Criminal profiling is the practice of inferring personality, behavioural, and demographic characteristics of criminals based on crime scene evidence. The use of this approach has grown dramatically over the last 20 years, and has also attracted a good deal of media attention.
Despite its popularity, its increasingly frequent application has not been matched by a similar body of research as to whether this approach actually "works". A recent review article¹ examines the evidence which supports this approach.
The authors first conducted a narrative review of 130 articles, and found that the support for this approach was mainly justified by way of "commonsense" rationales, rather than on empirical research findings which validated the technique statistically.
In the second part of their evaluation of this field, Snook et al conducted a meta-analysis of a small group of studies which actually sought to determine, under experimental conditions, whether criminal profilers were able to be more accurate when provided with real crime scene information which had resulted in arrests and convictions, than a variety of "naive" individuals such as students or psychologists with no criminal profiling, training or experience.
This approach yielded results which are somewhat short of an endorsement of criminal profiling as a major advance in law enforcement. While the profilers tended to perform somewhat better than comparison groups across the outcome measures, the degree of difference between the groups was relatively small, and fell short of what the authors consider acceptable performance of an "expert" which should decisively out-perform the non-expert or lay-person. The authors further go on to caution against the over-enthusiastic uptake of profiling in criminal investigations, particularly until better information is available as to the predictive accuracy of this approach.
While the study reported here is small scale, and the pieces of research on which it rests have their limitations, the authors do conclude that the profiling field generally relies on a weak standard of proof and that profilers do not decisively out-perform other groups when predicting the characteristics of an unknown criminal.
Got a story for Corrections News or want to request the print edition?
Email commdesk@corrections.govt.nz or phone (04) 460 3365.
ISSN 1178-8453