Spoth and colleagues (2003) evaluated Strengthening Families Program: For Parents and Youth 10–14 (SFP 10–14) program that was adapted for an African American sample randomly drawn from an urban site of a large, multisite longitudinal study (Cutrona et al. 2000). The original study used 1990 census data to identify block group areas with 10 percent or more African American residents and 20 percent or more low-income families. Schools identified 507 eligible adolescents in the fourth through sixth grades living in those block areas. A total of 77 percent (390) agreed to participate in the longitudinal study. Of the 348 families who completed the first wave of data collection, Spoth and colleagues randomly selected 200 families who were then randomly assigned to the SFP 10–14 group or a waitlist comparison group. Of these families, 85 (34 from the intervention group and 51 from the comparison group) were successfully contacted by phone, agreed to participate, and provided sufficient data for inclusion in the analyses. The mean age of participants was 10.5 years. Of the primary caregivers, 82.7 percent were the children’s biological mothers, 93.6 percent were female, and the mean age was 38.4 years. The caregivers had an average of 3.5 children.
The SFP 10–14 intervention group received a slightly revised program consisting of six rather than seven weekly 2-hour sessions. Content of the regular seventh session was incorporated into the sixth session for this evaluation. The first hour was spent in separate youth and caregiver groups; the second hour consisted of a joint family session. The culturally adapted teaching manuals, program videotapes, promotional videotapes, brochure, and all correspondence to families referred to the project as Harambee (a Swahili word meaning “pulling together”) and included depictions of African American participants and program implementers. The adapted program included all SFP 10–14 content and adhered to the original theoretical principles.
Telephone surveys of the self-report questionnaires used in earlier evaluations of SFP were used to collect data from caregivers and youths separately. Study outcomes included intervention-targeted parenting behaviors, number of family meetings, participation of youths in family meetings, parent–child affective quality, alcohol-related peer resistance, intervention-targeted child behaviors, general peer-resistance skills, and alcohol resistance skills.
Surveys were administered at three time points: time 1 at the beginning of the large, longitudinal study; time 2 after the intervention; and time 3 after the waitlist control group received the intervention. The groups differed on only one variable at baseline: the intervention group had a smaller average score on intervention-targeted child behaviors than the control group. This difference was addressed in the analyses, which included two sets of repeated-measures analyses of variance.
Spoth and colleagues (2004) used an experimental design to evaluate the SFP 10–14 program at 33 rural public schools, which were randomly assigned to three groups: the SFP 10–14 (called the Iowa Strengthening Families Program or ISFP), the Preparing for the Drug-Free Years (PDFY), or a minimal-contact control condition. Selected schools were located in rural communities with populations of fewer than 8,500 and a relatively high percentage of low-income families. All families with sixth graders were invited to participate. Of the 1,309 eligible families, 667 completed the pretests (238 ISFP group families, 221 PDFY group families, and 208 control group families).
Families in the SFP 10–14 group were provided seven weekly sessions. Fidelity measures were used to ensure high fidelity in program implementation. PDFY group families received five weekly sessions. Families in the control condition received a set of four parenting guidelines written by Cooperative Extension Service personnel.
This study analyzed data on substance use 6 years after baseline. Substance use data was taken from the National Survey of Delinquency and Drug Use.
Spoth, Randall, and Shin (2008) studied the long-term effects of SFP 10–14 on school success. Specifically, this study concentrated on the indirect effects of the intervention on school engagement in 8th grade and academic success in the 12th grade.
Participants were drawn from the earlier longitudinal study of families with adolescents in sixth grade at 33 rural schools in the Midwest. Blocked random assignment was used to create three groups: SFP 10–14, the Preparing for the Drug-Free Years, or a minimal-contact control condition. This 2008 study analyzed data from participants at the 22 schools assigned to the SFP 10–14 and control conditions only.
Of the 374 families in the two groups who completed the pretest, 308 completed the 6-year follow-up assessments. Of these families, 86 percent were dual-parent families and 98 percent were white. Most of the parents (98 percent of mothers and 95 percent of fathers) completed high school, and more than half reported some post–high school education. Average age was 37.2 years for mothers and 39.4 years for fathers.
Academic success was measured using multiple sources, including mother, father, and student reports. School engagement—or the connectedness of a student to the school—was measured by three indicators: an affective indicator (students’ feelings toward school), a cognitive indicator (students’ perceptions or beliefs related to school and self), and a behavioral component (students’ actions such as completing homework). Alcohol use was measured from students’ reports regarding initiation of use, attitudes toward use, and potential response to peer pressure for alcohol use. Four measures of parental competency were used: rules and consequences regarding alcohol use; parental efforts to involve the child in family activities and decisions; parental management of anger and strong emotion in the parent–child relationship; and parental activities to communicate understanding of children’s feeling and goals as well as parental intentions. Intent-to-treat structural equation modeling was used to analyze the data.