Teenagers don’t just influence each other – they learn from each other
Why are adolescents so sensitive to their peers’ behaviour? And is this sensitivity necessarily negative, as the term peer pressure often implies? In his PhD dissertation, UvA behavioural scientist Andrea Gradassi demonstrates that adolescents are particularly influenced by classmates to whom they are socially close, such as friends, as well as by peers who occupy high-status positions within the classroom social network.
‘The goal of this dissertation was to examine peer influence, particularly during adolescence,’ Gradassi explains. ‘Traditionally, psychology has tended to frame peer influence at this stage of life as passive and harmful.’
Gradassi challenges this narrow perspective by asking a more fundamental question: why are adolescents so responsive to their peers in the first place? From an evolutionary standpoint, learning from others is highly adaptive. Observing peers allows individuals to avoid costly trial-and-error learning. As Gradassi puts it: ‘If you look at what your peers are doing and they are successful, that behaviour may also be useful for you.’
Real experiments in the classroom
A distinctive feature of Gradassi’s research is its grounding in realistic social environments. Rather than relying solely on laboratory settings, he conducted experiments in Dutch secondary schools.
‘We went into schools and mapped the relationships among students,’ he explains. Using social network analysis, Gradassi identified friendships, popular students and socially central figures within each classroom. These data were then used to design experiments that measured peer influence systematically.
In one study, students completed tasks such as estimating the number of animals in an image. After submitting their initial response, they were shown the answer of a selected classmate and given the opportunity to revise their own.
The findings were clear: adolescents were more strongly influenced by friends than by socially distant classmates. ‘When the other person was a friend, students adjusted their answers more,’ Gradassi notes. This suggests that classroom learning is shaped not only by accuracy, but also by social closeness and trust.
Social status also played a significant role. Adolescents were more likely to incorporate information from peers who occupied central positions in the social network—students who were well connected and socially prominent.
Importantly, influence extended beyond popularity alone. Peers perceived as academically competent also exerted greater influence on others’ learning decisions. Together, these findings demonstrate that educational outcomes are shaped not only by individual ability, but also by the structure of social relationships within the classroom.
Older adolescents respond to more positive influence
One of the most striking findings in Gradassi’s dissertation concerns prosocial behaviour. In a large-scale study involving 456 Dutch adolescents, participants were given the option to donate money to charity or keep it for themselves. Based on prevailing theories, Gradassi expected selfish behaviour to increase with age.
Instead, the data revealed the opposite pattern: more responsive to positive peer influence. ‘When older adolescents saw someone donating money, they were more likely than younger adolescents to follow that example,’ Gradassi explains.
These findings suggest that peer influence can be a force for good. Positive role models may become increasingly influential during later adolescence, highlighting opportunities to harness peer dynamics in constructive ways.
Social networks and social media
Gradassi’s findings are particularly relevant in an era in which social influence is amplified online. Social media platforms continuously map social connections and can intensify peer dynamics.
Understanding who influences whom—and under what circumstances—is therefore crucial for educators, policymakers and adolescents themselves. ‘It is important to inform the public about how these social dynamics operate,’ Gradassi concludes.