducted using a RCT design, and assigned classes to conditions, were associated with the largest effect size for bullying perpetration, although the mean group for this subgroup was not statistically significant. Moreover, moderator analysis analogous to the ANOVA found that the difference in the mean effect size for RCT designs that assigned classes to experimental and control conditions were not statistically different to RCT designs that assigned schools to experimental and control conditions ( Q b   =  1.140, df  = 1, p  = .286 ) .

In relation to BAEC designs, evaluations that assigned students to experimental conditions were associated with the largest mean effect size, although the effect was not statistically significant. However, the difference between the mean effect for BAEC evaluations that assigned classes and those that assigned schools to conditions was statistically significant under the MVA model ( Q b   =  4.551, df  = 1, p  = .033).

For bullying victimization outcomes, studies where the unit of allocation was classes of participants were associated with the largest effect sizes, followed by schools and individual students under the MVA model. The difference between studies that allocated classes and studies that allocated schools was statistically significant ( Q b   =  12.450, df  = 1, p  < .001). This pattern was observed when all designs were included, and for the subgroup of RCT evaluations and the subgroup of BA/EC evaluations. Thus, when participants were assigned in classes the mean effect size for these RCT evaluations were significantly associated with larger effect sizes ( Q b   =  13.590, df  = 1, p  < .001) for reductions in bullying victimization than RCT evaluations that assigned schools. Yet the difference between the mean effect sizes for BA/EC evaluations that assigned classes were not statistically significant ( Q b   =  3.359, df  = 1, p  = .067) than BA/EC evaluations that assigned schools to experimental conditions.

9.5.6. Conflict of interest

COI was a categorical moderator variable with three levels: high‐risk (H), low‐risk (L), and possible‐risk (P). Moderator analysis analogous to the ANOVA was conducted so as to assess the differences between evaluations on each level. Studies categorized as possible‐risk on COI variable were excluded from subgroup comparisons to establish the differences between evaluations that were clearly high‐risk and evaluations that were clearly low‐risk. Table  15 outlines the mean summary effects for each group for both bullying perpetration and bullying victimization outcomes.

Moderator analyses results: Conflict of interest

Note : Four studies and six studies were excluded from the present moderator analysis for perpetration and victimization outcome respectively as not enough information was available.

Moderator analyses found that the difference between high‐risk and low‐risk studies on COI variable was statistically significant for bullying perpetration outcomes under both the MVA model ( Q B  = 50.129; df  = 1; p  < .001) and the random effects model ( Q B  = 4.900; df  = 1; p  = .027). This suggests that evaluations considered to have high COI were associated with larger overall effect sizes for bullying perpetration. Similarly, high‐risk COI studies were significantly associated with slightly larger effect sizes for bullying victimization in comparison to low‐risk COI studies when compared under both the MVA model ( Q B  = 16.127; df  = 1; p  < .001) and the random effects model ( Q B  = 4.449; df  = 1; p  = .035).

9.5.7. Program specificity

The majority of evaluations included in our meta‐analysis were of highly specific intervention programs, that is, those that targeted bullying behaviors and no other outcomes. Consistently across computational model and both perpetration and victimization outcomes these subgroups were associated with the largest mean effect sizes. These results are presented in Table  16 . Additionally, highly specific programs were the only subgroup of evaluations that gave a statistically significant mean summary effect under both the MVA model and the random effects model for bullying victimization outcomes. In relation to bullying perpetration outcomes, the subgroup of evaluations that were coded as “medium” on the program specificity moderator were associated with a statistically significant mean effect size under the MVA model ( p  < .001) and the random effects model ( p  = .036).

Moderator analyses results: Program specificity

10. DISCUSSION

10.1. summary of main findings.

Overall, our updated meta‐analysis found that school‐based antibullying programs are effective in reducing both school‐bullying perpetration and victimization. For school‐bullying perpetration the weighted mean OR = 1.324 under the MVA model, or OR = 1.309 under a random‐effects model (RE) were associated with reductions of approximately 19–20%. 9 In comparison, the weighted mean ORs for bullying victimization outcomes were 1.248 and 1.242 under the MVA model and the random effects model respectively. These mean effect sizes correspond to an approximate reduction in bullying victimization of 15–16%. These results suggest that the included interventions were slightly more effective at reducing school‐bullying perpetration than school‐bullying victimization.

