Updated on Dec 23, 2009

The psychological literature generally concurs that human aggression is any behavior intended to harm another living being who is motivated to avoid such behavior. Aggressive behavior can be physical, verbal, or relational; implies action or the threat of action; and can be characterized as either direct or indirect. Physical and verbal aggression, as their names suggest, describe physical harm or insults or threat of such actions. Relational aggression, by contrast, refers to behaviors that cause emotional harm by manipulating or damaging a victim's relationships with his or her peers or by injuring one's feelings of group acceptance (Dodge, Coie, & Lynam, 2006). Both physical and verbal aggression can be direct (e.g., a physical assault or a derogatory remark) or indirect (e.g., destroying another's prized possession or insulting a victim behind his or her back). Some, but not all research (see Bushman & Anderson, 2004) further distinguishes two types of aggression by the motives underlying the behavior. Proactive or instrumental aggression is a means to another end; the harm is directed toward attaining objects, privileges, or similar self-serving ends. Reactive, retaliatory, or hostile aggression is behavior motivated to harm another that is displayed in anger as a response to a perceived threat or provocation (Dodge, Coie, & Lynam, 2006).


Aggression in childhood and adolescence has typically been framed in one of a few prominent theories. Social learning theory (Bandura, 1973) is one of the earliest and most enduring theories of aggression that influenced thinking on children's aggressive behavior. Bandura theorized that children learned aggression by observing the behavior and the consequences of that behavior for others, a proposition made famous through the various “Bobo” studies that demonstrated the power of both live and filmed models to influence children to enthusiastically hit a Bobo doll. Closely related to modeling, this theory asserts that children who are positively reinforced or who observe others being positively reinforced for aggression are much more likely to persist in this behavior. Models provided through the mass media extend the power of observational learning far beyond the child's immediate environment. Importantly, media models are typically reinforced for aggression (Wilson et al., 2002), making them extremely powerful according to social learning theory.

Other recent theories have implicated family process in the development of aggression in childhood. Chief among them is coercion theory (Patterson, 2002), which postulates that the development of aggression is largely explained by coercive family processes in which parents and children mutually train one another's behavior. Children aggress, parents demand compliance, children escalate their aversive behavior, and parents escalate their demands but ultimately yield to the child's behavior, tacitly reinforcing children's aggressive behavior and perhaps modeling aggression in the process.

Social information processing theory (Dodge, Coie, & Lynam, 2006)) has been another generative model for the study of childhood aggression. Dodge (1986) originally developed a model of information processing that identified a hostile attributional bias as a foundation for the display of children's angry or reactive aggression in social situations His five-step model posited that children first encode and then interpret cues; highly aggressive youth may presume hostility from peers and selectively attend to cues that support that hostile attributional bias. The next steps include access or construction of a response, selecting a response, and enacting that response; for aggressive children with a hostile attributional bias, that response is most often aggressive. A further elaboration included an additional intermediate step of selecting a goal or preferred outcome of a response. As well, reciprocal effects between the child and the child's social environment (e.g., peer influences and reactions) and existing cognitive structures (e.g., memory stores, social schemas) that the child brings to the situation, were incorporated into the model.

The addition of constructs from social psychology (e.g., schemas) foregrounded linkages between this and another social information processing model of childhood aggression (Huesmann, 1988) that has influenced the study of media effects and aggression. Huesmann and colleagues argued that a child acquires aggressive scripts through observational learning (in both the proximal environment and the media) and perceived reinforcement of aggressive behavior. The accumulation and subsequent networking of multiple aggressive scripts into cognitive schemas results in social behavior that emphasizes aggression. Normative beliefs, or cognitions about what is right for the self, are also key cognitive structures in the Huesmann model. Aggressive children are presumed to have a greater store of aggressive scripts as well as normative beliefs that condone more aggression.

These information processing theories have been blended into a unified model (Huesmann, 1988) that highlights four decision points. A youth first perceives danger from the environment and next searches for and retrieves scripts from memory that are relevant to the situation. The youth then evaluates scripts stored in memory to decide what actions are acceptable, what actions lead to the most desired goal, and which actions are actually feasible. Finally, the youth evaluates the expected responses to any action. An aggressive child will selectively perceive cues and inappropriately attribute hostility when none exists. That child will also have a larger store of aggressive scripts from which to select and a greater propensity to positively evaluate aggressive scripts. This combination typically leads to a display of aggression.

