A Controlled Parenting Skills Outcome Study Examining Individual Difference and Attendance Effects

Abstract

Recent reviews of family-focused interventions designed to prevent juvenile conduct and substance use problems reveal that these interventions are widely disseminated, but only infrequently evaluated (e.g., Institute of Medicine, 1994; U.S. Department of Justice, 1992). Moreover, the limited number of evaluation studies reported in the literature have often revealed methodological problems (U.S. Department of Justice, 1992; Wiese, 1992, Yoshikawa, 1994). Problems evident in evaluation studies have typically involved deficits in study and intervention design such as lack of experimental control, small sample sizes, highly specific or unrepresentative samples, exclusive reliance on self-report measures, lack of a strong empirical or theoretical base, and failure to evaluate intervention fidelity. Additionally. there has been limited attention to the effects of parent or family characteristics and intervention parameters on outcomes (Kazdin, 1993; Spoth & Redmond. 1994). These deficits have collectively imposed substantial constraints on the validity and generalizability of findings. To address these deficits, the present investigation incorporates an experimental test of a theory-based, universal intervention with a reasonably large (n = 209) sample, multimethod measurement, implementation fidelity checks, and modeling of individual difference and intervention attendance effects on outcomes. The intervention program evaluated in this study, Preparing for the Drug (Free) Years, is based upon the social development model (Catalano & Hawkins, in press; Hawkins, Catalano, & Miller, 1992). This model draws upon social control theory (Hirschi, 1969) through its emphasis on the role of bonding to prosocial others (e.g., family members) as protection against the development of juvenile substance abuse and conduct problems. It also integrates concepts from social learning theory (Akers, 1977; Bandura, 1977b), postulating that bonding is enhanced through the provision of (a) opportunities for child involvement in prosocial interactions and activities, (b) child skills training for prosocial involvements, and (c) consistent rewards for child prosocial involvement and punishments for violations of prosocial norms. Because the Preparing for the Drug (Free) Years intervention is guided by the social development model, it emphasizes the enhancement of protective processes in the family. The outcomes serving as dependent variables in this study are parenting behaviors that can serve a protective role in the prevention of adolescent substance-related problems. Parent behaviors targeted by the program have been described as protective because they potentially reduce either the likelihood of child exposure to risk factors or the deleterious consequences of risk exposure (Benard, 1991; Catalano & Hawkins, in press; Coie et al., 1993; Farrington et al., 1990; Masten, 1994; Masten, Best, & Garmezy, 1991; Rutter, 1990; Spoth & Redmond, 1994; Yoshikawa, 1994). Training in effective parenting skills can improve parents' ability to serve as protective factors or to buffer the risks to which their children are exposed; it can also teach parents to avoid ineffective child management behaviors that can directly increase children's risk for substance abuse and conduct problems (Hawkins et al., 1992). Enhancing parents' ability to engage in protective parenting behaviors can be particularly important as children enter the preadolescent, middle school years. During these years, peers gain increased influence as socialization forces, and the transition from elementary to middle school environments can also increase child exposure to other risk factors (Catalano & Hawkins, in press; Eccles, Midgely, & Wigfield, 1993; Simmons & Blyth, 1987). Moreover, adolescents in rural families experiencing economic stress are more likely to be exposed to family-related risks (e.g., poor parenting practices) than are those in rural families without economic stress (Conger et al. …