The rationale for targeting both substance use and gender beliefs in a single intervention rests on a growing body of evidence that the two behaviors are deeply intertwined. Across Southern Africa, an estimated 20 to 40 percent of young people report using at least one psychoactive substance, often as a coping strategy against unemployment, fragile health systems, political instability, and climate-related stressors. Early substance use elevates lifetime risks of mortality and morbidity, and is linked to HIV acquisition, traffic accidents, intimate partner violence, unemployment, and progression to substance use disorders. Meanwhile, the internalization of traditional gender norms — what researchers call inequitable gender beliefs — is established early in adolescence and reinforced throughout youth, and it frequently co-occurs with substance use. Both are independently associated with intimate partner violence, sexually transmitted infections, and poor mental health, suggesting that they may share underlying social drivers that a well-designed intervention could address together rather than one at a time.

What makes PEGISUS distinctive is its insistence on treating the peer group, not the individual, as the unit of change. Participants are recruited through an index case who enrolls two to five of their own peers with whom they already use substances — specifically, people who report weekly substance use or at least monthly binge drinking with peers over the previous three months. This deliberate social architecture reflects decades of research showing that peers play a decisive role in the initiation and escalation of substance use, and that most existing interventions, which focus narrowly on individual behavior, neglect the social and environmental forces that sustain risky patterns. Recent reviews suggest peer-focused mental health interventions have rarely been tested at all in low- and middle-income countries, and a prior South African friendship-group study showed decreases in certain gender-related practices and alcohol use only among female groups, leaving a substantial evidentiary gap that this new trial aims to fill.

The trial’s setting within vocational training programs is equally consequential. In South Africa, 35 percent of youth aged 15 to 24 are neither employed nor in school; in Zimbabwe, the figure is 36 percent, and in Zambia 29 percent. Vocational training offers an alternative pathway to economic opportunity for young people who have not completed formal schooling, but substance use and inequitable gender beliefs may undermine those benefits through their associations with dropout, unemployment, and violence. The study partners — Build It International in Zambia, Masvingo Polytechnic in Zimbabwe, and the Sozo Foundation in South Africa — offer training in fields ranging from bricklaying and welding to hairstyling and coding. By embedding the intervention into these programs and having program staff deliver it, the researchers hope to solve two chronic problems in global mental health: the enormous treatment gap, with an estimated 95 percent of young people needing substance use care not receiving it, and the sustainability problem of interventions that vanish once research funding ends.

Technically, the intervention builds on two established evidence-based programs: RAD-PAL, a brief substance use intervention, and Manhood 2.0, a gender equity curriculum. The eight sessions, each roughly ninety minutes, are activity-based and delivered to entire peer groups together, covering substance use, gender, and building healthy relationships. Formative work shaped the content through twelve focus group discussions with 97 young people, feedback from youth advisory boards in each country, and a pilot with 18 peer groups involving 57 participants. Interventionists — employees of the vocational training programs themselves — undergo training in good clinical practice, motivational interviewing, group facilitation, and handling distressed participants, supported by two levels of ongoing supervision including clinical oversight designed to prevent burnout among non-specialist providers. All sessions are audio-recorded, and fidelity is assessed with a session-specific checklist rating each component on a three-point scale.

The trial’s methodological rigor is a selling point in itself. Peer groups are randomized 1:1, stratified by country, using concealed allocation in REDCap, with the comparison arm receiving the standard vocational training program plus a referral for substance use services. Assessments occur at baseline, 12 weeks, and 24 weeks, with the primary outcome being the between-group difference in change on the total substance use involvement score of the World Health Organization’s Alcohol, Smoking and Substance Involvement Screening Test, or ASSIST. Younger participants aged 16 and 17 complete an adapted version, the ASSIST-Y. Secondary outcomes include substance-specific ASSIST scores, changes in gender equitable beliefs measured by the 23-item Gender Equitable Men’s Scale, and — notably — vocational training attendance and completion rates, employment status, and income earned at 24 weeks. The sample size calculations assume a standardized net mean difference of −0.5, drawn from a meta-analysis of brief behavioral interventions for youth substance use, an intra-class correlation of 0.08, and an average cluster size of four, yielding 95 percent power at a two-sided significance level of 0.05.

Analysis will follow CONSORT guidelines and modified intention-to-treat principles using linear mixed-effects models with cluster-specific random effects, adjusted for baseline scores, sex, and age. Superiority will be demonstrated if the lower bound of the 95 percent confidence interval for the between-group mean difference exceeds zero. An exploratory interaction test will probe whether intervention effects differ across the three countries, though the researchers caution that such tests are typically underpowered. Sensitivity analyses with multiple imputation are planned if more than roughly five percent of primary endpoints are missing. The team’s prior trials with similar populations report attrition below 10 percent, aided by rapport-building with community-based research assistants who remain blinded to intervention arm.