The results of this meta‐analysis are consistent with findings from most of previous reviews that indicate that antibullying programs have a small but significant effect, with some variations in overall results being attributable to methodological differences in inclusion and exclusion criteria (Ttofi et al.,  2014 ). Our mean effect sizes are also consistent with the earlier review (Farrington & Ttofi, 2009 ; Ttofi & Farrington,  2011 ), although the differences further outline that moderator variables such as methodological design may be responsible for variability. For example, the weighted mean effect sizes for both bullying perpetration and bullying victimization outcomes estimated in the earlier Campbell report were larger than those estimated in the present report.

Yet, we included publication year as a categorical moderator variable in the present analysis. We found that more recent studies (i.e., those that were not included by Farrington & Ttofi,  2009 ) were significantly different to studies that were included in the earlier review. Namely, recent studies were actually associated with significantly larger effect sizes for both bullying perpetration and victimization outcomes (see Section 8.5.3).

Therefore, as we excluded studies considered to have utilized less scientifically rigorous methodological designs this may explain the differences in the weighted mean effect sizes. Specifically, we excluded evaluations conducted using “other experimental‐control designs,” described in the earlier review as evaluations in which participants were assigned to experimental and control conditions but bullying outcomes were only measured after implementation of the intervention. Thus, attributing any change in behaviors to the intervention is potentially risky because there may be other reasons why a positive effect of the intervention was observed. For example, the experimental and control groups were not comparable at baseline, but this remains unknown as no measure of bullying was obtained.

Thus, the inclusion of these less methodologically rigorous evaluations may explain why the weighted mean effects sizes reported in the earlier review were larger than those reported in the current report, but our moderator analysis found a contradictory pattern. The following sections of this report will aim to discuss the findings obtained by our moderator analyses and also the strengths and limitations of the current analysis and potential avenues for future research. The heterogeneity in this meta‐analysis was very large for both bullying perpetration and victimization outcomes. This may suggest that there was a wide range of effects across programs and we may not be able to explain differences using moderator analysis.

10.2. Moderator analyses

10.2.1. evaluation method.

Under both the MVA and random effects models, evaluations conducted using age cohort designs were identified to be, collectively, the most effective, or at least associated with the largest mean effect sizes. This is consistent with Farrington and Ttofi's ( 2009 ) review. This methodological design was first introduced as an evaluation design for the OBPP (Olweus,  1991 ). This approach has been criticized for the potential threats to internal validity, history and testing effects (Farrington & Ttofi,  2009 , p. 15). It has been suggested that this design avoids the threats of aging and maturation effects, as individuals within the same school act as a control group for same‐aged experimental participants (Olweus,  2005a ). However, this design is vulnerable to cross‐contamination between experimental and control participants which would impact the overall effectiveness. Notably, intervention researchers have tested the OBPP with other methodological designs (e.g., Bauer et al.,  2007 ) which resulted in smaller effects.

Interestingly, the pattern between RCTs and BA/EC designs was less clear. In relation to bullying victimization outcomes, evaluations using BA/EC designs appear to be more effective than evaluations using RCT designs. However, for bullying perpetration outcomes, evaluations using RCT designs appear to be more effective than evaluations that utilized BA/EC designs. Further research is needed to understand these effects. However, the nature of these analyses is correlational and the differences between effect sizes are marginal. Thus, no concrete conclusion can be drawn in relation to the association between randomized and nonrandomized quasi experimental designs and effect size in the present context.

10.2.2. Unit of allocation/randomization

In theory, RCTs are the best method of evaluation of interventions because random allocation ensures that any observed differences between experimental and control groups occurs as a result of experimental manipulation, thus giving the best possible internal validity (Farrington,  1983 ,  2003 ). However, the unit of random allocation can have an impact on internal validity. For example, we assume that individuals are randomly assigned to experimental and control conditions, so that RCT designs adequately account for the random variation that occurs in real‐world research (Weisburd,  2003 ).

However, in practice, evaluations of antibullying programs may be more likely to assign groups of individuals, for example in terms of classrooms or schools, to experimental conditions rather than individual students. This is true for both randomized (e.g., classrooms, Chaux et al.,  2016 ; or schools, Espelage et al.,  2015 ) and nonrandomized (e.g., classrooms, Ortega‐Ruiz et al.,  2012 ; or schools, Rawana et al.,  2011 ) methodologies. When this is the case, we need larger numbers to ensure adequate statistical conclusion validity and avoid issues of selection effects and differential attrition (Farrington & Ttofi,  2009 ; Ttofi & Farrington,  2011 ). There was a lot of variation in the unit of allocation in our primary studies, which may explain why we did not find that one methodological design was more effective than another.

