Beyond the individual: A consideration of context for the development of aggression

Beyond the individual: A consideration of context for the development of aggression

Aggression and Violent Behavior 11 (2006) 341 – 351 Beyond the individual: A consideration of context for the development of aggression Robert Cohen ...

155KB Sizes 0 Downloads 9 Views

Aggression and Violent Behavior 11 (2006) 341 – 351

Beyond the individual: A consideration of context for the development of aggression Robert Cohen a,⁎, Yeh Hsueh a , Kathryn M. Russell a , Glen E. Ray b a

Department of Psychology, The University of Memphis, Memphis, TN 38152-6400, United States b Auburn University at Montgomery, United States Received 27 October 2005; accepted 31 October 2005 Available online 22 December 2005

Abstract Historically, the focus for research on children’s aggression primarily has been on the characteristics of aggressors with a relatively recent interest in the characteristics of the victims of aggression as well. The present paper urges for the examination of aggression at multiple levels of social complexity particularly in terms of relationships and groups. A conceptual framework for engaging in this analysis is offered and a selective review of research is provided to demonstrate the utility of considering broader contextual factors in the study of children’s aggression. © 2005 Elsevier Ltd. All rights reserved. Keywords: Aggression; Context; Relationships; Group

Contents 1. Defining aggression . . . . . . . . . . . . . . . . . . . . . . 2. From subtypes to forms and functions of aggression . . . . . 3. Beyond the individual: a conceptual framework . . . . . . . . 4. Research on aggression beyond the individual: aggression and 5. Research on aggression beyond the individual: aggression and 6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . relationships . the peer group . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

342 343 344 345 347 348 349 349

Children's aggression has a long history in developmental psychology and a large body of research literature documents the intrapersonal characteristics of aggressors and the social and developmental consequences of being aggressive (see Coie & Dodge, 1998). Recently, victims of aggression and harassment have become a focus of interest. Researchers have begun to catalogue the intrapersonal characteristics of victims and the social and developmental ⁎ Corresponding author. Fax: +1 901 6782579. E-mail address: [email protected] (R. Cohen). 1359-1789/$ - see front matter © 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.avb.2005.10.004

342

R. Cohen et al. / Aggression and Violent Behavior 11 (2006) 341–351

consequences of being a victim (e.g., Juvonen & Graham, 2001). Although this emphasis on individuals has been quite fruitful, a number of researchers has advocated moving beyond the individual as the unit of analysis for the study of human development in general (e.g., Hinde, 1992) and of aggression in particular (Pierce & Cohen, 1995; Rubin, Bukowski, & Parker, 1998). Aggressors and victims exist within relationships and furthermore these relationships are embedded with the social nature of the peer group. The goal of the present review is to further promote this call for moving research efforts beyond the individual. We share the belief that much is to be gained not only analyzing different levels of social functioning (i.e., intrapersonal characteristics, relationships, groups) but also analyzing influences among these levels. This paper is divided into three major sections, with a selective review of the research literature in each section. The first section sets the stage by reviewing conceptual issues in the definition of aggression and types of aggression. We discuss some conceptual difficulties with the study of aggression and delineate a scheme for disentangling these difficulties. In the second section we offer a comprehensive framework for the study of aggression beyond the study of individual characteristics. In the third section we review literature that examines aggression at the dyadic (relationship) and peer group levels. 1. Defining aggression Coie and Dodge (1998) noted that aggression has proven to be difficult for researchers to unambiguously define. They offered the four component definition from Braine (1994), who defined aggressive behaviors as: (a) intentional acts, with (b) the potential for harm, (c) committed by an individual in an aroused physical state, and (d) perceived as aversive by the victim. By including intentions, this definition highlights cognitive components of aggressive acts and thus excludes behaviors such as “rough-and-tumble” play. By including the potential for harm and not just a damaging outcome, this definition avoids problems with defining aggression solely in terms of its consequences. Finally, both overt and covert types of aggressive behaviors would fit this definition. A second, and related, issue concerns measuring aggression, including who are the “best” informants about who is aggressive. Direct observation has been used (e.g., Price & Dodge, 1989), but as aggressive acts occur with a fairly low frequency, this has not been a particularly common approach. Most often researchers have relied on peer reports (e.g., Courtney & Cohen, 1996; Crick, 1996; David & Kistner, 2000; Warman & Cohen, 2000) and teacher reports e.g., Brendgen, Vitaro, Tremblay & Lavoie, 2001; Poulin & Boivin, 2000a). Studies also have used self-reports (Little, Jones, Henrich, & Hawley, 2003; Prinstein, Boergers, & Vernberg, 2001) and parent reports (Poulin & Boivin, 2000a). In their meta-analytic review of cross-informant agreement of behavioral and emotional problems, Achenbach, McConaughly, and Howell (1987) reported low agreement between peer and self reported problems (r = .26) and between teacher and self reported problems (r = .20). More recently, Little, Brauner, Jones, Nock, and Hawley (2003) reported low to moderate latent correlations between peer, teacher, and self reported overt and relational aggression (r's ranged from .13 to .52). Although these correlations are at best moderate, it is important to recognize that different informants have different access to the settings in which aggressive behaviors are occurring and thus each set of informants provides unique information from a unique perspective (Achenbach et al., 1987). Not only are the settings different, but also, as we will argue later, there are different levels of social functioning to consider when moving beyond the individual, and different informants have different perspectives on these levels. A third issue for the study of aggression has been proliferation of subtypes, many of which overlap (see Coie & Dodge, 1998; Underwood, Galen, & Paquette, 2001, for reviews). Among these overlapping constructs are physical/ verbal/overt/direct aggression, indirect/social/relational aggression, reactive/hostile aggression and proactive/ instrumental aggression. Despite this rather imprecise use of terms, distinctions that have garnered the greatest attention in recent years, and which are relevant for the present paper, are reactive and proactive aggression, and overt and relational aggression. The reactive and proactive aggression distinction (Dodge, 1991; Dodge & Coie, 1987; Price & Dodge, 1989) was developed from two different theoretical backgrounds. From Bandura's social learning theory, proactive aggression was defined as an acquired aggressive behavior used to influence or coerce others. Due to its instrumental nature, proactive aggression often is expressed without anger. Reactive aggression was derived from frustration–aggression theory and described aggressive behavior as a response to a perceived threat. The core of reactive aggression is retaliation, thus reactive aggression often is accompanied with a display of anger. Historically, most researchers define aggression as harm through verbal or physical acts, collectively termed overt aggression (Huesmann, Eron, Lefkowitz, & Walder, 1984). Consequently, researchers consistently found that boys

