Research Proposal – Evaluating a Whole-School Cyberbullying Intervention

Introduction

Cyberbullying has become a routine feature of adolescent life in Australia rather than an occasional aberration. The eSafety Commissioner, the national online safety regulator established under the Online Safety Act 2021, reports that roughly one in seven young Australians aged 12 to 17 experienced cyberbullying in a twelve-month period, and that most affected students did not formally report it (eSafety Commissioner, 2024). Australian longitudinal research shows that traditional and online victimisation frequently co-occur and track together across the secondary years, so cyberbullying is best understood as an extension of the peer aggression that also plays out in the schoolyard rather than a wholly separate phenomenon (Hemphill et al., 2015).

The harms are well documented and are not trivial. Victimisation is associated with depression, anxiety, self-harm, school avoidance and reduced academic engagement, and several of these associations persist after adjustment for prior mental health (Australian Institute of Health and Welfare [AIHW], 2024). Because online aggression can follow a young person home, repeat through sharing, and reach a large audience quickly, its reach and permanence can intensify distress relative to offline forms. National wellbeing data indicate that adolescents already report rising psychological distress, so any additional burden from online victimisation lands on a population that is under strain (AIHW, 2024).

Australian schools are expected to respond. Online safety and respectful relationships are embedded in the Health and Physical Education learning area of the Australian Curriculum (Australian Curriculum, Assessment and Reporting Authority [ACARA], 2022), and the legacy of the National Safe Schools Framework continues to shape the whole-school wellbeing policies that most schools now maintain. Whole-school approaches, which coordinate curriculum, peer and classroom practices, school policy and family engagement rather than relying on a single lesson, are the internationally endorsed model for bullying prevention (Campbell & Bauman, 2018). Yet Australian secondary schools still lack rigorous local evidence on whether a coordinated whole-school program reduces cyberbullying and, importantly, whether it does so by changing what bystanders do.

Problem Statement

Two gaps motivate this proposal. First, although meta-analyses show that school-based programs can reduce bullying, the average effects are modest, highly variable, and weaker for cyberbullying and for secondary students than for primary-aged children (Gaffney et al., 2019). Much of the strongest evidence derives from primary settings and from overseas jurisdictions, and the small number of Australian secondary-school trials leaves practitioners extrapolating from contexts with different curricula, platforms and reporting systems. Second, the mechanism through which whole-school programs are thought to work, namely the mobilisation of the bystander majority to intervene constructively, is frequently assumed but seldom measured directly in Australian evaluations (Salmivalli, 2010; Polanin et al., 2012). Without measuring bystander behaviour as an outcome in its own right, an evaluation cannot distinguish a program that genuinely shifts peer culture from one that produces only a short-lived change in self-reported victimisation. This proposal addresses both gaps by evaluating a whole-school program in Australian secondary schools with bystander behaviour as a co-primary outcome.

Aim and Research Questions

The study aims to evaluate the effect of an eighteen-month whole-school cyberbullying intervention on victimisation and bystander behaviour among Year 8 students in Australian secondary schools, using a quasi-experimental controlled design. Three research questions are posed:

1. Does the whole-school intervention reduce self-reported cyberbullying victimisation from baseline to post-intervention and follow-up, relative to comparison schools?
2. Does the intervention increase constructive bystander (defending) behaviour and reduce passive or reinforcing bystander responses over the same period?
3. Do changes in bystander behaviour account for changes in victimisation, and are any effects maintained at six-month follow-up?

Literature Review

Socio-ecological understandings of bullying

Contemporary bullying scholarship rejects the view that aggression is simply a property of individual “bullies” and locates it instead within nested social systems. Ecological theory holds that a young person’s behaviour is shaped simultaneously by individual characteristics, peer and classroom dynamics, the school as an organisation, and family and community influences, including the online environments in which adolescents now spend substantial time (Espelage, 2014). The practical implication is that interventions targeting only one level, such as a stand-alone curriculum unit, tend to be overwhelmed by contrary influences at the others. Whole-school programs are the applied expression of this theory, seeking coherence across levels so that classroom messages are reinforced by supervision practices, school policy, staff modelling and parent communication (Campbell & Bauman, 2018).

Bystander behaviour as a lever for change

Most cyberbullying occurs before an audience of peers who are neither the target nor the primary aggressor, and the responses of these bystanders shape whether an incident escalates or is contained. Salmivalli’s (2010) participant-role framework distinguishes assistants and reinforcers, who amplify aggression, from outsiders, who withdraw, and defenders, who support the target or challenge the behaviour. Increasing the proportion of defenders is theorised to remove the social reward that sustains bullying. Meta-analytic evidence supports bystander behaviour as a malleable and worthwhile target: programs that explicitly train bystander skills produce measurable increases in intervention behaviour, although effects on witnessing peers are uneven (Polanin et al., 2012). In the online context, defending can take low-risk forms such as reporting content, supporting the target privately, or refusing to forward material, which are teachable and consistent with the safety mechanisms of major platforms and the cyberbullying complaints scheme administered by the eSafety Commissioner.

