Criminal Profiling

Granfalloons and Gobbledygook

B R E N T S N O O K , P A U L G E N D R E A U ,
R A I G B E N N E L L , A N D P A U L J . T A Y L O R

READING THE OAIMS OF CRIMINAL PROFILERS and so on.3 (To view a profile, see http://www.
and watching popular television programs like brgov.convTaskForce/pdf/profile.pdf.) Notwith-
Criminal Minds can leave one with the standing these developments, the core fcxrus of
impression that Criminal Profiling (CP)—the task CP remains the derivation of inferences alx3ut
of infening demographic and personality détails an unknown offender's characteristics. Yet, a
of an offender from his or her crime scene 2001 study regarding the content of criminal
actions—is a well-practiced and reliable inves- profiles found tliat only 25% of statements in
tigative teclinique. Over the past three decades, profiles were inferences about offender cliarac-
CP has gained tremendous popularity as a teristics. Of that 25%, 82% of the inferences
media topic, an academic area of study, and a were unsubstantiated, 55% were unverifiable,
tool for police investigations worldwide. 24% were ambiguous, and 6% contained
However, as we demonstrate in this article, oppasing alternatives.'*
the acceptance of CP by many police officers, The specific process that profilers use
profilers, and the public is at odds with the to make their inferences appears to be shaped
absence of scientific evidence to confirm its relia- by their training. Profilers who emphasize a clini-
bility or validity. We think this confusion has ads- cal/psychological perspective draw on their psy-
en for two related reasons. The first is that chological training, knowledge and experience
people have developed a biased picture of CP with criminiil l:)ehavior, and possibly their intu-
because they typically hear only about its glow- ition, as they make their inferences. At its worse,
ing successes. The second, related, reason relates this type of CP appears to differ little from what
to what we know about cognition and the man- "psychic detectives" allegedly do when helping
ner by which people process information, which law enforcement agencies catch criminals or find
typically serves to support the credibility of CP. missing persons.'' In fact, you can probably take
any article or Ixxïk written on psychic detectives
The 5 W's of Criminal Profiling and replace the temi "psychic detective" with
1. What is profiling.-'"S^hen CP was originally "criminal profiler" and the a i l m e n t would cx5n-
popularized by the FBI, a profile consisted tinue to make perfect sense. By contrast, statisti-
primarily of a list of very basic characteristics cally oriented profilers claim to tese their infer-
(e.g., age, previous convictions) that were likely ences on the statistical analysis of data, which
to be possessed by the unknown offender of the comes from offenders wiio have previously com-
crime{s) under cor^ideration. ' Profiles were gen- mitted crimes that are similar to those being
erally used to narrow a list of potential suspects, investigated.
focus investigations, and construct interview 2. Who are profilers? Surprisingly, there is no
strategies.^ In more recent years, the potential coasensTJS about who Is qualified to be a profiler.
forms that a profile can take and the ways in Some have maintained that a profiler is anyone
which it can be used within a criminal investiga- who labels themselves a profiler and has engaged
tion have expanded to include suggestions in the practice of constructing a profile for a crim-
re^rding resource prioritization, case manage- inal investigadon,"^ whereas others have argued
ment, strategies for dealing with the media. that only individuals who have considerable

fi- V O L U M E 14 NUMBER 2 2008

investigative experience should be profilers," reason why police officers use CP is that they
Although some attempts have been made to reg- believe it "works". Indeed, survey results indicate
ulate and accredit profilers (e.g.. The International that some officers believe profiles are operational-
Criminal Investigative Analysis Fellowship), there ly useful, often because they reinforce their own
is no recognized regulatory body that provides a opinions, further their understanding of the
professional CP designation. Thus, those present- offender, and/or focus the investigation.'-^ Of
ing themselves as profiles may vary widely in course, it is also possible that some officers may
their level of experience and education. use CP simply because they believe they have
5- When is profiling used? The use of profil- "nothing to lose" by consulting a profiler, and/or
ers has typically been limited to certain low-vol- they are forced to do so in order to satisfy judicial
ume crimes such as sexual assaults committed requirements to exploit all available investigative
by strangers and homicides that appear to lack options to solve the crime.