R. Cohen et al. / Aggression and Violent Behavior 11 (2006) 341–351

343

were more aggressive than girls and much of the later aggression research often excluded girls as participants. Recognizing that harmful acts may not always be overt or physically damaging, researchers identified a subtype of aggression believed to be relevant for girls. Building upon research on indirect aggression (Bjorkqvist, Lagerspetz, & Kaukiainen, 1992; Lagerspetz, Bjorkqvist, & Peltonen, 1988) and social aggression (Cairns, Cairns, Neckerman, Ferguson, & Gariepy, 1989), Crick and Grotpeter (1995) defined “relational” aggression as a form of aggression in which the aggressor harms the victim through the manipulation or damage of the victim's social relationships. Although relational aggression may be overt and “face to face,” it also includes covert acts. Researchers have defined these two sets of subtypes (reactive vs. proactive and overt vs. relational) with the expectation that they specify separate (or at least different) kinds of aggression, but many have reported moderate to high correlations between them. Reactive and proactive aggression, based on teacher ratings of boys behavior, typically show a moderate to high degree of correlation, ranging from r = .68 to .90 (Brendgen et al., 2001; Poulin & Boivin, 2000a,b; Price & Dodge, 1989). Studies using peer reports or teacher reports of overt and relational aggression have found a similarly wide range of correlations, with peer report correlations ranging from r = .37 to r = .87 (Crick, 1996; Crick, Casas, & Mosher, 1997; Crick & Grotpeter, 1995; David & Kistner, 2000; Tomada & Schneider, 1997) and teacher report correlations ranging from r = .22 to r = .77 (Crick, 1996; Crick et al., 1997; Tomada & Schneider, 1997). Despite positive correlations between reactive and proactive aggression, and between overt and relational aggression, researchers have argued for keeping them as distinct subtypes. In the case of the reactive/proactive dichotomy, the rationale for maintaining them as two distinct subtypes emphasizes the different outcomes associated with each, such as peer rejection and later violent behavior. Past research has consistently linked reactive aggression with peer rejection (Dodge & Coie, 1987; Poulin & Boivin, 2000b). Although proactive aggression is sometimes associated with peer rejection, Poulin and Boivin (2000b) found that proactive aggression was positively related to peer social preference after controlling for the effects of reactive aggression. In a longitudinal study, proactive aggression, after controlling for reactive aggression, was predictive of later delinquency related physical violence. Reactive aggression, after controlling for proactive aggression, was predictive of later dating violence (Brendgen et al., 2001). Unfortunately, after Dodge and Coie's (1987) initial study showing proactive and reactive aggression as more common for boys, studies have tended to use only male samples. Thus, it is difficult to draw conclusions about both the prevalence of these subtypes between the two genders and the outcomes of each for girls. Although research has identified outcome differences between overt and relational aggression, these have typically been moderated by gender. For example, there is some evidence that for girls, relational aggression predicts peer rejection even after accounting for physical aggression (Crick, 1996; Crick & Grotpeter, 1995; Rys & Bear, 1997) and overt aggression predicts peer rejection for both boys and girls (Crick, 1996; Dodge, 1983; Ray, Cohen, Secrist, & Duncan, 1997; Rys & Bear, 1997). Additionally, Crick and colleagues have consistently reported that boys are more overtly aggressive than girls, and girls are more relationally aggressive than boys (Crick, 1996; Crick & Grotpeter, 1995; Crick, Bigbee, & Howes, 1996). Thus gender has become a driving rationale for maintaining overt and relational aggression as distinct subtypes. 2. From subtypes to forms and functions of aggression Recently, there has been the suggestion that the varied types of aggression may be captured in better ways than by using subtypes. Little, Jones, et al. (2003) offered the useful framework of forms and functions of aggression. Forms describe the actual behaviors used by the child. Functions can be thought of as the purpose or motivation behind the aggression. Forms describe “what” the behavior was whereas functions answer “why” the behavior was enacted. Overt and relational aggression represents two different forms of aggression. They differ in the method of harm: overt aggression is physical and/or verbal harm and relational aggression is harm directed at social relationships. Reactive and proactive aggression represents two separate functions of aggressive behavior. They differ in the purpose for harm: proactive aggression serves to attain some goal and reactive aggression serves as retaliation. The research challenge for integrating forms and functions of aggression is to overcome the “overt bias” in the reactive and proactive measure. For example, one of the original proactive aggression items is, “This child uses physical force in order to dominate other kids” (Dodge & Coie, 1987). This item was designed to measure the proactive function (“dominate other kids”) but is nested in an overt form (“uses physical force”). In fact, reactive/proactive aggression has been studied exclusively in terms of overt aggression to the exclusion of covert aggression. In other words, functions of aggression must be balanced across forms of aggression. To accomplish this, Little, Jones, et al. (2003) designed a new