The most comprehensive of the developmental theories of aggression, the biopsychosocial model (Dodge, Coie, & Lynam, 2006) is a transactional developmental model that incorporates biological dispositions; sociocul-tural contexts; and experiences with parents, peers, and social institutions. This model proposes that genetic bases and prenatal insults; life experiences that involve harsh treatment, rejection, and failure; family experiences that include poverty, neighborhood and family instability, harsh discipline, and limited parental education; and excessive early exposure to media violence all unfold in a transactional relationship during the child's development. Many of these factors also represent negative social experiences that can lead to dysfunctional patterns of social information processing, and these information processing patterns link the broad life experiences to the individual display of aggression. This comprehensive model most effectively demonstrates the multiple points of convergence (e.g., observational learning, social information processing, reinforcement) in all of the major theories of aggression in childhood and adolescence.


One common measurement strategy is checklists or rating scales that are completed by any combination of teachers, parents, and the child. A typical instrument widely used in research, schools, and clinical practice is the Achenbach Child Behavior Checklist, a scale with separate forms for parents (CBCL), teachers (TRF), and students (YSR). Scores on the multiple forms can be compiled to examine distinct problem areas across informants, including aggression and delinquent behavior. Other multiple informant rating scales for aggression in childhood and adolescence include the Social Skills Rating System, the Eyberg Child Behavior Inventory and Sutter-Eyberg Student Behavior Inventory, and the Behavioral Assessment Scale for Children (Buros Institute, 2007).

Direct behavioral observation, often described as naturalistic observation, is considered by some (Hudley, 2006) to be the most effective strategy for assessing aggression in childhood and adolescence. Behavior does not occur in a vacuum, and direct observations are able to capture the interactive environment in which the behavior exists, including the antecedent conditions and the consequences that elicit and maintain aggressive behavior. Naturalistic observations include recording behaviors in their natural setting and a descriptive coding system that requires a minimum of inference from observers and coders. Observational codes can measure behavior in several ways. Event recording simply tallies the frequency of a given behavior during the observation period. Interval recording similarly captures frequency, but divides the observation period into segments and counts the number of segments in which the target behavior is displayed, either throughout the interval or at a particular time point in the interval. Duration recording measures the length time a behavior lasts. Functional behavior assessment, an observational strategy, assesses antecedents, frequency, duration, and consequences of the aggressive behavior for the target child and others in the environment to determine the functions that the aggressive behavior serves for the child. In spite of obvious benefits of direct observation, the strategy can be limited by several problems (Hudley, 2006). Behaviors must be clearly defined, and observers must be trained to fully understand the exact behaviors that are to be captured. Observer bias or the tendency to see what one expects to see is especially troublesome in direct observation of aggression. In addition, particularly with adolescents, participant reactivity to the presence of observers can change or eliminate the exact behaviors that are the target of the observation. Finally, in school settings direct observations can be labor intensive and exceed the resources that are available on site.

Sociometric assessments continue to be used widely in research on childhood aggression, although their use in educational and clinical practice has declined dramatically (Hutton, Dubes, & Muir, 1992). The most common sociometric strategies involve children providing assessments of peers in school. These strategies have demonstrated substantial predictive validity for future negative outcomes. Peer nomination assessments typically ask students to nominate classmates or students in their grade who fit certain characteristics (e.g., starts fights), students they prefer (like most), or reject (like least) (see Hudley & Graham, 1993). The Class Play (Masten, Morrison, & Pelligrini, 1985), a variant of peer nomination techniques, directs children to cast their classmates in a variety of roles, either positive or negative, in a hypothetical play that the class will perform. In contrast, peer-rating procedures allow each child to rate every other student in the class on specified characteristics, preferences, or rejection. For young children, responses can be recorded with a series of faces rather than a numerical rating scale. Ethical concerns surrounding children expressing negative opinions about their peers have been raised, but no harmful reactions from participation in sociometric procedures have been documented (Iverson & Iverson, 1996).


Physical aggression in childhood is a relatively stable phenomenon. Highly aggressive boys and girls in middle childhood often continue to be aggressive in adolescence and adulthood, although the links between early and later behavior may not be as strong for girls as they are for boys (Dodge, Coie, & Lynam, 2006). Certainly not all highly aggressive children are aggressive and violent as they grow older, but such children are overrepresented in the population of aggressive and violent adolescents and adults. For example, longitudinal research has found that children at age 8 who were rated by their peers as highly aggressive self-reported high rates of aggression at age 18 and physical aggression toward spouses and children at age 30. Most troubling, those who were parents by age 30 reported high aggression for their children (Hues-mann, Eron, Lefkowitz, & Walder, 1984).

