Network Analysis in Criminology Research Paper

This sample Network Analysis in Criminology Research Paper is published for educational and informational purposes only. If you need help writing your assignment, please use our research paper writing service and buy a paper on any topic at affordable price. Also check our tips on how to write a research paper, see the lists of criminal justice research paper topics, and browse research paper examples.

Criminological theories have focused on different aspects of social relationships to explain the role peers play in shaping delinquency. Control theory, as presented by Hirschi (1969), focuses on the strength of attachment between individuals to explain variation in delinquent outcomes. Conversely, influence theories, such as differential association theory (Sutherland and Cressey 1960), propose that delinquency is learned within intimate relationships that expose individuals to definitions favorable to law and norm violation. Unfortunately, most empirical tests of these competing perspectives rely on data from random samples that do not allow for the direct measurement of peer characteristics (e.g., peer delinquency) or the assessment of the structural characteristics (e.g., cohesion) of peer and friendship groups. Overcoming these limitations, a network approach to crime/delinquency takes into account the structure of social relations, as well as the characteristics of actors comprising the network, to explain peer influence processes. This research paper outlines contemporary approaches to understanding the peer-delinquency association, focusing primarily on the ways that researchers operationalize key elements of influence and control theories, such as exposure to delinquent peers and peer group cohesion, with social network data. The paper also describes how criminologists are advancing the understanding of peer influence by studying behavioral development and network dynamics over time and by expanding the study of peer influence to include romantic partners, those to whom actors are indirectly tied, as well as associations occurring outside of school contexts.

Friend and peer influence has long been central to explanations of crime, delinquency, and other problem behaviors. Compared to children and adults, adolescents attribute greater importance to friends, spend more time socializing with friends, and are more strongly influenced by the behaviors and attitudes of their friends. While peer relationships have been less central in explaining adult offending, research documents the role of social influence from friends and romantic partners throughout the life course (Giordano et al. 2003). Not surprisingly then, the finding that individuals with delinquent friends are likely to be delinquent/criminal themselves is one of the most consistent and strongest findings in the criminological literature. Consistently robust associations between peer and individual delinquency have led some to argue that peer influence is one of the most important processes in explaining delinquent outcomes, regardless of whether the focus is on substance use, minor property offenses, status offenses, or violent crimes.

Prominent explanations of crime and delinquency disagree with regard to the mechanisms through which peers influence individual offending. While some of the debate is informed by theoretical concerns (e.g., “selection” versus “influence” as the driving mechanism), debate also stems from inadequate and inconsistent measurement of peer characteristics across research designs. In order to advance the understanding of peers’ impact on delinquent outcomes, this research paper provides a broad overview of past, present, and emerging techniques for measuring peer characteristics in criminological research. Particular attention is directed at social network processes and methods.

This research paper begins with a discussion of two prominent theoretical perspectives on delinquency, namely, control and influence theories. Close attention is paid to the manner in which peers are purported to influence crime and delinquency in these theoretical models. Next, social network approaches to peer influence on delinquency are discussed. The Section “Measuring Peer Influence: Past and Present” describes past and current methods of measuring peer characteristics in criminological research. This section first focuses on perceptual or self-report measures of peer characteristics, highlighting the methodological shortcomings of this approach. Next, network approaches to measuring the content (e.g., peer delinquency) and form (i.e., structural characteristics) of adolescent peer networks are identified. In addition, this section discusses manners in which network analysis has been used to help resolve competing claims regarding the role of peers in delinquent behavior in criminological theories. In section “New Directions in Peer Influence”, more recent innovations in social network approaches to peer influence are reviewed, highlighting how these methodological advancements can promote the understanding of peer influence processes. Finally, a conclusion provides a discussion of future directions in the study of peer influence that have great potential for advancing the current understanding of network processes as they relate to crime and delinquency.

