Content uploaded by Craig Bennell
Author content
All content in this area was uploaded by Craig Bennell on Jan 12, 2022
Content may be subject to copyright.
THE PERSONALITY PARADOX IN
OFFENDER PROFILING
A Theoretical Review of the Processes Involved
in Deriving Background Characteristics
From Crime Scene Actions
Laurence Alison, Craig Bennell,
and Andreas Mokros
University of Liverpool
David Ormerod
University of Hull
Most approaches to offender profiling depend on a naı¨ve trait perspective, in which
the task of predicting personality characteristics from crime scene actions relies on
a model that is nomothetic, deterministic, and nonsituationist. These approaches rest
on two basic premises: behavioral consistency across offenses and stable relation-
ships between configurations of offense behaviors and background characteristics.
Research supports the former premise but not the latter. Contemporary trait psy-
chology reveals that this is probably due to the fact that Person ⫻Situation
interactions have an effect on offense behavior. When profiling reports rely on a
naı¨ve trait approach, such reports should be used with caution in criminal investi-
gations and not at all as evidence in court until research demonstrates its predictive
validity.
Many of the individuals who have been involved directly in providing
offender profiles for investigations highlight the notion that profiling involves the
derivation of personality characteristics from crime scene actions. They claim that
the aim of offender profiling is to derive an offender’s likely characteristics from
the way in which he or she committed a particular crime, thus helping the police
to identify the perpetrator (Blau, 1994). For example, Douglas, Ressler, Burgess,
and Hartman (1986) described this as “a technique for identifying the major
personality and behavioral characteristics of an individual based upon an analysis
of the crimes he or she has committed” (p. 405). According to Pinizzotto and
Finkel (1990), an offender profile “focuses attention on individuals with person-
ality traits that parallel traits of others who have committed similar offences” (p.
216). Turvey (1999) referred to offender profiling as the “process of inferring
distinctive personality characteristics of individuals responsible for committing
criminal acts” (p. 1). Finally, Rossmo (2000) claimed that the profiling process is
based on the premise that the “interpretation of crime scene evidence can indicate
the personality type of the individual(s) who committed the offence” (p. 68).
Previous evaluations of offender profiling have focused on post hoc assess-
ments of accuracy based on the degree to which police officers claimed they were
Laurence Alison, Craig Bennell, and Andreas Mokros, Department of Psychology, University
of Liverpool; David Ormerod, Law School, University of Hull.
This article was supported by an Overseas Research Scholarship awarded to Craig Bennell by
the Overseas Research Students Awards Scheme.
Correspondence concerning this article should be addressed to Laurence Alison, Department of
Psychology, University of Liverpool, Eleanor Rathbone Building, Bedford Street South, Liverpool
L69 7ZA, United Kingdom. E-mail: alisonl@liverpool.ac.uk
Psychology, Public Policy, and Law
2002, Vol. 8, No. 1, 115–135 Copyright 2002 by the American Psychological Association, Inc.
1076-8971/02/$5.00 DOI: 10.1037//1076-8971.8.1.115
115
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
satisfied with the advice that they received. For example, a number of reviews of
profiling have been carried out primarily on the basis of opinions of detectives
about its utility (Britton, 1992; Copson, 1995; Douglas, 1981; Goldblatt, 1992;
Jackson, Van Koppen, & Herbrink, 1993). Broadly speaking, all of these reviews
concluded that investigators approved of the advice given and found it useful in
various ways. However, none of these reviews are definitive and all were some-
what limited in the extent to which they were able to obtain a representative
sample of profiles. They were also limited by the fact that they relied almost
entirely on the subjective opinions of the investigators subsequent to the conclu-
sion of the case.
In contrast to these evaluations of offender profiling that have focused
primarily on product-related issues, few studies or theoretical reviews have
considered the profiling process itself or the premises on which profiling methods
are founded. Therefore, the goal in this article is to review the validity of the
assumptions integral to the process of profiling. In particular, we demonstrate that
much of what is considered standard practice in offender profiling falls short of
current understanding in psychology about various psychological processes and
principles, and therefore profiling practices do not meet basic expert witness
standards under Federal Rule of Evidence 702, as recently interpreted.
The following points highlight the general argument advanced in this article
as to why the derivation of background characteristics from crime scene actions
is unlikely to be a valid and reliable process: (a) Most current profiling methods
rely on a naı¨ ve and outdated understanding of personality and the trait approach;
(b) global traits, or broad personality types, are unlikely to be useful in predicting
criminal behavior; (c) it is unlikely that the classification of offenders into broad
personality types would enable the profiler to relate clusters of sociodemographic
characteristics to different types; (d) a theoretical framework that emphasizes the
importance of Person ⫻Situation interactions in generating behavior may lead to
a more productive research endeavor; and (e) profiling should be used with
extreme caution in criminal investigations, and not at all as evidence in court, until
research demonstrates its predictive validity.
Points (a) through (c) demonstrate how common profiling standards and
practices would preclude such material being used in court in cases in which the
profile is adduced to infer that, because the defendant is a particular type of
individual, it is more probable that this defendant committed the crime. Points (a)
through (c) reflect the contemporary personality theorist’s perspective that such
inferences are not possible. Moreover, other reliability factors, some of which
correspond to criteria set out in Daubert v. Merrell Dow Pharmaceuticals, Inc.