a motive. Profiles are seen to be most useful in
these types of cases because offenders are Police Officers' Opinions of Criminal Profiling
more likely to exhibit psychopathology such as The few surveys that have asses.sed police offi-
psychopathy, schizoid thinking, and sadism.** cers' opinions about CP suggest they generally
This is assumed to increase the degree to find CP useiul for their investigations. An early
which offenders behave consistently across survey found that solving cases was attributed to
their crimes and other aspects of their lives.^ It CP advice in 46% of the 192 instances where
is also the case that a profiler may be consulted FBI profiling was requested.^'* Similarly, a 1993
at various stages of the investigation.'^ study found that 5 out of 6 surveyed police offi-
4. Where is profiling used? It appears that the cers in Tlie Netlierlands reported some degree
majority of CP occurs in the United States of usefulness for advice given by an FBI trained
through the FBI, with the most recent estimates profiler.i^ Likewise, a 1995 study found that 83%
indicating that CP is being applied in approxi- of a sample of 184 police officers in the United
mately 1000 cases per year.^* CP is also being Kingdom claimed that CP was operationally use-
used heavily in the United Kingdom, with 242 fiil and 92% reported that they would seek CP
instances of CP advice toeing reported between advice again.^^ Consistent with these results, a
1981 and 1994.12 Although exact estimates of CP 2001 study showed that a significant portion of
prevalence in other countries are not directly police officers in the United States believe that
available, its use has been documented in CP has value.1^ Finally, a 2007 survey of
Sweden, Finland, New Zealand, South Africa, Canadian police officers found that 66% of the
Germany, Canada, Ireland, Malaysia, Russia, officers believed that it contributed to their
Zimbabwe, and The Netherlands. investigation. Moreover, most officers reported
5. Why isprofUif^ used^ The most obvious that the profiler made accurate predictions.'^

WWW.SKEPTIC.COM
Putting CP to the Test general, profilers seem to ignore this empirical
Despite the fact that pwlice officers hold these research.
views, a review of the CP literature reveals that: Profilers also appear to be oblivious to
(a) the majority of CP approaches are based on research in closely related fields. For example,
an outdated theory of personality that lacks despite a massive effort to identify predictors of
strong empirical support, and (b) professional consistency in offender samples within commu-
profilers have a dismal performance record nity and prison settings, research has failed to
when the accuracy of their profiles have been turn up anything of value to criminal profilers.
examined. While it is possible to make reasonably accurate
Is CP based on an empirically supported theory? predictioas of criminal behavior with respea to
In a similar way to a theory of personality (the reddivism,^'' these inferences are liased on the
classic trait theory) that was popular in personali- analysis of behaviors beyond thase exhibited at
ty psychology up until the late 196Os.^'^ the over- an offender's crime scene. Indeed, the well-estab-
whelming majority of CP approaches assume that lished predictors of criminal l^havior (e.g., anti-
criminal behavior is determined by underly- social attitudes, cognition) are not the sorts of
ing dispositions (i.e., traits) within offenders variables typically focused on by profilers (e.g.,
that make them behave in a particular way.^^o crime scene behaviors), which raises unaaswered
The assumptions tliat emerge from this theory questions about why profilers might expect that
are fundamental to CP. For example, the trait the- behaviorally-liased profiling approaches will be
ory leads to an assumption that offenders will effective.
