Journal of Child and Adolescent Behavior
Open Access

Adolescence is a time in life when friends play an important role in decision-making. Adolescent social influence narratives frequently revolve around risky and maladaptive decisions, such as driving under the influence and using illegal substances. However, research has shown that socio-cultural influence can lead to an increase in pro-social behaviours as well as a decrease in risk-taking. While many studies support the idea that adolescents are much more responsive to friends influence than children or adults, the developmental processes that underpin this sensitivity are still unknown. One important reason for this lack of understanding, we argue, is the lack of precisely formulated models.

We then demonstrate how these could be transformed into different models, and how these scientific models can explain experimental design and assist in the identification of developmental processes. Eventually, by implementing our formal models to existing datasets, we demonstrate the utility of formalization by synthesizing seemingly disparate results from different studies. We conclude with a discussion of how formal modeling can be used to better understand peer influence development in adolescence.

Friends have an impact on almost every aspect of adolescent lifestyles, from the trivial, such as music and clothing preferences, to the much more threatening, including the use of illicit substances like drugs. These latter, riskier, preferences may have long-term implications for the adolescent as well as significant societal costs. Although it is well established empirically that the presence of peers influences risky behaviour in adolescence, the underlying developmental processes remain unidentified. Understanding these processes, on the other hand, is critical for at least two reasons. First, empowering adolescents to make more informed decisions will be more effective if we can personalize intervention strategies to their neurodevelopment affordances. Second, we will only be able to identify these affordances if we can successfully link adolescent neurological and intellectual development with adolescent behaviour across different social contexts.

This relation, we contend, cannot be decided to make without models of adolescent social influence. Thus, the goal of this article is to take the first steps toward developing a quantifiable and testable framework of adolescent social influence. Adolescence is characterized by a number of developmental changes that provide multiple biological explanations for the social influence on adolescent decision making. The contemporary theoretical perspectives on these changes are referred to as "verbal models." In contrast to formal models, verbal models do not make specific estimates. We first evaluate previous verbal models and the affiliated empirical studies about social influence in adolescents, concentrating on risky decision-making, in order to assert formal models that really make quantitative predictions.

We describe three verbal models of social influence that are malleable and can be formalized by suspending them in expected utility theory. Following that, we demonstrate that our formal models can be recovered reliably and thus can be used to contrast hypotheses using quantitative model comparison. Finally, we apply these models to established data to uncover previously unknown patterns of peer influence.

We conclude by discussing how the precision provided by this formal way contributes to a better comprehension of the neurodevelopment processes underlying social influence. We identify three major families of verbal models, which we will relate as the following:

• Social motivation model

• Reward sensitivity model

• Distraction model

These three models are based on two distinguishable neurodevelopmental explanations of adolescent decision-making. The importance of developing the "social brain" is emphasized in social motivation verbal models.