Paper
Comparing youth outcomes for group care and treatment foster care: An application of propensity score matching
- issue: Issue 3 / 2008
- authors: Bethany Lee, Ron Thompson
- keywords: Usa, outcomes, peer contagion, group care
- views: 4622
- downloaded: 0
- download pdf (69 Kb)
abstract
Background. Often youth with the most challenging needs are served in congregate care settings like group homes and residential treatment centers. However, concerns have been raised about whether youths who receive services in groups with other troubled youths experience negative effects. As children age and differentiate from their family of origin, peer groups have a growing impact on a youth's attitudes and behaviors. Peer contagion, the idea that problem behaviors increase due to the influence of anti-social peers, is a key criticism that has been leveled at group-based interventions like residential care.
By putting troubled youth together in an intervention setting, some researchers have found that youth problem behaviors increase. While a number of such studies have looked at peer influences in outpatient group therapy programs or in school classrooms, few efforts have assessed how living with other troubled youth in a group placement setting may impact an individual youth's behavior and outcome. Treatment foster care (Tfc) has been developed as an alternative to group care that should avoid peer contagion. In treatment foster care, a youth is placed individually in a family home with foster parents who receive additional training and support to manage the youth's challenging behaviors.
Purpose. This study compares outcomes for youths who received services in a state-of-the-art group care setting with similar youths who were served individually in treatment foster care homes. Administrative data from Girls and Boys Town, a large youth-serving agency in the United States, was utilized. This agency provides both group care and treatment foster care services. Both programs follow the Teaching-Family model, a therapeutic approach with some evidence of its effectiveness. The Teaching-Family model emphasizes social skills learning and relationship-building in a family-style living environment.
Propensity score matching was used to balance differences in 18 background variables for the samples of group care youth (N=744) and Tfc youth (N=125). Propensity score matching is a technique used to reduce bias in comparing two groups that have not been randomly assigned. A propensity score is computed for each individual and reflects his/her probability to be assigned to the treatment group given the observed covariates (demographics, placement history, behavior problems, maltreatment history, and family problems). By comparing individuals from different treatment groups with the same or similar propensity scores, causal effects of the treatments can be assessed.
Four matching techniques were utilized in comparing five youth outcomes. After matching, the groups did not differ significantly in background characteristics and the rates of achieving desired outcomes were compared. Two outcomes were assessed at discharge (favorable discharge from placement and return home [with biological/ adoptive parents or other relatives]); three outcomes were assessed at six months post-discharge (legal involvement, subsequent formal placement, and being in a home-like setting at follow-up).
Key findings. Although four propensity score matching algorithms were used to compare group care youths with Tfc youths, substantive results remained the same across method. This offers additional support for the findings. Specifically, results found that group care youths were about 20% more likely to be favourably discharged from care, compared to youth in treatment foster homes. Additionally, only 40% of treatment foster youth returned home to live with family following care, a positive outcome experienced by most (56-69%) group care youths. Youth in treatment foster care were significantly more likely to experience a subsequent formal out-of-home placement in the six months following care. Approximately 70% of TFC youths had a subsequent placement while only about one-fourth of group care youth went on to experience another formal placement. In regards to involvement in the legal system, there were no significant group differences as both groups experienced this outcome for just over 10% of the population. There were also no differences found in relation to being in a homelike setting at follow-up. Both groups were equally likely to be in a setting other than institutional care at six-months post-discharge.
Implications. This study improved upon previous research by holding the treatment model constant in comparing treatment foster care with group care. Contrary to previous research, findings from this study suggested that group care may be more effective than Tfc in achieving positive youth outcomes. While these unique results may be partly due to methodological limitations, it is also important to note that the group care program used here is not typical of most group care programs in the United States. The Teaching-Family model utilized in this study is more resource-intensive and evidence-based than most eclectic group care interventions. As such, this model shows promise in minimizing peer contagion as assessed in this study.
More research is needed on group care practice models to determine effective approaches in working with youths in out-of-home care. Although policy dictates least restrictive placement assignments, it is likely that some youths will continue to require residential treatment at least for short-term interventions. High-quality group care programs should be available to meet the challenging needs of these youths.
Subsequent research efforts should identify the critical elements of effective group programs to improve the quality of services for youths in residential settings. The Teaching-Family model considered here is unique in its emphasis on family-style living with live-in house parents who are trained in behavioral techniques as well as relationship development. Positive youth-staff relationships, opportunities for positive youth development within the placement setting, and a well-defined therapeutic approach may all be key program features that promote the effectiveness of group care.
Key references
An extended version of this article will be published in Children and Youth Services Review, B. R. Lee & R. Thompson, Comparing outcomes for youth in treatment foster care and family-style group care, Copyright Elsevier, 2008.
Guo, S., Barth, R. P., & Gibbons, C. (2005). Propensity score matching strategies for evaluating substance abuse services for child welfare clients. Children and Youth Services Review, 28(4), 357-383.
Wolf, M. M., Pillips, E. L., Fixsen, D. L., Braukmann, C. J., Kirigin, K. A., Willner, A. G., & Schumaker, J. (1976). Achievement Place: The Teaching-Family model. Child Care Quarterly, 5(2), 92-103.
Contacts: Bethany R. Lee, Ph.D., University of Maryland Baltimore School of Social Work, 525 W. Redwood St., Baltimore, MD 21201 USA, blee@ssw.umaryland.edu, Phone 410-706-3809.