The Role of Internal Control and Peer Relationships

1

Late Adolescent Violent Behavior:

The Role of Internal Control and Peer Relationships

Maria J. Whitmore

Distinguished Majors Thesis

University of Virginia

April, 2001

Advisor: Joseph P. Allen

Second Reader: N. Dickon Reppucci

Acknowledgements

I would like to acknowledge some of the people who made completion of this paper possible. First, I thank my advisor and mentor, Joseph P. Allen, for his support, input, and feedback throughout this process. I am grateful to Christy McFarland for giving of her time, energy, and guidance to this project, as well as to N. Dickon Reppucci for his input on this project. I would also like to express my appreciation to the dedicated graduate and undergraduate research assistants who spent time entering, cleaning, and coding the data presented herein. To Bruce, Lottie, and Sarah for their willingness to be sounding boards to my ideas. Lastly, I am forever indebted to my family for their constant love and support.

Abstract

Adolescence is a crucial time in the development of both peer relationships and violent and aggressive behavior. Given the shared salience of peers and deviant behavior in adolescence, it is important to discover how these two developmental phenomena interact and affect each other. Previous research has suggested that one factor that may influence both peer relationships and violent behavior outcomes is internal cognitive control on impulses and behaviors, yet no previous study has examined all three factors simultaneously. The current study sought to address four questions: (1) Does a lack of internal control predict peer relational difficulty? (2) Does peer relational difficulty predict adolescent violent behavior? (3) Does a lack of internal control alone predict adolescent violent behavior?, and (4) Does peer relational difficulty moderate and/or mediate the relationship between internal control and adolescent violent behavior? These questions were addressed using multiple methods and multiple measures of each construct. Individual and peer predictors of violent behavior were assessed within a socioeconomically diverse, at-risk sample of 127 adolescents over a two –year period. Internal control was defined as cognitive and affective resources that support regulated responses and behaviors. It was operationalized in terms of ego development, self-restraint, and social competence expectations. Peer relational factors examined in this study included conflict resolution skills, total attachment to friends, and broad social problems. Four major findings emerged. First, there was equivocal evidence that internal control predicted peer relational difficulty. Second, peer relations somewhat predicted violent behavior. The third finding was that internal control strongly predicted violent behavior. Lastly, conflict resolution skills were found to moderate the relationship between social competence expectations and violent behavior. Implications of these findings for the construct of internal control, the importance of peer relationships, and possible intervention strategies are discussed.

Introduction

Violent behavior by adolescents is a serious phenomenon, and one that has the potential to create serious costs for society. For example, in 1997, juveniles were involved in 27% of all serious violent victimizations. These violent crimes by juveniles have become increasingly lethal in recent years, as young people have more access to guns (Blumstein, 1995). While overall homicide rates in the U. S. have remained constant over the last three decades, youth homicide rates have risen (Garbarino, 1999). In 1997, more than 1,400 murders were determined to have involved a juvenile (U. S. Department of Justice, 1999). Even more recently, over the past two years, school shootings have left our schools as modern killing fields for a generation of children (Garbarino, 1999). It is no wonder then that 4% of high school students nationwide missed at least one day of school in the past month due to feeling unsafe at school (U. S. Department of Justice, 1999). Not only is society paying the price for juvenile violence in lives and in fear, but it is also an economic issue. The cost of allowing just one adolescent to leave high school for a life of crime is estimated at between 1.7 and 2.3 million dollars (U. S. Department of Justice, 1999).

