NCJ Number
31977
Journal
Journal of Police Science and Administration Volume: 3 Issue: 3 Dated: (SEPTEMBER 1975) Pages: 301-304
Date Published
1975
Length
4 pages
Annotation
THE TECHNIQUES INVOLVED IN THE USE OF ALGORITHMS IN THE FORM OF ASSOCIATION COEFFICIENTS TO QUANTITATIVELY ESTIMATE THE STRENGTH OF INTERCASE RESEMBLANCE IS DESCRIBED.
Abstract
IN MANY INVESTIGATIVE AGENCIES, A MODUS OPERANDI FILE IS MAINTAINED TO BRING INTERCASE RESEMBLANCE INTO THE TOTAL INVESTIGATIVE EFFORT. IN MANY INSTANCES, HOWEVER, SIMILARITY IN METHOD IS AN ENTIRELY SUBJECTIVE DETERMINATION. THIS OBJECTIVE METHOD RELIES ON THE CODING OF CASES BY THE PRESENCE OR ABSENCE OF A SET NUMBER OF CASE FEATURES. FOR EXAMPLE, THE CODE COULD INDICATE WHETHER THE OFFENSE TOOK PLACE AT NIGHT OR WHETHER THE OFFENDER MADE AN ATTEMPT TO DEFEAT THE ALARM SYSTEM. THE CODING IS DICHOTOMOUS, AND ONLY ONE RESPONSE IS PERMITTED FOR EACH ATTRIBUTE. WHEN THE ARRAY OF FEATURES IS COMPILED, IT MAY BE COMPARED WITH A CORRESPONDING ARRAY ABSTRACTED FROM ANOTHER CASE. THIS COMPARISON MAY BE SUMMARIZED BY USE OF A 2X2 FREQUENCY MATRIX. THE UPPER LEFT QUADRANT WOULD INDICATE FEATURES PRESENT IN BOTH CASES, WHILE THE LOWER RIGHT QUADRANT WOULD INDICATE FEATURES ABSENT IN BOTH CASES. THE OTHER TWO QUADRANTS WOULD INDICATE FEATURES IN WHICH THE TWO CASES DO NOT AGREE. FROM THIS DATA MATRIX, COEFFICIENTS OF ASSOCIATION MAY BE COMPLETED. FOUR METHODS OF COMPUTING THIS COEFFICIENT ARE PRESENTED. BENEFITS OF THIS METHOD ARE SEEN AS: OBJECTIVITY; QUANTITATIVE RESULTS; INTER-RATER RELIABILITY; AND EASY APPLICATION TO COMPUTER ANALYSIS.