Whole-school programs and the quality of the evidence

The most rigorous Australian contribution is the Cyber Friendly Schools program, a cluster-randomised trial in Western Australian secondary schools that produced short-term reductions in cyber-victimisation and perpetration, with effects that attenuated once active implementation ceased (Cross et al., 2016). This finding, that gains depend on sustained delivery, recurs across the international literature and underlines the importance of a delayed follow-up measurement point. More broadly, updated meta-analysis confirms that anti-bullying programs work on average but that effect sizes are small and are moderated heavily by implementation fidelity, program intensity and student age (Gaffney et al., 2019). Two limitations of the existing evidence base are salient for the present design. Many trials measure only victimisation and perpetration, omitting the bystander mechanism, and many rely on immediate post-tests that cannot establish whether change endures. A design that measures bystander behaviour directly and includes a delayed follow-up therefore addresses recognised weaknesses rather than merely replicating prior work.

Conceptual Framework

The intervention is grounded in a socio-ecological program logic. Each component of the program is mapped to a level of the student’s social ecology, and all components are theorised to converge on two proximal outcomes, more constructive bystander behaviour and lower victimisation, which in turn support the distal outcomes of improved wellbeing and school connectedness. A feedback path recognises that delivery is iterative: fidelity and outcome data are returned to the school team so that activities can be adjusted, consistent with the continuous-improvement orientation of Australian school wellbeing policy. Figure 1 presents the model.

Methodology

Design

A quasi-experimental controlled design is proposed, with three intervention schools and three matched comparison schools measured at three time points: baseline (T0), post-intervention (T1) at the end of active delivery, and follow-up (T2) approximately six months later. Random allocation of schools was considered but judged impractical at the required scale, and participating systems prefer to nominate their readiness to implement; schools are therefore allocated non-randomly and matched, which defines the design as quasi-experimental. Matching on observable characteristics and statistical adjustment for baseline differences are used to strengthen causal inference in the absence of randomisation.

Setting and participants

Six co-educational secondary schools across metropolitan and regional New South Wales and Victoria will be recruited, paired on sector, socio-educational advantage as indexed by the Index of Community Socio-Educational Advantage, enrolment size and baseline cyberbullying prevalence, with one school in each pair allocated to the intervention. The participant cohort is the Year 8 population, chosen because victimisation peaks in the early secondary years and because a single-cohort focus allows the program to be embedded within one curriculum stage. With an average of approximately 150 Year 8 students per school, the target sample is around 900 students. This sample provides adequate power to detect small individual-level effects, a standardised difference of about 0.30, on victimisation and bystander outcomes, while the modest number of clusters limits power for school-level contrasts, a constraint that is addressed in the analysis and acknowledged as a limitation.

Intervention

The eighteen-month program operationalises the framework in Figure 1 through four coordinated components. At the individual level, six curriculum lessons aligned to the Health and Physical Education learning area (ACARA, 2022) build knowledge of online safety, help-seeking and coping. At the peer and classroom level, structured activities rehearse constructive bystander responses and establish prosocial norms. At the school level, staff professional learning, a reviewed anti-bullying policy and consistent response procedures embed the program in daily practice, consistent with the whole-school orientation inherited from the National Safe Schools Framework. At the family and online level, parent information sessions and links to eSafety Commissioner resources extend the messages beyond the school gate. Comparison schools continue their usual wellbeing provision, and implementation fidelity is monitored through delivery logs and structured observation.

Measures

Constructs are assessed with validated self-report instruments wherever available, supplemented by investigator-developed items, as summarised in Table 1. The student survey is estimated to take 25 minutes and is administered under teacher supervision on school devices.

Table 1: Constructs, instruments and their role in the analytic model

Note. The PECK was developed and validated with Australian school students, supporting the local relevance of victimisation estimates (Hunt et al., 2012).

Procedure and analysis

Surveys are administered at T0, T1 and T2 using a secure platform hosted on Australian servers, with a self-generated matching code enabling longitudinal linkage without recording names. Analysis follows intention-to-treat principles. Because measurement occasions are nested within students and students within schools, three-level linear mixed models are specified, with time, condition and their interaction as fixed effects and random intercepts for students and schools; the time by condition interaction tests the intervention effect for Research Questions 1 and 2. Research Question 3 is examined through multilevel mediation, estimating whether the effect on victimisation operates through change in bystander behaviour. Missing data are handled with full-information maximum likelihood, and sensitivity analyses probe the influence of attrition and of clustering.

Ethical Considerations and Safeguarding

The project will be submitted for full review by the administering university’s Human Research Ethics Committee and conducted in accordance with the National Statement on Ethical Conduct in Human Research (National Health and Medical Research Council [NHMRC], 2023). Research with minors on a sensitive topic is not treated as low risk, and four safeguards are built into the design.

