Improving Probation Decisions through Statistical Training

Publisher Website
Google Scholar
Abstract
Much research in cognitive and social psychology has demonstrated that people often fail to incorporate statistical principles in making judgments and decisions. For example, people generally assign too little weight to base-rate information and too much weight to case-specific information. Probation officers are no exception. Two studies examined whether statistical training could increase the use of statistical principles in probation decisions. Study 1 found that novice probation officers trained on the law of large numbers were more likely than untrained officers to employ statistical principles in solving open-ended probation problems. Study 2 found that training novice probation officers on the law of large numbers increased their use of base-rate information on predictions about recidivism for 10 realistic probation cases, but only when the cases involved high risk. These results suggest that statistical training may improve probation decisions and judgments.

This publication has 7 references indexed in Scilit: