NCJ Number
11384
Date Published
1973
Length
211 pages
Annotation
DEVELOPMENT AND VALIDATION OF A CRIME-MODELING TECHNIQUE TO BE USED IN ESTIMATING EXPECTED CRIME LEVELS IN EVALUATIVE STUDIES.
Abstract
THE AUTHOR BEGINS WITH A GENERAL DISCUSSION OF RESEARCH EFFORTS IN THE AREA OF LAW ENFORCEMENT EFFECTIVENESS. DEFICIENCIES IN THIS TYPE OF RESEARCH ARE NOTED. ONE AREA OF DEFICIENCY IN EVALUATING CRIME CONTROL PROGRAMS IS THE INABILITY TO ESTIMATE ACCURATELY THE LEVELS OF CRIME WHICH WOULD HAVE EXISTED WITHIN AN EXPERIMENTAL AREA IN THE ABSENCE OF THE PROGRAM. SUCH A MEASURE IS ESSENTIAL TO THE SUCCESSFUL EVALUATION OF THE PROGRAM'S IMPACT. THIS STUDY DESCRIBES THE EVOLUTION AND DEVELOPMENT OF THE AUTHOR'S OWN MULTIPLE REGRESSION CRIME ESTIMATION TECHNIQUE-THE CRIME-CORRELATED AREA MODEL. THE ASSUMPTION UNDERLYING THE MODEL IS THAT A SET OF UNIVERSAL CRIME INFLUENCES OPERATE UPON CITIES IN SUCH A WAY AS TO CAUSE LEVELS OF CRIME WITHIN VARIOUS AREAS TO FLUCTUATE IN A SIMILAR MANNER. THE CRIME-CORRELATED AREA MODEL BASES ITS ESTIMATES OF CRIME LEVELS UPON THE LEVELS WHICH OCCUR IN OTHER GEOGRAPHICAL AREAS. RESULTS OF A COMPARATIVE ANALYSIS OF THE CRIME-CORRELATED AREA MODEL AND THREE TRADITIONAL CRIME ESTIMATING MODELS ARE PRESENTED. THE ANALYSIS REVEALED THAT THE AUTHOR'S PROPOSED MODEL WAS THE MOST ACCURATE FOR THE SAMPLE DATA SELECTED. THE DOCUMENT DESCRIBES A MANPOWER EXPERIMENT, CONDUCTED BY THE WASHINGTON METROPOLITAN POLICE DEPARTMENT, IN WHICH THE CRIME CORRELATED AREA MODEL WAS USED IN AN ANALYSIS OF THE RESULTS. A SELECTED BIBLIOGRAPHY IS INCLUDED. (AUTHOR ABSTRACT MODIFIED)