Health & Environmental Research Online (HERO)


Print Feedback Export to File
504939 
Journal Article 
How can the recent research on risk assessment improve predictions of criminal recidivism? 
Nedopil, N 
2006 
1 
Neuropsychiatrie
ISSN: 0948-6259 
20 
1 
15-22 
German 
Predictions of criminal recidivism are required more and. more often in most Western countries in recent years. Their validity has to be high, especially when they lead to a release from secure institutions. At the same time research on criminal recidivism has also increased substantially and produced a number of instruments and a refinement of methods. Today the forensic expert should answer the following questions: Who will when, under which circumstances, with which crime reoffend, and how can this be prevented. With these questions in mind prediction becomes a process, in which the parameters relevant for committing, treating, loosening and releasing have to be considered independently and in their interaction. In this paper both an overview over the current instruments and methods for risk assessment and the results of an own study on long term outcome of offenders using the ILRV are presented. Reoffences are best predicted by the base-rate of recidivism for the specific crime which then is individually modified and by static risk factors. Clinical variables and factors pertaining to the social environment in which the individual is to be released did not correlate significantly with outcome. Predictions, using current instruments for risk assessment were more accurate for violent than for non-violent crimes. The results are in concordance with other published data, except that previously little attention has been paid to crime specific base rates in prediction research. 
risk assessment; criminal recidivism; prediction; assessment; instruments; ILRV; violence