As empirical psychologists Snook, Cullen et al. (2008) ask, with no scientific evidence that profiling is effective, why do police (or the public) continue to believe in it or police to ask for profiling help? The answer lies in the way the profiling message is delivered, in conjunction with certain types of thinking errors present in the mind of audience members. First, there is a heavy emphasis in the writings on CP on personal experiences, or anecdotes. However, no matter, how interesting such tales, or how effective they are in convincing people of the truth of their claims, anecdotes have no scientific worth. Personal experiences cannot be verified by anyone else and, unlike the specifics of case histories, they cannot lead to valuable generalizations as in an inductive approach. In addition, these personal tales always yield successful conclusions. The FBI, for instance, never reveals its failures, or allows others to look at their data. Instead, they report the number of correct predictions, when they should be reporting the proportion of correct predictions (Snook, Cullen et al., 2008).
The likelihood of one's believing a message is always increased by its repetition, by claims of its growing use and claims of its successes. And nothing has been more frequently told than we have been about the FBI's assistance to other nations and purported successes in solving a variety of complex cases. The more times people hear that something works, the more apt the claim is to gain credence, without having to show any actual evidence of its effectiveness. An example of this thinking bias can be seen in the work of social scientists Shipley and Arrigo (in Kocsis, Ed., 2008, pgs. 120-121). On one page they admit that there has been no empirical validation for any of the serial murder typologies, but on the next they offer, without comment, FBI agents Ressler's et al.'s (1986) claim that the crime scene "can be used much like a fingerprint to help identify the murderer."
In selling themselves as experts, profilers oversell their successes, the extent and variety of their areas of expertise and they emphasize their years of specialized training, although each of these have been shown to be an exaggeration, untrue or unhelpful. Training in Investigative Psychology, as an example, has not enabled its practitioners to provide police with the types of assistance they need. It may also be the case that acting as expert witnesses in trials has given the profilers a bump. But legal scholars (e.g. Ormerod, 1996) cited by Snook, Cullen et al. (2008, see pg. 1266) lodge a number of arguments against this idea. e.g. profiling is not a generally accepted scientific technique; it is not reliable; it cannot pinpoint a killer and it cannot explain any facet of human behavior that is beyond the understanding of a jury. Occasionally, profilers have not been allowed to participate in trials because, it is claimed, they practice "junk science."
Snook, Cullen et al. (2008) also explain how the mind works such that unverified messages are likely to be believed. Briefly, one reason people are easily misled about the worth of profiling is the ambiguous nature of profiles. This is particularly true when an ambiguous profile is used after the apprehension of the offender (which is called hindsight bias). If it appears that whenever profilers show up, cases are subsequently successfully resolved, this may seem to be a causal pattern. Seeing a pattern where none exists is called an illusory correlation.
Perhaps the most relevant bias that occurs with profiling is that the desire to find evidence that supports belief in profiling may unwittingly lead us to see evidence and patterns where none exist and to ignore evidence that disconfirms what we want to learn. But what happens when the predictions from actual cases are evaluated with scientific methods? Empirical psychologists Snook, Cullen et al. (2008) provide this example from the literature. Alison, Smith, Eastman, et al. (2003) examined 21 profiles that had been used in major criminal investigations. There were 880 statements that predicted characteristics of an unknown offender. They found that 82 percent were wrong, 55 percent were not verifiable, 24 percent were too ambiguous for testing and only 28 percent could be falsified (verified). Thus, in the vast majority of cases, the vast majority of predictions made by profilers were unsubstantiated, and most of the remaining predictions could not even be tested.
The work conducted by Snook et al (2007, 2008) is highly recommended and any deficiencies in the author's articles are likely to be clarified by reviewing their work.
In summary, it is unlikely that serial murder will ever be curtailed by continuing reliance on criminal profiling. Further, profilers do more than fail to follow behavioral science rules or achieve scientific credibility. They have taken over the entire enterprise of providing police with new solutions to unsolved crimes so that nothing else is being tried. They provided a single, untested, tool to fight a war with unknown enemies and unknown dimensions. They picked a battle against sex murderers whose numbers were hidden in the "unknown" motives category yet their interest in motives surpasses all else. Police have turned to profilers because they have had nothing else. Because at least some have been made to feel as if they have less to offer than profilers or other outside experts. It is time that they take back their power and that they are given help, not in the form of experts, but in the form of knowledge about what works.
And much is already known. Wellford and Cronin (1999) found that most homicides can be solved by the actions of police and sufficient manpower, used in specified, common-sense ways. Their work should be viewed as a manual and implemented in cities across this nation. However, eventually curtailing serial murder requires a focus on deterrence and prevention (i.e. increasing apprehension, conviction and the certainty of penalties, along with de-glorification of serial killers in the media.) Both England and Germany have reduced serial murder waves (Jenkins, 1988; Ritter, 2001) and maintained high homicide clearance rates, and we can learn from their success. If we still have the will and a concern for the lives of our citizens.