Theoretical Explanations Of “Peer Effects”

Control theories, such as those presented by Hirschi (1969) and Gottfredson and Hirschi (1990), and influence theories, such as those presented by Sutherland (Sutherland and Cressey 1960) and Akers (2009), are prevailing perspectives that explain the importance of peers for adolescent delinquency. These perspectives offer opposing explanations with regard to the ways in which peers influence individual delinquency. Accordingly, criminologists have focused on different features of peer contexts in order to assess the plausibility of the theoretical mechanisms in empirical research. This section briefly outlines the predominant control and influence theories of crime, and makes note of the relative importance of the form and content of peer networks in each perspective and then discusses the network approach to peer influence, and identifies how it helps advance the understanding of peer selection and influence processes.

Control Explanations Of Crime

Hirschi’s (1969) social control theory proposes that connections to others serve as the primary constraints against delinquent impulses and actions. With regard to peer networks, social control theory predicts that strong friendship bonds are invariably and negatively associated with adolescent delinquency. This is because strong friendships entail attachment to conventional peers that would be weak or absent among more selfish and delinquent individuals. Hirschi explains the peer-delinquency association by suggesting that delinquent youth cannot form genuine friendships with peers. This implies that social ties among delinquent youth lack high levels of attachment.

Gottfredson and Hirschi (1990) general theory of crime elaborates the process through which individuals self-select into delinquent peer groups. Gottfredson and Hirschi argue that peers have no direct influence on individual offending. Rather, individual self-control (i.e., the ability to control impulsive behavior), which is relatively stable by early adolescence, shapes how adolescents cluster together in peer settings. Delinquent adolescents with low self-control are likely to befriend other delinquents as a result of their similar levels of self-control. Apart from determining the types of friends one makes, low self-control is also a primary cause of delinquent behavior; thus delinquency and associations with delinquent others are both directly caused by low self-control.

Importantly, Hirschi and Gottfredson and Hirschi’s control models propose that the association between delinquent peers and individual offending is spurious. Delinquency is instead explained by other processes such as inadequate social control caused by weak attachment to others or low self-control. These models explain any peer-delinquency association resulting from selection processes, and not because friends influence individual delinquency.

With regard to measuring peer effects on individual offending, social control theory primarily focuses on attachment within adolescent networks. Both control models also predict that delinquent peer groups will be less cohesive than those consisting of nondelinquent individuals with high self-control. Accordingly, measures that capture the strength of friendship ties, as well as the configuration of ties linking individuals within friendship groups, are crucial elements in tests of control theories. Such measures include subjective assessments of peer attachment, which capture mutual trust and involvement among friends, and network density, which measures the probability that a tie will exist between two members of friendship groups. Conversely, measures of peer characteristics, such as delinquency, are less central because the peer-delinquency association is believed to be spurious in these models. In support of both variants of control theory, weak attachments and low self-control are consistently associated with delinquency in research (Pratt and Cullen 2000).

However, research has also found that delinquent adolescents are no less attached to friends than their nondelinquent counterparts (Giordano et al. 1986). Additionally, Young et al. (2011) found self-control has a negligible impact on adolescent friendship formation after triad closure, or the tendency for individuals who share friends to become friends themselves, and other individual characteristics are taken into account. Similarly, Kreager et al. (2011), found group-level delinquency has no impact on the structural characteristics of friendship groups when group composition (i.e., average socioeconomic status and proportion male) is taken into account. Perhaps most importantly, friendships with delinquent peers is positively, and consistently, associated with delinquent behavior in criminological research (Haynie 2001), even after controlling for self-control and peer attachment (Pratt and Cullen 2000). Although friendship selection and social and self-control most likely shape delinquent behavior, peer influence continues to play an important part in explaining delinquent outcomes.

In contrast to control theories, influence theories argue that peer characteristics are consequential for individual offending. Accordingly, empirical tests of these theories have in turn largely focused on the content of social networks to explain individual offending.

