This study applied item response theory analysis to the Structured Assessment of Violence Risk in Youth (SAVRY) and found that the SAVRY items were not equally discriminating JIY with varying levels of the latent trait.
This study used item response theory analysis to investigate item- and test-level functioning of the Structured Assessment of Violence Risk in Youth (SAVRY) and differential item functioning (DIF) across gender and race/ethnicity in justice-involved youth (JIY). Results obtained from the application of the graded response model showed that the SAVRY items were not equally discriminating JIY with varying levels of the latent trait, with “Poor compliance” as the most discriminating item and “History of self-harm or suicide attempts” as the least discriminating item. At the test level, the SAVRY provided precise (reliable) information about the latent trait for the majority of JIY whose latent trait between two standard deviations below and above the mean. Results of DIF revealed that six items operated inconsistently between White, Black, and Hispanic JIY, among which two items also functioned differentially across gender. Participants were 868 JIY (23.7% female; 26.9% White, 50.9% Black, and 22.2% Hispanic) in pre-trial detention centers in Connecticut. (Published Abstract Provided)
Downloads
Similar Publications
- Assessment of the Effectiveness of Emergency Lighting, Retroreflective Markings, and Paint Color on Policing and Law Enforcement Safety
- A Method to Detect AAC Audio Forgery
- A Review of the Evolution of the NCS-NCVS Police Reporting and Response Questions and Their Application to Older Women Experiencing Violent Victimization