Principal Investigator: The Conduct Problem Prevention Research Group (Karen Bierman, John D. Coie, Kenneth A. Dodge, Mark T. Greenberg, John E. Lochman, and Robert J. McMahon)
Level of Intervention: Universal, Selected & Indicated components
Target Population: Elementary school-aged children at-risk for conduct disorder and other negative adolescent outcomes (e.g. school drop-out, delinquency) due to elevated teacher and parent-rated conduct problems at school entry.
References: CPPRG (1992, 1998, 1999a, 1999b).
Theory (Risk & Protective Factors Targeted):
Based on developmental theory of the development of conduct disorder and other adolescent problem outcomes. Multiple influences are thought to interact to increase likelihood of development of disorder. Combination of environmental risks (e.g. poverty, high crime neighborhood), pre and post-natal difficulties, individual differences in attention and activity level, and family factors of high conflict and instability that contribute to ineffective parenting.
Children at risk for developing conduct disorder often enter school unprepared for the social, emotional, and cognitive demands of that context. Their problem behaviors carry over to the school setting and they often experience academic failure. A weak bond between home and school, and a lack of support for parents and teachers, contributes to childrens adjustment problems in this context. Additionally, in at-risk neighborhoods, it is often the case that the density of children with problems is higher in the average classroom.
Research and past intervention efforts have shown that prevention should be intense during major transitions (such as school entry and the transition to middle school) and should involve the multiple contexts and sources of influence (i.e. home, school, & peers).
Description of Intervention:
The PATHS school-based curriculum (Kuche & Greenberg, 1994) was implemented in grades 1 through 5 for all students in the intervention and control schools (see program description under Universal Programs).
The enrichment program consisted of 5 additional components in grade 1: 1) parent training groups designed to promote the development of positive family-school relationships and to teach parents behavior management skills, particularly in the use of praise, time-out, and self-restraint, 2) home visits for the purpose of fostering parents problem-solving skills, self efficacy, and life management, 3) child social skills training groups, 4) child tutoring in reading, and 5) dyadic child friendship enhancement activities during the school day (peer-pairing).
In grade 2, parent and child groups met twice monthly and then shifted to a monthly schedule for all succeeding years of the project. Home visiting, tutoring, and child case management activities followed a criterion-based schedule in succeeding years.
The total sample across 1st to 3rd grade consists of three cohorts of children with 898 high-risk (intervention and control subjects) and 385 normative comparisons. The sample is 66% male and 34% female with 51% African-American subjects, 47% Caucasian subjects, and 2% belonging to another racial group.
Schools within four sites (Durham, NC; rural Central Pennsylvania; Nashville, TN; and Seattle, WA) were selected as high risk based on crime and poverty statistics. The identified schools were then divided into two matched sets and the sets were randomly assigned to intervention or control conditions.
Multistage screening of all kindergarten children from the schools were conducted using teacher ratings of disruptive behavior (TOCA-R) and parent ratings on similar items. The students with the highest the combined teacher-parent scores (approximately the top 10%) were identified as the selected target population and assigned to intervention (enrichment program) or control based on the school they entered in first grade.
Outcomes reported are from post-intervention (1st grade) and two follow-up points (at the end of 2nd and 3rd grades). Linear growth curve analyses indicated that across the first three years of the program, parent ratings of oppositional-aggressive behavior (Parent Daily Report; Chamberlain & Reid, 1987) declined over time for both groups but significantly more for intervention children compared to controls. Conduct problem measures assessed at end of each grade indicated that for teacher-rated behavior, intervention students exhibited significantly lower conduct problems (Teacher Observations of Classroom Adjustment-Revised; Werthamer-Larsson, Kellam, & Oueson-McGregor, 1990) than control subjects. Results of survival function analyses indicted that significantly more control subjects were identified as in need of special education services compared to controls. Specifically, there was a 26% reduction in the rate of special education assignment for the intervention students.
Strengths & Limitations:
FAST Track is a well designed randomized clinical trial of a preventive intervention that incorporates all of the current knowledge on the development of antisocial behavior. FAST Track is unique because of the size and diversity of its sample, and the fact that the same program is being conducted in both urban and rural settings. As such, it actually incorporates replications within its design and allows for comparisons across diverse settings and populations. The inclusion of such a diverse and well-representative sample will also allow for a great deal of generalization of the programs outcomes. In addition, very few programs integrate universal and selected components or target multiple risk and protective factors simultaneously across multiple settings and multiple socialization agents.
The initial findings indicating behavioral improvements at home and school, and reductions in special education referrals, are promising and provide hope for the prospect of reducing antisocial behavior and its associated poor outcomes in the long term. These results need to be replicated over time though, and with diverse measurement sources. This will be possible given the longitudinal design of the program and its inclusion of multiple data sources. In addition, given the multiple settings where the program is being conducted it is extremely important that implementation analyses be conducted to establish the fidelity of the program and rule out any differences between sites that may have contributed to the positive outcomes. Results for grade 1 are in press; results at grade 3 have only been presented at conferences.