Principal Investigator: Cynthia Hudley

Level of Intervention: Indicated

Target Population: Aggressive 10-12 year old African-American boys

References: Hudley & Graham (1993, 1995)

Theory (Risk & Protective Factors Targeted):
Childhood aggression is very stable over time and predicts a number of poor adolescent outcomes. Many of the social and economic correlates of childhood aggression are more prevalent among ethnic minorities. The authors developed their prevention program for aggressive youth based on an attribution theory of aggression, whereby the cognitive inability to appropriately interpret the intent of others actions may promote reactive aggression. The authors were also interested in the role of emotion in attribution theory and anger as a critical link between social cognition and aggression.

Description of Intervention:

Attributional Intervention:
12-lesson social-cognitive school based intervention was designed to teach aggressive students not to infer hostile peer intent in negative social interactions of ambiguous social origin through role play, story reading, and discussion of personal experiences. The intervention consisted of three components. The first focused on helping the boys accurately detect intentionality in social situations. The second was designed to increase the boys’ use of non-hostile attributions when interpreting the intent of others. The third focused on teaching the students appropriate responses to ambiguously caused negative outcomes. The program sessions were held twice-weekly for 6 weeks, in locations away from the regular classroom. A typical session lasted 40-60 minutes. Each group consisted of 6 students, four aggressive students and two non-aggressive students.

Attention Training:
The attention training condition consisted of a 12-session, nonsocial problem solving program based on the Building Thinking Skills program (Black & Black, 1984). The instructional format was similar to the attributional intervention.

Research Subjects:
Subjects were 120 poor (30% received reduced or free lunch), African American boys (from two schools in Los Angeles). From this pool, 78 boys were classified as aggressive and 42 were classified as non-aggressive. Aggressive students were identified with teacher ratings and peer sociometric nominations (positive, negative, aggressive behavior, prosocial behavior). To be aggressive, subjects had to have > median on the aggression subscale of the Teacher Checklist (Coie, 1990; Coie & Dodge, 1988), social preference scores < 0, and have at least twice as many aggressive as prosocial peer nominations. To be non-aggressive, subjects had to be at or below the median on the aggressive subscale of the Teacher Checklist, have a social preference score greater than 0, and have received at least twice as many prosocial as aggressive nominations.

Research Design:
Subjects (72 aggressive and 36 non-aggressive) were randomly assigned to one of three groups: treatment, attention training, or no-treatment control. Intervention and attention-training groups were divided into groups with four aggressive subjects and two non-aggressive subjects in each. The control group also contained aggressive and non-aggressive subjects.

After subject attrition the final sample consisted of 66 boys.

At post-test, the aggressive boys in the intervention group were rated as significantly less aggressive by teachers (p<.05) compared to the attention training or control groups. This treatment effect was also evident on specific items of the teacher ratings related to reactive aggression (p<.05).

Aggressive subjects’ judgements of intent, feelings of anger, and behavioral tendencies were assessed in four different types of hypothetical peer provocation situations (i.e. prosocial, accidental, ambiguous, and hostile). The overall intervention effect was significant (p<.001) and these effects were evident on all three variables, but only in ambiguous scenarios. Treatment group membership accounted for 33% of the multivariate variance. Boys who participated in the attributional program perceived significantly less hostile intent (p<.001), reported significantly less anger (p<.001), and endorsed less hostile behavior (p<.05) compared to the other two groups. There were no significant group differences in disciplinary office referrals (of any type).

In an analog task designed to assess subjects responses in actual peer interactions, aggressive boys who had participated in the intervention were significantly less likely to infer hostile intent (p<.001) compared to the other groups. Verbalizations during the task were coded and boys in the experimental condition received significantly higher scores (p<.01) than the other two groups. Higher scores indicated more neutral verbal behavior as opposed to aggressive verbal behavior.

Strengths & Limitations:
The Attributional Intervention Program targeted specific social-cognitive risk factors associated with childhood aggression by providing an intervention at the individual child level. In a randomized trial, support was found for the effectiveness of the program in altering the types of attributions and the emotional responses that the participants provided in ambiguous situations. At post-test, subjects were also rated as less aggressive by teachers and observed as less hostile in an actual ambiguous social interaction. These findings are promising given the quality of the evaluation design, the use of multiple sources of information including behavioral observations, and the fact that the teachers who provided student assessments were unaware of the child’s status.

A strength of the evaluation was that the authors addressed the issues of dosage, staff training, and program fidelity. All students were required to attend a minimum of 10 sessions and all subjects met this requirement. Group leaders were two African-American women with backgrounds in education. The instructors participated in 16 hours of training with the curriculum developer and conducted 3 experimental and 3 attention-training groups each. Weekly supervision was provided to the group leaders to monitor and discuss implementation integrity. It is important to consider however, that the sample size was relatively small, the subjects were all male, and the majority of the participants were African-American, which limit the generalizability of the findings. In addition, the absence of follow-up data leaves in question the long-term impact of the program.

The author compared the aggressive and non-aggressive subjects to determine the clinical significance of the study findings. Although aggressive boys in the intervention were rated by teachers as less aggressive after participating in the attributional program, they were still rated as significantly more aggressive than the non-aggressive subjects at post-test.

back to text     back to table of contents     external links to program     contact investigator