Moreover, the majority of included evaluations did not use the same unit for allocation and analysis, thus, posing a threat to our results. We approach the results therefore with caution, favouring more conservative estimates. Furthermore, the relationship between the unit of randomization/allocation moderator variable and the effect sizes for bullying perpetration and victimization outcomes was unclear. Whether or not the differences between subgroups of evaluations that assigned classes or schools to experimental conditions were statistically significant or not depended on the computational model used and the bullying outcome in question. For bullying perpetration, the differences between studies based on unit of allocation were not statistically significant for randomized and nonrandomized studies. For bullying victimization outcomes, studies where classes were the unit of allocation were associated with the largest effect sizes when all designs where included and for randomized evaluations, but not for nonrandomized evaluations, separately.

Risk of bias analysis also found that a large number of RCT studies were categorized as being high risk for allocation‐related items on the EPOC tool. Therefore, the differences observed between primary evaluations in our meta‐analysis may be due to the observation that largely the unit of allocation and the unit of analysis were not the same in primary studies. However, further analysis and investigation is needed to better understand these results.

10.2.3. Location of intervention

Overall, the results of our meta‐analysis are consistent with previous findings and show that school‐based antibullying programs have a modest but significant effect in reducing bullying behaviors. However, our meta‐analysis included evaluations of antibullying programs from a wide range of countries and specific intervention programs, far more than previous meta‐analyses (e.g., Cantone et al.,  2015 ; Chalamandaris & Piette,  2015 ; Evans et al.,  2014 ; Jiménez‐Barbero et al.,  2012 ,  2016 ). As a result, the results of this meta‐analysis are robust and have implications for bullying research globally.

Our analysis identifies that antibullying programs worldwide are effective in reducing school‐bullying perpetration and victimization by significant amounts. Moreover, evaluations in different countries appear to vary in effectiveness. In Greece, where evaluations included in our meta‐analysis were associated with the largest effect sizes, school‐bullying perpetration behaviors were reduced by approximately 40%. Evaluations conducted in the Norway, Italy and the United States were also effective in reducing bullying perpetration by approximately 21–25%.

Antibullying programs implemented and evaluated in Italy were associated with the largest reduction in school‐bullying victimization in our meta‐analysis, with the odds ratio effect size corresponding to an approximate reduction of 31%. Moreover, evaluations conducted in Spain and Norway reduced school‐bullying victimization by approximately 28% and 23%, respectively. Evaluations conducted in Finland, Germany and the UK were also significantly effective, although less so, reducing school‐bullying victimization by approximately 8–12%.

There are many potential explanations for the differences in effectiveness observed between countries. For example, definitions of school‐bullying, and behaviors that constitute bullying, differ between countries. Previous research conducted by Smith et al. (2000) showed that school‐bullying is perceived differently across different countries and cultures and this may explain variability in bullying reporting. Definitions of school bullying, and behaviors that constitute bullying, differ between countries. For example, Smith et al. ( 2016 ) showed that school bullying in Eastern cultures manifests more often as exclusion or isolation of an individual victim. In comparison, school bullying in Western cultures comprises a wider range of physical, verbal and relational forms of aggression.

Our meta‐analysis included several examples of cases where the same intervention program was evaluated in different countries (e.g., KiVa program in Finland (Kärnä et al.,  2013 ) and in Italy (Nocentini & Menesini, 2016)). While societal practices, educational systems, and individual lifestyles may differ greatly, some argue that there may be some support for the cross‐national applicability of specific intervention programs. However, there is a current lack of existing research comparing the effectiveness of specific interventions in specific countries.

Previous research has indicated that are also cultural differences in bullying behaviors among adolescents (e.g., Smith et al.,  2016 ). As such, an antibullying program to reduce these behaviors may be impacted by these differences. This is particularly evident when we observe the variations in effect sizes for the OBPP (Olweus,  1993 ) and the KiVa antibullying program. These programs may be the most well‐known antibullying programs that are commercially available, and as such as the only examples in our review of interventions evaluated in completely different locations.