exhibit similar behaviors across their offenses Can professiotial profilers mak^ accurate
because traits, rather than situational factors, are inference? Within the CP domain, negligible
the determinants of their behavior. Perhaps more quantitative differences have been found
important for the practice of CP, the theory also between the predictive ability of "professional
suggests that offenders V.TU display similar behav- profilers" and "non-profilers". The accuracy of
iors in their crimes and in other aspects of their profiler inferences has been tested by compar-
lives (e.g., in their interpersonal relationships). ing the performance of so-called professional
The sole reliance on trait-based models of profilers with that of non-profiler groups in
profiling is fundamentally flawed. Criminal profil- mock profiling scénarios.^"* In a typical experi-
ers do not seem to recognize that a consensus ment, profilers and non-profiler groups are
began to emerge in the psychological litera- asked to review details of a solved crime and
ture some 40 years ago that it was a mistake to make inferences about the likely offender (via a
rely on traits as the primary explanation for multiple choice questionnaire). Inferences are
behavior.21 Situational factors contribute as much typically divided \r\Xo four categories; cognitive
to the prediction of behavior as personality dis- processes (e.g., whether or not offender exliibits
positions. This is equally true when predicting remorse), physical attributes (e.g., presence of
criminal Inehavior. facial liair), offense behaviors (e.g., whether the
Tlie importance of situational factors is appar- offender removed items from the crime scene),
ent when one considers research in die profiling and social history/habits (e.g., alcohol consump-
domain. For example, offenders rarely display tion). The results from these four categories are
higti levels of behavioral consistency aaoss the also combined to form an overall profile per-
crimes they commit.-^ A similar picture emei^s formance measure. The accuracy of these infer-
when evaluating the degree to wliich offenders ences is then checked against the actual perpe-
exhibit consistency across their crimes and other trator's physical characteristics, thoughts, and
aspects of their lives. At best, small pockets of behaviors.
consistency have been identified, whereby a spe- Two 2(X)7 meta-analyses of these studies
cific crime scene behavior is found to relate to a were revealing.^*» The first analysis compared the
specific background characteristic. For example, predictive accuracy of a group of self-labeled
a 1997 study found that rapists who forced entry profilers and experienced investigators against
into premises were four times more likely to non-profilers (e.g., college students and psychol-
have prior convictions for property offenses than ogists). The profilers/investigators were found to
those who did not engage in tiiat behavior.^^ In be more accurate than non-police personnel on

V O L U M E 14 N U M B E R 2 2 0 0 8
 T
an overall measure of profile accuracy (r = .24) appears to have helped investigators. Second,
and on tlie physical attribute category (r -= .10). anecdotal evidence from any source may exag-
On the other hand, the predictive accuracy gerate the actual usefulness of a profile in vari-
of the profilers/investigators was marginally ous ways. Tlilrd, profiling anecdotes are prone
worse or no better than the non-profilers to be distorted in some way to make them
when it came to inferences of cognitive more entertaining and informative.
processes (r = -,06), offense behaviors (r = .00), 2. Repetition of the message that "profiling
and social history/habits (r= -.09). works." Repeating the message that "CP is an
in the second analysis, the experienced inves- effective investigative tool" or "police officers
tigators were included in the non-profiler group. .seek profilers' input" can contribute to the CP
In this analysis, the results favored the profilers illusion because people tend to telieve messages
across all five predictor categories, but the differ- they hear repeatedly. The 2007 study noted
ences were not lai^e enough to be statistically above found that the message "profiling works"
significant. The best result came when the overall is clearly stated in 52% of the 130 profiling
profile was considered (r = .32). However, even articles reviewed, whereas only 3% of articles
if this most optimistic of results could be replicat- unequivocally stated that profiling does not
ed, it warrants consideration that many variables work. As previously argued, that positive mes-
included in this analysis of profilers' expertise are sage is unsupported by the research on the pre-
well known in the criminológica! literature (e.g., dictive ability of profilers.^
the likelihood that a serial offender will be of a 5. Counting the hits and discounting the miss-
particular age. have particular convictions, suffer es. Profilers create the impression that their infer-
sLibstance abuse problems, etc.). This means, and ences are highly accurate by over-emphasizing
we hasten to emphasize this point, that any their correct inferences.^ When all the necessary
police professional with a good knowledge of and pertinent information is not explicitly repc^rt-
the criminological literature should be able to ed, readers may form beliefs based solely upon
achieve this level of success simply by relying on the information that is presented to them.