First, consent is layered. Written consent is sought from a parent or guardian for each student, and students provide their own informed assent on the day, with a clear statement that they may skip any question or withdraw without consequence. Second, a written distress and referral protocol operates throughout data collection. Support information for Kids Helpline, headspace and Lifeline appears on the survey and on a take-home card, and any student who shows distress is supported by a nominated school wellbeing staff member through the school’s existing pathways. Where a response or disclosure indicates risk of harm, staff follow mandatory reporting obligations and, for serious online abuse, the reporting mechanisms of the eSafety Commissioner. Third, because the survey asks about recent victimisation, items are framed carefully to avoid prompting rumination, and no student is identified to peers as a participant of interest. Fourth, data are stored de-identified on encrypted Australian servers, the linkage key is held separately and destroyed after T2, and results are reported only at group level so that individuals and small schools cannot be identified.

Project Timeline

The project runs over eighteen months, as set out in Table 2. Ethics approval gates all data collection, and the follow-up wave is timed for approximately six months after active delivery concludes so that maintenance of effects can be assessed.

Table 2: Eighteen-month project timeline by phase, activity and month

Significance

The study offers three contributions. Empirically, it will provide Australian secondary schools with controlled evidence on whether a coordinated whole-school program reduces cyberbullying, addressing the shortage of local trials at this level. Methodologically, by measuring bystander behaviour as a co-primary outcome and testing it as a mediator, the study moves beyond counting incidents to examine the peer-culture mechanism that whole-school approaches are meant to activate (Polanin et al., 2012; Salmivalli, 2010). Practically, findings will inform how schools, curriculum authorities and the youth wellbeing sector, including services such as headspace to which affected students are commonly referred, invest in prevention, and the socio-ecological program logic in Figure 1 offers a transferable template for schools designing their own responses.

Limitations

Several limitations are acknowledged in advance. Non-random allocation leaves the design vulnerable to selection effects, mitigated but not eliminated by matching and statistical adjustment. Six schools provide limited power at the cluster level, so school-level moderators are treated as exploratory rather than confirmatory. All outcomes are self-reported and therefore subject to social-desirability and recall biases, although anonymity within the linkage system reduces this risk. The single-cohort, single-region focus supports internal coherence but limits generalisation to other year levels and states, and to the rapidly changing online platforms on which cyberbullying occurs. Finally, an eighteen-month window captures maintenance at six months but cannot establish longer-term durability.

Conclusion

Cyberbullying is a common and consequential experience for Australian secondary students, and whole-school approaches are the recommended response, yet local controlled evidence and direct measurement of the bystander mechanism remain scarce (Cross et al., 2016; Gaffney et al., 2019). By evaluating a coordinated whole-school program across six schools, with victimisation and bystander behaviour as co-primary outcomes and a delayed follow-up, the proposed study aims to establish not only whether the program works but how it works. The intention is to give Australian schools evidence they can act on, and to treat online harm among young people with the seriousness that national data and the experiences of students warrant.

References

Australian Curriculum, Assessment and Reporting Authority. (2022). Australian Curriculum: Health and Physical Education (Version 9.0). ACARA.

Australian Institute of Health and Welfare. (2024). Australia’s youth: Bullying and mental health. AIHW.

Campbell, M. A., & Bauman, S. (2018). Reducing cyberbullying in schools: International evidence-based best practices. Academic Press.

Cross, D., Shaw, T., Hadwen, K., Cardoso, P., Slee, P., Roberts, C., Thomas, L., & Barnes, A. (2016). Longitudinal impact of the Cyber Friendly Schools program on adolescents’ cyberbullying behaviour. Aggressive Behavior, 42(2), 166-180.

Del Rey, R., Casas, J. A., Ortega-Ruiz, R., Schultze-Krumbholz, A., Scheithauer, H., Smith, P., Thompson, F., Barkoukis, V., Tsorbatzoudis, H., Brighi, A., Guarini, A., Pyzalski, J., & Plichta, P. (2015). Structural validation and cross-cultural robustness of the European Cyberbullying Intervention Project Questionnaire. Computers in Human Behavior, 50, 141-147.

eSafety Commissioner. (2024). The digital lives of young Australians. Australian Government.

Espelage, D. L. (2014). Ecological theory: Preventing youth bullying, aggression, and victimisation. Theory Into Practice, 53(4), 257-264.

Gaffney, H., Ttofi, M. M., & Farrington, D. P. (2019). Evaluating the effectiveness of school-bullying prevention programs: An updated meta-analytical review. Aggression and Violent Behavior, 45, 111-133.

Goodman, R. (1997). The Strengths and Difficulties Questionnaire: A research note. Journal of Child Psychology and Psychiatry, 38(5), 581-586.

Hemphill, S. A., Tollit, M., Kotevski, A., & Heerde, J. A. (2015). Predictors of traditional and cyber-bullying victimisation: A longitudinal study of Australian secondary school students. Journal of Interpersonal Violence, 30(15), 2567-2590.

Hunt, C., Peters, L., & Rapee, R. M. (2012). Development of a measure of the experience of being bullied in youth. Psychological Assessment, 24(1), 156-165.

National Health and Medical Research Council. (2023). National statement on ethical conduct in human research. NHMRC.

Polanin, J. R., Espelage, D. L., & Pigott, T. D. (2012). A meta-analysis of school-based bullying prevention programs’ effects on bystander intervention behaviour. School Psychology Review, 41(1), 47-65.

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