The OBPP program was originally designed and implemented in Norway, and it is therefore not surprising that the OBPP program appears to be effective in reducing both school‐bullying perpetration and victimization when evaluated in Norway, compared to evaluations in the United States (see Table  13 ). While the program was still significantly effective in the United States, the percentage decrease in school‐bullying perpetration was roughly 25% and in school‐bullying victimization was roughly 11%. These figures are lesser in comparison to the decreases in bullying behaviors seen in Norwegian evaluations (35% perpetration; 29% victimization). These differences could be attributed to different evaluation methodologies (see Gaffney et al., 2019), however, they most likely reflect cultural and societal differences between youth in Norway and youth in the United States.

Interestingly, the opposite is observed with the KiVa program. When KiVa was evaluated in Finnish samples, the program was effective in reducing school‐bullying perpetration by approximately 4–5% and school‐bullying victimization by approximately 6% (Kärnä et al.,  2011a ,  2011b ,  2013 ). However, when evaluated in Italian primary and secondary schools, the effect sizes were much larger. Nocentini and Mensini (2016) found that KiVa was effective in reducing school‐bullying perpetration by approximately 15–20% and school‐bullying victimization by approximately 25%.

In the case of KiVa, each of the evaluations used the same methodology (i.e., RCT), but varied greatly in the sample size. Thus, further research is needed to explain why some interventions (e.g., OBPP or KiVa) appear to be more effective in some samples compared to others. The programs are still effective, but the variation in effect size could be attributable to a number of different methodological and implementation factors that warrant further exploration.

10.2.4. Intervention program

Following this logic, we also explored the effectiveness of the specific antibullying programs. Out of the four most widely disseminated antibullying programs included in our review (i.e., KiVA, NoTrap!, OBPP, ViSC), the OBPP was collectively the most effective in reducing school bullying perpetration of these. Across 11 evaluations, the OBPP reduced bullying perpetration by approximately 26%, which was larger than any other widely disseminated program.

In relation to school‐bullying victimization outcomes, the NoTrap! program was the most effective, reducing victimization by around 37%. NoTrap! also reduced bullying perpetration by a considerable amount, approximately 22%, but this effect was not statistically significant. The KiVA program, significantly reduced school bullying perpetration by approximately 9% and school bullying victimization by approximately 11%. The ViSC program was the only program to increase bullying perpetration (by roughly 4%) and bullying victimization (by roughly 4%) although these effects were not statistically significant.

Another moderator we used to code differences between included evaluations was the specificity of the intervention program. In other words, we evaluated each intervention program on how specific it related to bullying behaviors. Unsurprisingly, our findings suggest that antibullying programs gave the largest overall effect sizes. While the significance of the differences between subgroups was not computed due to the large discrepancies between the numbers of evaluations included in each subgroup.

However, our inclusion criteria for the current report was strictly concerned with school‐bullying intervention programs and behavioral outcomes of bullying. As such, we may have overlooked effective programs that only included nonbehavioral outcomes of bullying (e.g., attitudes toward bullying, awareness of bullying) or other problem behaviors (e.g., peer aggression or victimization, mental health issues, juvenile delinquency, etc.) that occur among young people in schools. Changes in these behaviors may also impact bullying, either directly or indirectly, yet, more research is needed to understand this potential effect. Most obvious in the present report is how programs that target specifically school‐bullying may impact cyber‐bullying, and vice versa, given the significant overlap in the prevalence of these behaviors (Baldry et al.,  2017 ).

Further research is also needed to better understand specifically “what works” in these “specific interventions.” In the previous review, (Farrington and Ttofi  2009 ; Ttofi & Farrington,  2011 ) conducted detailed coding of interventions and evaluations and analyzed how effect sizes varied between components and features of primary studies. For example, parent training, playground supervision, and more intense and longer programs were significantly correlated with larger reductions in bullying perpetration (Ttofi & Farrington,  2011 ). Moreover, several intervention components were associated with larger reductions in bullying victimization (e.g., videos, disciplinary methods, co‐operative group work and more intense and longer programs). Therefore, an important avenue for future research is to assess the differences in effectiveness of antibullying programs according to specific intervention components across the 100 evaluations included in our meta‐analysis. Such research would have important implications for policy and the development of future antibullying programs.