The ethical architecture of the study reflects the sensitivity of working with minors, illegal substance use, and gender-related content in contexts where same-sex relations remain criminalized. Participants under 18 require caregiver consent before providing assent, and parental consent forms deliberately frame the study around general substance use risk rather than disclosing individual selection criteria, to reduce the risk of family conflict or violence. Screening occurs separately for individuals within peer groups to ensure voluntary participation free of peer coercion, and group ground rules plus facilitator training aim to manage disagreements that may arise when discussing gender beliefs in mixed groups. In Zambia and Zimbabwe, participants are asked not to disclose sexual or gender minority identities, with interventionists trained to navigate these topics with cultural sensitivity while prioritizing participant safety and dignity within local legal frameworks. Adverse events must be reported within 24 hours and assessed against standardized criteria, with referrals available to local clinicians at every site.

If PEGISUS proves effective, the implications could ripple far beyond the 264 participants. The design tackles socioeconomic adversity head-on by pairing behavioral change with tangible employment pathways, and its delivery model — run by existing program staff within institutions that already serve the target population — offers a plausible route to scale and sustainability that most clinic-based interventions lack. The researchers acknowledge limitations, including potential homophily from index participants recruiting similar peers, the impossibility of blinding participants in a behavioral intervention, and a six-month follow-up window too short to capture longer-term employment effects. Yet the trial represents the first known study in Southern Africa to test an integrated intervention for substance use and gender equity beliefs delivered through vocational training. For a continent where more than half the population is under 24, and where young people’s coping mechanisms are shaped by structural forces as much as personal choices, the outcome of this trial may help determine whether the next generation of prevention efforts treats the whole social world of young people — their friends, their beliefs, and their economic futures — as one inseparable system.

Subject of Research: A cluster randomized controlled trial testing a peer group intervention to reduce substance use and inequitable gender beliefs among young people in vocational training programs in Southern Africa

Article Title: Protocol for a cluster randomized controlled trial among young people and their peers to reduce substance use and inequitable gender beliefs in Southern Africa

Article References: Protocol for a cluster randomized controlled trial among young people and their peers to reduce substance use and inequitable gender beliefs in Southern Africa. (n.d.). https://doi.org/10.1186/s13722-026-00718-7

Image Credits: AI Generated

DOI: 10.1186/s13722-026-00718-7

Keywords: substance use, gender norms, peer intervention, young people, vocational training, cluster randomized trial, Southern Africa, Zambia, Zimbabwe, South Africa, ASSIST, gender equity

The intellectual lineage of the study reaches back to the 1960s, when Bibb Latané and John Darley conducted their now-legendary experiments on bystander intervention in the United States. Sparked in part by public outrage over the murder of Kitty Genovese in New York — a case later revealed to have been badly misreported — their work identified diffusion of responsibility as a central mechanism of inaction: the more witnesses present, the less any single individual feels personally obligated to help. Their Adult Decision Tree to Determine Helping became a foundational framework in social psychology, describing a sequence of judgments about noticing an event, interpreting it as an emergency, assuming responsibility, knowing how to help, and deciding to act. What remained strikingly underexplored, however, was whether this model transfers to adolescents navigating the morally and socially treacherous terrain of school bullying.

The timing of the two research traditions is itself telling. Bullying first emerged as an internationally recognized research topic in the 1970s, propelled by Dan Olweus’s pioneering work in Scandinavia on what he originally termed ‘mobbing’ among schoolchildren. Olweus went on to develop the Bullying Circle, a model that maps the distinct roles students occupy in bullying episodes — from the bully and the victim to defenders, supporters of the bully, passive onlookers and outsiders. Meanwhile, American experimental psychology was independently building its science of adult emergency helping. For decades these two literatures ran largely in parallel. The new study deliberately brings them together, asking whether the decision architecture mapped for adults in emergencies also describes how teenagers reason about bullying — a chronic, socially embedded form of victimization rather than a sudden, unambiguous crisis.

To answer that question, the researchers conducted a qualitative, phenomenological, in-depth case study at a large public high school in the Midwest — a Class A institution serving more than 885 students. Fifty-five adolescents aged 14 to 19 participated. Rather than relying on a single instrument, the team triangulated their data across participant essays, individual interviews and focus groups, all centered on students’ lived experiences of bullying and bystander behavior. Participants were introduced to the Olweus Bullying Circle and asked to discuss the roles students play in bullying situations, and their responses were then systematically compared against Latané and Darley’s adult decision model. The phenomenological approach, grounded in established traditions of qualitative inquiry, sought to capture not just what students did but how they experienced and made sense of those moments from the inside.

The results show that adolescents do indeed traverse territory familiar from the adult literature, but with important additions. Like adults, students in the study reported considering whether intervening would expose them to negative judgment, what methods of intervention were available to them, whether their actions would actually be effective, and whether they could act without endangering themselves. These four considerations — judgment, methods, efficacy and safety — parallel the classic decision tree’s stages of assessing the situation, assuming responsibility and choosing a course of action. In this sense, the core logic of helping behavior identified in adults appears to be recognizably present in teenage cognition when bullying unfolds before their eyes.