(1993), all militate against profiling advice being used in court. These factors
include the extent to which the principle has been tested, the extent to which (a)
the principle has been published and subjected to peer review and (b) the
witness’s specialized field of knowledge has gained acceptance with the general
scientific, technical, or specialized community. It is appropriate for psychologists
trained in areas such as personality psychology and social psychology to form part
of a relevant scientific community to evaluate such reports because many state-
ments in offender profiles purportedly rely on various psychological principles
and processes.
116 ALISON, BENNELL, MOKROS, AND ORMEROD
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
The Profiling Process Relies on a Naı¨ ve Trait Approach
Many statements contained within offender profiles tend to attribute behaviors
to underlying, relatively context-free dispositional constructs within the offender.
This practice bears a strong resemblance to traditional trait theories of personality
that were common in psychology before the 1970s (Mischel, 1968). As is the case
with traditional trait theories, the theory underpinning most forms of offender
profiling is nomothetic in its attempt to make general predictions about offenders.
It is also deterministic in its assumption that all offenders’behaviors are affected
in predictable ways. Finally, it is largely nonsituationist in its belief that behavior
is thought to remain stable in the face of different environmental influences.
The Profiling Process Explained
The assumption regarding primary traits is that they are stable and general in
that they determine a person’s inclination to act consistently in a particular way
(stable) across a variety of situations (general). As the notion of behavioral
dispositions implies, traits are not directly observable. Rather, they are inferred
from behavior (Mischel, 1999). In the case of profiling, these latent phenomena
are inferred from crime scene actions. An example illustrates how tautological this
argument can be. If a crime is particularly violent, this leads to the conclusion that
the offender is particularly aggressive. Similarly, aggressive offenders commit
any given crime in a particularly violent way. Thus, traits are both inferred from
and explained by behavior.
A more sophisticated but similar example comes from the work that Turvey
(1999) referred to in his discussion of motivational typologies, in which he
outlined a variety of rape types (power reassurance, power assertive, anger
retaliatory, anger excitation, and profit) on the basis of previous research by Groth
(1979) and Burgess and Hazelwood (1995). He stated, for example, that “power
assertive” offender actions “suggest an underlying lack of confidence and a sense
of personal inadequacy, that are expressed through control, mastery and humili-
ation of the victim, while demonstrating the offender’s sense of authority”
(Turvey, 1999, p. 173). These offenders, according to Turvey, “may begin to do
things that might lead to their identification” (p. 174). Although Turvey added the
appropriate caveat that, “all too often, investigators and criminal profilers use . . .
offender classifications to label a rapist’s behavior with a single investigative
‘diagnosis’” (p. 181), such classification systems have been used and remain in
use for that very purpose.
For example, in a report prepared in the case of the Scarborough rapist (Paul
Bernardo), a Federal Bureau of Investigation (FBI) profiler claimed that in
observing the first eight offenses of the unknown offender, the offender could be
classified as a “sexual sadist” (http://www.corpus-delicti.com/court_bernardo_
warrant.html). This led the profiler to conclude that the following character-
istics (and many others) could be predicted on the basis of such a classification
system: male, White, parental infidelity or divorce, married at the time of the
offense, known for cross-dressing, military experience, and a fascination with
police work. Similar examples are abundant in some of the advice given to police
forces in the United Kingdom, although the most significant case in the last
117THE PERSONALITY PARADOX IN OFFENDER PROFILING
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
decade was the advice provided by a psychologist in the Rachel Nickell murder
enquiry.
In the Rachel Nickell enquiry, a clinical psychologist provided the police with
an offender profile claiming that he was able to deduce from the crime scene both
the murderer’s sexual fantasies and a range of background characteristics. Exam-
ples of statements from the profile include the notion that the offender would be
a lonely individual who was socially and sexually inept, would live nearby and
probably alone, and that the offender would be interested in unusual and isolated
hobbies (Britton, 1997). When the police eventually focused on a suspect whom
they believed fit the profile, a man who had also been identified by a number of
witnesses as being near the scene of the crime at the time of the murder, a covert
operation known as Operation Edzell was set up.
This operation involved a female undercover officer befriending the suspect
through a lonely hearts club the suspect was known to have joined prior to the
murder. The operation was purportedly designed to establish whether the suspect
would divulge sexual fantasies to his newly found partner that corresponded with
those predicted in the psychologist’s profile of the murderer. After an exchange of
numerous letters between the undercover officer and the suspect, which slowly
became more sexually explicit on both parts, the enquiry team were convinced
that the sexual fantasies expressed in the suspect’s letters fit the profile allegedly
constructed at the beginning of the enquiry (the profiler in this case wrote his
original profile on a white board, the details of which were erased, and therefore
it is impossible to ascertain whether the original profile matched the profile that
was subsequently written down). When the suspect was prosecuted for the
murder, however, the judge ruled the evidence of the undercover operation
inadmissible. Subsequent to this decision, no further evidence was offered against
the suspect. Although there was no attempt to use the profile as the basis for
evidence, it is difficult to ignore the possibility that it may have shaped the
sequence of events in the enquiry and the way in which the undercover operation
was conducted. (See Ormerod, 1996a, 1996b, for further details of this case.)
