Characterizing CVI Street Outreach Participants and Service Dosage: Implications for Measurement and Evaluation (WP-25-08)
Marisa Ross, Susan Burtner, and Andrew Papachristos
Community violence intervention street outreach (CVI-SO) strategies are growing in popularity as non-punitive approaches to solving the public health problem of community gun violence. Evidence on the effectiveness of CVI-SO on rates of violence is mixed and faces challenges due to concerns with documentation and data privacy, intentional selection bias in program design, and variation in participant risk and needs. Effective evaluation requires methods that accurately capture the scope and delivery of services, starting with a greater understanding of the services CVI participants receive and how they vary based on individual characteristics. This study explores the services that participants received from a coalition of Chicago CVI organizations from 2017-2023. Considering administrative and programmatic data from over 4,000 participants’ nearly 200,000 interactions with providers, the researchers examine patterns in demographics, network-based risk factors, and service provision and dosage. They then use descriptive and latent class analyses to characterize the “typical” participant in Chicago. Results show that CVI work relies heavily on long-term mentoring relationships to change individual behavior. Service patterns show that latent groups exist with varying dosage: Higher dosage participants with higher risk for gun violence receive more frequent contacts over longer periods, demonstrating how organizations adjust their approach based on participant needs. Classes that primarily receive behavioral or workforce-related services emerge, demonstrating a relationship to risk that becomes less clear over time. Findings underscore the need for evaluation frameworks that capture both the strategic variation in service delivery and the multiple pathways through which CVI programs influence participant outcomes.