Influence Perspectives

Influence theories argue that criminal and other “antisocial” behaviors are learned from intimate social relationships with others where attitudes or “definitions” favorable to law violation are acquired. The social transmission of crime/delinquency primarily occurs within peer networks through the transference of favorable attitudes and definitions that encourage criminal behavior. Akers’ (2009) extension to differential reinforcement theory emphasizes behavioral modeling and operant conditioning processes. This model assumes that the adoption of criminal behavior occurs through the imitation of friends’ behavior and the observation of its consequences, either positive or negative. Sutherland and Akers’ theories represent prototypical influence theories in that they argue that crime and delinquency, like any behavior, are learned in intimate relationships, including peer friendships.

The characteristics of peers within adolescent friendship groups have been at the forefront of empirical tests of differential association and social learning theories. For example, the principle of differential association states, “A person becomes delinquent because of an excess of definitions favorable to violation of law over definitions unfavorable to violation of law” (Sutherland and Cressey 1960, p. 78). In empirical research, differential association has been most commonly operationalized by asking respondents how many or what proportion of their friends commit deviant acts. These “perceptual” measures of peer delinquency have been remarkably strong and consistent predictors of individual offending in delinquency research (Pratt et al. 2010). Both the consistency and robustness of the association between peer and individual offending attest to the validity of peer influence theories.

While perceptual measures of peers’ criminal involvement are some of the most robust predictors of individual offending, tests of social learning and differential association theories have been hindered by oft-used self-report measures of delinquency in criminological research. Perceptual measures of peer delinquency are susceptible to report bias, which likely results from delinquent individuals ascribing their own behaviors to their friends (Gottfredson and Hirschi 1990). Indeed, recent research indicates that perceptual measures of peer delinquency both overestimate (Haynie and Osgood 2005) and underestimate (Weerman and Smeenk 2005) the actual level of delinquency among peers.

One drawback of social learning and differential association perspectives is that they offer little theoretical and empirical attention towards understanding how the form, or structural characteristics, of social networks factors into peer influence processes. These oversights are significant, given that exposure to and circulation of norms and behaviors conducive to offending are central features of the theories. For example, Sutherland proposes that differential associations vary in their “frequency, duration, priority, and intensity” (Sutherland and Cressey 1960, p. 78). Unfortunately, the exact meaning of frequency, duration, and intensity of differential associations has been underdeveloped and left largely unmeasured in empirical tests of influence theories (Haynie 2001). In contrast, the social network perspective more fully explains the mechanisms through which social relations shape the peer-delinquency relationship. Social network data, in turn, more fully capture both the form and content of peer groups than nonrelational data, which enable more thorough examinations of peer influence processes.

A Social Network Perspective

The social network perspective offers a unique view of humans in its basic premise that individuals are interdependent and, more importantly, that these interdependencies have important consequences for behavior. Therefore, it emphasizes both the configuration of ties connecting individuals in a social structure and the characteristics of actors in that structure to explain delinquent outcomes. Scholars interested in applying a social network perspective to understand individual offending require data on the ties or connections among individuals within a particular setting. While these ties can take on different forms, researchers interested in adolescent behavior most often examine the form of friendship ties among adolescents. This suggests that the behavior and structure of adolescent networks offer powerful explanatory power for delinquency and other social behaviors. Network perspectives also allow researchers to test competing theoretical hypotheses of control and influence theories, while also advancing new concepts – such as centrality and density – and methods that expand our knowledge of delinquent contexts.

Friendship networks are particularly important for understanding adolescent involvement in crime/delinquency for a number of reasons. Integration into friendship groups where adolescents form bonds and spend time with peers may facilitate or inhibit delinquent behavior depending upon the norms, values, and behaviors within the network. Also, while adolescents may discount a friends’ evaluation, they are less likely to discount a group evaluation. This suggests that embeddedness within friendship networks takes on additional influence as it generates expectations for behavior while reinforcing group social norms and beliefs. Adherence to group norms is especially likely during adolescence because acquiring peer acceptance is of central importance during this stage of the life course. Furthermore, friendships are crucial in establishing role identity throughout adolescence. As a result, adolescent peer networks are especially effective at directing and constraining individual members’ behavior.

