An Investigation of How Change in Dynamic Risk and Protective Factors Affects the Prediction of Imminent Criminal Recidivism

Dynamic risk and protective factors are changeable, psychosocial variables associated with an increased or decreased likelihood of future criminal behaviour. These variables have an important role in correctional psychology. In particular, they are increasingly central to the management and supervision of individuals released from prison. The changeable nature of these variables means that, with frequent reassessment, the likelihood of recidivism can be monitored during the release period, and intervention can be more carefully targeted to an individual’s needs. However, research has yet to clearly demonstrate that reassessment of dynamic risk and protective factors can accurately track the likelihood of recidivism over time. Further, relatively little is known about how these variables change over time, and how change is associated with recidivism. This thesis set out to investigate whether reassessment of a dynamic risk assessment tool—the Dynamic Risk Assessment for Offender Re-entry (DRAOR; Serin, 2007; Serin, Mailloux, & Wilson, 2012)—would enhance the prediction of imminent recidivism among a large sample of high-risk men (n = 966) released from prison on parole in New Zealand. The analyses addressing this question were divided into three primary sections: 1) an investigation of whether a single proximal assessment was a more accurate predictor of imminent recidivism than a single baseline assessment completed shortly after release; 2) an investigation of whether a single proximal assessment was a more accurate predictor of recidivism than a series of aggregated measures across multiple time points, and; 3) an investigation of whether several different measures of intra-individual change demonstrated incremental predictive validity over the most proximal assessment. This approach represented a replication and extension of the framework set out by Lloyd (2015) in a recent thesis for testing whether reassessment of dynamic risk and protective factors enhances the prediction of imminent recidivism. Across all three sections, results provided consistent evidence that the most proximal assessment was the most accurate predictor of imminent recidivism. The most proximal assessment was a significantly more accurate predictor than a baseline assessment, and neither aggregation nor measures of intra-individual change clearly improved predictive accuracy. These results highlight the importance of reassessment for monitoring changes in the likelihood of recidivism over time and have important implications for community correctional agencies who are tasked with managing individuals released from prison, particularly those deemed to be the highest risk of recidivism. The results also have theoretical implications for the concepts of dynamic risk and protective factors and their role in the process leading to recidivism. A better understanding of the recidivism process should lead to intervention strategies that are more effective at reducing recidivism