Improving Measurement of Youth and Young Adult Delinquency Risk: Evidence from the Eastern and Southern Caribbean

Developing effective risk screens to identify at-risk youth is crucial for the success of these programs. However, existing risk screens often face challenges such as inaccurate predictions, lengthy administration times, and the need for specialized expertise to score results. A study conducted in 2020 as part of the Community, Family and Youth Resilience (CFYR) Family Matters program in Guyana, St. Lucia, and St. Kitts and Nevis identified limitations in the existing pre-program risk screen. The screen did not accurately distinguish between high and low-risk youth and required nearly an hour to administer. These limitations hindered program targeting and impact. To address these challenges, the United States Agency for International Development (USAID) commissioned a follow-up study to develop a streamlined and potentially more accurate risk assessment tool. The study used machine learning algorithms to analyze longitudinal survey data from 2,393 potentially at-risk youth in the Caribbean and Central America. The analysis identified the strongest predictors of delinquency, including antisocial tendencies, weak parental supervision, impulsive behavior, negative peer influence, and past delinquency. The resulting tool, the Youth Risk Screen (Y-RISC), is designed for both youth (ages 11 to 17) and young adults (ages 18 to 29). The Y-RISC includes 39 questions and takes approximately 15 minutes to administer. The tool is intended to be easy to use and score, making it a valuable resource for program administrators. Pre-testing of the Y-RISC was conducted on a sample of 90 young adults from five Caribbean countries in May 2021. The results showed high levels of respondent attention and comprehension throughout the survey. Supplemental analysis also provided evidence that sensitive questions about self and peer delinquency elicited truthful responses. The Y-RISC is currently available for use by researchers, implementers, and policymakers. Additional data collection is required to fully validate the tool's accuracy and identify any necessary adaptations for use outside the Caribbean.