In this evaluation of the Multidimensional Treatment Foster Care–Adolescents (MTFC–A) program conducted by Chamberlain and Reid (1998), 85 boys referred for community placement in the juvenile justice system were randomly assigned to either MTFC–A or Group Care (GC). Six boys dropped out for lack of parental consent, and five boys assessed at baseline failed to be placed and were sent home. Boys were 12–17 years old (mean=14.9 years), with histories of serious chronic delinquency averaging 14 previous criminal referrals and more than four felonies. The mean age at first criminal referral was 12.6 years. All were mandated to out-of-home care by a committee of juvenile court, and all lived in the Pacific Northwest. The large majority (85 percent) were white, 6 percent were African American, 5 percent were Hispanic, and 3 percent were Native American. There were no significant differences between the two groups on any of the baseline variables.
Treatment fidelity was assessed at 3 months through the use of an onsite interview examining practices at each placement site.
For the intervention group, daily phone contact with MTFC parents and a 2-hour weekly supervision meeting for MTFC–A parents were provided. Case managers and individual and family therapists were supervised in weekly 2-hour meetings with the project director and clinical consultants. Individual and family therapy sessions were videotaped and reviewed in these meetings. MTFC parents were taught how to implement an individualized plan for each youth. A three-level system was used in which the boy’s privileges and level of supervision were based on his compliance with program rules, adjustment in school, and general progress. Each boy attended weekly individual therapy emphasizing skill building in solving problems, social perspective-taking, and nonaggressive methods of self-expression. All boys were enrolled in public school, and 45 percent attended some special education classes. Boys carried a card to each class, and teachers had to sign off on attendance, homework completion, and attitude. Consequences were delivered for even minor rule infractions (such as being 2 minutes late to class). These included loss of points and privileges, extra chores, or in extreme cases stays in detention.
In the GC condition, the programs concentrated on establishing prosocial norms through therapeutic group work, during which youth confronted one another about negative behavior and participated in discipline and decision-making.
Outcome measures included number of days each month a youth was in care, on the run, in detention, or in State training school. Delinquent and criminal activities were assessed from official criminal referral data recorded by the Oregon Youth Authority, which included all officially reported misdemeanor and felony offenses on the youth’s record from 1 year prebaseline until 1 year postprogram, discharge, or expulsion. In addition, all boys completed the Elliott Behavior Checklist, a self-report questionnaire about criminal or delinquent behaviors during a specified time period. Three subscales were examined: general delinquency, index offenses, and felony assault.
The subjects for this study by Chamberlain, Leve, and DeGarmo (2007) were 103 girls referred by Juvenile Court judges in Oregon between 1997 and 2002. The girls had been mandated to out-of-home care because of chronic delinquency and were randomly assigned to the experimental condition (n=37) or the control condition (n=44). The average length of stay in the randomized placement was 174 days, and the average time between baseline and intervention was 47 days. The girls were 13 to 17 years old at baseline. Seventy-four percent were white, 12 percent Native American, 9 percent Hispanic, and 2 percent African American, 1 percent Asian American, and 2 percent were designated as “other.”
Foster parents of the girls assigned to the intervention group received daily phone contact, during which the Parent Daily Report Checklist was completed. Weekly fidelity data was collected on parent implementation of an individualized, in-home daily reinforcement system for the girls. Weekly foster parent group training, supervision, and support meetings were led by experienced program supervisors. School functioning was closely monitored with a daily school card signed by teachers.
In the control condition, most programs reported endorsing a specific treatment model, with the primary philosophy being behavioral (70 percent), eclectic (26 percent), or family style (4 percent) Seventy percent of the programs reported delivering therapeutic services at least weekly.
Outcomes included the number of criminal referrals; girls’ self report of total days spent in detention, correctional facilities, jail, or prison; and girls’ self report of delinquency measured by the Elliott General Delinquency Scale. Delinquency was based on a composite of the three measured variables. Structural equation modeling and latent growth curve models were used to analyze the data.