The methods and propositions described in these last two examples, that
clusters of specific characteristics about the offender can be derived from exam-
ining crime scene actions, are not unique. Indeed, many of the most widely
recognized and often-used experts in the United States, United Kingdom, and
several other European countries have made similar claims (Åsgard, 1998; Boon,
1997; Douglas, Burgess, Burgess, & Ressler, 1992; Douglas et al., 1986). They
reflect the belief that the same behavioral dispositions that determine the style of
crime scene behavior are reflected in more general, non-offense patterns in the
individual’s life. In fact, some profiling advice claims that these dispositions are
linked directly with certain demographic features. Figure 1 exemplifies the sup-
posed relationship between offense behavior, latent traits, and demographic
features.
The most frequently cited theory of offense behavior in terms of this simple
model is exemplified in the FBI’s proposition of differences between “organized”
and “disorganized” offenders (Douglas et al., 1986). The implication for adher-
ence to such a simple model is that, for example, offenders will vary according to
the degree of “organizational control” over a victim, and this will also reflect a
stable trait emergent in other features of the offender’s life. Therefore, factors
118 ALISON, BENNELL, MOKROS, AND ORMEROD
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
such as social maturity, intelligence, and previous convictions can all be “pro-
filed” on the basis of the level of organization observed at the crime scene.
This process is illustrated in a quotation from Douglas et al. (1992) in which
they stated that “[t]he crime scene is presumed to reflect the murderer’s behavior
and personality in much the same way as furnishings reveal the homeowner’s
character” (p. 21). However, for the over 200 classification categories of murder,
rape, and arson that they presented in their Crime Classification Manual, it was
also conceded by the authors that “at present, there have been no systematic
efforts to validate these profile derived classifications” (Douglas et al., 1992, p.
22). Despite their acknowledgment, this theory and its derivations enjoy a con-
sistent presence in the literature on serial killing and sexual assault (e.g., Badcock,
1997; Britton, 1997; Hickey, 1997; Holmes & DeBurger, 1988). Furthermore,
they have been referred to in the advice given by profilers in police investigations
(Alison, Smith, & Eastman, 2001) and, although fairly consistently rejected on
appeal in the United States (Ormerod, 1996a), attempts have been made to use the
process in criminal proceedings.
Research on the Profiling Process
In a comparative study of profilers, groups of homicide detectives, psychol-
ogists, and students, Pinizzotto and Finkel (1990) described the steps that lead to
profiling inferences. They stated that professional profilers: (a) assess the type of
criminal act with reference to individuals who have committed similar acts
previously, (b) thoroughly analyze the crime scene, (c) scrutinize the background
Figure 1. Implicit model of offender profiling.
119THE PERSONALITY PARADOX IN OFFENDER PROFILING
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
of the victim as well as any possible suspects, and (d) establish the likely
motivations of all parties involved. Finally, a description of the perpetrator is
generated, as the fifth step, from the characteristics supposedly connected with
such an individual’s “psychological make-up” (p. 216).
It is argued that the inferential process accomplished in the five steps de-
scribed above can be represented in the question series, “What?” to “Why?” to
“Who?” (Pinizzotto & Finkel, 1990). On the basis of crime scene material (What),
a particular motivation for the offense behavior is attributed to the perpetrator
(Why). This, in turn, leads to the description of the perpetrator’s likely charac-
teristics (Who). This simple “What” to “Why” to “Who” inference assumes that
the supposed specific motivations that drive the initiation of the offense are
consistently associated with specific types of background characteristics of the
offender (e.g., “If Motivation X, then Characteristics A, B, C, and D”). This
practice is problematic because it is not clear how a profiler moves from one point
to the next (i.e., what rules of thumb connect each inferential leap; Pinizzotto &
Finkel, 1990). Moreover, profilers commonly do not specify which (if any)
behavioral, correlational, or psychological principles they rely on. Therefore,
labels such as “organized” and “disorganized” (Ressler, Burgess, Douglas, Hart-
man, & D’Agostino, 1986) may simply be the result of a readiness to attribute a
latent trait to a set of crime scene behaviors despite the lack of clear evidence for
the existence of such factors.
Furthermore, Prentky and Burgess (2000) criticized the practice of developing
offender profiles by identifying two potential problems: (a) the difficulty involved
in the correct classification of the offender based on extremely limited material
and (b) the inability to draw reliable conclusions based on “nothing more than
what one knows about a particular subtype” (p. 194). As yet, there have been very
few attempts to evaluate the evidence for the feasibility of such inferences in
relation to offender profiling. One recent study that has attempted to do this
demonstrated that there was no evidence to support the hypothesis that subsets or
clusters of behaviors were associated with subsets or clusters of offender back-
ground characteristics (Mokros & Alison, in press).
The Personality Paradox
The readiness to invoke dispositions rather than explain behavior in terms of
situational influences has been described by some as a disparity between intuition
and empiricism (i.e., an intuitive belief within most people that behavior is
cross-situationally consistent vs. much empirical evidence that suggests it is not;
Bem & Allen, 1974). Bem and Allen (1974) have called this disparity “the
personality paradox” and, according to their observations, individuals are prone to
infer stable dispositions from behavior even though evidence has consistently
demonstrated that global trait constructs fail to accurately predict behavior over
time and across specific situations. It is possible, therefore, that this inappropriate
heuristic has perpetuated the use of such inappropriate methods in offender
profiling because many of the individuals who currently provide profiling advice
have little academic understanding of personality and of the difficulties associated
with inferring behavior from global traits. In some cases, profilers openly admit
120 ALISON, BENNELL, MOKROS, AND ORMEROD
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
that opinions about offenders are based on intuition and experience rather than
scientific evidence.