Additionally, it appears that since 2009 several large‐scale antibullying programs have been implemented and evaluated (e.g., KiVa; Kärnä et al.,  2013 ; NoTrap!; Menesini et al.,  2012 ; Palladino et al.,  2016 ). Because there is typically more information available on the specific components of these programs, we may be able to code more specific details in future analyses. For example, many studies may fit the criteria for “parent training,” but there is a significant difference between the intensity of parental involvement. For example, some studies may include parents merely by sending letters home with participant children (e.g., Brown et al.,  2011 ), while others include parents more actively by holding information evenings or requiring children to complete take‐home tasks with parental involvement (e.g., Berry & Hunt,  2009 ; Domino,  2013 ).

Earlier research highlighted how varying levels of implementation of each intervention component may explain variability in intervention outcomes (Bloom et al.,  2003 ). Interestingly, a narrative review by Smith et al. ( 2003 ) reported that although 14 whole‐school antibullying programs obtained modest effects overall, those that monitored implementation obtained twice the mean effects on self‐reported rates of bullying and victimization than those that did not monitor implementation. Thus, additional analyses are required to better understand specifically what works in existing antibullying programs and the underlying mechanisms of behavioral change

10.2.5. COI and publication type

Possibly the most conclusive results from our moderator analyses were observed in relation to COI and publication type. First, across both computational models and outcomes, studies that were categorized as being high‐risk for COI were associated with significantly larger reductions in bullying perpetration and victimization. Second, under the MVA model of meta‐analysis, non‐peer‐reviewed evaluations were associated with significantly larger reductions in both bullying perpetration and victimization outcomes. However, the same results were not observed under the random effects.

We examined COI in terms of the involvement of the program developer in the evaluation. Our results may indicate possible sources of biases. For example, it may be that when the individual, or team, that are credited with developing an antibullying program are also involved in the evaluation of said intervention, biases such as confirmation bias may impact the results. However, it may not be a perceivably “negative” source of bias. Perhaps, when the program developer is involved in the implementation of the program, the intervention is simply delivered better and more effectively. There are a number of other factors that could also be affected and in turn impact the effect size, such as teacher and staff efficacy and motivation to participate the in the program.

There are more sophisticated measures of COI (e.g., Eisner et al.,  2012 ) that include elements such as whether or not the evaluator could potentially benefit financially from the intervention program. Further indicators of COI are thus needed to better understand the impact on evaluation results. For example, our findings in relation to COI and larger effect sizes may be explained as: evaluations in which the program developer was included appear to be more effective because of the expertise and intricate knowledge of the developer. Therefore, the results may reflect differences in the quality of program implementation rather than troublesome biases. Additional research is needed.

10.3. Limitations and avenues for future research

Like most meta‐analyses, the current report is largely limited by the lack of understanding as to what is the “true effect.” When comparing mean effect sizes between moderators for example, it is difficult to determine the validity of the result. Throughout our discussion of result we discuss that one subgroup of studies was associated with larger or smaller effect sizes than another, and the statistical significance of these differences. Thus, we avoid saying studies in subgroup A (e.g., evaluations conducted in Greece) are more effective than studies in subgroup B (e.g., evaluations conducted in Italy). Due to the correlational nature of our moderator analyses we cannot make causal inferences. In addition to this limitation, and those previously discussed (Section  9.2 ), the following section of this report discusses some further limitations.

10.3.1. Measurement of bullying

Experts in the area of school‐bullying research have outlined how there still remain issues of comparability in the assessment of school‐bullying perpetration and victimization (Volk et al.,  2017 ). Studies included in the present meta‐analysis used a wide variety of quantitative measures of school‐bullying behaviors, including self‐report measures (e.g., the Revised Olweus Bully/Victim Questionnaire—Olweus,  1986 ,  1996 ), or peer‐report measures (e.g., the Participant Role Questionnaire—Salmivalli et al.,  1996 ). One issue that arises is that the timeframe within which participants are required to indicate the frequency of bullying can vary greatly. One scale may ask about bullying experiences within the last 3 months, while another may ask about ever having experienced, or participated in, school‐bullying. Moreover, included studies utilized a mixture of continuous or dichotomous measures of school‐bullying, and the cut‐off points used to categorize someone as either a bully, victim, or not‐involved also varied.

Furthermore, the majority of evaluations included in our analysis reported bullying outcomes at different time points, largely, before implementation, after implementation, with a possible additional follow‐up time point. However, we computed effect sizes using measures of bullying taken before implementation and immediately post implementation of the intervention. Therefore, we cannot generalize results to the long‐term effectiveness of antibullying programs, or any potential influence of dose‐response effect. Future research should aim to examine the longitudinal effectiveness of interventions to reduce bullying perpetration and victimization in the long‐term.

