School-based interventions are more likely to work when they target the right behaviour at the right age and combine trusted pupils with support from adults, a review finds.
Peer-led school programmes are most effective when they match the way adolescent influence actually works, according to René Veenstra of the University of Groningen. Writing in the International Journal of Behavioral Development, Veenstra argues that many schemes fail because they assume any well-liked student can sway classmates, when in fact peer influence depends on age, topic and the structure of the social network around the young people involved. Evidence from social network studies and school-based trials suggests that what spreads among students is often more selective, and more limited, than programme designers expect.
The review draws on research across bullying, mental health, substance use, healthy lifestyles and sexual behaviour. It brings together longitudinal network studies, which try to separate choosing similar friends from being shaped by them, and randomised trials of peer-led interventions, which typically recruit student leaders by nomination. Findings from recent work on adolescence show that different peer relationships can matter in different ways: best friends may shape some behaviours, while status-driven norms among popular students may matter more in others. A separate meta-analysis of experimental studies also suggests that susceptibility to peer influence varies by behaviour and by the method used to exert that influence.
Smoking is one of the clearest examples of timing at work. Veenstra notes that peer influence on smoking is strongest before the age of 15, after which similarity between friends more often reflects selection than influence. That helps explain why ASSIST, a programme that trained 12- to 15-year-olds to talk informally with friends, reduced the odds of smoking, while Towards No Drug Abuse, aimed at older teenagers, did not. Other examples point in the same direction: ROOTS lowered conflict in school and benefited from input by popular students, while Sources of Strength improved attitudes towards help-seeking and strengthened links between students and adults. By contrast, four recent programmes focused on physical activity and obesity showed no meaningful effect.
Veenstra’s broader conclusion is that peer programmes work best when they are narrow, well timed and built around behaviours that are genuinely contagious within friendship groups. He argues that a programme needs enough student reach to matter, but not so much complexity that messages lose force as they spread. The most effective models, he suggests, usually combine peer leaders with adults, rather than relying on students alone. That lesson matters for mentoring and prevention work alike: peer mentors should be chosen for warmth and standing within their group, not popularity alone, because influence can be harmful when it is driven by self-interest rather than care for others. A recent vaping-prevention programme, he notes, found that students connected to both peers and adults were less likely to vape, reinforcing the case for joined-up support.
Disclaimer: This content is for informational purposes only and is not intended to be a substitute for professional medical judgment, advice, diagnosis, or treatment.
