Study 1
Pantin and colleagues (2003) used an experimental design to evaluate the effectiveness of Familias Unidas. Study participants were recruited from three middle schools in low-income areas of Miami, Fla. As the intervention targets Hispanic immigrants, only Hispanic sixth- and seventh-grade students were eligible to be included in the study. Participants were excluded if they had any history of psychiatric hospitalization or if they did not live within, or were planning on moving outside of, the catchment area during the study.
The total sample consisted of 167 adolescents and their families. The Familias Unidas intervention group included 96 adolescents and their families, and the control group had 71 participants. A majority of the sample was Cuban (39 percent), followed by Central and South American (29 percent and 17 percent, respectively). There was a small portion of Puerto Ricans/Dominicans (5 percent), and the remainder of the sample identified as “other” Hispanic. Almost all of the parents (94 percent) and about half of the adolescents (49 percent) were born outside of the United States.
The Familias Unidas intervention was delivered through family-centered, multiple-parent groups (10 to 12 parents) that met weekly for about 9 months. Parents attended a mean of 24 group sessions, which were led by a trained facilitator. Each session lasted about 1 hour, and parents received $10 for attending. The intervention was delivered in three stages: engagement, promoting parental investment, and fostering parenting skills. Within each of these stages were specific objectives, such as communicating with school personnel, monitoring homework, setting goals for their children, and building partnerships with their peers’ parents. The control group received no intervention. They were only contacted for scheduling and completing of assessments.
The study had five data collection points, with the first at baseline and the other four at 3-month intervals from baseline (3, 6, 9, and 12 months). After completion of the baseline assessment, families were stratified by their adolescent’s school and grade (sixth or seventh grade) and were randomly assigned to either the Familias Unidas intervention or to a control condition. The intervention lasted 9 months and culminated with the 9-month assessment. The 12-month assessment functioned as a 3-month follow-up after the completion of the intervention.
Data collection and assessments were used to measure three broad areas: parental investment, adolescent behavior problems, and adolescent school bonding/achievement. Parental investment was measured with items from the Parenting Practices Scale and the Social Support Appraisal Scale. Adolescent behavior was measured with items from the Revised Behavior Problem Checklist, the Conners–Wells Self-Report Scale, the Behavior Scale Part I, and the Interpersonal Competence Inventory. Lastly, school bonding and academic achievement was measured by the Adolescent Competence Scale, the School Attitudes/Bonding Scale, and the Piers–Harris Children’s Self-Concept Scale. All questionnaires and instruments have very high levels of reliability and validity.
To determine the effectiveness of the Familias Unidas intervention on parental investment, adolescent behavior problems, and school achievement, a mixed-model ANOVA analytic technique was used. This helps account for some of the errors and issues that arise when dealing with nested observations and repeated measures.No subgroup analyses were conducted.
Study 2
Pantin and colleagues (2009) used random assignment to evaluate the effectiveness of Familias Unidas. Three large middle schools in a low-income urban area of Miami–Dade County were selected for this study. School counselors were asked to select Hispanic 8th grade students that exhibited signs of conduct disorder, socialized aggression, and attention problems. These three problem behaviors come from subscales of the Revised Behavior Problem Checklist (RBPC). Selected students were given a letter to bring home to their parents describing the study. Interested parents contacted the researchers to schedule screening and conduct baseline assessments. The study was conducted between January 2004 and July 2008. Students were followed from the 8th grade through high school.
Only adolescents rated as having behavior problems by their parents and by the RBPC checklist were included in the study. Students also had to be of Hispanic immigrant origin (at least one parent born in a Spanish-speaking country of the Americas), had one parent willing to participate in the study, and lived within the selected school districts of Miami–Dade County. Students were excluded if their family was going to move out of the selected school districts during the intervention or out of South Florida in the remaining 3 years of the study. Additionally, if students did not want to participate or if parents’ work schedules restricted them from group sessions, those students were excluded.
The selection process resulted in 213 adolescents (136 boys and 77 girls) and 213 parents/caregivers (27 men and 186 women). The majority of families (86.9 percent) earned less than $30,000 a year. Roughly half (56.1 percent) of the students were born in the United States. Immigrant youth and parents were primarily born in Honduras (26.9 percent), Cuba (20.4 percent), and Nicaragua (16.1 percent).
The treatment group, 109 children and parents, received the Familias Unidas intervention. This consisted of 9 group sessions lasting for 2 hours and 10 family visits lasting for 1 hour. At follow-up intervals, families attended a 1-hour booster session. The follow-up intervals were 10, 16, 22, and 28 months from baseline and initial recruitment into the study.
The control group, 104 children and parents, received three referrals to agencies within the Miami–Dade study area that serve adolescents with behavior problems. Researchers contacted these families again for follow-up assessment activities and surveys. Aside from that, the control group families had no contact with the study or any elements of Familias Unidas. No data was collected on the families that used services from the referral agencies.
Data was collected on family functioning, substance use, sexual risk behaviors, and externalizing disorders. Family functioning information was obtained by using the Parenting Practices Scale, Parent–Adolescent Communication Scale, Family Relations Scale, and the Parent Relationship with Peer Group Scale. Substance use was measured with items similar to those in the Monitoring the Future Study: Youth were asked whether they had ever smoked, drunk alcohol, or used illegal drugs in their lifetime and 30 days before assessment. Adolescents who answered yes to any of those questions were asked follow-up probes to determine the frequency and variety of drugs used.
Sexual behavior was measured using the Sexual Behavior instrument. Similar to the substance use questions, youth were asked at every assessment whether they had ever had sex (all forms) in their lifetime and in the 90 days before assessment. Those that stated they had engaged in sex were asked how often condoms were used, on a Likert scale ranging from “never” to “always.” Lastly, behavior problems such as conduct disorders and aggression were assessed using the Diagnostic Interview for Children (or DISC). These scales are not formal psychiatric diagnoses, but have been shown to be predictive of disorders.
Assessments were completed at baseline, and at 6, 18, and 30 months after baseline. Families were paid $20, $25, $30, and $35 for completing each of the 4 assessment periods, as well as given an additional $30 to cover transportation costs.
Growth curve modeling was used to estimate individual trajectories of change and to test for slope differences over time. Data from the four assessment points was used to determine the outcomes for the three behaviors addressed by the program: substance use, risky sexual behavior, and antisocial behavior. Confirmatory factor analysis and Analysis of Variance (or ANOVA) were used to determine the impact of Familias Unidas on family functioning. No subgroup analyses were conducted.