When conducting our systematic searches for the present review, we did not set restrictions based on measurement issues, other than including quantitative measures of school‐bullying behaviors. However, types of reports, for example, could influence the overall effectiveness effect size. This may possibly explain why our meta‐analysis found that programs are more effective in reducing bullying perpetration outcomes. For example, if programs are concerned with raising awareness about bullying and the associated negative impact on victims, participants who reported bullying perpetration before the intervention may be less likely to self‐report bullying behaviors after completing the program. As a result, the intervention may be perceived as being effective, but the change in reports of bullying may have been a result of social desirability responding (He et al.,  2015 ; Rigby & Johnson,  2006 ). Conversely, raising awareness on the negative impact of school bullying may lead to increased reporting of victimization due to sensitization effects (Stevens et al.,  2000 ). Notably, sensitization effects due to raised awareness may affect not only self‐report data but also peer nomination data and teacher reports (Smith et al.,  2003 , p. 597). Therefore, future research could aim to examine whether the style of report used, differing cut‐off points and varying timeframes affect estimations of intervention effectiveness.

10.3.2. Cyberbullying behaviors

Another key limitation of the present review is the omission of cyberbullying behaviors. Prominent researchers in the area have argued that cyberbullying behaviors do not warrant a completely separate line of study, because of the significant overlap between offline and online bullying (Olweus & Limber,  2017 ). A recent meta‐analysis of cyberbullying intervention and prevention programs found that, out of studies assessing various facets of cyberbullying, a large number were concerned with this overlap (Gaffney et al., 2019). The Gaffney et al. (2019) meta‐analysis concluded that anticyberbullying programs were effective in reducing cyberbullying perpetration by roughly 9–15% and cyberbullying victimization by roughly 14–15%. As illustrated in that other review, there is a need for future research to assess the effectiveness of intervention programs that target both online and offline bullying concurrently. As a result of the significant overlap (e.g., Waasdorp & Bradshaw, 2015), it is important for policy makers, researchers, and program developers to know whether or not these forms of aggressive behaviors should be targeted together or individually. Future research should aim to examine the effectiveness of programs designed to reduce school‐bullying on cyberbullying outcomes, and vice versa. Additional analysis to examine the differences between programs that target offline and online behaviors concurrently in terms of effectiveness to reduce both school‐ and cyber‐bullying is also needed.

10.3.3. Models of meta‐analyses

The current report presents findings using two computational models of meta‐analyses: the random effects model and the multiplicative variance adjustment model. While, the random effects model is often suggested as the preferred model for meta‐analyses in social sciences, for reasons already discussed (Section  7.3 ), this approach is also limited. However, even though many meta‐analyses in medical sciences (e.g., Ayieko et al.,  2014 ; Dorjee et al.,  2018 ; Woolf‐King et al.,  2013 ) have used the MVA model as an alternative method of accounting for between‐study heterogeneity in weighted mean effect sizes, this model is yet to be widely accepted in behavioral sciences. A number of recent publications (e.g., Portnoy & Farrington,  2015 ; Zych et al.,  2019 ) have begun to use the MVA model.

It is evident in the current report that the results are influenced by the computational model used. The overall mean effect sizes for bullying perpetration and victimization were not that different under both models but the results of moderator analyses were greatly influenced by how we accounted for the between‐study heterogeneity. Further research is needed in order to examine the reasons for this and also evaluate how best to choose an appropriate computational model when conducting a meta‐analysis.

10.4. Concluding remarks

This report presents an updated systematic and meta‐analytical review of the effectiveness of school‐bullying intervention and prevention programs. Overall, our review found that school‐based antibullying programs are effective in reducing both bullying perpetration and bullying victimization, and that effect sizes can vary according to several moderator variables. However, further research is needed to better understand the reasons for variation in observed effect sizes. Research is needed to investigate the specific components of antibullying programs that work best to reduce bullying behaviors. The results of our meta‐analysis have important implications for policy and the development of future antibullying programs, but future research should aim to better understand the effective mechanisms in bullying intervention and prevention.

11. TECHNICAL APPENDICES

11.1. calculating the before‐after intervention effect.

