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

How You Can Tell If the Simulations in Computational Criminology Are Any Good

NCJ Number
224431
Journal
Journal of Experimental Criminology Volume: 4 Issue: 3 Dated: September 2008 Pages: 289-308
Author(s)
Richard Berk
Date Published
September 2008
Length
20 pages
Annotation
This work discusses the validity of simulations in computational criminology.
Abstract
The work notes that computational criminology applies computer simulations to study topics of interest for criminologists, and that the validity of the simulations ultimately depends on whether they are able to reproduce empirical phenomena with sufficient accuracy. The sole means cited to determine this is by comparing model output to real observations. This paper provides an overview of how such model evaluations can be undertaken. The work provides background on the types of credibility, citing qualitative, internal qualitative, and external qualitative, along with a detailed discussion of data-based computer model evaluation. These include: model specification; determining evaluation criteria; data collection; constructing model approximations; analyzing model output; and revision of the model. The authors concluded that there is no substitute for testing a computer model against data, as it is a well-understood and widely accepted principle in the scientific disciplines that make extensive use of simulations. It is recommended that criminology settle for nothing less. Figure and references

Downloads

No download available

Availability