Geographic Profiling and its Application in Crime Investigation

Article Overview

  • Title: Geographic profiling survey: A preliminary examination of geographic profilers’ views and experiences

  • Authors:

    • Karla Emeno, University of Ontario Institute of Technology, Canada

    • Craig Bennell, Carleton University, Canada

    • Brent Snook, Memorial University, Canada

    • Paul J Taylor, Lancaster University, UK

  • Abstract:

    • Geographic profiling (GP) is an investigative technique used to predict a serial offender’s home location based on the locations of crimes committed.

    • Growing use of GP in policing, but limited understanding of its procedures and applications.

    • The study involved an international survey of police professionals who provide GP advice.

    • Survey focused on:

    • Construction of geographic profiles

    • Perceived usefulness and accuracy of GP

    • Examination of core GP conditions prior to profile construction

    • Types of cases using GP

    • Results indicate wide usage in various operational settings, even with violation of GP conditions. Respondents perceive computerized GP systems as more accurate than manual methods.

    • The research advances the understanding of GP's application and sets the stage for future inquiries.

  • Keywords: Geographic profiling, serial offenders, crime investigation, investigative psychology

Introduction to Geographic Profiling

  • Definition of Geographic Profiling (GP):

    • Defined as "a criminal investigative methodology that analyses the locations of a connected series of crimes to determine the most probable area of offender residence" (Rossmo, 2012: 144).

  • Functional Purpose:

    • Often used to prioritize suspects based on proximity to the predicted residence, focusing on those living closest first (Rossmo, 2000).

  • Theoretical Assumptions of GP:

    • Most serial offenders do not travel far from home (distance decay).

    • Most offenders live within the area of their criminal activities (domocentricity).

Methods of Geographic Profiling

  • Classification of GP Strategies:

    • Spatial Distribution Strategies:

    • Utilize the distribution of crime locations to estimate a central point representing the offender’s residence.

      • Examples include:

      • Centroid: Mean of x- and y-coordinates of crime locations.

      • Center of Minimum Distance: Defined by the farthest two crime points in a series.

    • Probability Distance Strategies:

    • Apply mathematical functions (linear, lognormal, truncated negative exponential) to predict the offender’s residence based on crime locations.

    • Generates a probability surface indicating how likely it is for the offender to reside in areas covered by their criminal activities.

  • Current Trends:

    • Computerized systems implementing probability distance strategies (e.g., Rigel, CrimeStat, Dragnet) are the most common methods for conducting GP today.

Accuracy of Geographic Profiling

  • Research Findings on Accuracy:

    • Rossmo (2000) evaluated Rigel using FBI serial murder case data and found a mean hit score percentage of 6%, indicating the area needed to be searched was small relative to the total area.

    • Canter et al. (2000) assessed Dragnet using disposal locations of 79 serial killers with an 11% average hit score.

    • Studies comparing complexities of GP strategies showed no correlation between complexity and accuracy (Snook, Zito et al., 2005).

    • Some studies (e.g., Paulsen, 2006a/b) indicated simpler strategies sometimes yield accuracy comparable to complex methods.

  • Measuring GP Performance:

    • Hit Score Percentage: Indicates how much of the area must be searched to locate the offender's home.

    • Error Distance: Measures the distance between predicted and actual home locations.

  • Debate Over Measurement Methods:

    • Critics argue that error distance may not reflect GP accuracy accurately, advocating for hit score percentage evaluation (Rossmo, 2011).

Core Conditions for Effective GP

  • Identified Conditions by Rossmo (2000, 2005a):

    1. Offender must have committed at least five crimes.

    2. Crimes must be linked to the same offender; series should be complete.

    3. Offender must not commute into the investigation area.

    4. Offender must not change anchor points during the crime series.

    5. Suitable target distribution should be uniform around the offender’s residence.

  • Decision-Making Challenges:

    • Determining factors like commuting vs. marauder status (Canter and Larkin, 1993) can introduce errors in profiling accuracy.

Application of GP Across Crime Types

  • Versatility of GP:

    • Initially developed for serial murder investigations, GP is applicable to other serial crimes:

    • Rape

    • Arson

    • Robbery

    • Burglary

    • Fraud

    • Auto theft

    • Kidnapping (Rossmo, 2012).

  • Accuracy Variation by Crime Type:

    • Paulsen (2006a) highlighted that GP accuracy varies with crime type, with specific crimes yielding better profiles than others (e.g., auto theft performing better than commercial robbery).

Methodology of Current Study

  • Survey Overview:

    • Conducted using SurveyMonkey from international police professionals involved in GP.

    • Participants were recruited via email, police mailing lists, and conferences.

    • The survey consisted of 47 questions designed primarily to elicit closed responses assessing key GP aspects.

  • Sample Size and Composition:

    • 35 began the survey; 22 completed it after filtering out non-qualified participants.

    • Composition: 16 males, 6 females; mean age 44.6 (range 29-65).

    • Majority had GP training (77%).

Results of the Survey

  • Profile Generation Practices:

    • 91% of respondents develop GP individually.

    • 77% used computerized systems; Rigel was the most utilized (53%).

    • GP leads to increased patrols in likely offender areas (77% usage).

  • Perceived Accuracy and Usefulness:

    • Average of 53.2% usefulness reported in moving investigations forward.

    • 75% found computerized systems the most accurate; educated guessing perceived as least effective.

  • Adherence to Core Conditions:

    • Some respondents use GP even when core conditions are violated.

    • A significant number (96%) still use GP when conditions regarding commuting status and anchor points are not checked.

Discussion

  • Insights Gained on GP Usage:

    • Computerized GP systems dominate in profile generation, and trained individuals show higher usage of these systems.

    • Positive perceptions of GP’s accuracy suggest an ongoing commitment to the methodology despite inherent challenges.

  • Considerations for Future Research:

    • Exploration of GP practices in varied contexts and the development of streamlined methodologies for participant recruitment in surveys.

    • Address limitations experienced in the current research, including sample size and methodological issues such as questionnaire length.

Conclusion

  • Despite limitations, the study provides preliminary insights into GP usage and perceptions in law enforcement. The findings underpin the ongoing relevance and application of GP as an investigative tool in police work.