Williams et al. ( 2015 ) evaluated the effectiveness of the Start Strong program based on students' self‐reported experiences of bullying victimization. The primary study found that, at baseline, 23% of participants in the experimental group ( N  = 717) reported bullying victimization, while 23% of participants in the control group ( N  = 800) also reported bullying victimization at baseline. Hence, the baseline OR was calculated as follows (Table  17 ):

Data used to estimate baseline odds ratio

Thus, the OR before  = 0.999, Ln OR before  = −0.002, and var Ln OR before  = 0.015. Williams et al. ( 2015 ) report that after implementation of the Start Strong program, bullying victimization was reported by 28% of experimental participants and 34% of control participants. Accordingly, the posttest OR was calculated as follows (Table  18 ):

Data used to estimate postintervention odds ratio

Thus, the OR after  = 1.323; Ln OR after  = 0.28; and var Ln OR after  = 0.013. Employing these figures, the ln OR for the intervention effect of the Start Strong program was calculated as:

The ln OR change is computed as the difference between the before and after effect size and the variance of this new estimate is adjusted by multiplying the sum of the variances of before and after variances by 0.75. This is an approximation of the assumed correlation between before and after effect sizes. The ln OR change and the SE of ln OR change were then entered into CMA as an estimation of the intervention effect.

11.2. Multiplicative variance adjustment

In the present meta‐analysis, the summary effect size estimated for bullying perpetration was OR = 1.324 with 95% confidence intervals of 1.298–1.351 under a fixed effects model. The effect size in the MVA model is the same as the effect size in the fixed effects model. The variance of the effect size in the MVA model is calculated as follows:

Therefore, in the above example of the summary effect size for bullying perpetration outcomes, the FE var is 0.000104. Therefore, with Q  = 458.555 and df  = 109, the MVA adjustment for fixed effects is 0.02098, calculated as:

Therefore, the adjusted standard error is 0.0209. In this example thus, the MVA fixed effect is OR = 1.324, and the 95% confidence intervals are 1.271–1.380.

11.3. Odds ratio to percentage conversion

The conversion from weighted mean odds ratio to percentage value is also described in the previous Campbell report (see Farrington & Ttofi, 2009 ). The formula involves assuming equal allocation of participants to experimental and control conditions and that the % of bullies and/or victims was lesser in the experimental condition than in the control condition (as supported by our overall positive mean effect size).

For example, if there are 200 participants in each experimental condition and approximately 30% of participants report bullying victimization in the control condition and 25% victims in the experimental condition, the numbers of victims and nonvictims would be as follows: (Table  19 ).

Data used to convert odds ratio to percentage

Therefore using the previously described formula for estimating an odds ratio, the following data would correspond to an odds ratio of 1.286 (i.e., [150 × 60]/[140 × 50]). Moreover, the percentage decrease would be approximately 16.67% (i.e., (10/60) × 100).

Using this basic formula, we can manipulate the % and number of victims in each experimental condition in order to achieve a odds ratio that corresponds to our weighted mean effect size (i.e., MVA: OR = 1.324 and RE: OR = 1.309 for bullying perpetration; MVA: OR = 1.248 and RE: OR = 1.242 for bullying victimization). Using the n values that give the closest possible mean effect size we can thus estimate the corresponding percentage reduction in either bullying perpetration or victimization outcomes.

APPENDIX A. 

Appendix: full search syntax, database: web of science.

Bully* AND Intervention AND Evaluation

Anti‐Bullying AND School AND Program* AND Evaluation

Anti‐Bully* AND Program* AND Outcome

Bully‐victim AND Prevention AND Evaluation

Bully* AND School AND Intervention

Bully* AND School AND Prevention

Database: Scopus

Bully* AND School AND Program*

Bully* AND School AND Evaluation

Bully* AND School AND Intervention AND Evaluation

Bully* AND School AND Prevention AND Evaluation

Anti‐bullying AND Program* AND Evaluation

Database: National Criminal Justice Reference Service

Bully* AND Prevention AND Evaluation

Anti‐bullying AND Program* AND Effect*

Database: PsycINFO

Bully* AND Intervention AND Program* AND Evaluation

Bully* AND Prevention AND Program* AND Effect*

Database: Cochrane Controlled Trials Register

Bully* AND Intervention AND Program*

Bully* AND Prevention AND Program AND Evaluation

Database: British Education Index

Bully* AND Prevention AND Program* AND Evaluation

Bully* AND Intervention AND Program* AND Effect*

Database: Embase

Database: medline, database: eric & criminal justice abstracts.

www.scholar.google.co.uk

APPENDIX B. 

Appendix: risk of bias results for included studies.