While a social network perspective offers a particularly useful tool for understanding peer influence, it is only recently that the full benefit of a network approach for crime/delinquency research has been revealed. The following section describes past and recent methods of measuring peer characteristics. Although conventional, nonnetworked, “perceptual” methods are discussed, this section primarily focuses on network approaches to assessing peer influence.

Measuring Peer Influence: Past And Present

Prior studies examining the effect of peer influence on delinquency generally ask adolescents to think about their friends and to report whether their friends have participated in a particular illegal behavior or set of illegal behaviors. Respondents usually report that “none,” “some,” or “all” of their friends participated in the behavior being discussed. Studies relying on such perceptual measures have found robust associations between peer delinquency and respondent offending (Pratt et al. 2010). However, the consistent association between perceived peer deviance and respondent delinquency provides only limited support for influence theories such as differential association theory for a number of reasons. First, social psychologists refer to the tendency to assume that friends behave as do you, as assumed similarity or projection. Therefore, the perceptual approach to measuring peer characteristics contains a same-source bias that inflates similarity in behavior between peers. In fact, prior research demonstrates self-report measures of peer delinquency tend to be not only inaccurate but are systematically biased (Young et al. 2011). Findings such as these support Gottfredson and Hirschi’s claim that individual characteristics, such as self-control, bias perceptual measures of peer delinquency and other behaviors. Accordingly, the extent to which perceptual studies of peer delinquency can actually support influence theories is limited by the extent to which they are unbiased by individual characteristics. Furthermore, misspecification of peer delinquency is especially problematic in studies that test competing claims of influence and control theories, as biased measures of peer delinquency likely results in the upwardly biased estimates of peer effects and downwardly biased estimates of self-control effects (Meldrum et al. 2009).

Apart from mounting evidence indicating systematic misspecification of peer delinquency, perceptual measures of peer characteristics are limited in a number of other ways. For example, past research employs imprecise definitions of the friendship group in which behavioral influence is thought to occur. As a result, the number of friends considered by the adolescent when responding to questions regarding peer delinquency is often unspecified. Accordingly, the structure and composition of an adolescent’s delinquent friendship network remains relatively unclear.

Perhaps as importantly, perpetual studies of peer influence fail to consider how structural properties of network – that is, how adolescents are connected to one another via friendship ties – condition peer influence processes. In doing so, past research assumes that everyone in the friendship network is affected by friends’ behavior in the same manner. Non network approaches overlook individual position within the network (e.g., central versus peripheral), the cohesiveness of the network (i.e., the interconnections among network members), and the adolescent’s status (e.g., popularity) within the network. These structural characteristics likely shape the degree to which adolescents are influenced by group behavior (Haynie 2001). A network perspective is guided by the assumption that both behaviors exhibited by network members and the structure of the network have important consequences for understanding subsequent behavior. With regard to delinquency, this suggests that exposure to pro or anti-delinquent behaviors will depend upon the structure of the network, the adolescent’s position within the network, and the behaviors exhibited in the network.

Measuring Peer Delinquency With Network Data

Recent data collection efforts, such as the National Longitudinal Study of Adolescent Health (Add Health), have made relational data available to criminologists interested in applying network theories and methods to the understanding of peer influence. The Add Health data are unique in that they provide an opportunity to analyze complete social networks of adolescents attending a representative sample of schools in the United States. To achieve this, researchers collected data on every student attending school on the day the questionnaire was administered. As part of this data collection effort, respondents were asked to nominate up to ten of their closest friends (five male and five female friends) from school rosters listing all schoolmates. Because every student present completed the questionnaire, it is possible to observe almost all friendship ties among students in the school. Add Health provides the largest and most comprehensive portrayal of adolescents’ school-based friendship networks to date. Other smaller-scale efforts, such as the Netherlands Institute for the Study of Crime and Law Enforcement (NSCR) “School Study,” the Context of Adolescent Substance Use Study, and the Promoting School Community-University Partnerships to Enhance Resilience (PROSPER) study have utilized similar data collection techniques to directly measure peer and network structural characteristics.

