NCJ Number
186955
Journal
Journal of Quantitative Criminology Volume: 16 Issue: 4 Dated: December 2000 Pages: 385-414
Date Published
December 2000
Length
30 pages
Annotation
This study develops a statistical procedure to analyze recidivism in samples that are subject to the presence of desisters and to multiple modes of reconviction.
Abstract
The statistical procedure allows for a more accurate study of individuals' transition and hazard in the type and timing of offenses following a specific type of conviction. The study shows use of a nonparametric approach for investigating failure in the presence of other acting causes, obtains initial estimators of the probabilities of reconviction for different types of offenses, and demonstrates that the method can be used both to display the data and to choose an appropriate parametric family for the survival times. The study presents an exponential mixture model for competing risks in such a way that it allows adjustment for concomitant variables and assessment of their effects on the probabilities both of reconviction for predetermined types of offenses and desistance, and of the hazards of reconviction. It also suggests a method for assessing calibration of predicted survival probabilities. The study finds a high probability of sexual reconviction for individuals with previous sexual convictions and evidence of diversity and a raised hazard of reconviction for young chronic offenders. Figures, tables, references