344

R. Cohen et al. / Aggression and Violent Behavior 11 (2006) 341–351

measure using self reports of aggression. Structural equation modeling confirmed the form and function distinction: reactive aggression was composed of both reactive overt and reactive relational aggression and proactive aggression was composed of both proactive overt and proactive relational aggression. Of particular interest were the correlations between the two forms of aggression and between the two functions of aggression in this model. A strong positive correlation was reported between overt and relational aggression (which was slightly higher than those reported in the literature reviewed above). In contrast to past findings, the correlation between reactive and proactive aggression was small and negative. The authors credit this independence of reactive and proactive aggression to the removal of the “form” variance from these two constructs. Thus, although the function of aggressive behavior (proactive or reactive) is likely to be relatively consistent for an individual, the form of their behavior (overt and relational) is likely to vary. As noted at the outset, the majority of research on children's aggression has focused on intrapersonal characteristics of aggressors (and victims). We believe that the fascination with categorizing and subcategorizing aggressive behaviors is indicative of this focus. The research and conceptual contributions by Little, Jones, et al. (2003) serve two purposes for the present paper. First, Little's work clarifies the relation among two of the most common distinctions currently drawn concerning types of children's aggression (proactive versus reactive aggression; overt versus relational aggression). Aggressive children tend to rely on a single function or motive (proactive or reactive) driving their aggression, but they are likely to use a variety of forms (overt and relational) in the service of those motives. A second contribution from Little, Jones, et al. (2003a), and important to the remainder of this paper, is that emphasizing the functions of aggressive acts, we believe, paves the way for introducing multiple levels of social functioning for the study of aggression and thus going beyond an analysis of intrapersonal characteristics. Considerations of forms and functions of aggression imply more than characteristics of children who engage in aggressive acts. A full understanding of forms and functions relies on context and a full understanding of context requires consideration of multiple social levels (e.g., dyadic and group) in both proximal and distal settings (see, Cohen & Siegel, 1991, for a consideration of facets of context). Before turning to research relevant to this extension of interest, the next section describes a conceptual framework for considering aggression within the context of social factors (and levels) beyond the individual. 3. Beyond the individual: a conceptual framework Several current theories in the social sciences are relevant for expanding our study of children's aggression beyond consideration of individual characteristics of aggressors and victims. These theories share the assumption that an individual is embedded within social systems which have both direct and indirect influences on both behavior and development. These theories differ in the emphasis they place on particular social systems for their levels of analysis and on what (and how) processes function to promote change or stability. It is beyond the scope of this paper to review these theories; rather, we choose to present a conceptualization in terms of a hierarchy of social systems, drawing largely from Hinde (1992) and Rubin et al. (1998), that we feel best suits the study of aggression. Hinde (1992) offered a hierarchy of social complexity that runs from physiological factors, to individual, to interactions between individuals, to relationships, to groups, to society. The analyses of (psycho) physiological factors and of individuals, of course, have long been within the domain of psychological research. Distinctions among interactions, relationships, and groups are important for the present paper. Interactions are defined as interdependent behaviors in which the behavior of one person serves, in turn, as a stimulus for, and a response to, the behavior of others. Relationships are more complex than interactions in that they connote a history of successive interactions between two people, as well as behavioral expectations for the future. Groups comprise multiple interacting individuals and multiple relationships and involve processes such as cohesion, and the construction of social hierarchies. There is no indication of reductionism here; one level of social complexity within the hierarchy is not better than another. Each level is depicted as dynamically connected to the level above and below it bidirectionally therefore each level is influenced by adjacent levels. A child's aggressor–victim relationship is influenced by the range and nature of social interactions experienced by each member of the dyad as well as being influenced by the nature of the peer group in which it is embedded. Furthermore, each level has unique properties associated with it, requiring distinctive descriptive and explanatory concepts. Thus, a researcher cannot completely understand functioning at one level by comprehensively studying only lower levels.

R. Cohen et al. / Aggression and Violent Behavior 11 (2006) 341–351

345

As an additional set of constructs, each of the six levels of social complexity is influenced by the physical environment and by sociocultural values, beliefs, and institutions. Also, Hinde (1992) proposes what he terms an objective versus subjective distinction for each level and for the physical environment and sociocultural influences. That is, there is always an insider perspective and an outsider perspective. For example, an observer can detail many objective characteristics of children's aggressor–victim relationships but there are aspects of those relationships that are known only by the participants. The reader is directed to the excellent review by Rubin et al. (1998) which draws upon Hinde's hierarchy for research on children's peer relations. Also, Pepler, Craig, and O'Connell (1999) offer an analysis of bullying from a dynamic systems perspective that complements this approach, portions of which are presented later in this paper. Finally, Cohen and Siegel's (1991) analysis of four facets of context for the study of development is consistent with Hinde's views. An important issue presented in Hinde (1992), that seems particularly pertinent to the study of children's aggression, is how an individual's characteristics are affected by relationships. This is a difficult issue. Not only do individual characteristics vary in their susceptibility to being influenced by relationships, but also this may change developmentally. For example, Hanish, Ryan, Martin, and Fabes (2005) reported that having a supportive friend was irrelevant to a preschool child's probability of becoming a victim in an aggressor–victim relationship. However, by kindergarten age, having a supportive friend served as an important buffer against this type of relationship forming. As another example, although social psychological research has investigated how characteristics such as similarity and complementarity relate to how individuals come to form and maintain relationships, Auhagen and Hinde (1997) argued that researchers must confront the very large number of dimensions that individuals vary on, despite a general lack of agreement about these dimensions. There are two important implications of Hinde's (1992) perspective that are of interest to the present paper. First, there is the clear recognition of multiple (hierarchical) and unique levels of social complexity, or social functioning, that are all important for a full understanding of behavior and development. Second, and related, findings from research at a particular level do not necessarily directly translate to, nor inform, interpretations of behavior at other levels. For example, having comprehensive knowledge of aggressors and of victims (individual level) will not necessarily give us a clear understanding of how a particular aggressor and a particular victim will interact. Further, having comprehensive knowledge of how a particular aggressor and a particular victim interact, will not necessarily give us a clear understanding of a particular aggressor–victim relationship. Research must be conducted at multiple levels of social complexity–individual, interaction, relationship, and group–to provide the data necessary to integrate the understanding of aggression. A corollary to this second implication is that we must construct measures that are relevant to the level of complexity we are studying and further we must be cautious in generalizing conclusions to one level based on data from measures designed for a different level. As an illustration of the value of these implications, we turn now to a consideration of research that evaluates relationship and group effects for the study of children's aggression. 4. Research on aggression beyond the individual: aggression and relationships An aggressive child directs aggressive behaviors to a target or victim. One strategy for studying this phenomenon would be to understand the physical, cognitive, and social characteristics of children who perform these behaviors as well as the physical, cognitive, and social characteristics of children who are the recipients. Another fruitful approach is to recognize that aggressors and victims are involved in a relationship. As discussed more fully later, a relationship, as opposed to interactions, connotes a history between members as well as expectations concerning current and subsequent encounters. Clinical notions of displaced aggression aside, for our purposes, aggression is always an interpersonal phenomenon; there is always harm and/or the intent to harm another person(s). Studying the interactions between aggressors and their victims is important. Following our discussion of Hinde (1992), studying aggressor–victim relationships is equally important. The call for examining aggression at the relationship level is not a new perspective in psychology (e.g., Mummendey, Linnewebber, & Loschper, 1984; Pierce & Cohen, 1995) although there has not been a great deal of empirical attention to this, with perhaps the exception of interest in bully–victim relationships. Using behavior observation of boys' dyadic play groups, Dodge, Price, Coie, and Christopoulos (1990) reported that 50% of the aggressive acts occurred between 20% of the dyads. Some of the dyads were mutually aggressive dyads (both members high in aggression) and some were asymmetrically aggressive dyads (only one member high on aggression). Aggression in mutually aggressive dyads was predominately reactive, whereas aggression in asymmetrical

