Many aggressive children and adolescents have grown up in violent homes and neighborhoods, but some become violent for other reasons, as in the student shootings at Columbine High School and other schools. In particular, are there psychological or family variables that constrain children toward such a developmental pathway? Many analysts assume that demographic and socioeconomic factors account for most child aggression. Past studies on antecedents of aggression have reported that in general there is no one single cause of aggression. Specifically, there are probably multiple pathways by which individuals become aggressive and/or violent.
When this study was started, the focus in the aggression literature was on the environmental antecedents/risk factors of aggression, with little on psychological and personality factors. This study attempts to delve into some of the latter issues while revisiting the environmental factors previously described in the literature.
The study, which was funded by the National Institute of Child Health and Human Development, was presented at the biannual conference of the Society for Research in Child Development and the European Conference of Developmental Psychology.
The seven-year study showed that two characteristics strongly predicted the development of aggression (assessed by physical violence and externalizing behaviors): 1) physical punishment/ aggression by parents against the children and 2) inhibited temperament of the child. Further analyses in progress indicate that a negative self-representation may also predict later aggression and violence.
The most powerful predictor was parents’ physical punishment and violence, which also related to exposure to victimization and violent fantasies. Inhibited temperament was the only personality characteristic that predicted aggression, which suggests possible connections with the isolated, alienated children and adolescents who have committed school attacks. The important factors for predicting aggression seemed to be fearfulness in general and apprehension/lack of confidence about school, not simple shyness. Inhibited children may react to problems with peers and schools by making negative attributions against others and thus becoming more likely to act aggressively against others.
Inhibited children in this study were characterized as socially withdrawn, uncomfortable or distressed in new situations, and anxious about making new friends or trying new activities. Examples of aggressive behavior included fighting and lashing out at their peers both physically and verbally, insulting them, hitting and pushing them, attacking them with weapons, and, in extreme cases, criminal aggression, including murder.
In summary, the following factors did predict aggression:
- harsh physical punishment in the family
- peer victimization predicted aggressive fantasy, and aggressive fantasy predicted amount of aggression
- low self-esteem predicted amount of aggression, but not as strongly as the other factors
Most demographic and socioeconomic factors made little contribution to predicting aggression. Although aggression increased with age as expected, there was little effect of socioeconomic status, race, ethnicity, or gender. Also, many potential psychological and environmental predictors such as family environment (other than parental violence), sensation seeking, self-worth, and degree of fantasy played little role.
In summary, the following factors did not predict aggression:
- racial and ethnic group
- socioeconomic status sensation seeking—boys and older children were higher sensation seekers, but sensation seeking did not predict aggression
Findings indicate that most of the study children were not very aggressive, fitting national norms. Specifically, 25 out of the 440 children in the original sample were above the clinical line on the CBCL. So far, two children have been jailed for severe criminal aggression, including murder.
Representative sampling was conducted to include 440 children between the ages of 7 and 13 in Springfield, Massachusetts, a small city which has a population similar to national norms and represents a microcosm of the diversity and problems found in the U.S. Certain zip codes were oversampled to obtain equal numbers of Caucasians, African Americans and Hispanic/Latinos. An approximately equal number of male and female and Hispanic/Latino, Caucasian and African-American children were included from all socioeconomic levels. Assessment began at 7 to 13 years and has continued in four waves to 12 to 18 years of age.
Investigators include Malcolm Watson (Brandeis University), Kurt Fischer (Harvard Graduate School of Education), and the New England Research Institutes. Trained interviewers went into participants’ homes and conducted a series of two-hour interviews with both the study children and their mothers.
A wide range of demographic, family, and psychological factors were assessed in interviews, with standardized instruments including the Child Behavior Checklist (CBCL), Self in Relationships, the Slaby Fighting Scale and the Harter Self Esteem Scale, Conflict Tactics Scale, Reznick’s Childhood Inhibition Scale, Moos’s Family Environment Scale, Harter’s Perceived Competence Scale, Fischer’s Self-in-Relationships Interview, the Adolescent Health and YRB surveys, and the WISC, as well as additional interview items and assessments of neighborhood and household characteristics. These instruments helped researchers examine a number of psychosocial factors including:
Temperament: several kinds of temperament were measured, with the focus on what Jerome Kagan calls “inhibition.”
Fantasy Thought: the frequency of children’s day dreaming and the content of their thoughts were evaluated for aggressive tendencies.
Connectedness and Disconnectedness in Families: measures included factors associated with cohesion among family members, attachment styles, and children’s representations of themselves in important relationships.
Sensation Seeking: children were asked to what extent they seek out high sensation oriented activities.
Researchers used regression and structural equation modeling. Data was collected at four separate time points.
About the Researchers
Kurt Fischer is the Charles Bigelow Professor of Education at the Harvard Graduate School of Education. He is an expert on cognitive and emotional development and directs the school’s Mind, Brain, and Education program. Malcolm Watson is a professor of psychology at Brandeis University, where he studies aggression in children and adolescents, art and aesthetic concepts in children, and social adjustment as it relates to fantasy play and family conflicts.