With regard to the peer-delinquency association, criminologists have operationalized peer delinquency in a number of ways. Theoretical concerns should be, and for the most part, are primary in informing approaches to measuring peer delinquency. For example, the principle of differential association theory maintains that individuals become delinquent because they experience excess of definitions favorable to law violation, compared to definitions that favor observing law. In testing this claim, Haynie (2002) constructed a “relative” measure of peer delinquency which consisted of the proportion of friends that are delinquent within individuals’ peer networks. Results indicated adolescents’ delinquent behavior most resembled that of their peer groups when they were enmeshed in friendship networks with greatest behavioral consensus (i.e., all friends are either delinquent or nondelinquent). Another more common way of operationalizing peer delinquency is to measure the average amount of delinquency across one’s peer group (Haynie 2001). This operationalization allows researchers to focus more on the consequences of total exposure, rather than relative exposure, to delinquent peers.

Recently, McGloin (2009) expanded the operationalization of peer delinquency by considering how peer deviance relative to one’s own, rather than simply exposure to delinquent peers, shapes individual delinquency. McGloin points out that one aim of interactive relationships is to achieve behavioral congruence, or homeostasis with one’s close associations. One hypothesis that stems from this perspective is that adolescents will seek behavioral congruence with their closest friends. In order to test this assertion, McGloin constructed a measure of delinquency balance, which consisted of the difference between respondents’ delinquency and that of their best friends. Results indicate adolescents tend to commit less delinquency over time when their delinquency was previously higher than their best friend; conversely, respondents tend to become more delinquent when they were initially less delinquent than their best friend.

Influence theories in large part emphasize the importance of learning that takes place within intimate peer groups. Accordingly, most empirical tests and influence theories have primarily focused on the influence of peers to whom one is directly tied. However, influence theories stipulate that peer influence also takes place through vicarious learning, in which individuals observe certain behavior and others’ reactions to that behavior. Individuals are more likely to perform observed behavior that is positively reinforced by others throughout social interactions. This suggests that peer influence processes likely also occur between individuals who are only indirectly tied.

Network approaches to peer influence allow for the testing of processes that extend beyond those to whom individuals are directly connected. For example, Payne and Cornwell (2007) found that the delinquency of indirect associates (i.e., friends of a friend who are not my friends) is positively associated with respondent delinquency, even after taking into account the delinquency of those to whom one is directly tied. However, the authors also found that the influence of indirect associations diminished according to the extent to which the overall delinquency of one’s indirect ties diverges from that of the immediate friendship group. Similarly, Kreager and Haynie (2011), demonstrate that behavior among those to whom one is indirectly tied through a dating partner (i.e., friends of my partner who are not my friends) influence individual behavior. Interestingly, the authors find evidence that the drinking of friends of partners have stronger associations with a dater’s future behavior than do his or her own friends or the romantic partner.

Social proximity to delinquent individuals also factors into peer influence processes. In their examination of peer and social network influences on adolescent substance use, Ennett et al. (2006) found that the shortest distance to the nearest substance user, as measured by the number of ties needed to link respondents to the nearest substance user within a school, is negatively associated with substance use after controlling for the substance use within friendship networks.

Recently developed methods for identifying peer “clusters” allow researchers to further examine peer influence occurring at multiple levels and between individuals who are only indirectly tied. Peer clusters consist of densely connected groups in which ties are more likely to occur between members of the same group than members of different groups. Peer clusters also represent “meso levels” of social organization, which are larger in size than micro level friendship groups, but smaller in scale than other levels of aggregation, such as schools (Magino 2009). These conglomerations of friendship ties, which have discernable boundaries, may be used to capture larger social circles or “cliques,” within which peer influence likely occurs.