346

R. Cohen et al. / Aggression and Violent Behavior 11 (2006) 341–351

dyads was primarily proactive. These findings demonstrate that aggression occurs in some relationships that elicit anger and mutual hostility and occurs in other relationships due to cues that aggression can lead to a successful instrumental act. The boys in Dodge et al. (1990), by design, were boys low in social standing in the peer group and were unacquainted with one another. As an extension, Coie et al. (1999), in an article aptly titled, “It Takes Two to Fight,” specifically evaluated the influence of relationship history of aggressive dyads, using play groups of third-grade boys made up of aggressive and non-aggressive classmates. Coie et al. (1999) compared mutually aggressive dyads to randomly selected dyads among classmates. Using Kenny's (1994) social relations model, variance in dyadic aggression was partitioned into individual member effects (actor and target) and into relationship effects. Results demonstrated that relational factors were as important as individual member effects for determining the frequency and the quality of both total aggression and proactive aggression. Coie et al. (1999) suggested that mutually aggressive boys may have been more likely to make attributions of hostile intent about each other than other dyads, and the attributions of hostile intent may have been a major cause of the aggression. In short, the dyadic relationship history was a major contributor to the aggression among the boys. Indeed, others have suggested that the hostile attribution bias of aggressive children, well documented in research on individual characteristics of aggressors (see Crick & Dodge, 1994), may indeed be best understood as a dyadic, relationship process (Perry, Williard, & Perry, 1990). Hubbard, Dodge, Cillessen, Coie, and Schwartz (2001) directly evaluated the hypothesis that social cognitive attribution biases of relationships were related to aggression. Using the same play groups of familiar boys methodology as Coie et al. (1999), Hubbard et al. (2001) included social cognition interviews about the peer. Hostile attribution biases toward a particular peer were related to actual observed reactive aggression toward that peer even when controlling statistically for actor and target effects. Social cognitive evaluations of outcome expectancies were related to occurrences of proactive aggression. These findings provide fairly direct support for the proposition that social cognitive evaluations are indeed related to not only aggression but to the function of aggression, and, importantly, these evaluations appear to be influenced by relationship information and not just social cognitive characteristics of the aggressor. Research has documented how forms of aggression may influence the quality of friendships relationships. Grotpeter and Crick (1996) reported that friends of overtly aggressive third through sixth graders were likely to cooperate with the aggressive friend in aggressive acts. Relationally aggressive children were likely to report higher levels of intimacy within their friendships than other children reported. Presumably based on this greater intimacy knowledge, relationally aggressive children were likely to engage in the use of relational aggression against the friend. Thus the form of the aggression that children used in the peer group influenced the quality of their friendships, with overtly aggressive children drawing their friends into aggressive acts within the peer group and children who were relationally aggressive using that aggression within the relationship. Both forms of aggression served to control the activities of the friend and the relationship. It should be noted that this pattern was evident for both boys and girls. Although a clear example of the relevance of relationship as a level of analysis, the research by Grotpeter and Crick (1996) is also relevant to analyzing the interplay between relationships and groups. As a final note, other research documents the importance of relationships, beyond the aggressor–victim relationship, within the peer group for the study of aggression and victimization. For example, stability of aggression over a one-year period has been shown to be related to the aggression level of mutual friends (Warman & Cohen, 2000). Likewise, stability of victimization was related to the victimization level of mutual friends (Browning, Cohen, & Warman, 2003). A number of researchers have shown that having a mutual friend reduces subsequent harassment (e.g., Hodges, Boivin, Vitaro, & Bukowski, 1999). In summary, the research reviewed above serves two purposes for the present review. First, this research supports the distinctions of forms and functions of aggression presented in the first section and extends these distinctions to aggression and relationships. Second, this research documents that although there may be continuity in constructs across levels of social functioning (i.e., hostile attribution bias operates at the individual and at the relationship level), these levels clearly interact with one another and each level brings with it unique considerations. Questions remain: How do children evaluate the characteristics of others in making relationship–relevant decisions? How might characteristics of children influence the particular nature of aggressor–victim relationships? Finally, since relationships are embedded within other social systems, in particular the peer group, how do other relationships in the group affect the aggressor and the victim? How does the aggressor–victim relationship affect the operation of the peer group? We turn next to research relevant for understanding aggression in relation to the peer group.