The stability of aggressive behavior from childhood through adulthood has been well-documented (Olweus, 1979; Pettit, 1997; Loeber & Hay, 1997). Past aggressive behavior is considered the best predictor of future aggressive behavior (Moffitt, 1993). In spite of this stability, however, there is also evidence for change in aggressive behavior. Adolescence has been identified as a critical time in the lifespan development of aggression and violence. Moffitt has identified two general patterns of the life course development of criminal behavior, childhood-onset and adolescent-onset (1993). This dichotomy has come to be generally recognized, and is reflected in the current edition of the American Psychiatric Association’s Diagnostic and Statistical Manual which specifies two types of conduct problems: childhood-onset type and adolescent onset type (DSM-IV, 1994) The adolescent-onset pattern has been found to be much more common, as antisocial behavior tends to rise and peak in adolescence (around age 17). These rates drop sharply as participants grow into early adulthood (Moffitt, Caspi, Dickson, Silva, & Stanton, 1996). It seems that many adolescents, then, “outgrow” aggressive behavior (Allen, Moore, & Kuperminc, 1997). A significant number of highly aggressive adolescents greatly reduce their levels of aggression in late adolescence, and desist entirely from this behavior as they mature (Blumstein, Cohen, & Farrington, 1988). Approximately only 5% of offenders are responsible for half of all crimes, so clearly most adolescents are not persistent in their offending throughout their lives (Moffitt, 1993). This instability of adolescent aggression provides us with hope that natural processes are in place which affect change on aggression. Understanding of these processes could lead to effective interventions with violent youth.

Children’s understanding of their social environment has been shown to be a major contributor to the development and change in their aggressive behavior (Dodge, 1980; Rubin, Bream, & Rose-Krasnor, 1991; Egan, Monson, & Perry, 1998). There are several conceptual models which have been included in this type of work, but all make the assumption that environmental conditions promote aggression-inducing cognitions (Egan, Monson, & Perry, 1998). For example, some researchers believe that certain types of coercive parenting lead children to infer that aggression is an effective way of getting needs met (Patterson, Reid, & Dishion, 1992). Others believe that a history of abusive parenting may cause a child to perceive hostile intent in other people (Weiss, Dodge, Bates, & Pettit, 1992). Still others suggest that it is a combination of low internal cognitive control (dysregulation, poorly modulated behavior, impulsivity) which interacts with the social environment to produce aggression in children (Pope & Bierman, 1999).

The concept of internal control will be further examined in this study. It has been variously described and operationalized in the literature. In general, though, internal control refers to regulation of behavior through internal schemas and resources rather than external forces such as social interaction. One common element across descriptions of internal control is that it is considered to be a “master” construct that encompasses other personality traits (Gottfredson & Hirschi, 1990). Weinberger, for example, suggested that self-restraint is comprised of four subscales of self-control including impulse control, suppression of aggression, consideration of others, and responsibility (Feldman & Weinberger, 1994). A second element that is common in descriptions of internal control is impulse suppression. Those with low self-control have been described as having a “here and now” orientation that does not account for long-term goals and consequences (Gottfredson & Hirschi, 1990).

In 1999, Tate tested a model of internal control, parenting practices and violent behavior using a sample of academically at-risk adolescents in a longitudinal study. He defined internal control as “cognitive and affective resources that support regulated responses and behaviors” (p. 6), and operationalized low internal control as a lack of self-restraint and ego development as well as social information processing difficulties such as hostile attributional biases and beliefs about the value of aggression. He found that low internal control in mid-adolescence significantly predicted violent behavior in late adolescence. Tate also found effects of maternal internal controls on their children’s internal controls and violent behavior, such that mothers’ lack of internal control at Time 1 predicted adolescent violence two years later. Maternal parenting practices were also found to have an effect on the development of aggression, but these effects were mediated by adolescent internal controls. Finally, Tate’s results showed that changes over time in violent behavior were predicted by the interaction of maternal internal controls and adolescent internal controls. (Tate, 1999)

Clearly then, a lack of internal controls is important to the prediction of adolescent violent behavior and the continuity and change in that violent behavior over time. However, there is evidence that low internal controls also have a negative impact on children’s peer relationships (Stormshak et al., 1999; Pope & Bierman, 1999; Cillessen, van Ijzendoorn, van Lieshout, & Hartup, 1992). Forty-eight percent of peer-rejected boys show a pattern of behavior which is aggressive, impulsive, disruptive, and noncooperative (Cillessen, et al., 1992). The rejection of this type of behavior by children seems to be stable across contexts. In one study of childhood peer sociometrics, inattention/hyperactivity had a negative effect on peer preference across classrooms, with no effect for the amount of such behavior in the classroom in general. Other types of problem behavior (such as aggression and withdrawal) varied in their lack of classroom acceptance depending upon peer group norms in particular classes (Stormshak et al., 1999). Additionally, adolescents who engage in withdrawn or aggressive behavior seem at risk for peer rejection and victimization only when these behaviors are also accompanied by an irritable-inattentive pattern (Pope & Bierman, 1999).

