Research That Matters (January 17 - 20, 2008)


Hampton Ballroom (Omni Shoreham)

Under Construction: Measuring Fidelity to an Evolving Model of Child Welfare Reform

Lisa Lewis, MS, MSW, Indiana University - Purdue University, Indianapolis, Gail Folaron, PhD, Indiana University - Purdue University, Indianapolis, and Sabrina Williamson, PhD, Indiana University.

Background and Purpose.

Thirty-five court actions from 1995 to 2005 mandated changes in public child welfare (Kosanovich, & Joseph, 2005). Indiana began implementing a bottom up reform to their system as the only state to do so without a court order. As a result, this has become a topic of national interest. Indiana is including a five step practice model that was successfully piloted in Alabama (Brazelton, 1998) and later attempted in Florida (CSSP, 2003) & Utah (Utah DCFS, 2006 ). The new practice model includes protocols for family engagement and teaming similar to the wrap around model originally introduced in criminal justice (Braithewaite, 1989) and subsequently applied to child welfare (Adams & Chandler, 2004).

Conducting a field evaluation during the initial program implementation was challenging. Three research questions are addressed:

1. How faithfully did workers implement the new practice model?

2. What factors support/hinder workers' ability to adopt the new model?

3. How have community perceptions of the child welfare agency changed over time?

Methods.

Eleven counties that piloted the reform served as the sample. The fidelity data instrument was modeled after the Wraparound Fidelity Index 4.0 (Bruns, Suter, Force, Sather & Leverentz-Brady, 2006). A supplementary parent satisfaction survey assessed the effectiveness of family engagement. Data was collected for questions 1 & 2 through telephone interviews with parents and staff. Data collection was triangulated by observations during family engagement and family team meetings from a random sample of cases.

Focus groups, formal and informal individual interviews, and surveys were utilized to determine changing perceptions of community (question 3). Exploratory factor analysis and correlations were used to analyze the quantitative data. Qualitative data were audio recorded, transcribed, entered into Atlas-Ti, coded and analyzed by two researchers. Themes and patterns were identified using the constant comparative method (Glaser & Strauss, 1968).

Results.

Regarding fidelity (question 1), workers carefully followed the protocol but required up to15 teaming and engaging sessions to become skilled with the process. A comprehensive, internal support network around each worker provided a strong “transfer of learning” component (Macauley & Cree, 1999) that pushed workers into action and provided feedback for skill development.

Barriers to implementation (question 2) included resistance from workers, varied learning curves, staff turnover, lack of supervisory support, policies incompatible with goals of the model, and family inability to participate in the teaming process. Supports included comprehensive mentoring for workers and the perceived value of the model.

Initial concern about the reform from community agencies (question 3) generally dissipated; some agencies changed their practice to support the reform.

Conclusions and Implications.

This presentation introduces promising practices for moving a child welfare system from a criminal justice orientation to a social work practice model. Child welfare educators, administrators and staff will benefit from a discussion of system change, barriers and supports that can impact movement. Researchers will enjoy hearing the challenges inherent in studying an evolving system along with the lessons learned.