base rate information. In other words, success in Findings from psychological research suggest that
CP does not appear to be based on specialized the exclusive presentation of correa inferences
knowledge of the peculiarities and idiosyncrasies can lead people to overestimate the accuracy
found at a given crime scene. and potential utility of profiles. It Ls tlierefore not
surprising that reading articles about profiling
Why do People Believe that Criminal Profiling Works? might lead people to conclude that it is a viable
Given tlie state of affairs with CP research, one tool.
can only wonder why police officers and the 4. Profilers are not 'experts." Experts are peo-
public would have faith in such a dubious tech- ple who have professional competence in a spe-
nique. Below are eight potential reasons. The cialized area. People have a tendency to accept
first four relate to how infomiation about CP is information that is reported to them by supposed
presented to p)eople. The second four relate to experts. However, problems can arise when peo-
how people might process that information. ple wholeheartedly believe in the power of an
1. The power of atiecdotes. CP accounts in expert s "specialized knowledge" when that
books, magazines, law enforcement bulletins, knowledge has little foundation. In practice, pro-
and peer-reviewed journal articles often rely filers present themselves as experts by implying
entirely on a "case in point', ''case study", that they possess accuniulated wisdom, investiga-
"actual case," or "success story" to illustrate tive and behavioral science experience, and train-
how profiling is useful in catching a criminal. ing and/or knowledge of abnormal behavior that
For instance, a 2(K)7 study found that 60% of provide them witli the necessary skills to collea
the CP literature relied on anecdotes as a and analyze crime scene information and peek
source of evidence.^^ But anecdotes are inade- inside the criminal mind. In addition, research
quate for effectively validating CP for at least has shown that police officers tend to believe
three reasons. First, in attempting to convince that profiles written by supposed experts are
others that profiling works, a profiler can surely more accurate than those written by other
find at least one anecdote in wliich a profile consultants, even when identical information

W W W . S K E P T I C . C O M f,
appeared in both profiles.^« The problem with 7. Imitation. People tend to believe things or
this state of affairs is that there is little evidence do things a certain way because they were
supporting the propasition that profilers' possess believed or done that way by others in the
specialized skills that warrant labeling them as past.^ In fact, a large amount of what we know
experts.^i is naturally acquired from other people's beliav-
5. Humans are pattern-seekers. Humans ior and instructions. Thus, those who observe
attempt to find order and meaning in the uncer- other people using CP are likely to lx)th use it
tain world and then form beliefs tliat can guide and believe it works, even if the initial user dcxís
future behaviors. ^2 In attempting to find useful not hold this belief Police officers may believe
patterns, however, people sometimes find pat- CP is a good investigative technique íxícaiise
terns that are meaningle.ss. When information is they ol")serve other police officers using it. Police
presented in such a way as to make us believe officers spend time with other officers, communi-
that CP works, it is no wonder that this Ls tlie cating various skills and proper policing behav-
conclusion that is reached. The information, iors through lx)tli fomial and infomial teachings.
however, may be biased in several ways. Through police culture, profiling advocates (e.g.,
Profilers may wish to inflate their own useftil- tliose officers trained to use CP) can directly
ness (self-serving bias) and may actually be and/or indirectly instruct other officers that CP is
more confident in their abilities than is warrant- effective. In any case, it is unlikely that any of
ed (over<onfidence); people might believe tliat the otlier officers would have access to all of the
a profiler's advice solved an investigation information needed to properly determine
because they are unaware of, or do not consid- whether CP works.
er, the rest of the police work that was involved 8. Mistakingfictionfor fact. Because people
in the case (attril^ution error); profilers, police are generally intrigued by the criminal mind, pro-
officers, and the public are prone to make errors filing activities tend to generate a lot of public
whenever they partake in after-the-fact reason- fascination. The increasing numl^er of books.