Human aggression and violence is by definition a social phenomenon, as it occurs within the context of a social interaction (Baron & Richardson, 1994). No social interaction is more important to the adolescent than that with the peers. It is within the context of those relationships that young adolescents may begin to learn about socially-acceptable regulation of emotions and anger (Underwood, Hurley, Johansen, & Mosley, 1999). One of the most important developmental tasks of adolescence is gradual differentiation from parents in the attainment of autonomy. As teens begin to experiment with ways of living which are different from those of their parents, the judgement of peers becomes increasingly important (Allen, Moore, & Kuperminc, 1997). Peers provide the adolescent not only with new perspective, but also support through the transition to adulthood. Therefore, problematic peer relationships or the total lack of peer relationships can have a profound negative impact on adolescent social and emotional development (Bagwell, Newcomb, & Bukowski, 1998). Rejection by peers as early as grade three has been shown to be predictive of both internalizing and externalizing behavior problems in adolescence (Coie, Terry, Lenox, Lochman, & Hyman, 1995). Later in life, preadolescent peer rejection and the lack of friends have both been associated with poor academic adjustment and a variety of psychopathological symptoms in adulthood (DeRosier, Kupersmidt, & Patterson, 1994; Bagwell, Newcomb, & Bukowski, 1998).

Social relationships have been found to be especially strongly related to aggressive and delinquent outcomes. However, the exact nature of the relationship is disputed. Some researchers argue that preexisting aggression in the child or adolescent results in peer relational problems and rejection (Pettit, 1997; Dishion, Andrews, & Crosby, 1995; Pope & Bierman, 1999). Studies have shown that aggressive children, particularly those displaying a pattern of cognitive dysregulation (such as that mentioned above), are more likely to be rejected by their peers than others (Pope & Beirman, 1999). Dishion et al. studied the friendships of antisocial boys in early adolescence and found that these relationships were dominated by bossiness and coercive behavior and were perceived by the boys as only marginally satisfactory.

Other researchers suggest that it is rejection by mainstream peers which leads adolescents into association with a deviant peer group in which aggression is considered normative (Allen, Moore, & Kuperminc,1997; Patterson, Reid, & Dishion, 1992). Moffitt (1993) argues that deviant peer networks may be attractive to these adolescents because of the glamour associated with delinquency, particularly at a time when they are striving for differentiation from their parents and the status quo.

A third group of researchers argue that peer relational difficulties are the cause of aggression and not the result (Kupersmidt, Burchinal, & Patterson, 1995). In a study of third- through seventh-grade children, Kupersmidt et al. found that both dyadic and group peer relational problems were significant risk factors for aggression and delinquency.

The present study will seek to add to the existing literature concerning the relationship between low internal control and peer relational difficulty, and the relationship between such difficulty and violent behavior. We will attempt to address the question of the nature and direction of the relationship between peer problems and violence: Do peer problems lead to violent behavior, or does violence contribute to peer problems? Are both effects present? Unlike past studies, which have tended to use aggression or antisocial behavior as outcomes, the present study will focus particularly on violence. Violence is a severe form of aggressive behavior (Baron & Richardson, 1994), and it is very costly to both the adolescent and society at large (U. S. Department of Justice, 1999).

Many past studies of peer relationships and aggression and violence have used children or young adolescents (e.g. ages 12-13). However, this study will focus on older adolescents. In so doing, we hope to target a very critical developmental stage in which teens rely on peer relationships in their strivings for autonomy from their parents.

The present study will use as a starting point Tate’s model of violent adolescent behavior (1999). Tate found that low internal control is a significant predictor of adolescent violence, and that this particular cognitive deficit results from interactions with parents. We are interested in what happens to adolescents who have not acquired these internal controls as they attempt to form relationships with peers and break away from their parents. We will attempt to answer the following questions using an at-risk adolescent sample, examining outcomes when the teens are aged 16 and at age 18: Will low internal controls result in difficulty with peers in adolescence? Will peer relational difficulty result in violent behavior? How will peer relationships mediate the effects of low internal control on violence?