Search
Program Calendar
Browse By Day
Browse By Room
Search Tips
Virtual Exhibit Hall
Personal Schedule
Sign In
For the past 40 years, Canada has been developing its knowledge of Evidence-Based Practices (EBP) used with offenders. However, these EBPs, although promoted by the scientific community, are slowly implemented in Belgium. Training issues are a major obstacle to their implementation (Cook et al., 2009). Indeed, teaching EBPs is complex and tacit (Delany & Golding, 2014), and experience is not sufficient to master them (Gigante, 2013). It is then necessary to couple a high-quality initial training with the possibility for professionals to test their learning, and systematic, and consistent feedback on their performance. Consequently, Canadian, and Belgian researchers set up a collaboration to design Autonomous Virtual Agents (AVA) simulating perpetrators of violence with whom participants can interact to train in the assessment of risk, needs, and protective factors with a transdisciplinary approach. This research is divided into four parts. First, a database of risk interview questions was collected from second-year master’s students in forensic psychological sciences. Second, questions were classified into intentions through a thematic analysis. Intentions group together different formulations of the same question (e.g. “How is the relationship with your parents?” and “How do you get along with your parents?”). In June 2023, Belgian perpetrators of violence will be interviewed to construct identities and obtain responses to the intentions. The collected responses will be integrated into the AVAs. The third part will consist of the integration of the AVAs into the training platform. Finally, a study of the usefulness of the AVAs (qualitative and quantitative approaches) will be carried out. Training on risk assessment using AVA will be offered to psychologists in prison. At the end, their advice will be collected to evaluate their feelings about the AVA. The present presentation will discuss in detail the development process of the most realistic AVA possible.