Clustering algorithms, such as Moody’s (1999) CROWDS or Frank’s (Frank et al. 2008) Kliquefinder, allow researchers to detect dense friendship groups in social networks. Resulting subgroups in turn may be used as distinct levels of analysis in the study of peer influence processes. For example, Frank et al. (2008) used the Kliquefinder algorithm to identify “local positions,” represented by clusters of adolescents based on their shared academic course taking among Add Health respondents. The authors found that girls were more likely to subsequently take higher-level courses when they were previously embedded in peer group clusters that include girls who also took advanced math courses. While this finding does not directly pertain to delinquency, it nevertheless points to the likelihood that peer influence processes extend beyond individual ties. Although peer group clustering techniques have not been widely employed in delinquency research (see Kreager et al. 2011; Ennett et al. 2006 for notable exceptions), the study of peer influence processes occurring within larger peer groups represents an interesting future direction for research to pursue.

Exactly how structural characteristics of peer networks condition the peer-delinquency association has also been largely overlooked in prior delinquency research. This oversight is unfortunate, as differential association theory proposes that while criminal behavior is in large part shaped by exposure to delinquent peers, differential associations vary in their “frequency, duration, priority, and intensity” (Sutherland and Cressey 1960, p. 78). Additionally, the form, or structural characteristics of peer networks, likely shapes peer influence processes by facilitating, among other things, behavioral reinforcement, circulation of definitions of behavior and attitudes conducive to offending, and exposure to nonredundant information (Granovetter 1973), which may facilitate co-offending opportunities and contagion of delinquent behavior.

Recognizing that the form of adolescent networks likely shapes the relationship between the content of peer networks and delinquent outcomes, Haynie (2001) evaluated whether network characteristics condition the effect of peer delinquency on individual offending. Results indicated that network centrality, density, and popularity accentuated the positive association between delinquent peers and individual delinquency. Further research examining how structural features of social networks shape the association between peer characteristics and risk behavior will help advance the understanding of how the effects of differential associations on individual outcomes vary according to network characteristics.

New Directions In Peer Influence

Emerging and exciting areas related to social networks and adolescent offending have accompanied increased access to high-quality social network data. For instance, actor-based models have recently been introduced to analyze behavioral development and network dynamics over time. Developed by Snijders and colleagues (Steglich et al. 2010), the SIENA statistical package uses simulation and Markov chain models to estimate changes in friendship ties and behavior over multiple waves of network cross sections or panels. Recent applications of SIENA in criminological research are advancing the understanding of peer influence by distinguishing selection from influence effects, while also adjusting for the structural characteristics of the network.

Researchers interested in peer influence are also expanding their empirical focus to associations other than friends. Increased inter-gender contact in mid-to-late adolescence entails increased opportunities for forming romantic relationships (Warr 2002). Recognizing the potential for romantic partners to impact offending, Haynie et al. (2005) demonstrate that romantic partner’s delinquency exerts a unique effect on adolescents’ delinquency, over and beyond the delinquency of one’s friends. Giordano et al. (2010) revisit Hirschi’s “cold and brittle” relationship hypothesis with regard to delinquents and romantic involvement. Using data from the Toledo Adolescent Relationships Study, they found little difference in the quality and longevity of delinquents’ and nondelinquents’ romantic relationships.

While this research paper has largely focused on how peers influence individual delinquency, most criminological theories recognize that delinquency often occurs in groups. Accordingly, peers also impact individual delinquency by providing co-offending opportunities, the effect of which is analytically distinct from peer influence (Warr 2002). Unfortunately co-offending remains a relatively understudied topic in criminology. Notable exceptions (McCarthy et al. 1998) provide strong evidence that co-offending is an important mechanism through which peers impact individual crime deviance.

Innovative network studies continue to advance the understanding of peer influence processes. For example, the PROSPER Peers data, emanating from Pennsylvania State University under the supervision of Osgood and colleagues (Kreager et al. 2011) follows two successive sixth-grade cohorts in 28 rural Iowa and Pennsylvania communities. Social network data were collected yearly from the 6th to the 12th grade, resulting in an impressive number of complete school networks ( 400) over a 7-year span within each school cohort. With detailed delinquency and substance use measures and SIENA analyses, studies from PROSPER and similar longitudinal network datasets promise to expand our understanding of peer selection and influence processes.