From early infancy to the early school years, children's aggression is typically expressed through temper tantrums and direct physical means (hitting, pulling, pushing, etc.). In the general population, simple displays of direct aggression peak between 2 and 3 years of age and then decline, largely due to children's developing social and cognitive abilities. However, some children remain aggressive through adolescence and beyond, and the form of their aggressive behavior changes across development. These children may show any combination of increasingly dysfunctional social cognitions, increasing aggression in interpersonal situations, and steady declines in prosocial behavior (Dodge, Coie, & Lynam, 2006). As well, children shift the site of their behaviors from early childhood through adolescence, as increasing age allows them to spend more time in the community and less time under direct adult supervision. As the setting for behavior changes, new forms of behavior may emerge; the 8 year old who fights at school may engage in physical mugging at 17. Finally, the intensity or force of aggression also changes over time for some aggressive children, in a trajectory that moves from simple hitting and pushing in early childhood to physical assaults with deadly objects in late adolescence and early adulthood.


Childhood aggression carries a host of negative developmental consequences that persist and accumulate over time, including delinquency and criminality, peer rejection, poor school adjustment, and mental health concerns. Although highly, overtly aggressive elementary age children are often rejected by their peers, not all aggressive children are rejected; those most likely to be rejected are socially incompetent and retaliate aggressively at times when peers find the behavior unwarranted and in violation of peer norms. Children and adolescents who are both rejected and aggressive in elementary school experience significantly higher rates of self reported depression and lower rates of peer rated friendship than their average peers. Aggressive children therefore may find themselves part of a deviant peer group composed of other, similar children who reinforce aggression, delinquency, and other behaviors (Dodge, Coie, & Lynam, 2006).

Another robust consequence of early aggressive behavior is adolescent delinquency and adult criminality, particularly among men. Longitudinal data find that men convicted of a violent crime by age 30 were more than three times as likely to have been rated by teachers and parents as highly aggressive in childhood or early adolescence than a comparison group of men who were not convicted of such crimes (Loeber et al., 2005). Highly aggressive students are also perceived as generally less academically successful, more behaviorally disruptive, and less motivated in class (e.g., off task, not doing homework) in comparison to their nonaggressive peers (Cairns, Cairns, & Neckerman, 1989). Finally, high levels of aggression have been cited as among the primary reasons that children and adolescents are referred to mental health services (Dodge, Coie, & Lynam, 2006).


Because cognitive mediators have been firmly linked to children's aggressive behavior, school based programs have often focused on cognitive-behavioral interventions. Cognitive-behavioral strategies address specific cognitive distortions that cause the display of aggression. One such program, the BrainPower Program (Hudley, 2003; Hud-ley et al., 1998), modifies attributional bias that supports childhood aggression, as described earlier. This program teaches aggressive children to recognize that negative social outcomes with peers may sometimes be caused by accidental rather than intentionally hostile causes. Students are taught to effectively search for social cues, initially attribute ambiguous negative outcomes to accidental causes, and develop less verbally and physically aggressive behavioral responses.

However, given the overwhelming evidence that multiple interpersonal processes regulate childhood aggression (Dodge, Coie, & Lynam, 2006), aggression reduction programs are most effective as one part of a comprehensive intervention to support the healthy development of children, families, and communities. One example of many such programs, Families and Schools Together (FAST) (McDonald & Sayger, 1998), incorporates the full range of relationships and settings impacted by youths' antisocial behavior. FAST builds, sustains, and enhances relationships between youth and their families, peers, teachers, and other community members. FAST brings a group of families from the same community together for weekly activities and two years of monthly school-community meetings to facilitate the development of mutual support networks in the community and the school.

