Language Adaptation and Perceptions of Provider Cultural Competence and Support As Predictors of Participant Responsiveness in the Bridges Program

Multiple dimensions of implementation (e.g., quality, adaptation, and participant responsiveness) determine program effectiveness. While most studies rely on direct effects analyses, new research suggests these dimensions interrelate to produce outcomes. Providers’ implementation behaviors may influence participant responsiveness, the most proximal dimension for determining program effects. Prevention researchers have voiced a concern for increasing responsiveness among Latinos. Bridges is a family-based program supporting Latino adolescent mental health and school success. Despite the fact that Bridges was based on research with the target audience, it is up to individual providers to implement the program with cultural competence, of which an important piece is language. As in other cities with large Latino populations, a blend of Spanish and English is often used. In Bridges, families could select to participate in Spanish or English speaking groups. Yet, as it is not always practical to control the language spoken, providers adapted the language spoken to facilitate understanding.

 We hypothesize:

• Language adaptation predicts perceptions of provider behaviors (cultural competence, support, and therapeutic alliance). Because preliminary analyses indicated this was more common for adolescents (46%) than parents (4%), this is tested for adolescents only.
• Perceptions of provider behaviors predict participant responsiveness (attendance and home practice of program skills)
• Parents’ responsiveness predict changes in proximal parenting outcomes (support, discipline, and monitoring) from pretest to posttest
• Adolescents’ responsiveness predict changes in proximal adolescent outcomes (coping) from pretest to posttest
• Proximal outcomes predict changes in distal outcomes (adolescent school engagement and substance use)

Data come from the 24 intervention group families (500 parents and 353 children) who participated in the Bridges efficacy trial (Gonzales, et al., 2007).

• Language adaptation ratings by independent observers (occurring in 46% of groups)
• Attendance (Mean attendance across each session was 183 for mothers, 91 for fathers, and 218 for children)
• Participant ratings on weekly skills practice worksheets (currently coding)
• Participant perceptions of provider behavior (11 items assessing cultural competence, support, and therapeutic alliance) collected at post-test [αs = .82 (mother), .81 (father), and .86 (child)]
• Pre- and post-test assessments of proximal outcomes
• Distal outcomes at the long-term follow-up

The multi-level model will be tested in Mplus. Results will indicate the importance of provider behaviors in understanding how actively Latino participants (mothers, fathers, and adolescents) participate in family-based prevention programs.