R. Cohen et al. / Aggression and Violent Behavior 11 (2006) 341–351

347

5. Research on aggression beyond the individual: aggression and the peer group Displays of aggression are influenced not only by relationships but by the peer group as well. A child who is socially rejected by the peer group will tend to become more aggressive over time (Bierman & Wargo, 1995). O'Connell, Pepler, and Craig (1999) reported that peers were present in 85% of the bullying episodes recorded on the school playground and that the presence of an audience was related to the persistence of the bullying behaviors. The audience for bullying and aggression is largely made-up of bystanders who do not interfere. Furthermore, Graham and Juvenon (2001) reported that 52% of sixth and seventh graders believe victims are victims because of factors within their control; i.e., they bring on the aggression. These data reinforce the notion that aggression, and aggressive relationships, influence and are influenced by the peer group. A fairly well studied area relevant to this discussion concerns aggression and peer rejection. Aggression and peer rejection have been shown to be highly related. For example, more than 50% of aggressive children are also sociometrically rejected (Bierman & Smoot, 1991) and approximately 50% of sociometrically rejected children are characterized as being aggressive (French, 1988). The current and long-term consequences of being rejected by peers are well documented (see Asher & Coie, 1990). Dodge et al. (2003) offered an extensive longitudinal analysis of the relation between peer rejection and aggression, following a sample of children from grades 1 to 3 to grades 5 to 7 and a second sample from kindergarten to third grade. Two findings are relevant to the present discussion. First, peer rejection was related to later aggression, controlling for early aggression. Second, this finding held only for children identified as high (above the median) in aggression when they were young. The authors conclude, “. . . social rejection by peers acts as a social stressor that increases a tendency to react aggressively among children who are so disposed (p. 391).” In the language of the present paper, group effects (peer rejection) and individual effects (aggressive characteristics) are dynamically related. It would be interesting, of course, to evaluate how aggressor–victim relationships fit into this association as well. However, it has also been noted that although most studies find a negative relation between aggression and peer acceptance, some research reports no association (e.g., Phillipsen, Bridges, McLemore, & Saponaro, 1999) and some report a positive association (e.g., Salmivalli, Kaukiainen, & Lagerspetz, 2000). Chang (2004), assessing Chinese adolescents in 82 different classrooms, found that classroom norms for aggression were strongly related to the expression of aggression in the classroom. Thus, the congruence of an individual's behaviors with the norms groups form must be considered. This social context model from Chang (2004) is completely in line with the levels of social complexity approach offered in the present paper. It should be noted in passing that peer groups, of course, are composed of subgroups which also have an impact on how interactions will be assessed. For example, Bellmore, Witkow, Graham, and Juvonen (2004), found that being aggressed against when part of the racial majority in a classroom was associated with greater loneliness than being aggressed against when part of the racial minority in a classroom of six graders. Consistent with Chang (2004), subgroups within groups create social norms which influence interactions, and also influence consequences at the individual level. These is a growing body of research documenting the dynamic interplay between individual characteristics of aggressors and characteristics of the peer group. It is certainly misleading to assume that peer groups are homogeneous in their response to aggressive acts. It is instructive to mention the work by Salmivilli (2001) on bully–victim group effects in this regard. There are what Salmivilli terms “participant roles” for peers leading up to, during, and after a bully–victim episode. Salmivalli, Lagerspetz, Bjorkqvist, Osterman, and Kaukiainen (1996) developed a peer-evaluation questionnaire to identify these participant roles. Five participant roles were identified: bully, assistant to the bully, reinforcer to the bully, defender of the victim, and outsider. Bullies (approximately 8– 11% of the children sampled) are active, initiative-taking children who start bullying episodes, are leaders in bullying situations, and often times make others join in the bullying behavior (the ring-leader). Assistants to the bully (approximately 7–12% of the children sampled) also are active participants in the bully–victim episode but are followers as opposed to leaders. Reinforcers to the bully (approximately 15–20% of the children sampled) are bystanders who encourage the bully either by providing direct feedback to the bully or by enjoying the situation; reinforcers serve as an important audience for the bully. Defenders of the victim (approximately 17–20% of the children sampled) are peers who offer supportive words to the victim, actively attempt to put a stop to the bullying, or alert adults to the situation. Outsiders (approximately 24–32% of the children sampled) are peers not present during the bully–victim episode, or move away or ignore the situation altogether. Next we discuss research that

