Abstract
The prediction of recidivism is of growing importance in the criminal justice system, and authorities have recommended the use of statistical prediction devices along with case-specific information. However, research has shown that laypeople and experts have difficulty dealing with statistical information, and tend to ignore it in the presence of case-specific information. A study was conducted in which subjects were given one of three different types of base-rate information about known recidivism rates in a population, and a sample of several case descriptions, and asked to predict recidivism for the sample cases. No use of base-rate information occurred when case-specific information was available. The implications for decisions and training in the criminal justice system are discussed.
Funding Information
  • National Science Foundation (SOC75-18061)
  • John W. Payne

This publication has 15 references indexed in Scilit: