Date
9-17-2026
Department
Helms School of Government
Degree
Doctor of Philosophy in Criminal Justice (PhD)
Chair
Gregory Koehle
Keywords
risk-need-responsivity, juvenile probation, Positive Achievement Change Tool (PACT), dynamic risk, protective factors, reassessment, recidivism
Disciplines
Social and Behavioral Sciences
Recommended Citation
Sattler, David Wayne, "Dynamic Domain Score Changes as Predictors of Juvenile Recidivism: Evidence From Washington State's PACT" (2026). Doctoral Dissertations and Projects. 8870.
https://digitalcommons.liberty.edu/doctoral/8870
Abstract
Juvenile probation agencies routinely administer risk-need assessments. However, evidence remains limited regarding the association between changes in scores from repeated risk-need assessments and recidivism among justice-involved youth. Grounded in general personality and cognitive social learning theory and the risk-need-responsivity model, this quantitative study examined whether changes in dynamic scores on the Positive Achievement Change Tool were associated with recidivism among youth on juvenile community supervision in Washington State. Using 24,929 paired initial and final Positive Achievement Change Tool assessments, the study analyzed total and domain-level dynamic score changes with recidivism measured during the 24 months following the final assessment. Logistic regression models controlled for age, gender, race/ethnicity, initial risk level, baseline dynamic scores, and length of supervision. Reductions in total dynamic scores were significantly associated with lower odds of any, misdemeanor, felony, and against-person recidivism, although the effects were small. Domain-level findings varied by outcome. Reductions in the relationships and alcohol and drugs domains showed the most consistent associations, while school, family, employment, and aggression were outcome-specific. Score increases were generally more strongly associated with higher recidivism than score decreases were with lower recidivism. Moderation effects were limited, with some variation by age, high-risk classification, and race/ethnicity. The findings supported the use of reassessment as a case management tool and informed the development of a five-step model for setting improvement targets, selecting priority domains, monitoring worsening scores, addressing responsivity factors, and evaluating programs through dynamic risk reduction.
