Static and Dynamic Predictive Utility of the Level of Service Inventory

• In Event: Predicting Behavior and Pursuing Profit: Offender Reintegration, Recidivism, and Puppy Mills

Abstract

The Level of Service Inventory Revised (LSI-R) assessment has become a common actuarial assessment used by correctional institutions across the country to determine the risk of recidivism for offenders once released from prison. It is composed of 54 questions falling within 10 domains that are considered crucial for determining needs for prison rehabilitative interventions. Multiple studies have determined the instrument to be a valid predictor for recidivism. However, the use of the instrument, originally created for community corrections, has potential drawbacks. This research will determine if the LSI-R is an appropriate determinate of risk to recidivate for offenders and scrutinize whether all static or dynamic factors from the LSI-R are necessary for inclusion. Many of the static factors indicative of an individual’s socioeconomic background, employment history, and family factors can be considered culturally biased. If not used appropriately, individuals from impoverished backgrounds are unnecessarily although unintentionally harmed. This study will use an ex post facto design on archival data to compare groups of offenders with similar LSI risk scores determine whether additional factors from pre-sentence investigation reports and incarceration disciplinary reports enhanced information on risk to recidivate more than the LSI-R assessment alone.

Author

• Janeena Wing, Idaho Department of Correction