348

R. Cohen et al. / Aggression and Violent Behavior 11 (2006) 341–351

directly evaluated observer's reactions to aggression, research that considered several of the levels of social complexity at the heart of this analysis. Within scenarios of hypothetical characters peer and situations that included provocation, Ray and Cohen (1997) manipulated the relationship between two peers (best friend, acquaintance, and enemy) and the intent behind the provocation (accidental, hostile, ambiguous). Eight- and 11-year-old children were asked questions about their interpretation of the events as well as their evaluation of the participants and their predictions about responses. Relationship information did not matter when intention clues were clear (accidental or hostile intent). That is, both the intentions of the aggressor and the responses of the victim were evaluated as positive if the cues were clearly accidental and were evaluated as negative if the cues were clearly hostile. Relationship information mattered if the cues as to intention were unclear. When intention cues were ambiguous, children generally viewed the interaction as positive for best friends and acquaintances, and as negative for enemies. Interestingly, and regardless of type of intention cue, children believed that future interactions would be affected by the provocation episode. They believed that best friends and acquaintances (but not enemies) would like each other less following the encounter. These data demonstrate the bidirectional influence of interaction and relationship data and hint at the implications at the group level. As observers (group level) children evaluated the aggression in a here-and-now mode; that is, intention cues were more salient than relationship information. In the absence of these cues, relationship information influenced the observer's decisions. However the observer integrated the here-and-now into their relationship constructs, believing that aggression would be harmful to the relationship thereafter. As an extension to Ray and Cohen (1997), Ray, Norman, Sadowski, and Cohen (1999) again gave children provocation scenarios and the intent behind the provocation again was manipulated (accidental, hostile, ambiguous). Also manipulated was the presumed relationship, this time to the target of the aggression (best friend, acquaintance, enemy). Intention cues generally operated as in Ray and Cohen (1997); aggressor's intention and victim's behavioral response were believed to be positive with accidental cues and to be negative with hostile cues. Ambiguous intention cues did not influence findings here with a presumed relationship between victim and observer, as they did when there was a presumed relationship between the aggressor and the victim. Observer– victim relationship information influenced the observer's evaluation of the victim after the provocation situation. Before the conflict, the observer expressed greater liking for the best friend than the acquaintance and greater liking for the acquaintance than the enemy. Following the aggressive encounter, the observer's liking for the best friend was unchanged. However, children reported liking the acquaintance and the enemy more. Perhaps the children were simply being sympathetic to the plight of the victim. It should be noted that this change in liking for enemies, although a statistically significant finding, still resulted in very low liking scores (less than 2.5 on a 5-point scale). As a final note for this section on aggression and the peer group, mention can be made of other researchers who show clear group effects on children's aggression. Stormshak, Bierman, Bruschi, Dodge and Coie (1999) empirically demonstrated with first graders that some behaviors are relatively uninfluenced by group social context and other behaviors are quite sensitive to group social context. Prosocial behaviors were associated with positive outcomes and inattentive–disruptive behaviors were associated with negative outcomes across different classroom social contexts. The consequences for an individual of aggressive behaviors and of withdrawal behaviors depended upon the group social context in which these behaviors were displayed. Aggression (and withdrawal) was associated with positive social consequences in some classrooms and with negative social consequences in others. To complicate this a bit more, aggression may relate to different consequences in different ways. Cillessen and Mayeux (2004) reported that both relational and physical aggression were negatively related to sociometric popularity (based on peer nominations for liking) of fifth through ninth graders but were positively associated with perceived popularity (peer nominations for popularity). Furthermore, this divergence increased over time for relational aggression and decreased over time for physical aggression. Stormshak et al. (1999) and Cillessen and Mayeux (2004) highlight the importance of group context for the evaluation of aggressive behaviors generally, as well as the evaluation of different forms of aggressive behaviors. 6. Conclusions The primary goal of the present review was to promote the call for research on children's aggression beyond an analysis of the individual. We subscribe to a definition of aggression that includes consideration of cognitive activity

R. Cohen et al. / Aggression and Violent Behavior 11 (2006) 341–351

349

and cognitive abilities. Rather than a continued proliferation of categories and subtypes of aggression, we support an integration of the conceptualization of aggression that distinguishes among forms and functions of aggression. Consistent with this form/function distinction, we strongly advocate for a systemic analysis of children's aggression, preferring a scheme like that of Hinde (1992), particularly in terms of analyzing individual, interaction, relationship, and group effects. We believe there is much to be gained with this multilayer social context approach. An appreciation for these various levels certainly changes both the scope of research projects as well as interpretation of findings. Research should be designed to carefully address unique contributions of each level. Investigations into the characteristics of aggressors and victims should continue, although we would also recommend the simultaneous evaluation of multiple levels. Clearly, care must be taken in drawing conclusions about one level from data collected in research designed to evaluate another level. Again, we believe that research that specifically addresses the dialectical interplay between levels will prove the most interesting and enlightening. For the developmental researcher, evaluating how these levels change in relation to age-related changes is of paramount interest. As noted, aggression is a context-sensitive behavior. The prevalence, intensity, nature, and consequences of being aggressive depend upon individual characteristics, previous interactions, social relationships, and the peer group. As a final consideration, we suggest that the study of context can be profitably expanded even beyond the scope of the conceptualization offered here. The conceptualization of context espoused here is fairly grounded in the immediate world of the child. Context can be expanded to include distal social systems as well (see Cohen & Siegel, 1991). That is, research relating individual, relationship, and group functioning, for example at home, to individual, relationship, and group functioning at school can further add to our understanding of contextual effects on children's aggression. As another elaboration, also from Cohen and Siegel (1991) research on children's aggression would benefit from a more detailed analysis of the opportunities and constraints offered by physical environments in which children engage. As a final expansion for consideration, examining cultural effects on the prevalence, intensity, nature, and consequences would also contribute to our understanding of children's aggression. Should aggression prove to have negative consequences across cultures, it would be important to evaluate the meaning and display of aggression as examined through the medium of culture. Acknowledgement This project was supported, in part, by a Center of Excellence Grant from the state of Tennessee to the Center for Applied Psychological Research, Department of Psychology, The University of Memphis. References Achenbach, T. M., McConaughly, S. H., & Howell, C. T. (1987). Child/adolescent behavioral and emotional problems: Implications of crossinformant correlations for situational specificity. Psychological Bulletin, 101, 213−232. Asher, S. R., & Coie, J. D. (1990). Peer rejection in childhood. New York: Cambridge University Press. Auhagen, A. E., & Hinde, R. A. (1997). Individual characteristics and personal relationships. Personal Relationships, 4, 63−84. Bellmore, A. D., Witkow, M. R., Graham, S., & Juvonen, J. (2004). Beyond the individual: The impact of ethnic context and classroom behavioral norms on victim's adjustment. Developmental Psychology, 40, 1159−1172. Bierman, K. L., & Smoot, D. L. (1991). Linking family characteristics with poor peer relations: The mediating role of conduct problems. Journal of Abnormal Psychology, 19, 341−356. Bierman, K. L., & Wargo, J. B. (1995). Predicting the longitudinal course associated with aggressive-rejected, aggressive (non-rejected), and rejected (non-aggressive) status. Development and Psychopathology, 7, 669−682. Bjorkqvist, K., Lagerspetz, K. M. J., & Kaukiainen, A. (1992). Do girls manipulate and boys fight? Developmental trends in regard to direct and indirect aggression. Aggressive Behavior, 18, 117−127. Braine, P. F. (1994). Hormonal aspects of aggression and violence. In A. J. ReisJr., & J. A. Roth (Eds.), Understanding and control of biobehavioral influences on violence, Vol. 2. Washington, DC: National Academy Press. Brendgen, M., Vitaro, F., Tremblay, R. E., & Lavoie, F. (2001). Reactive and proactive aggression: Predictions to physical violence in different contexts and moderating effects of parental monitoring and caregiving behavior. Journal of Abnormal Child Psychology, 29(4), 293−304. Browning, C., Cohen, R., & Warman, D. M. (2003). Peer social competence and the stability of victimization. Child Study Journal, 33, 73−90. Cairns, R. B., Cairns, B. D., Neckerman, H. J., Ferguson, L. L., & Gariepy, J. (1989). Growth and aggression: 1. Childhood to early adolescence. Developmental Psychology, 25(2), 320−330. Chang, L. (2004). The role of classroom norms in contextualizing the relations of children's social behaviors to peer acceptance. Developmental Psychology, 40, 691−702.