ing; and, perhaps most important, tliere may be films, and television programs that deal with pro-
a tendency to seek evidence that supports an filing, as well as tlie recent growth in college and
existing belief that CP works and ignore or filter university' courses that acidress profiling issues,
out evidence tliat contradicts such a belief (con- supports this observation. Exposure to primarily
firmation bias). fictional accounts of CP unfortunately means
6. Vague profilesfitany case. The inferences people may base their beliefs upon those
in some profiles are so ambiguous, vague, accounts; especially since people are not very
and/or general that the profile (like horoscopes) adept at remembering the source of information
can appear to describe any suspect.33 This is diat they acquire during routine daily activities.^5
problematic for both practice and research.
For example, in a case with multiple sus- Conclusion
pects, profiles that contain many ambiguous There is a growing I-»elief that profilers can accu-
inferences may not assist in xhe elimination of rately and conslstentiy predict a criminars charac-
the innocent. It is also possible that interpreting teristics based on crime scene evidence. This
ambiguous statements (and sub.sequently using i:)elief is evident from tlie fact that CP is becom-
that interpretation to guide investigative decision- ing increasingly prevalent as an investigative
making) may contribute to the anest of an inno- technique and that positive opinions of CP are
cent suspect and thus the release of, or the ces- being communicated in the published literature.
sation of a search for, the actual criminal. In Such a loeÜef is premature because tlie teclinique
this latter regard, readers should be reminded has yet to be theoretically or empirically support-
of the fi^quent reporting in tlie media of wrong- ed. Belief in this unscientific policing practice
ftil convictions. Regarding re.search, it Is difficult appears to be due to the erroneous infomiation
to retrospectively determine and report the actual that police officers (and the public) receive about
accuracy of profiles if they can be interpreted to CP and the way that this information is
fit many individuals. Moreover, ambiguous processed. Sinc-e profiling has the potential to
inferences are not falsifiable, thus the profiler mislead criminal investigators, it is a practice that
can never be shown to be wrong. must be approached with die utmost caution. •

ff- V o L U M E 1 N U M B E R 2 2 0 0 8
 T
References Investigative Analysis: of Consumer a Stranger Rapist from
1. Hicks. S. J., & Sales. B. An Overview." In A. W. Satisfaction with his Offence Behaviour."
D. 2006. Criminal Burgess & R. R. Specific Profile Ana/ys/s Science and Justice,
Profiling: Developing an Hazelwood (Eds.), and investigative Advice 37.161-170.
Effective Science and Practical Aspects of Offered by the 24. Gendreau, P., Little, T.,
Practice. Washington. Rape investigation: A Scientific Research & Goggin. C. 1996. "A
DC: The American Multidisciplinary Advisory Unit of the Metaxinalysis of the
Psycholo^cal Approach. Boca Raton, National Criminal Predictors of Adult
Association. FL: CRC Press. Intelligence Division Offender Recidivism:
2. Douglas. J. E., Ressler. 8. Geberth, V. 1995. (CRil - The Netherlands. What Works!"
R. K.. Burgess, A. W., "Criminal Personality Leiden: NISCALE: Cnminoiogy. 31. 4 0 1 -
& Hartman. C. R. Profiling." Law and Netheriands Institute 433.
1986. "Crimina! Order, 43, 4 5 4 9 . for the Study of 25. Kocsis, R. N., Hayes,
Profiling from Crime Criminality and Law A. F.. & Irwin. H. J.
9. Pinizzotto. A. J.. &
Scene Analysis." Enforcement.
Rnkel. U. J. 1990. 2002. "Investigative
Behavioral Scienœs "Criminal Personality 16. Copson. op cit. Experience and
and the Law, 4, 401- Profiling: An Outcome 17. Trager, J.. & Brewster. Accuracy in
421. and Process Study." J. 2 0 0 1 . "The Psychological Profiling
3. Ainsworth, P. B. 2(X)1. Law and Human Effectiveness of of a Violent Crime."
Offender Profiling and Behavior, 14. 215-233. Psychological Profiles." Journal of Interpersonal
Cringe Analysis. 10. Copson, G. 1995. Journal of Police and Violence, 17, 8Í1S23.
