Abstract
This thesis examines properties of the Youth Level of Service/Case Management Inventory (YLS/CMI) within Norwegian Therapeutic Residential Care (TRC), a welfare-based and nonsecure system. Whereas the YLS/CMI was developed for use in juvenile justice, Norwegian TRC operates under the Child Welfare Act and serves adolescents with serious behavioral problems, complex developmental needs, and diverse offending histories. Consequently, the performance and practical utility of the YLS/CMI cannot be assumed in this setting and require systematic investigation.
The thesis addresses three central questions: (1) whether the YLS/CMI predicts later offending among adolescents in TRC, (2) whether the instrument can differentiate adolescents with diverse patterns of risk, need, and responsivity factors at intake, and (3) whether reassessments capture change in criminogenic needs during placement and whether such changes relate to subsequent criminal involvement. These questions were investigated in three empirical studies using national assessment data collected through standardized procedures and using various criminal outcomes variables.
Study 1 evaluated the predictive performance of the YLS/CMI among 646 adolescents placed in TRC. Intake scores demonstrated adequate accuracy in predicting subsequent offending behaviors (AUC = .62–.77), with higher performance among males than females. These findings align with international research but extend the evidence base to a welfare-oriented setting with heterogeneous offending histories and lower base rates of official justice involvement.
Study 2 analyzed intake assessments for 737 adolescents referred to TRC to examine whether systematic YLS/CMI subgroup differentiation reflects distinctive and meaningful risk and need profiles. Results showed that adolescents classified as high risk, low/moderate risk, or as having serious substance use problems differed across key criminogenic domains and demographic characteristics. Reflecting extant literature on residential care, assessed clinical responsivity considerations showed high prevalence of mental health related needs. However, these factors differed less between subgroups. The findings support the instrument’s utility for structured case formulation, placement decisions, and informing case planning.
Study 3 investigated dynamic properties of the YLS/CMI in a cohort of 313 adolescents reassessed at discharge. Total scores and several criminogenic domains showed significant reductions during placement. Importantly, decreases in dynamic risk factors, including education/employment problems, peer relations, substance use, personality/behavior, and antisocial attitudes, were associated with lower likelihood of police charges one year after discharge. These findings support the value of reassessment in updating risk estimates and informing aftercare planning.
Collectively, the results show that the YLS/CMI retains predictive, discriminative, and dynamic validity in Norwegian TRC, despite the instrument being applied outside its original justice-based context. The findings indicate that structured assessment can enhance decisionmaking, and support more individualized and proportionate treatment planning in a system where needs are complex and service pathways heterogeneous. At the same time, subgroup differences, contextual constraints, and limitations in available measurement data underscore the need for careful interpretation, gender- and context-sensitive application, and further refinement of specific domains and items.
The thesis concludes that structured criminogenic assessment with the YLS/CMI can serve as a valuable decision-support framework in TRC, strengthening both clinical practice and system-level planning. Future research should examine item-level functioning, the assessment of protective factors, trajectories of change, and how assessment results are integrated into facilities’ case management processes. Such work could be essential for advancing evidenceinformed approaches within welfare-based residential care and ensuring coherent, equitable, and developmentally attuned support for adolescents with complex behavioral needs.