350

R. Cohen et al. / Aggression and Violent Behavior 11 (2006) 341–351

Cillessen, A. H. N., & Mayeux, L. (2004). From Censure to reinforcement: Developmental changes in the association between aggression and social status. Child Development, 75, 147−163. Cohen, R., & Siegel, A. W. (Eds.). (1991). Context and development. Hillsdale, NJ: Lawrence Erlbaum Associates. Coie, J. D., Cillessen, A. H., Dodge, K. A., Hubbard, J. A., Schwartz, D., Lemerise, E. A., et al. (1999). It takes two to fight: A test of relational factors and a method for assessing aggressive dyads. Developmental Psychology, 35, 1179−1188. Coie, J. D., & Dodge, K. A. (1998). Aggression and antisocial behavior. In W. Damon, & N. Eisenberg (Eds.), Handbook of child psychology, 5th Ed. Social, emotional and personality development, vol. 3 (pp. 779–861). New York: Wiley. Courtney, M. L., & Cohen, R. (1996). Behavior segmentation by boys as a function of aggressiveness and prior information. Child Development, 67, 1034−1047. Crick, N. R. (1996). The role of overt aggression, relational aggression, and prosocial behavior in the prediction of children's future social adjustment. Child Development, 67, 2317−2327. Crick, N. R., Bigbee, M. A., & Howes, C. (1996). Gender differences in children's normative beliefs about aggression: How do I hurt thee? Let me count the ways. Child Development, 67(3), 1003−1014. Crick, N. R., Casas, J. F., & Mosher, M. (1997). Relational and overt aggression in preschool. Developmental Psychology, 33, 579−588. Crick, N. R., & Dodge, K. A. (1994). A review and reformulation of social information-processing mechanisms in children's social adjustment. Psychological Bulletin, 115, 74−101. Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social–psychological adjustment. Child Development, 66(3), 710−722. David, C. F., & Kistner, J. A. (2000). Do positive self perceptions have a “dark side”? Examination of the link between perceptual bias and aggression. Journal of Abnormal Child Psychology, 28, 327−337. Dodge, K. A. (1983). Behavioral antecedents of peer social status. Child Development, 54, 1386−1399. Dodge, K. A. (1991). The structure and function of reactive and proactive aggression. In K. J. Pepler, & K. H. Rubin (Eds.), The development and treatment of childhood aggression. Hillsdale, NJ: Lawrence Erlbaum Associates. Dodge, K. A., & Coie, J. D. (1987). Social-information-processing factors in reactive and proactive aggression in children's peer groups. Journal of Personality and Social Psychology, 53, 1146−1158. Dodge, K. A., Lansford, J. E., Burks, V. S., Bates, J. E., Pettit, G. S., Fontaine, R., et al. (2003). Peer rejection and social information processing factors in the development of aggressive behavior problems in children. Child Development, 74, 374−393. Dodge, K. A., Price, J. M., Coie, J. D., & Christopoulos, C. (1990). On the development of aggressive dyadic relationships in boys' peer groups. Human Development, 33, 260−270. French, D. C. (1988). Heterogeneity of peer-rejected boys: Aggressive and nonaggressive subtypes. Child Development, 59, 976−985. Graham, S., & Juvonen, J. (2001). An attributional approach to peer victimization. In J. Juvonen, & S. Graham (Eds.), Peer harassment in school (pp. 49−72). New York: Guilford Press. Grotpeter, J. K., & Crick, N. R. (1996). Relational aggression, overt aggression, and friendship. Child Development, 67, 2328−2338. Hanish, L. D., Ryan, P., Martin, C. L., & Fabes, R. A. (2005). The social context of young children's peer victimization. Social Development, 14, 2−19. Hinde, R. A. (1992). Developmental psychology in the context of other behavioral sciences. Developmental Psychology, 28, 1018−1029. Hodges, E. V. E., Boivin, M., Vitaro, F., & Bukowski, W. M. (1999). The power of friendship: Protection against an escalating cycle of peer victimization. Developmental Psychology, 35, 94−101. Hubbard, J. A., Dodge, K. A., Cillessen, A. H. N., Coie, J. D., & Schwartz, D. (2001). The dyadic nature of social information processing in boys' reactive and proactive aggression. Journal of Personality and Social Psychology, 80, 268−280. Huesmann, L. R., Eron, L. D., Lefkowitz, M. M., & Walder, L. O. (1984). Stability of aggression over time and generations. Developmental Psychology, 20(6), 1120−1134. Juvonen, J., & Graham, S. (Eds.). (2001). Peer harassment in school. New York: Guilford Press. Kenny, D. A. (1994). Interpersonal perception: A social relations analysis. New York: Guilford Press. Lagerspetz, K. M. J., Bjorkqvist, K., & Peltonen, T. (1988). Is indirect aggression typical of females? Gender differences in aggressiveness in 11- to 12-year-old children. Aggressive Behavior, 14, 403−414. Little, T. D., Brauner, J., Jones, S. M., Nock, M. K., & Hawley, P. H. (2003). Rethinking aggression: A typological examination of the functions of aggression. Merrill-Palmer Quarterly, 49(3), 343−369. Little, T. D., Jones, S. M., Henrich, C. C., & Hawley, P. H. (2003). Disentangling the “whys” from the “whats” of aggressive behaviour. International Journal of Behavioral Development, 27, 122−123. Mummendey, A., Linnewebber, V., & Loschper, G. (1984). Aggression: From act to interaction. In A. Mummendey (Ed.), Social psychology of aggression: From individual behavior to social interaction (pp. 69−106). New York: Springer. O'Connell, P., Pepler, D., & Craig, W. (1999). Peer Involvement in Bullying: Insights and challenges for intervention. Journal of Adolescence, 22, 437−452. Pepler, D., Craig, W. M., & O'Connell, P. (1999). Understanding bullying from a dynamic systems perspective. In A. Slater, & D. Muir (Eds.), The Blackwell reader in developmental psychology (pp. 440−451). Malden, MA: Blackwell. Perry, D. G., Williard, J. C., & Perry, L. C. (1990). Peers' perceptions of the consequences that victimized children provide aggressors. Child Development, 61, 1310−1325. Phillipsen, L. C., Bridges, S. K., McLemore, T. C., & Saponaro, L. A. (1999). Perceptions of social behavior and peer acceptance in kindergarten. Journal of Research in Childhood Education, 14, 68−77. Pierce, K., & Cohen, R. (1995). Aggressor's and their victims: Toward a contextual framework for understanding children's aggressor–victim relationships. Developmental Review, 15, 292−310.