This personality paradox is exemplified in the definitions of profiling outlined
at the outset of this article and in statements such as the following:
[P]rofiling rests on the assumption that at least certain offenders have consistent
behavioral traits. This consistency is thought to persist from crime to crime and
also to affect various non-criminal aspects of their personality and lifestyle, thus
making them, to some extent, identifiable. (Homant & Kennedy, 1998, p. 328)
Similarly, the traditional view of personality dispositions leads to the assump-
tion that “individuals are characterized by stable and broadly generalized dispo-
sitions that endure over long periods of time and that generate consistencies in
their social behavior across a wide range of situations” (Mischel, 1990, p. 112).
Empirical tests, however, generally yield little supportive evidence for this
traditional trait view (Mischel, 1968). In a number of classic studies carried out in
the 1920s and 1930s (Dudycha, 1936; Hartshorne & May, 1928; Newcomb,
1929), behavioral consistency was tested by observing people’s social behavior as
it occurred across a variety of natural settings. The intercorrelations among
behaviors composing a particular trait concept rarely exceeded .30, leading many
theorists to question not only popular trait theories (e.g., Mischel, 1968; Peterson,
1968; Vernon, 1964) but also the very concept of personality (Epstein, 1979).
More recent studies, such as the often-cited Carleton College study, also support
these earlier findings insofar as the two global trait concepts of friendliness and
conscientiousness mostly failed to allow predictions of behavior across specific
situations (Mischel & Peake, 1982). As in the earlier studies, the mean intercor-
relations of items were positive but very low (.13 and .08, respectively).
Beyond the lack of empirical support for high levels of behavioral consis-
tency, the traditional trait perspective has been criticized on other grounds.
Briefly, these include the following: (a) analysis of variance studies (summarized
in Bowers, 1973) demonstrating that much of the variation in behavior across
situations can be attributed to situational influences or Person ⫻Situation inter-
actions (however, see Golding, 1975, and Olweus, 1977, for criticisms of these
studies); (b) the circularity in reasoning that emerges when one tries to explain
behavior from a traditional trait perspective (Bandura, 1969; Cervone & Shoda,
1999a, 1999b); and finally, (c) recent findings suggesting that there exist impor-
tant forms of personality coherence that traditional trait models fail to capture, in
particular, forms of personality coherence that are manifest at the individual
(idiographic) level but not at a general (nomothetic) level (Cervone & Shoda,
1999a, 1999b; Zelli & Dodge, 1999).
On the basis of evidence concerning the traditional trait approach, one would
not expect a task such as offender profiling, in which global traits are derived from
specific actions (or vice versa) to be possible. Moreover, the profilers’task is even
more ambitious than this: Many profilers make inferences about characteristics
that are not appropriate for a psychological definition of traits. Rather, profilers
include features such as the offender’s age, gender, ethnicity, marital status,
degree of sexual maturity, and likely reaction to police questioning (Annon, 1995;
Ault & Reese, 1980; Grubin, 1995; Homant & Kennedy, 1998).
121THE PERSONALITY PARADOX IN OFFENDER PROFILING
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
Assumptions Underlying Offender Profiling
Two basic assumptions underpin the profiling methods discussed so far. First,
the actions of any given offender are consistent across offenses, or, in statistical
terms, intraindividual behavioral variation across offenses is smaller than inter-
individual behavioral variation (the consistency assumption). Clearly, if the
behavior of the same offender varied more than between offenders, it would not
be logical to look for a common denominator that summarized the behavioral
tendencies of one offender. The second assumption is that similar offense styles
have to be associated with similar background characteristics. For example,
aggressive individuals are aggressive both in their offense behavior and in the way
they act in other (non-offense) situations (the homology assumption). These two
assumptions relate to the necessary (consistency) and sufficient (homology)
conditions for offender profiling to be valid and useful.
The Consistency of Offense Behavior
Offenders’behavioral consistency has been examined in a variety of ways.
Traditionally, criminological research has adopted a molar approach, defining
behavioral consistency as the probability that an individual will repeatedly com-
mit similar types of offenses (Farrington, 1997). Being less occupied with legal
nomenclature, psychological research, in contrast, has emphasized a molecular
analysis of criminal behavior. In this context, behavioral consistency has been
defined as the repetition of particular aspects of behavior if the same offender
engages in the same type of offense again (Canter, 1995).
Numerous studies have provided some support for the notion of offender
consistency. For example, one study examined the consistency of behaviors
displayed by different burglars of residential properties (Green, Booth, & Bider-
man, 1976). On the basis of 14 aspects of the crimes, such as “location of entry,”
“method of entry,” and “value of property taken,” Green and his colleagues were
able to use cluster analysis to accurately assign 14 out of 15 cases of burglary to
the three actual perpetrators. Similarly, studies have suggested that behavioral
consistency exists in the crime scene behaviors of serial rapists, though only to a
limited degree (Canter, 1995; Craik & Patrick, 1994; Grubin, Kelly, & Brunsdon,
2001; Wilson, Jack, & Butterworth, 1996). Thus, there does appear to be some,
albeit limited, evidence supporting the consistency assumption in offender
behavior.
