U.S. flag

An official website of the United States government, Department of Justice.

NCJRS Virtual Library

The Virtual Library houses over 235,000 criminal justice resources, including all known OJP works.
Click here to search the NCJRS Virtual Library

Parameterization of Recidivism

NCJ Number
81011
Journal
Evaluation Review Volume: 5 Issue: 6 Dated: (December 1981) Pages: 810-821
Author(s)
H Wainer
Date Published
1981
Length
12 pages
Annotation
A model of recidivism proposed by Maltz and McCleary (1977) is modified to better fit the recidivism data used in the Maltz and McCleary study.
Abstract

Maltz and McCleary present a model to describe recidivism based on two premises: that the parolee population will divide according to those who will recidivate and those who will not and that the likelihood of recidivism can be described by a negative exponential function. They contrast their model with that given previously by Stollmack and Harris (1974) and show that the difference between the two is the first premise. This yields the following two parameterizations: P(t) = 1 - exp(-at) (Stollmack-Harris) and P (t) = rx 1 - exp(-at) (Maltz-McCleary). P(t) is the probability that a particular parolee will be returned to prison before time t; a is the failure rate, and r is the proportion of the parolee population that will ever fail. Maltz and McCleary present data (from Taylor, 1971) which shows that neither model fits the data very well (although Maltz and McCleary do not conclude this). An extension of the Maltz-McCleary model is proposed. The extension derives from the assumption that the speed of recidivism is proportional to the number of persons left to recidivate. This yields an additive constant to the exponent of the Maltz-McCleary model that allows the model to fit Taylor's data rather well. Uses for the model are briefly discussed. Tabular and graphic data and six references are provided.