One area that remains particularly underdeveloped in peer influence research relates to the impact of nonschool friends and peers. Out-of-school peers are likely important in shaping delinquency because they are more likely to be delinquent, older, and less connected to conventional institutions than school-based peers. Understanding the content and structural properties of nonschool friendship networks, may provide important clues for the etiology of delinquency. Collecting such data requires researchers to move beyond the bounds of schools and focus on communities, which is logistically more demanding and likely to be more costly, but capable of providing new insights into the peerdelinquency association.

Papachristos et al. (2012) present an innovative approach to measuring non-school-based networks among high-risk individuals in Cape Verdean, a low-income, predominately African-American community in Boston. To construct the network of high-risk individuals, the authors first identified the population of Cape Verdean gang members who were known by the Boston Police Department (BPD). Next, the authors generated a list of gang members’ immediate associates from Field Intelligence Observation (FIO) data which were collected through observations conducted by the BPD. This respondent-driven approach resulted in a network in which ties between individuals occurred when they were observed in each other’s presence by the police and recorded in the FIO data. This process was repeated for the associates of the gang members. From there, the authors constructed a network of high-risk individuals that consisted of gang members, their associates, and the associates of the non-gang members. Using matched data on fatal and nonfatal gunshot injuries, the authors demonstrate that the probability of experiencing gunshot injury was related to one’s distance from other gunshot victims.

The limited availability of community-based network data is largely a result of the astronomical costs of obtaining population samples needed to construct network-based measures across several communities. One innovative approach to N constructing nonschool social networks entails constructing affiliation, or “two-mode” networks, which, in the context of community research, consist of ties between individuals and activities or locations. Recently, Browning (2011) constructed two-mode networks among 65 neighborhoods with data from the Los Angeles Family and Neighborhood Study (LAFANS). Browning demonstrates that triad closure within networks based on individual overlap in respondents’ routine “activity spaces” is negatively associated with adolescent risk taking. One primary advantage of two-mode approaches to community networks is that affiliation network data can be collected through clustered random samples. This lowers the cost of constructing measures of global network structure and individual embeddedness within community-based networks. Frank’s Kliquefinder algorithm may also be applied to two-mode networks to construct adolescent group clusters, within which peer influence may also take place (Frank et al. 2008).

Summary

This research paper aimed to provide an overview of network approaches to the measurement of peer effects in criminological research. After outlining the merits and limitations of control and influence theories, the paper advocated a social network approach to understanding the role of peers for two primary reasons. First, the network framework described in the paper emphasized the importance of the social connections among individuals within a social setting. Capturing the relations among actors within social contexts allows criminologists to assess the interconnectedness of actors, which is an important element of both influence and control perspectives. Second, the network perspective facilitates direct measurement of peer characteristics, which allows for more stringent tests of influence and control theories. In the end, the paper intended to demonstrate the value for a network approach to the study of the peerdelinquency association which incorporates characteristics of the friendship groups in which adolescents are enmeshed.

This research paper also addressed emerging areas related to social networks and peer influence. Such areas include changes in friendship structure and composition, influence from actors to whom one is indirectly tied, romantic partner influence, and co-offending. Finally, the paper identified innovative approaches to capturing nonschool social networks, including a respondent-driven approach based on police observation data, and affiliation networks based on activity-location overlap.

The network approach to adolescent delinquency provides a coherent and promising framework for investigating the variety of ways that peers shape and influence involvement in delinquency and crime. This approach is consistent with the current emphasis on the significance of social contexts (e.g., neighborhood, school) and helps understand the intricate ways in which individuals’ behavior is shaped by the social networks in which they are embedded.

