Psycho-Geographical Profiling is concerned with the spatial analysis and psychological behavioral patterns of criminals. The technique employs a variety of methods, including distance to crime research, demographical analysis, environmental psychology, landscape analysis, geographical information systems, point pattern analysis, crime site residual analysis, and psychological criminal profiling. This process has both quantitative and qualitative (landscape analysis) dimensions to its application. Moreover, GPA seems to be a particular effective method for the needs of police investigators attempting to solve complex serial crimes. The technique examines the spatial data connected to a series of crime sites and in this study, more specifically, victims' body dump sites and point of fatal encounter sites.
Landscape Analysis - While several sciences specialize in such matters as crime-scene analysis, none better than geography in conjunction with environmental psychology can analyze the site and situation of the entire crime. Qualities of places and the relations among place are stock-in-trade for geographers. While the quantitative methods, at first at least, aim at general patterns of crimes in the aggregate, the qualitative method used here aims at reconnecting the known facets of the crimes to the actual landscape. In this phase, geo-forensic analysis attempts to reconstruct the course of travel of the killer and the victim from before the crime until well after it. Every reasonable effort is made to lay down a complete trail for each person in the crime (Newton, 1985).
The application of geographic predication is best described by Harries (1980 p: 120) "Prediction, depending on the level of its reliability, may be the most valuable end product of any social science investigation. Even some success in predication is often worthwhile, since it creates an awareness of tendencies within a system, without necessarily specifying either the parameters of those tendencies or underlying causes."
Geographical Profiling does not "solve" murder cases. Rather the method provides an additional avenue of scientific investigation that, with the many other forensic specialties, may provide some help for the investigation of a serial killing. The actual search for the killer remains completely in the orbit of the sworn. The geo-forensic techniques seem to provide an additional perspective that may assist police during the investigation of a serial murder in which bodies of victims are scattered over a large geographical area. The method seems also to be able to indicate that a sequence of murders belong in one serial, even when the police deny such an interpretation. Geographical Profiling may also be used to assist a defendant accused of one or more murders in a probable serial murder if his personal geography, because of the route patterns, made him logically unable to have committed the other murder in the apparent serial murder. The theories put forth in this research are in their infancy and further research is needed to test whether the applications are valid and robust. In the nonce, geographical pattern analysis of serial murder has so far confirmed the rationality and indeed the ordinariness of the killer.
Predator is Trade Mark under the U.S. and International laws. Any unauthorized use of this written material or aspects of Predator program is prohibited by law and is punishable.
Strengths of the Predator geographical profiling model compared to Rossmo's model and Canter's Dragnet system.
1: Predator does not assume that the crime area is only
circular but rather looks at the locations from all angular positions. The Predator
function calculates the two sides of two points using the Arctangent function (Atn).
Neither Rossmo's model or Dragnet considers the angles of crimes. On going research of 107
U.S. serial killers by Maurice Godwin, Ph.D. shows that crime locations of serial killers
fall somewhere between a 90 and 145 degrees. Data on which the Dragnet System is
based were collected from 'true-crime' magazines and books, which are not reliable
2: The Predator does not assume a 'buffer zone' but does,
however, make allowances for buffer zones, especially when analyzing the victims'
abduction sites. The is contrary to the Dragnet that does not make allowances for buffer
3: The Predator uses the more reliable UTM coordinates for plotting the crime locations.
4: The Predator uses the actual discrete crime locations
along with several distance-decay functions to geographically profile the crime series.
Predator does not use the rather aggregated 'mean inter-point distance' (MDS) that PlobPlot
uses. Dragnet assumes that the MID gives equal weight to all distances, however, this is
unlikely the case since it averages the distances even before the analysis is carried out.
Rather than rely on distances, the Predator uses the actual UTM coordinates to represent
the crime locations.
5: Such theoretical methods as the Dragnet employ presumptuous mean inter-point distances between crime locations that the creators assume represents a 'monotonous plain,' a theoretical surface only analogously related to the actual surface of the earth. To be sure, their monotonous plain lacks rivers, cliffs, streets, shopping centers, and so forth. However, the Predator Model employs two location-allocation theories. Firstly, it employs a method that looks at the location of central facilities and the allocation of flows to them. Such a method projects, theoretically, the probable trips of the offender to the center. Secondly, the Predator employs an analysis of a dispersing offender and reverses the directional assumption of the location-allocation model, while also keeping the assumptions of a monotonous plain and of optimization. Rather than assuming that the victim travels to the offender's home-base, it assumes that the offender travels outward from his or her home, so the 'allocation of flows' is 'from' rather than 'to' the center. This process allows the criminal's spatial movements to be modeled from all directions; no matter if they are moving in or out.
Note: Some have criticized me for not participating in a study that involved evaluating the methodology of various geographical profiling software programs back in 2004. In my opinion this study was biased from the start based on previous supporting comments from some of the participates on software that was to be evaluated. I actually proposed to the researchers a better alternative for evaluating the various software programs. I suggested that we pick an unsolved serial murder case where the victims were known to be linked to one offender for example through DNA or fingerprints. Then each owner of the software program would be provided the same exact geocoding information on crime locations associated with the unsolved murders. We then would submitted our results (maps) of the predicted offender's home anchor point to one researcher who would then maintain those results until the case is solved. Once the case was solved we then could review the geographical profiling results previously submitted to see whose predicted anchor point was the closest to the killer's home base location. However, I never heard back from those individuals carrying out the 2004 evaluation.
COPYRIGHT (2011): All rights reserved. No part of this may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from Dr. Maurice Godwin.
Back to Main Page