Building upon goal setting theory (Locke & Latham, 1990; 2002) and a solution-focused approach that holds a person accountable for solutions instead of focusing on problems, the Plumas Program used clients' self-determined goals as the primary venue of change for treating domestic violence offenders. The purpose of the study was to investigate the role of self-determined goals in predicting recidivism in treating domestic violence offenders. We hypothesized that Goal Specificity and Goal Agreement would predict Recidivism as mediated by Confidence to work on goal.

Data was collected from 88 court-mandated batterers who were offered the opportunity to avoid prosecution by completing the Plumas Program. Research participants included 70 males and 18 females, with their ages ranged from 19 to 74 years (mean = 37.5, SD = 9.8). Predictor variables of the model were: (1) “Goal specificity” that was measured by a 3-point Likert scale completed by the facilitators' at termination that evaluated participants' self-determined treatment goals as behaviorally described, positively stated, and stated as small steps and in process form. (2) “Goal agreement” that was measured by a 3-point Likert scale completed by facilitators at termination that evaluated the extent to which the goal was mutually agreed upon by the participant and the facilitator. Mediating variable was “Confidence” that was measured by participants' self-reported level of confidence to continue working on their goals upon completion of treatment program. The dependent variable was “Recidivism” that measured participants' re-offending behaviors after attending the treatment program based on accumulative statistics provided by the victim witness office, probation office, and district attorney's office between 1997-2004. Brain injury and experience of child abuse were entered as controlled variables in the model because of significant association between these variables with recidivism.

We utilized Mplus structural equation modeling program 2.12 (Muthen & Muthen, 2001) to test the relationships between predictor, mediator, and dependent variables. Mplus allows for the specification of endogenous variables (recidivism) as categorical. The recidivism rate for participants who completed the Plumas Program was 10.2%. Findings indicated that the hypothesized model was a good fit to the data, χ2 (5, n=88) = 4.72, p = .45, CFI = 1.0, RMSEA = 0. This model accounted for 58% of variance in recidivism. Goal Specificity and Goal Agreement significantly predicted Recidivism. There was a direct path from Goal Specificity to Recidivism (Beta = -.43) and an indirect path through “Confidence.” Goal Agreement had an indirect path to Recidivism mediated through “Confidence.” The controlled variables, brain injury and experience of childhood abuse, significantly predicted Recidivism.

Significance of the study: Different from most current batterer treatment programs that are dominated by a deficits perspective, this treatment program uses the language and symbols of “solution” and “self-determination” for treating domestic violence offenders. The design of the research also moved beyond the conventional input-output model in measuring outcomes of batterer programs and included an investigation of treatment components that account for outcomes.

Primary pedagogical venue will include a formal presentation and group discussion.