Study
Thompson and colleagues (1996) evaluated the short- and long-term educational effects of the Boys Town home campus program using a quasi-experimental, longitudinal research design. Study participants were recruited from all program admissions between May 1981 and June 1985. Children were referred by a juvenile justice agency, a social service agency, or treatment provider.
To be eligible for the study, children had to be eligible for admission to the residential program, had to have parental consent, and had to have the first interview within 36 hours of becoming a resident. The final study sample consisted of 587 children, with 503 in the treatment group and 84 in the comparison group. Children in the comparison group were those who met admission criteria, but did not enter the program either by choice or because of space limitations. The comparison group received treatment-as-usual, which included counseling or some other type of treatment.
The treatment group was predominantly male (92 percent), white (71 percent), and on average about 14.4 years old. The average length of stay in the program was 20 months. Similarly, the comparison group was predominately male (92 percent) and white (68 percent), and on average about 14.7 years old. There were no significant differences between the groups at the initial interview, and potential group differences at follow-up were controlled for in the statistical analysis.
Participants were interviewed when they entered the study, and then continuously every 3 months after until the spring of 1987, after which they were interviewed every 6 months until their 13th interview (the 13th interview was conducted for all youths by 1988).Final interviews were attempted for all participants in 1989. All interviews were conducted via the telephone. The CrimeSolutions review looked at the differences between the pre-intervention interview and the final interview. Outcome measures included grade point average, years of school completed, receiving a high school diploma or GED, and importance or chance of college. Most measures were analyzed using hierarchical linear modeling and the average length of stay was used as time of departure. The outcomes were examined at the initial interview, during placement, and after placement. No subgroup analyses were conducted.
