Addictive Behaviors, Vol. 25, No. 6, pp. 899–911, 2000
Copyright © 2000 Elsevier Science Ltd.
Printed in the USA. All rights reserved
0306-4603/00/$–see front matter

PII S0306-4603(00)00126-X

899

Pergamon

Session: Universal to Indicated Models of Prevention: Robert McMahon, Chair

A MULTILEVEL APPROACH TO FAMILY-CENTERED
PREVENTION IN SCHOOLS: PROCESS AND OUTCOME

THOMAS J. DISHION and KATHRYN KAVANAGH

University of Oregon

Abstract —

The Adolescent Transitions Program (ATP) is a multilevel approach to family-
based interventions within a middle-school setting. The intervention strategy is based on an
ecological framework for studying social and emotional development in children and adoles-
cents, emphasizing a network of contextual factors within which parenting is both directly and
indirectly influential on the development of problem behavior. The ATP model includes a
universal, selected, and indicated strategy for serving families with young adolescents. The
model is designed to address the needs of families of young adolescents that present with a
range of problem behavior and diverse developmental histories. The three interventions lev-
els are described, and outcome data are presented, that support the effectiveness of the ATP
model. This approach and the associated data are consistent with a broad literature support-
ing the effectiveness of family interventions, especially for high-risk youth. The effective im-
plementation of family interventions within a school context suggests that these interventions
can make a significant contribution to reducing problem behavior and substance use from a
public health perspective. © 2000 Elsevier Science Ltd.

Key Words.

Prevention, Families, Substance use, Problem behavior, Adolescents.

An ecological perspective on adolescent problem behavior uses a wide-angle lens to
study development and to design effective interventions (Bronfenbrenner, 1989; Dish-
ion, French, & Patterson, 1995). With respect to development, families have both a di-
rect and indirect effect on the etiology of adolescent problem behavior (see Fig. 1 for a
summary). The direct effect is clear, disrupted, and chaotic, and coercive parenting co-
varies with levels of antisocial behavior in preschool (Gardner, 1989; Shaw, Owens,
Vondra, Keenan, & Winslow, 1996) and childhood (Loeber & Dishion, 1983; McCord,
McCord, & Howard, 1960; Patterson, 1982; Patterson, Reid, & Dishion, 1992).
Antisocial behavior in childhood appears to provide a basis for accumulative risk
associated with peer rejection, poor academic skills, and eventual involvement in a de-
viant peer group (Dishion, Patterson, Stoolmiller, & Skinner, 1991; Patterson et al.,
1992). Because of this process of accumulative risk, we think that youth starting early
have a poorer prognosis than those who start late, with respect to persistence and seri-
ousness of future problem behaviors, such as delinquency and substance use (Dishion,
Capaldi, & Yoerger, 1999; Moffitt, 1993; Patterson, 1993). Parenting seems to have an
indirect effect by virtue of the role of family management, in general, and parent mon-

This project was supported by grant DA 07031 from the National Institute on Drug Abuse at the National
Institutes of Health to the first author, by grant MH 46690 from the National Institute of Mental Health to
John Reid, Ph.D., and by grant MH 37940 from the National Institute of Mental Health to Deborah
Capaldi, Ph.D.
Address correspondence to: Thomas J. Dishion, Ph.D., University of Oregon, Child and Family Center,
Department of Psychology, 195 West 12th Avenue, Eugene, OR 97401-3408; E-mail: tomd@darkwing.
uoregon.edu

900 T. J. DISHION and K. KAVANAGH

itoring, in particular, in modulating at-risk youth involvement and exposure to deviant
peers (Dishion & McMahon, 1998; Patterson, Reid, & Dishion, 1992).
The literature suggests that multiple risk factors disrupt parental functioning (see
Fig. 1), including poverty (Conger et al., 1992; Elder, Van Nguyen, & Caspi, 1985),
marital discord (Hetherington & Clingempeel, 1992), and parental substance use
(Chassin, Presson, & Sherman, 1990; Dishion, Reid, & Patterson, 1988). Additionally,
oppression and the lack of resources and stigmatization (Duran & Duran, 1995;
McLloyd, 1990) undermine the cross-generation transmission of socialization prac-
tices from parent to child, and eventually, from the child to the next generation.
Parenting practices can serve a protective function within a disrupted community
(Sampson & Laub, 1994). Parent supervision, in particular, may protect youth from
escalating patterns of problem behavior in high-risk neighborhoods (Dishion & Mc-
Mahon, 1998). Supporting the caregivers’ use of behavior management skills and
building strong parent-child relationships have been shown to reduce early opposi-
tional problems in the preschool years (Webster-Stratton, 1984), antisocial behavior in
middle childhood (Patterson, 1974; Patterson, Dishion, & Chamberlain, 1993), and
problem behavior and substance use in early adolescence (Dishion, Andrews, Ka-
vanagh, & Soberman, 1996; Henggeler, Melton, & Smith, 1992).
An ecological framework is useful for designing interventions that address mental
health problems as a public health issue (Biglan, 1995). A critical problem in child
clinical psychology is that interventions designed in university settings are simply not:
(a) accessible to the vast majority of families; or (b) replicable with respect to being ef-
fective in community settings. Recently Hoagwood and Koretz (1996) suggested that
intervention and prevention research would benefit policy if interventions were de-
Fig. 1. The ecology of parent influence.