The Puzzle of Adolescent Risk Taking: An Experimental-Longitudinal Investigation of Individual, Social and Cultural Influences

Abstract

Adolescents are known as stereotypical risk-takers, as they engage in disproportionate levels of risk-taking (e.g., binge drinking and delinquency). However, meta-analytic findings based on experimental studies using behavioral risky decision-making tasks revealed that adolescents do not always engage in heightened risk-taking compared to children and adults. Namely, although adolescents took more risks than adults on such tasks, overall adolescents took equal levels of risks as children. Moreover, on tasks that provided the opportunity to avoid risk-taking, adolescents actually avoided taking risks more often than children. Finally, early adolescents engaged in more risk-taking than mid-late adolescents. Such results on age differences in risk-taking on experimental risk-taking tasks are meaningful for understanding real-world risk-taking, as adolescents’ performance on an often employed risk-taking task was shown to predict self-reported adolescent alcohol use, delinquency and risky traffic behavior. As for social predictors of adolescent risk-taking on such experimental risk-taking tasks, mere peer presence did not have a main effect on adolescent’s risk-taking on such tasks, however boys took more risks than girls in the presence of peers, while boys and girls took equal levels of risks when alone. Furthermore, in another experimental study, earlier maturing girls engaged in heightened risk-taking in the presence of peers to a greater extent than earlier maturing boys. Longitudinal results further revealed that friends’ and (older) siblings’ externalizing problems (delinquency and aggression), and mother-adolescent negative relationship quality are significant predictors of adolescent externalizing problems. Additionally, when considering adolescent phase and gender, for early adolescent girls, delinquent peer norms (and not delinquent peer pressure) predicted adolescent delinquency one year later. However, for early adolescent boys, delinquent peer pressure predicted adolescent delinquency one year later, but only when there were higher levels of mother-adolescent negative relationship quality. Moreover, mother-adolescent negative relationship quality also predicted middle-adolescent girls’ delinquency one year later. Although parents’ own externalizing problem behaviors did not longitudinally predict adolescent externalizing problems, experimental findings showed that mere mother and father presence decreased adolescent risk-taking on a risky decision-making task. Interestingly, mere sibling presence also decreased adolescent risk-taking. As for adolescent smoking, deviant peer pressure is also important for increases in smoking from ages 12 to 17, even when significant individual factors such as impulsivity are taken into account. Next, both cultural similarities and differences were demonstrated for links between adolescent risk factors and risk behaviors. For example, intention to use alcohol predicted alcohol use one year later for adolescents both in the Netherlands and on St. Maarten. Also, an interesting cross-national difference was found showing that whereas alcohol use predicted cannabis use one year later in adolescents on St. Maarten, the reverse was true for adolescents in the Netherlands. Finally, no ethnicity differences existed in the link between delinquency and depressive symptoms over three years, however, at two of the three waves, ethnic minority Dutch adolescents reported higher levels of depressive symptoms than ethnic majority Dutch adolescents, and a gender by adolescent phase moderation effect was present. Hence, a plethora of individual, social and cultural factors contribute to adolescent risk-taking.