Yet the adolescents’ decision calculus was demonstrably richer and, in certain respects, more constraining than the adult model predicts. Three additional dimensions surfaced repeatedly in student accounts: social apprehension, social safety, and a concentration of responsibility. Social apprehension captures the acute fear of reputational damage, ridicule or social backlash that teenagers face when they break ranks with the peer group — a pressure that operates far more intensely in adolescent social ecologies, where status and belonging feel existential, than in the anonymous adult emergency scenarios of the 1960s experiments. Social safety extends this concern into the anticipated consequences of helping: students worried not merely about physical harm but about being targeted themselves, becoming the next victim, or losing friendships. And the concentration of responsibility describes how adolescents assess whether helping is truly their job, particularly when adults — teachers, counselors, administrators — are perceived as the appropriate authority figures to handle such matters.

These findings carry significant weight because bullying is not a marginal problem. Decades of research following Olweus have established that bullying is fundamentally a group process, in which peer bystanders are not neutral scenery but active participants whose reactions can either sustain or dismantle the bullying dynamic. Naturalistic observational studies have shown that when peers do intervene, bullying often stops quickly — yet intervention remains comparatively rare. Previous work has linked bystander defending to factors such as empathy, school connectedness, parental warmth and friendship quality, and newer studies are extending the question into cyberbullying environments, where the ‘good bystander’ concept is being re-examined for digital contexts. The present study adds a crucial piece: a granular, first-person account of the adolescent decision process itself, grounded in the voices of the students who actually face these choices daily.

Methodologically, the study also demonstrates the value of qualitative rigor in a field often dominated by surveys and experiments. The triangulation of essays, interviews and focus groups, combined with member checking to verify the accuracy of interpretations with participants, allowed the researchers to build trustworthiness into their design. The work was approved by the Oakland University Institutional Review Board, with informed consent obtained from parents and guardians and assent from the minor participants themselves, who were told they could withdraw at any point. The authors report no funding for the research and no conflicts of interest, and de-identified data are available upon request to protect participant confidentiality.

The implications reach beyond academic theory. The study explicitly connects its findings to United Nations Sustainable Development Goal 3, Good Health and Well-Being, and Goal 4, Quality Education, situating bystander behavior within a global agenda for adolescent mental health and safe schooling. Whole-school intervention programs, including the widely disseminated Olweus Bullying Prevention Program, have long emphasized mobilizing the peer majority; this research suggests such efforts should explicitly address the distinctive adolescent barriers of social apprehension, social safety concerns and diluted personal responsibility. If educators want students to move from passive onlookers to defenders, they must reduce the reputational cost of helping, teach concrete and effective intervention methods, and clarify that responsibility for stopping cruelty belongs to everyone in the room — not only to the adults at the front of it. Half a century after Latané and Darley asked why adults fail to help, American teenagers are telling researchers, with striking candor, exactly what it would take for them to step in.

The intellectual backdrop of the bystander tradition extends beyond the famous emergency experiments themselves. Related work from the same era explored how the presence of others shapes effort and accountability, including research on social loafing showing that individuals contribute less when working in groups, and studies demonstrating that a bystander’s perceived ability to help alters whether intervention occurs at all. Even the presence of children at an emergency scene was found to increase adults’ sense of responsibility, underscoring how situational cues continuously recalibrate the helping calculus that Latané and Darley first formalized.

The study’s qualitative design also reflects careful attention to established standards of trustworthiness in phenomenological and grounded-theory traditions. Techniques such as member checking, in which interpretations are returned to participants for verification, draw on longstanding methodological debates about whether such procedures genuinely enhance rigor or serve mainly as symbolic validation. By anchoring the analysis in multiple data sources and established qualitative frameworks, the researchers positioned student voice — a concept with its own scholarly history in educational improvement research — at the center of the inquiry rather than treating adolescents as passive subjects of adult theorizing.

The findings also resonate with a broader international evidence base. Systematic reviews and meta-analyses have linked empathy to bystander defending roles in children and adolescents, while whole-school interventions promoting mental health and preventing risk behaviors have shown measurable benefits across adolescence. Recent qualitative work on cyberbullying is extending the good bystander question into digital spaces, where motivations and victim perceptions take new forms, and studies of juvenile populations suggest bystander characteristics can even influence displaced aggression in provocative situations.

By framing the work within the United Nations 2030 Agenda, the authors align a seemingly small-scale school study with global policy commitments, positioning peer intervention as a lever for both well-being and educational quality worldwide.

Subject of Research: Adolescent bystander behavior and intervention decision-making in school bullying situations in the United States

Article Title: To help or not to help?: adolescent bystander behavior in bullying situations in the US

Article References: Vannest, R. J., & Anderson, K. (2026). To help or not to help?: adolescent bystander behavior in bullying situations in the US. SN Social Sciences, 6(9), Article 424. https://doi.org/10.1007/s43545-026-01691-3

Image Credits: AI Generated

DOI: 10.1007/s43545-026-01691-3

Keywords: bullying, bystander behavior, bystander intervention, adolescence, Olweus Bullying Circle, Latané and Darley, diffusion of responsibility, qualitative research, school safety, peer intervention, prosocial behavior, social psychology