The Homology of Offense Behavior and Offender Characteristics
The second premise is that the manner in which an offense is committed
corresponds with a particular configuration of background characteristics. This
differs from the more humble findings that relate to bivariate measures of
association, examples of which include Davies, Wittebrood, and Jackson’s (1998)
and House’s (1997) observations on the relationship between previous convic-
tions and certain crime scene actions of rapists. Davies et al. used odds ratios and
base rates (i.e., probabilities derived from percentages), and House reported
percentages, to describe the relationship between pairs of crime scene actions and
previous convictions. For example, Davies et al. noted that offenders who display
122 ALISON, BENNELL, MOKROS, AND ORMEROD
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
awareness of forensic procedures by destroying or removing semen have a
likelihood of a previous conviction for a sexual offense that is almost four times
higher (exact odds ratio ⫽3.96) than those offenders who do not take such
precautions.
However, these simple relationships between a given action and a given
characteristic are very different from the far more ambitious accounts that are
often referred to by practitioners and researchers in the profiling field (e.g., see
Blau, 1994; Douglas et al., 1986; Pinizzotto & Finkel, 1990). In these cases,
attempts were made to profile clusters of background features from crime scene
actions to develop a psychological portrait of the offender. They were concerned
with multivariate forms of prediction in which particular configurations or sets of
actions were linked to particular sets of characteristics. When tested, however, the
results were not very promising.
In the study by Davies et al. (1998), the integration of a range of crime scene
actions as predictors within logistic regression models failed to show a substantial
improvement over the information obtained through simple base rates in the
majority of instances. Similarly, in the study by House (1997), the 50 rapists in his
sample appeared relatively homogeneous with respect to their criminal histories,
regardless of whether they acted in a primarily aggressive, pseudo-intimate,
instrumental/criminal, or sadistic manner during the sexual assault. More recent
studies have also failed to discover any relationships between specific behavioral
themes of rape behavior and subsets of offender background characteristics
(Mokros & Alison, in press).
Toward a Contemporary Model of Offender Profiling
Most varieties of offense behavior for which profiling may be of relevance
involve intense, relatively short-lived, and potentially traumatic interactions that
are generally characterized by the diametrically opposed interests of the offender
and victim. Therefore, the influence of situational factors and the role of the
victim should not be neglected. A theoretical framework that emphasizes the
importance of Person ⫻Situation interactions in generating behavior may provide
a more productive model for offender profiling.
Unlike traditional trait approaches that conceptualize traits as stable attributes
within each person that determine how we behave, contemporary trait approaches
view traits as probable “if . . . , then . . . ” relations between clusters of behaviors
and clusters of situations (Mischel, 1990; Wright & Mischel, 1987a). The con-
temporary view on traits, therefore, is not a challenge “to the existence of
dispositions, but to the assumptions about their nature” (Mischel, 1990, p. 131).
The central problem with traditional trait approaches, as perceived by contempo-
rary trait psychologists, is the “limited utility of inferring broad, context-free
dispositions from behavioral signs as the basis for trying to explain the phenom-
enon of personality, and for predicting an individual’s specific behavior in specific
situations” (Mischel, 1990, p. 131).
Thus, the analysis of behavior from a contemporary trait perspective extends
beyond the idea of drawing on underlying context-free dispositional constructs in
an attempt to explain and predict behavior. Instead, the approach takes account of
contextual details. The primary unit of observation is “not the unconditional
123THE PERSONALITY PARADOX IN OFFENDER PROFILING
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
probability of trait-relevant behavior, p(B), . . . rather, it is the conditional prob-
ability of a certain behavior or category of behaviors given a certain condition or
set of conditions has occurred, p(B|C)” (Wright & Mischel, 1987a, p. 1161).
Conditional trait theorists also recognize that behavior categories and situa-
tion categories will often vary in the degree to which they are “well defined or
fuzzy” (Wright & Mischel, 1987a, p. 1161). In the criminal domain, for example,
behavior categories (e.g., hostility) will often consist of a range of specific offense
behaviors, with some being highly typical of the category (e.g., the use of physical
violence) and some being more peripheral (e.g., the use of verbal violence;
Bennell, Alison, Stein, Alison, & Canter, 2001). Likewise, situation categories
(e.g., a high-risk event) in the criminal domain will consist of various contexts that
are highly typical (e.g., a victim screaming for help when the attack is taking place
in public) as well as contexts that are more peripheral (e.g., a victim screaming for
help when the attack is taking place indoors). Even the links between behavior
categories and situation categories will vary “in the degree to which they are
necessary and sufficient [e.g., if encountered by serious resistance, physical
violence will result] versus probabilistic [e.g., if encountered by serious resis-
tance, physical violence might result, with probability p]”; Wright & Mischel,
1987a, p. 1161). The probabilistic nature of these links will, in part, determine
what affects the behavioral consistency of offenders across their crime series and
how consistent offenders are between their criminal and noncriminal life.
Challenges for a Contemporary Model of Offender Profiling
One of the primary challenges for contemporary trait psychologists is to
understand very clearly the relationship between people’s behavior, on the one
hand, and the situations they typically encounter, on the other. In the words of
Wright and Mischel (1987a), contemporary trait psychologists must “identify the
categories of conditions in which predictable behaviors relevant to some dispo-
sitional domain are most likely to be observed” (p. 1162). This has proven to be
an extremely difficult task in the noncriminal domain, but it would be even more
difficult to accomplish in the investigative domain. In the investigative domain,
very little is ever known about the context within which any given crime takes
place beyond the time of occurrence, the location where the offense was com-
mitted, and some information about the victim—contextual features that profilers
are often encouraged to examine (Douglas et al., 1992).