Cullompton, Devon: Coals to Newcastle? Criminal Psycholc^, 26. Snook, et al., op cit.
Willan. Part 1: A Study of 16. 20-28. 27. Ibid.
4. Alison, L. J.. Snnith, M. Offender Profiling. 18. Snook, et al. opcrt. 28. Ibid.
D.. Eastman. 0., & London: Home Office. B., Eastwood. J.. 29. Doutas, op cit.
Rainbow, L. 2003. Police Research Gendreau, P.. Gcggin, 30. Kocsis and Hayes,
"Toulmin's Philosophy Group; Snook. B.. C.,&Culten. R. M. 2004, op cit.
of Argument and its Haines. A., Taylor. P. 2007. "Taking Stock of 3 1 . Snook. Eastwood, et
Relevance to Offender J.. & Benneil. C. Criminal Profiling: A al., op cit.
Profiling." Psychology. 2007. "Criminal Narrative Review and 32. Shermer, M. 2002.
Crime, and Law, 9. Profiling Belief and Meta-analysis." Why People Beiieve
173-183. Use: A Survey of Criminal Justice and Weird Things:
5. Lyons, A. & Taizzi, M. Canadian Police Behavior. 34. 437-453. Pseudoscienœ,
1991. The Blue Sense: Officer Opinion." 19. Mischel. W. 1968. Duperstition, and Other
Psychic Detectives and Canadian Journal of Personality and Confusions of Our
Crime. The Mysterious Police and Security Assessment. New York; Time. New York: Henry
Press: New York. Services. 5, 169-179. Wiley. Hott and Company.
6. Kocsis, R. N. 2004. 1 1 . Witkin, G.1996. "How 20. Homant, R. J., & 33. Alison, L. J.. Smith,
"Psychological Profiling the FBI Paints Portraits Kennedy, D. B. 1998. M.. & Morgan, K.
of Serial Arson Skills: of the Nations Most "Psychological Aspects 2003. "Interpreting the
An Assessment of Wanted." U.S. News of Crime Scene Accuracy of Offender
Skills and Accuracy." and World Report, 120, Profiling." Criminal Profiles." Psychology.
Criminal Justice and April 22. 32. Justice and Behavior, Crime, and Law. 9,
Behavior. 31. 341-361; 12. Copson. op cit. 25. 319-343. 18S195.
Kocsis. R. N., & Hayes. 13. Copson, op cit; 2 1 . Mischel, op cit. 34. Dawkins. R. 2003. A
A. F. 2004. "Believing Snook, et al. op cit. 22. Bateman. A. L., & Devil's Chaplain:
is Seeing? Investigating 14. Pinizzotto, A. J. 1984. Salfati. C. G. 2007. Reflections on Home.
t i e Perceived Accuracy "Forensic Psychology: "An Examination of Lies. Science, and
of Criminal Criminal Personality Behavioral Consistency Love. New York:
Psychological Profiles." Profiling." Joumal of Using Individual Houghton Mifflin
International Joumal of Police Science and Behaviors or Groups of Company.
Offender Therapy and Administration. 12, 32- Behaviors in Serial 35. Johnson, M. K. 1997.
Comparative 40. Homicide." Behavioral "Source Monitoring and
Criminology. 48.149- 15. Jackson, J. L. van Sciences and the Law. Memory Distortion."
160. Koppen. R J.. & 25. 527-544. Philosophical
7. Hazelwood. R. R., Herbrink, J. C. M. 23. Davies, A., Wittebrood. Transactions of the
Ressler. R. K.. Depue, 1993. Dœs the K., & Jackson, J. L Royal Society of
R. L. & Douglas, J. E. Service Meet the 1997. "Predicting the London B, 352,1733-
1995. "Criminal Needs? An Evaiuation Criminal Antecedents of 1745.

W W W . S K E P T I C . C O M