VI. Future Directions
This research paper highlights some of the research questions that have addressed the relationship between peer networks and delinquency, but it is also important to consider future research directions that need to be explored further.
Although this research paper has emphasized the important context of adolescent friendships, future research would benefit greatly by incorporating multiple dimensions of potential influence. In particular, the delinquency involvement of other key individuals in youths’ networks, such as romantic partners, siblings, parents, and neighbors, may add to our understanding of influence processes. Because research has highlighted the importance of competing prosocial and deviant friendships that make up the bulk of adolescents’ friendships, it likely that other influential persons beyond friends could tilt the ratio of definitions favorable versus unfavorable to delinquency involvement. Incorporating these multiple contexts of adolescents’ lives into future analyses would also increase our understanding of the relative risk factors that adolescents face and potentially provide avenues toward reducing these risk factors.
Comparing the strength of influence across relational contexts also can provide unique insight into adolescents’ overall susceptibility to delinquent patters. One example of research that incorporated romantic partner behavior is that by Haynie, Giordano, Manning, and Longmore (2005), which shows that romantic partners’ delinquency exerts a unique effect on adolescents’ delinquency, over and beyond that of friends’ delinquency and control variables. In addition, recent work has compared the influence of “best friends” to that of youth considered “close friends” (Weerman & Smeenk, 2005). Taking this further, it would be interesting to also consider influence deriving from that of the overall school network in which adolescents are enmeshed. Along these lines, future research could identify the most popular students in the school to determine whether their behavior is especially influential for other individuals located in the school (who may or may not be tied to the most popular students).
Future research should also consider whether and how peer influence varies across demographic groups, such as by gender and race. On the basis of studies of homophily in friendship choice and evidence that race is one of the most important characteristics that influences which friendships form, we might expect to find that African American youth are more likely to be found in mixed networks where definitions toward delinquency are less clearcut. In terms of susceptibility to delinquency by race, prior evidence suggests mixed findings. In terms of gender, research suggests that although girls place more emphasis on close friendships incorporating intimacy and closeness than do boys, there is some evidence that boys are more susceptible to peer influence (Giordano et al., 1986). Other research suggests that gender differences in peer influence depend on the sex composition of the friendship network (Haynie, Steffensmeier, & Bell, 2007).
Another avenue for future research involves incorporating the school and neighborhood context to better understand how social environmentsmake unique contributions to the levels and severity of delinquency found among individuals and in their networks. Neighborhood and school environments are especially likely to determine the exposure adolescents have to prosocial or delinquent others. In addition, school factors such as school size, school disciplinary practices, school climate, school resources, and school policies such as tracking are likely to produce environments more or less conducive to delinquency and/or to place delinquent youth in closer proximity to other delinquent youth. This information would allow researchers to ask whether delinquent behavior among friends is more likely to occur in disadvantaged or disorganized schools, for example.
In addition, researchers may be interested in whether and how characteristics of the overall global school network (e.g., the density of ties, the racial heterogeneity of ties) influence levels of delinquency in the school. For instance, this type of information can be used to identify school characteristics that are most likely to suppress delinquency and/or violent behavior in the school, reduce the transmission of delinquent behavior, and/or decrease opportunities for high-risk youth to cluster together. In sum, future research should pay more explicit attention to the ways that neighborhoods and schools shape adolescent friendship networks, which in turn provide the contexts in which peer influences appear to flourish.
Future research should begin to examine how friendship networks and behavior change over time in school contexts. For example, researchers should consider the question of what predicts the dissolution of friendship ties over time. According to socialization theories, an individual’s behavior is shaped by the group norms to which youth are exposed. In the case of friendships, what happens when the behavior in question is not displayed by all members of the group? The normative influence process could sway the group’s behavior in favor of or against the behavior in question. Perhaps there is a tipping point at which it becomes more likely for the group to adapt the behavior in question or members who are not displaying the behavior to select out of the group (i.e., the tie is dissolved). These are interesting questions that could be addressed using longitudinal network data available.
Finally, future research needs to attempt to identify the mechanisms responsible for transmitting peer behavior to individuals. Although socialization theories suggest a variety of mechanisms that are potentially responsible, research has yet to clearly identify the specific ways that networks influence behavior. For instance, does the effect of peers on subsequent behavior result from social capital generated in the group, the modeling of group processes, increased opportunities for delinquency, deterrence factors, or a mixture of these mechanisms? Precise identification of the mechanism underlying behavioral similarity may require a different methodological approach.