However, possible ways of accomplishing this task in the noncriminal domain
are slowly beginning to emerge and may prove useful in the profiling context. For
example, it is possible to conduct interviews with people to determine directly
from them the situational features they consider relevant to particular behaviors
they exhibit. In one study, adults and children were asked to generate “if . . . , then
...” contingencies to describe why they think people behave the way they do
(e.g., “If Sally is teased, then she becomes withdrawn”; Wright & Mischel,
1987b). Cluster analyzing these “if . . . , then . . . ” contingencies proved to be a
useful way of identifying important situational features and relevant situation–
behavior links in a noncriminal context (Wright & Mischel, 1987b). This ap-
proach has the potential to reduce the number of “if . . . , then . . . ” statements
needed to predict behavior by showing how it is possible to generalize from one
124 ALISON, BENNELL, MOKROS, AND ORMEROD
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
“if . . . , then . . . ” contingency to another one located within the same taxonomic
group. This is extremely important, considering the potentially large number of
“if’s” that would be needed to represent every situation an individual may
encounter (Shoda, 1999).
A similar approach could also be applied in the profiling context by inter-
viewing offenders about the circumstances they consider relevant to behaviors
they exhibit (e.g., “If I encounter a victim who physically resists me, then I
become very hostile,” or “If I find myself in a situation where there is a high risk
of getting caught, then I become very controlling”). The problem is that previous
research in the noncriminal domain suggests that these classification systems may
not account for every individual, nor may they be “sensitive enough to the
important nuances of each situation that may critically affect their psychological
meaning” (Shoda, 1999, p. 162). The same sorts of problems would be likely to
arise in the profiling context as well.
In addition to identifying relevant “if’s” and “then’s,” recent studies have
begun to indicate that the nature of the “if’s” in these situation–behavior contin-
gencies is extremely important to consider. For example, studies that have
examined how surface level categorizations of situations affect behavior have
yielded poor results (Lord, 1982; Shoda, Mischel, & Wright, 1994). Situational
similarity, defined purely by the physical features of situations (e.g., during
dinner, in the office, at the park, etc.), does not appear to influence noncriminal
behavioral consistency (Shoda, 1999). Instead, the psychological meaning of
specific situations to particular individuals must be established to generate any
valid inferences (Shoda, 1999). In one study, focusing on the types of interper-
sonal interactions taking place across situations has proved productive (Shoda et
al., 1994), whereas, in another, focusing on the competency demands character-
istic of particular situations proved a valid and useful way of defining situations
(Wright & Mischel, 1987a). Thus, an important development would be to assess
the psychological relevance of situational properties that might emerge in criminal
activity. When doing this, it might not be sufficient to simply know that a crime
was committed against a young woman, outdoors, during the night. Instead, the
psychologically active ingredients (Shoda et al., 1994) making up these criminal
situations may need to be identified and taken into account.
Some Cautionary Notes
Although contemporary trait approaches may be more suitable for making
inferences about personality features on the basis of crime scene actions, we
should add some cautionary notes about the extent to which this approach may or
may not be useful in the course of a criminal investigation or in court.
One would have to consider what is commonly termed the bandwidth–fidelity
or abstraction issue (Hampson, 1995). This originated with Cronbach and Gles-
er’s (1957) discussion of the bandwidth–fidelity trade-off, which refers to “the
differences in precision of [personality] measurement at different levels of ab-
straction” (Hampson, 1995, p. 95). As Hampson (1995) stated:
Where an abstract personality trait is involved [e.g., conscientiousness], then
measurement will be imprecise in the sense that the prediction of specific behav-
iors [e.g., punctuality] will be relatively poor. Nevertheless, predictions can be
125THE PERSONALITY PARADOX IN OFFENDER PROFILING
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
made for a wide range of behavior. . . . In contrast, where a less abstract trait is
involved, such as punctuality, then prediction of specific punctual behaviors will
be good. However, this fidelity is gained at the expense of bandwidth: knowing a
person’s punctuality is only predictive of a small range of behaviors. (p. 95)
The bandwidth–fidelity trade-off connects directly with current concerns
regarding levels of differentiation in offending behavior (Canter, 2000). Exam-
ining offense behaviors at a very general level (e.g., in terms of broad psycho-
logical themes underlying offender–victim interactions) may be productive in the
sense that one might be able to predict very general information about an
offender’s background. For example, there has been some success in demonstrat-
ing how, in cases of arson, particular configurations of actions (or behavioral
themes) are associated with different underlying themes within an offender’s
background (e.g., whether the arsonist has a general history of psychiatric prob-
lems, or whether the offender has a long history of previous offenses; Canter &
Fritzon, 1998). However, evaluating offense behavior at this aggregate level
probably means that profilers will not be able to predict very specific background
characteristics. For example, Douglas et al.’s (1992) accounts of the offender
having an “uncharacteristically detailed, precise, airtight alibi” (p. 25) or that the
offender will be “considered odd by those who know him” (p. 130) are probably
a little ambitious. Thus bandwidth is gained at the expense of fidelity.
