Date of Award


Level of Access Assigned by Author

Open-Access Dissertation

Degree Name

Doctor of Philosophy (PhD)




Douglas W. Nangle

Second Committee Member

Cynthia Erdley

Third Committee Member

Jeffrey Hecker


The present study was designed to evaluate the effectiveness of a social-cognitive skills training curriculum targeting aggression in Head Start preschoolers. Developmental psychopathology research clearly suggests that early intervention with aggressive preschoolers is critical. Despite the fact that aggressive behavior becomes stable by the elementary school years, there has been relatively little interest in intervening with preschool-aged children. In the present study, the COMPASS program is a 12-session curriculum teaching basic individual social skills (e.g., sharing, questioning, listening). Following the work of Mize and Ladd (1983; 1990), the curriculum was puppet-facilitated and employed the components of instruction, modeling, rehearsal, and evaluation feedback. Two Head Start centers comprised the treatment group and two centers comprised the control group, with a total of 80 preschoolers participating. Assessments were made at pre-treatment, post-treatment, and three-month follow-up. The assessment battery consisted of the Social Skills Rating System-Teacher Form, teacher-rated relational aggression, Aggression subscale of the Achenbach Caregiver-Teacher Report Form, peer ratings of acceptance, and the Enactive Social Knowledge Interview. It was hypothesized that aggression would decrease, while prosocial skills and peer acceptance would increase, more so in the treatment than the control group. The SSRS-T, relational aggression ratings, and aggregate peer acceptance ratings had significant site differences at pre-treatment, which were controlled for statistically. No significant results emerged with respect to aggression and social skill levels. With respect to peer acceptance ratings, both the treatment and control groups decreased between post-treatment and follow-up, although the treatment group decreased slightly, but significantly, more than the control group. Several post-hoc hypotheses are offered as explanation for the generally null findings, including applicability of the Mize and Ladd model to a Head Start population, inability to adequately control for site differences, length of the treatment, and difficulty adding to the effectiveness of the Head Start program alone. Several directions for future research are discussed, including the addition of more sessions to the curriculum, a parent-involvement component, and a contingency management program. It would also be helpful to evaluate COMPASS in a larger number of Head Start centers and to extend follow-up assessment beyond three months.