Effects of Peer Observation on Risky Decision-Making in Adolescence: A Meta-Analytic Review

Real-world health and crime statistics indicate that adolescents are prone to engage in risks in the presence of peers. Although this effect has been documented in several lab studies, existing evidence varies and the psychological mechanisms that give rise to peer observation-induced shifts in adolescent risky decision making remain poorly understood. We conducted a systematic literature review and meta-analysis to quantify the magnitude of the effect of direct peer observation on risky decision-making in adolescents. Across 186 effect sizes, representing data from 53 distinct research reports and over 5,000 participants, we found evidence that during adolescence, observation by peers increases decisions to take risks relative to decisions made while unobserved, with a small mean effect size (Hedges' g = 0.16). We also found high effect size heterogeneity (I-2 = 82.63% and tau(2) = 0.078), motivating analysis of moderation. We evaluated whether variables hypothesized by theory and prior research to amplify or reduce risk taking in the presence of peers systematically moderated the size of this effect, including factors related to the decision context, the peer context, and the experimental design. The overall effect was moderated by peers' expression of risk-seeking preferences, such that the effect of peer observation was only significant when peers were also expressing pro-risk attitudes. Evidence for publication bias was not consistently observed. Taken together, this work supports the notion that mere peer observation increases adolescent risky decision-making, but this effect is extremely small unless the peers are additionally expressing pro-risk preferences. Moreover, this work provokes questions regarding whether the field's approach to studying peer influence is optimal at conceptual and practical levels, and whether it is maximally translatable to real-world contexts. We offer suggestions for future work that could lead to a clearer understanding of peer observation effects during adolescence.