There are many studies of sexual offenders that support the aggregate level of
research. For example, it has been demonstrated in one study that rapists display
significantly diverse facets of personality disorder depending on their level of
physical violence as demonstrated in their crime scene behavior (Proulx, Aubut,
Perron, & McKibben, 1994). Among the findings reported in this study was the
observation that more violent offenders score significantly higher on the histri-
onic, narcissistic, antisocial, and paranoid subscales than do the less violent
offenders. Similar results have been reported for another sample of rapists
(Langevin, Paitich, & Russon, 1985). In addition, when Knight, Warren, Rebous-
sin, and Soley (1998) derived a set of crime-scene behavior variables from a
sample of rapists and then applied these items to predict motivational character-
istics for a second sample of rapists, they were somewhat successful in predicting
the individuals who could be classified as expressive, aggressive, antisocial, and
sadistic types (Knight & Prentky, 1990).
In a more recent study, rapists were assigned to one of three groups (sadistic,
opportunistic, and anger rapists) according to their respective modus operandi
(Proulx, St.-Yves, Guay, & Ouimet, 1999). In this study, they found that sub-
stantial differences existed between the sadistic and the opportunistic types with
respect to personality disorders. The sadistic offenders were more likely to have
avoidant, schizoid, and dependent tendencies, whereas the opportunistic offenders
were characterized as narcissistic, paranoid, and antisocial. This study indicates
that it may be possible to discriminate between rapists on the basis of their crime
scene actions and that such differentiation may be reflected in personality. Such
a procedure could properly be referred to as a psychological profile because it
refers exclusively to particular psychological constructs. In contrast, the current
use of the term psychological profiling to refer to demographic characteristics is
126 ALISON, BENNELL, MOKROS, AND ORMEROD
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
a misnomer. Whether the former types of profiling could prove valuable to law
enforcement agencies remains to be seen.
Therefore, three main objectives for future research emerge. The first objec-
tive is to explore the relationship between personality, as defined through dispo-
sitional concepts, and crime scene actions. These dispositional concepts need to
take the form of measurable psychological constructs, such as aggressiveness or
hostility, rather than demographic features. Only then is the second step possible:
assessing whether these dispositions function as moderators of offender back-
ground characteristics. Finally, an alternative theoretical framework for exploring
the relationship between crime scene actions and characteristics of offenders can
be explored that avoids the previous major flaws of the traditional, global trait
perspective, such as its neglect of situational influences. In this regard, contem-
porary trait approaches may provide a more useful model for offender profiling.
Investigative and Legal Implications
In this article, we illustrate why the evidence for relying on a nomothetic,
deterministic, and nonsituationist model of offender profiling is not compelling.
Thus, any claim involving the derivation of specific clusters of characteristics of
the offender on the basis of an evaluation of the crime scene must be treated with
a degree of skepticism. Despite this, and despite the admission by many profilers
that their work is little more than educated guesswork, the utility of profiling
seems to be generally accepted as valid. Relatively recent reports suggest that the
FBI currently has 12 full-time profilers who collectively are involved in about
1,000 cases per year (Witkin, 1996). Few studies have considered whether such
profiling advice may actually be an impediment to an enquiry, and there are some
commentaries on Regina v. Stagg (1994) as to how that may have been the case
(Alison & Canter, 1999; Ormerod, 1996a). In contrast, most of the evaluation
studies have suggested that police officers were relatively happy with the advice
received and found it useful despite the fact that it very rarely led to the
identification of the offender (Copson, 1995).
The arguments that investigators commonly present for the utility of profiling
include statements such as, “It helps narrow the suspect search,”“It gives another
point of view,” and “It helps us understand the offender” (Alison et al., 2001).
These comments highlight the possibility that profiling can be of some help and,
clearly, with the attendant caveats about profiling standards and appropriate
knowledge of how to use such advice, investigations may well benefit from
external input of this sort. However, investigators should be wary of (a) reports
that rely on broad generalizations about behavior, (b) very specific statements
about the likely characteristics of the offender, and (c) claims about the motiva-
tions behind the offense. This may lead to an inappropriate focus on a subpopu-
lation of possible suspects, the mismanagement of resources, and a delay in the
successful resolution of the case. The argument is not that all profiling advice is
poor but, rather, that many of the most influential claims about offenders in the
last few decades have been based on intuition, outdated and ill-informed concepts
about personality, or on very small-scale studies that are methodologically con-
troversial. Thus, commonly used terms such as organized/disorganized (Ressler et
al., 1986), visionary type murderers (Holmes & DeBurger, 1988), and power
127THE PERSONALITY PARADOX IN OFFENDER PROFILING
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
reassurance rapists (Burgess & Hazelwood, 1995) should be treated with extreme
caution, particularly in cases in which the profiler claims that these typologies are
associated with specific sets of background characteristics.
In terms of using offender profiles in court, there are more persuasive reasons
to exclude testimony that relies on such simple classification systems. Although
there have been no cases to date in the United Kingdom that have incorporated
such advice, there have been several attempts to do so in the United States
(Ormerod, 1996a, 1999). The majority of these attempts have involved expert
witnesses providing evidence in relation to the consistency assumption discussed
earlier rather than the homology assumption, which is also integral to the process
of profiling. Recent examples include a number of cases in which attempts were
made by former FBI agents to link offenses to a particular offender on the basis
of so-called signature aspects of offense behavior (Douglas & Munn, 1992).