R. Cohen et al. / Aggression and Violent Behavior 11 (2006) 341–351

351

Poulin, F., & Boivin, M. (2000a). The role of proactive and reactive aggression in the formation and development of boys' friendships. Developmental Psychology, 36(2), 233−240. Poulin, F., & Boivin, M. (2000b). Reactive and proactive aggression: Evidence of a two-factor model. Psychological Assessment, 12(2), 115−122. Price, J. M., & Dodge, K. A. (1989). Reactive and proactive aggression in childhood: Relations to peer status and social context dimensions. Journal of Abnormal Child Psychology, 17(4), 455−471. Prinstein, M. J., Boergers, J., & Vernberg, E. M. (2001). Overt and relational aggression in adolescents: Social–psychological adjustment of aggressors and victims. Journal of Clinical Child Psychology, 30, 479−491. Ray, G. E., & Cohen, R. (1997). Children's evaluations of provocation between peers. Aggressive Behavior, 23, 417−431. Ray, G. E., Cohen, R., Secrist, M. E., & Duncan, M. K. (1997). Relating aggressive and victimization behaviors to children's sociometric status and friendships. Journal of Social and Personal Relationships, 14, 95−108. Ray, G. E., Norman, M., Sadowski, C. J., & Cohen, R. (1999). The role of evaluator–victim relationships in children's evaluations of peer conflict. Social Development, 8, 380−394. Rubin, K. H., Bukowski, W., & Parker, J. G. (1998). Peer interactions, relationships, and groups. In W. Damon, & N. Eisenberg (Eds.), Handbook of child psychology, 5th Ed. Social, emotional and personality development, vol. 3 (pp. 779–861). New York: Wiley. Rys, G. S., & Bear, G. G. (1997). Relational aggression and peer relations: Gender and developmental issues. Merrill-Palmer Quarterly, 43, 87−106. Salmivalli, C. (2001). Group view on victimization-empirical findings and their implications. In J. Juvonen, & S. Graham (Eds.), Peer harassment in school: The plight of the vulnerable and victimized (pp. 396−419). New York: Guilford Press. Salmivalli, C., Kaukiainen, A., & Lagerspetz, K. (2000). Aggression and sociometric status among peers: Do gender and type of aggression matter? Scandinavian Journal of Psychology, 41, 17−24. Salmivalli, C., Lagerspetz, K., Bjorkqvist, K., Osterman, K., & Kaukiainen, A. (1996). Bullying as a group process: Participant roles and their relations to social status within the group. Aggressive Behavior, 22, 1−15. Stormshak, E. A., Bierman, K. L., Bruschi, C., Dodge, K. A., & Coie, J. D. (1999). The relation between behavior problems and peer preference in different classroom contexts. Child Development, 70, 169−182. Tomada, G., & Schneider, B. H. (1997). Relational aggression, gender, and peer acceptance: Invariance across culture, stability over time, and concordance among informants. Developmental Psychology, 33, 601−609. Underwood, M. K., Galen, B. R., & Paquette, J. A. (2001). Top ten challenges for understanding gender and aggression in children: Why can't we all just get along? Social Development, 10, 248−266. Warman, D. M., & Cohen, R. (2000). Stability of aggressive behaviors and children's peer relationships. Aggressive Behavior, 26, 277−290.