Dr Matt Frize1, Ms Catriona (Keenie) Daly, Mr Mathew Gullotta
1Department of Families, Fairness & Housing, Melbourne, Australia
Biography:
Matt is the Director of Victoria’s Forensic Disability Services, within the Department of Families, Fairness & Housing. He is a registered supervising clinical, forensic and educational / developmental psychologist and holds a clinical associate position at Swinburne University. Matt has a Masters in Educational & Developmental Psychology, a Doctorate in Clinical Psychology and a PhD. Matt has previously led the NSW forensic disability service and is a co-author of the ARMIDILO (general) and ARMIDILO (sex) risk assessment tools. He has consulted internationally and has published a number of book chapters and peer reviewed articles on forensic disability.
Abstract:
The predictive validity of offender risk assessment measures tends to be assessed over a period of years given their use in sentencing and decisions for release from custody. Evaluation over a long time frame also improves statistical power. The short-term predictive power of a risk assessment tool, however, is a critical consideration in a range of scenarios, especially for services that support people with an Intellectual Disability, given a tool's ability to provide direction on risk management for services supporting an individual in the community. The current study aimed to assess the reliability, ecological validity and predictive validity of a suite of tailored and general risk assessment tools, over a three and six month period. The Assessment of Risk and Manageability of Individuals with Developmental and Intellectual Limitations who Offend – Generally (ARMIDILO-G), the Historical Clinical Risk – 20 (HCR-20), Level of Service Inventory – Revised (LSI-R), Current Risk of Violence (CuRV), Group Risk Assessment Model (GRAM) and Psychopathy Checklist Revised (PCL-R) were administered to 139 people with an Intellectual Disability who have a history of offending and were being supported by a NSW community based forensic disability service. The study explores the circumstances in which a tool may show ecological validity and explores features that promote improved predictive validity over short timeframes.