NCJ Number
136264
Journal
Criminology Volume: 30 Issue: 1 Dated: (February 1992) Pages: 149-155
Date Published
1992
Length
7 pages
Annotation
Current debate over different criminal career models, primarily those of Greenberg and Barnett, is discussed, and similarities and differences between models are examined.
Abstract
While Greenberg has made important contributions to research on criminal careers, Barnett claims to demonstrate shortcomings of Greenberg's model. Beyond examining the relative merits of analyses and tests to which the Greenberg and Barnett models have been subjected, the relevant question is whether it is possible to move beyond the current debate to the further development of stochastic criminal career models. Both Greenberg and Barnett contend that the probability of an individual characterized by mean offending rate is suitably characterized by a Poisson probability distribution. The models differ, however, in their specifications of the nature of the posited heterogeneity and the nature and significance of offender age or some other offender characteristic. Both models have three limitations: (1) they have been calibrated primarily on summary statistics derived from group aggregates; (2) individual-level covariates that should affect the mean offending rate have not been controlled in calibration exercises; and (3) no general class of models has been defined within which both the Greenberg and Barnett models can be compared for goodness-of-fit and other aspects of model evaluation. Analysts need to specify, estimate, and assess the individual-level counterparts of alternative criminal career models to further the development of stochastic models of criminal careers. 15 references