Bibliography:

Akers RL (2009) Social learning and social structure: a general theory of crime and deviance. Transaction Publishers, New Brunswick
Browning CR (2011) The spatial and social embeddedness of youth activities. Paper presented at the annual meeting of the Association of American Geographers, Seattle
Ennett ST et al (2006) The peer context of adolescent substance use: findings from social network analysis. J Res Adolescent 16:159–186
Frank KA et al (2008) The social dynamics of mathematics course taking in high school. Am J Sociol 113:1645–1696
Giordano PC, Cernkovich SA, Pugh MD (1986) Friendships and delinquency. Am J Sociol 91:1170–1202
Giordano PC, Cernkovich SA, Holland DD (2003) Changes in friendship relations over the life course: implications for desistance from crime. Criminology 41:293–328
Giordano PC, Lonardo RA, Longmore MA (2010) Adolescent romance and delinquency: a further exploration of Hirschi’s ‘cold and brittle’ relationships hypothesis. Criminology 48:919–946
Gottfredson MR, Hirschi T (1990) A general theory of crime. Stanford University Press, Stanford
Granovetter MS (1973) The strength of weak ties. Am J Sociol 78:1360–1380
Haynie DL (2001) Delinquent peers revisited: does network structure matter? Am J Sociol 106:1013–1057
Haynie DL (2002) Friendship networks and delinquency: the relative nature of peer delinquency. J Quant Criminol 18:99–134
Haynie DL, Osgood DW (2005) Reconsidering peers and delinquency: how do peers matter? Soc Forces 84:1109–1130
Haynie DL, Giordano PC, Manning WD, Longmore MA (2005) Adolescent romantic relationships and delinquency involvement. Criminology 43:177–210
Hirschi T (1969) Causes of delinquency. University of California Press, Berkeley
Kreager DA, Haynie DL (2011) Dangerous liaisons? Dating and drinking diffusion in adolescent peer networks. Am Sociol Rev 76(5):737–763
Kreager DA, Rullison K, Moody J (2011) Delinquency and the structure of adolescent peer groups. Criminology 49:95–127
Magino W (2009) The downside of social closure: brokerage, parental influence, and delinquency among African American Boys. Sociol Educ 82: 147–172
McCarthy B, Hagan J, Cohen LE (1998) Uncertainty, cooperation, and crime: understanding the decision to co-offend. Soc Forces 77:155–184
McGloin JM (2009) Delinquency balance: revisiting peer influence. Criminology 47:439–477
Meldrum RC, Young JTN, Weerman FM (2009) Reconsidering the effect of self-control and delinquent peers. J Res Crime Delinq 46:353–376
Moody J (1999) The structure of adolescent social relations: modeling friendship in dynamic social settings. University of North Carolina at Chapel Hill.UMI Dissertation Services, Ann Arbor
Papachristos AV, Braga AA, Hureau DM (2012) Social networks and the risk of gunshot injury. J Urban Health 89:992–1003
Payne D, Cornwell B (2007) Reconsidering peer influences on delinquency: do less proximate contacts matter. J Quant Criminol 23:127–149
Pratt TC, Cullen FT (2000) The empirical status of Gottfredson and Hirschi’s general theory of crime: a meta-analysis. Criminology 38:931–964
Pratt TC et al (2010) The empirical status of social learning theory: a meta-analysis. Justice Q 27: 765–802
Steglich C, Snijders TAB, Pearson M (2010) Dynamic networks and behavior: separating selection from influence. Sociol Methodol 40:329–393
Sutherland EH, Cressey DR (1960) Principles of criminology, 6th edn. J.B. Lippincott, Philadelphia
Warr M (2002) Companions in crime: the social aspects of criminal conduct. Cambridge University Press, Cambridge
Weerman FM, Smeenk WH (2005) Peer similarity in delinquency for different types of friends: a comparison using two measurement methods. Criminology 43:499–524
Young JTN, Barnes JC, Meldrum RC, Weerman FM (2011) Assessing and explaining misperceptions of peer delinquency. Criminology 49:599–630