Are Criminal Profilers “Any Better Than A Bartender?” Not Necessarily, Suggests Review Of 40 Years Of Relevant Research

GettyImages-136503301.jpgBy guest blogger Tomasz Witkowski

The profession of “criminal profiler” is one shrouded in secrecy, even giving off a hint of danger. Yet when the American psychiatrist James A. Brussel began profiling a particular suspect in the 1950s, law enforcement officers were not entirely inclined to trust him. However, it turned out Brussel accurately defined the suspect’s height, clothing and even religion. This spectacular success was the beginning of the profession of the profiler. The FBI formed its Behavioral Science Unit in 1974 to study serial predators. Since then, the art and craft of criminal profiling have become the subject of numerous books, TV shows and iconic films such as The Silence of the Lambs. Criminal profilers are not, however, just characters created to make interesting films and books – in the real world the accuracy of their expert opinions is often key to protecting the safety and lives of others.

Can we say, after the passage of 40 years since the job of offender profiling (OP) was established, that this profession is a craft worthy of trust, one whose practitioners make use of tried and tested tools, or rather would it be more accurate to describe it as an art-form grounded in intuition that supplies us with foggy, uncertain predictions? Answers to these questions are given by Bryanna Fox from the University of South Florida and David P. Farrington from the University of Cambridge in the December edition of Psychological Bulletin, where they present a systematic review and meta-analysis of 426 publications on OP from 1976 through 2016. 

While Fox and Farrington assure us that the field of offender profiling has made considerable improvement in the scientific rigor of its research, I believe a closer look at their results do not give us much cause for optimism. The problem is OP is a field whose beginnings had such little scientific rigor that even though things have improved there is still such a long way to go. 

For example, the proportion of OP publications using advanced statistical analyses has risen substantially, but while they were at 0 per cent four decades ago, the figure still stands at just 33 per cent. In the past ten years the number of peer-reviewed studies out-paced non-peer reviewed publications by a margin of three to one, which sounds like good news. However, the new review demonstrates that OP has been studied quite poorly and a science that fails after so long to uncover any meaningful effects is arguably nothing more than a cargo cult science.

For instance, despite being applied often in active police investigations, very few evaluations of the profiles’ accuracy or effectiveness have taken place. Less than 10 per cent of all publications contained evaluations of existing profiles or profiling strategies conducted to determine the level of accuracy of these profiles and methods. 

Also, no previous analysis has determined the specific recurring themes or criminal profile categories that exist within the body of publications on OP. Fox and Farrington took up this task, analysing 62 publications that have developed offender profiles. They found considerable variation in the number, characteristics and terminology used to describe otherwise similar profile types, which, by the way, have little in common with the pop culture depiction of the “Organized” and “Disorganized” serial killer profiles originally proposed by agents in the FBI’s Behavioral Science Unit in the 1980s. The recurring profile themes or category types found by Fox and Farrington in the literature included: expressive, instrumental, visionary, hedonistic, power/control, traveling and local homicide offenders.

The wide variety in methods and terminology used across the field is striking. Some authors proposed profiles on the basis of their investigative experience or the modus operandi at the crime scene, while others drew on their reviews of the OP literature or a statistical analysis of all available crime scene information. This wide variety makes synthesis of the findings of OP as a whole very difficult.

Far greater cause for optimism is given by the results of the meta-analysis of case-linkage analysis research from 2002 to 2016. Case linkage is the process of determining whether there are discrete connections, or distinctive behavioural factors, that associate two or more previously unrelated cases by means of crime scene analysis, such as victim selection/targeting, method of approach and attack, location type, nature and sequence of sexual acts, any verbal activity, modus operandi, signature behaviour, and the amount of time spent in the commission of the crime. Results indicate that, in general, case linkage analysis can be used successfully to link crimes in a series of offences committed by a single offender. In fact, well over half of all reported analyses of case-linkage accuracy were statistically in the moderate to strong range, whereas almost 20 per cent were very strong. 

Despite the tremendous strides made by OP in relation to where it began 40 years ago, the results of this new review suggest OP nevertheless mostly still fails to meet the “Daubert standard” – originating from a 1993 case and now employed widely in the USA, this standard defines the scientific credibility that must be passed for evidence to be presented in courtrooms. Specifically, there is no known reliability or error rate for the vast number of methodologies or profiles developed in the OP field. Yet Fox and Farrington remain optimistic. They believe that: ”OP appears to be on a positive trajectory in terms of the use of more scientific and statistical methods, data and analyses, and evaluations of our work. Continuation on this trajectory, particularly in the next 10 years, will help the field of OP transform from a ‘cocktail napkin psuedoscience’ into an evidence-based discipline that produces research used to help police, inform policy, and ultimately, make the world a safer place.”

Hopefully this will be the case. Today, however, our answer to the question posed in 1976 by British forensic psychiatrist Colin Campbell about criminal profiling in an article for Psychology Today  – “Are we better at this than a bartender?” – is a simple, “not necessarily.”

What Have We Learned From Offender Profiling? A Systematic Review and Meta-Analysis of 40 Years of Research

Post written by Dr Tomasz Witkowski for the BPS Research Digest. Tomasz is a psychologist and science writer who specialises in debunking pseudoscience in the field of psychology, psychotherapy and diagnosis. He has published over a dozen books, dozens of scientific papers and over 100 popular articles (some of them in Skeptical Inquirer). In 2016, his latest book Psychology Led Astray: Cargo Cult in Science and Therapy was published by BrownWalker Press. He blogs at Forbidden Psychology.

Article source: