Geo-Profiling

In an age when a computer-based solution is offered for nearly every problem, it's almost a knee-jerk reaction to look for an automated method to address difficult situations. This is as true in law enforcement as elsewhere, especially when every fourth booth at a police trade show seems to be offering some kind of computer or software. Recent research shows that the man or woman in Basic Blue is just as effective, if not more so, at some tasks where a computer solution is available.

Police experience and research indicates that most offenders commit the majority of their crimes in a geographical "comfort zone." Geographic profiling is the technique of predicting an unknown offender's residence, place of work or other location that could serve as a base of operation. When working a case with no leads, it is obviously advantageous to have an idea where the offender feels most comfortable. Computer software has been developed to help detectives locate this comfort zone.

It is the contention of five researchers from the United Kingdom (Craig Bennell, Brent Snook, Paul Taylor, Shevaun Corey and Julia Keyton) that officers who receive a minimal amount of training can predict the location of serial offenders as well as the computer software. The research, published in Criminal Justice and Behavior (January 2007), suggested that people using a simple mental template could come up with very accurate results.

Initially the theory was tested using students. Students were given 10 maps, each one marked with the locations of five crimes committed by a serial murderer; they were asked to guess where they thought the killer lived. One group of students then received two additional bits of information. One was that serial offenders often live near their crimes; the second was that offenders live within a circle created around where the crimes are committed. The group, armed with this minimal training, increased the accuracy of their predictions equal to that of the computer software "Dragnet." In a similar test, students made predictions as accurate as the geographic profiling tool in "CrimeStat." Some students were even more accurate than the software!

Would success with students translate to success with law enforcement officers? The researchers assembled 91 police officers from a large constabulary in the United Kingdom. None of the officers had a very good working knowledge of geographical profiling. The participants were provided with 36 maps pinpointing the crime scenes of serial burglaries in Newfoundland, Canada. About half the officers completed all 36 maps with no training. The other half of the officers looked at the first 18 maps and selected where they thought the perpetrator lived. These officers then received the same brief training the students received (serial offenders live near their crimes and in a circle) and completed the last 18 maps. The results between officers who had received the training, officers who had not received the training and the CrimeStat computer program were compared.

The initial results showed that for the most part, even untrained officers performed about as well as the CrimeStat program. With the training, the officers outperformed the computer software. The little bit of training, coupled with the officers' experience and intuition, proved to be a very effective way of locating where the serial offenders lived. The implications for policing are enormous.

Very brief, simple, and inexpensive training can have a remarkable impact on officers' performance. For one thing, this means that expensive computer models, beyond the means of many departments' budgets, are not necessary to predict accurately where predators may reside. It also means the average officer is capable of making predictions that will expedite the solving of crimes.