The Added Value of the YLS/CMI Subscales Beyond the Total Score Among Japanese Youth Offenders in Detention

Abstract

This study examined whether subscales of the Youth Level of Service Case Management Inventory (YLS/CMI) have added value beyond the total score. Including 238 detained Japanese juvenile offenders (ages 16–19), the proportional reduction in mean squared error (PRMSE) method was applied. Most subscales showed added value, except for Attitudes/Orientation. For the younger group, all subscales except Attitudes/Orientation showed added value. For older juveniles, Family Circumstances/Parenting, Education/Employment, and Attitudes/Orientation did not. Construct validity was assessed via correlations with prior delinquency indicators. While the total score showed consistent associations, some subscales—notably among older youth—did not, raising concerns about their relevance. These findings suggested that subscales should be interpreted according to both statistical distinctiveness and theoretical support. Subscale analysis may aid intervention planning, particularly for younger offenders, but its utility should be judged by both psychometric and practical standards.

Plain Language Summary

Do the YLS/CMI Subscales Help Understand Risk Among Japanese Youth Offenders in Detention?

This study explored how to improve risk assessments for young people held in detention centers in Japan. A tool called the Youth Level of Service/Case Management Inventory (YLS/CMI) was used to estimate how likely a young person is to reoffend. The tool provides a total score and also includes eight subcategories (called subscales) such as family issues, education, peer relationships, and attitudes. The study examined whether these subscales offer useful information beyond the total score. Data were analyzed from 238 Japanese male youth offenders, aged 16 to 19, who were held in protective custody. A statistical method was used to check if each subscale improved how well the tool explained or predicted behavior. The results showed that some subscales, such as prior offenses and peer relationships, provided helpful additional information, especially for younger youth. However, other subscales, such as attitudes and education, were less useful particularly for older youth. This suggests that the usefulness of different parts of the tool may depend on age or personal background. Understanding which parts of a risk assessment tool offer the most meaningful information can help professionals create better support and treatment plans for young people involved in the justice system, both in Japan and beyond.