Classrooms with lax and inconsistent discipline and schools that do not address problems of aggression and bullying see more physical aggression among students (Osher et al., 2004). Witnessing aggression (e.g., fighting, bullying, weapons) in school increases physically aggressive behavior, more so for girls than for boys (O'Keefe, 1997). However, zero tolerance discipline policies for physically aggressive behavior reduce attendance, motivation, and engagement in school among those students who do not run afoul of the policy. These negative policies typically do not reduce physical fighting in schools and actually promote aggression and youth violence in the community by expelling children who need the structure and education provided by school. The overall goal for schools must be the two-sided process of reducing aggressive behavior and promoting a peaceful, positive climate, and that responsibility can be shared by students and adults alike. School groups organized specifically to promote non-violence often benefit from improved school climate and reduced levels of aggressive, antisocial behavior, including bullying, fighting, and teasing (Office of Elementary and Secondary Education, 2002).


Bandura, A. (1973). Aggression: A social learning analysis. Englewood Cliffs, NJ: Prentice-Hall.

Barnow, S., Lucht, M., & Freyberger, H. (2005). Correlates of aggressive and delinquent conduct problems in adolescence. Aggressive Behavior, 31, 24–39.

Buros Institute (2007). Test Reviews Online. Retrieved April 4,

Cairns, R., Cairns, B, & Neckerman, H. (1989). Early school dropout: Configurations and determinants. Child Development, 60, 1437–1452.

Dodge, K. (1986). A social information processing model of social competence in children. In M. Perlmutter (Ed.), The Minnesota Symposium on Child Psychology (Vol. 18, pp. 77–125). Hillsdale, NJ: Erlbaum.

Dodge, K., Coie, J., & Lynam, D. (2006). Aggression and antisocial behavior in youth. In W. Damon & R. Lerner (Series Eds.) & N. Eisenberg (Vol. Ed.), Handbook of child psychology: Vol. 3. Social, emotional, and personality development (6th ed., pp.719–788). New York: Wiley.

Hudley, C. (2003). Cognitive-behavioral intervention with aggressive children. In M. Matson (Ed.), Neurobiology of aggression: Understanding and preventing violence (pp. 275–288). Totowa, NJ: Humana Press.

Hudley, C. (2006). Who's watching the watchers? The challenge of observing peer interactions on elementary school playgrounds. New Directions for Evaluation, 110, 73–85.

Hudley, C., Britsch, B., Wakefield, W., Smith, T., DeMorat, M., & Cho, S. (1998). An attribution retraining program to reduce aggression in elementary school students. Psychology in the Schools, 35, 271–282.

Huesmann, L. R. (1988). An information processing model for the development of aggression. Aggressive Behavior, 14, 13–24.

Huesmann, L. R., Eron, L., Lefkowitz, M., & Walder, L. (1984). Stability of aggression over time and generations. Developmental Psychology, 20, 1120–1134.

Hutton, J., Dubes, R., & Muir, S. (1992). Assessment practices of school psychologists: Ten years later. School Psychology Review, 21, 271–284.

Iverson, A., & Iverson, G. (1996). Children's long-term reactions to participating in sociometric assessment. Psychology in the Schools, 33, 103–112.

Loeber, R., Pardini, D., Homish, D., Wei, E., Crawford, A., Farrington, D., et al. (2005). The prediction of violence and homicide in young men. Journal of Consulting and Clinical Psychology, 73, 1074–1088.

Masten, A., Morrison, P., & Pelligrini, D. (1985). A revised class play method of peer assessment. Developmental Psychology, 21, 523–533.

McDonald, L., & Sayger, T. (1998). Impact of a family and school based prevention program on protective factors for high risk youth. Drugs and Society, 12, 61–86.

Office of Elementary and Secondary Education. (2002). Studentled crime prevention: A real resource with powerful promise. Washington DC: U.S. Department of Education, Safe and Drug-Free Schools Program.

O'Keefe, M. (1997). Adolescents' exposure to community and school violence: Prevalence and behavioral correlates. Journal of Adolescent Health, 20, 368–376.

Osher, D., VanAcker, R., Morrison, G., Gable, R., Dwyer, K., & Quinn, M. (2004). Warning signs of problems in schools: Ecological perspectives and effective practices for combating school aggression and violence. Journal of School Violence, 3, 13–37.

Patterson, G. (2002). The early developmental of coercive family process. In J. Reid, G. Patterson, & J. Snyder (Eds.), Antisocial behavior in children and adolescents: Developmental theories and models for intervention (pp. 25–44). Washington, DC: American Psychological Association.

Wilson, B., Smith, S., Potter, W. J., Kunkel, D., Linz, D., Colvin, C., et al. (2002). Violence in children's television programming: Assessing the risks. Journal of Communication, 52, 5–35.

Add your own comment