One such case involved 3 counts of murder in which substantial physical and
circumstantial evidence was available that linked the defendant to two of the
murders. In addition to this evidence, the trial court accepted testimony provided
by an FBI agent that each of the 3 murders in question had characteristic
behavioral similarities that suggested they were the work of the same offender
(Delaware v. Pennell, 1991). On appeal before the Delaware Supreme Court, the
defendant’s counsel argued that the testimony provided by the FBI agent did not
meet the standards of the test then imposed as a standard for all scientific evidence
to be adduced before the court—the Frye test—and therefore it should not be
admissible in court. The court affirmed the original ruling, however, stating that
the Frye test did not apply in this case because the FBI agent’s linkage analysis
was based solely on his own knowledge and experience and not on scientific tests
for which the Frye test would typically be applied.
In a similar case involving 4 counts of murder, evidence provided by an FBI
agent was admitted at the pretrial hearing to support the State’s position that
another set of crimes (consisting of 4 additional murders that the defendant was
also charged with) were, in fact, connected to the crimes the defendant was being
tried for. The FBI agent testified that, based on various ritual aspects of the
murders that were (in the agent’s opinion) particularly rare, the 8 murders were
the work of one person. The agent did not testify before the jury, but his testimony
was accepted by the trial judge as evidence on the issue of suspect identity. On
appeal before the Louisiana Supreme Court, the court found no error in the trial
court’s ruling to accept the FBI agent’s linkage analysis (Louisiana v. Code,
1993).
Unlike these two previous cases, in a more recent case (New Jersey v. Fortin,
2000), the Superior Court of New Jersey reversed an earlier ruling that permitted
a former FBI agent’s linkage analysis to be used as evidence. The case involved
1 murder and 1 attempted murder; the FBI agent provided testimony that due to
distinct similarities across the crimes in question, the two crimes were the work
of the same offender. Despite the fact that such evidence had been used in the
previously cited cases, the court was not persuaded that these techniques were
sufficiently reliable for use in this particular case because of the many differences
that existed between the crimes in question.
Finally, in a case that involved evidence being admitted that was related more
directly to the homology assumption, a former FBI agent provided testimony as
128 ALISON, BENNELL, MOKROS, AND ORMEROD
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
to the classification of offenses according to the organized/disorganized dichot-
omy. The former agent contended that the 9 crimes for which the defendant was
being charged fit a mixed category, displaying elements of both themes. The
defendant’s counsel highlighted the disorganized characteristics in an effort to
prove the defendant’s diminished responsibility. However, in cross-examination,
the State elicited testimony from the expert that some of the crimes displayed
signs of organization, thereby indicating a lack of mental illness. Under appeal to
the Supreme Court of North Carolina, the defendant’s counsel contended that the
cross-examination was improper as it was prejudicial and had no probative value
(North Carolina v. Wallace, 2000). However, the court concluded that the
cross-examination was permissible, as the jury had been given proper instructions
to limit their consideration of the expert’s testimony.
Thus, in some cases, advice regarding the consistency of offense behavior has
been considered sufficiently reliable and relevant to be helpful to the jury, and in
others it has been rejected. However, as far as we are aware, in no case has
profiling evidence been successfully used to provide evidence probative of guilt
or innocence. Ormerod (1996a, 1999) has comprehensively outlined the number
of hurdles that the expert testimony would have to overcome to be able to
introduce such evidence, at least in U.K. courts. Much of Ormerod’s discussion is
in relation to Regina v. Stagg (1994), but the points made are applicable to any
case in which attempts are made to introduce profiling evidence in court. There
are questions over relevance, whether the advice is expert evidence, whether it can
be used as similar fact evidence, and whether its potential for prejudicial influence
outweighs its probative value.
Ormerod (1999) noted that for a profile about the guilt of a specific defendant
to be legally relevant, it must render the facts at issue (the commission of the
crime by this defendant) more probable. Thus, in Ormerod’s view, it is not
sufficient to state that the defendant possesses the qualities of the type of
individual who could have committed the offense because the jury must evaluate
whether the defendant matching the description is the offender. Even if the
profiling advice were based on very specific details, such as those in the Paul
Bernardo case outlined earlier, using these details to prove the guilt or innocence
of the offender would not be possible because there is no reliable evidence to
suggest that such forms of profiling are possible. Indeed, the arguments against
the homology assumption suggest that it is very unlikely that such details could be
inferred from crime scene characteristics.
With hard scientific or technical evidence, one must consider the reliability of
the evidence. A ballistics expert is allowed to testify about the likelihood of the
match between the offense bullet and the defendant’s gun only because he or she
can demonstrate by reliable means that the tests are accurate. This is quite
different from the evidence of a fortune-teller, for example. If the police sought
to rely on the fortune-teller’s evidence to demonstrate that it was more probable
that the defendant committed the crime, the court would reject the evidence as
irrelevant. There is no reliability in the evidence, therefore, it has no foundation
on which to make its claim. Where does that leave a profiler who relies not on the
hard science of ballistics but on spurious and outdated psychological principles?
Given the arguments outlined in this article regarding the inappropriateness of the
homology assumption, it is improbable that a profiler would be able to demon-
129THE PERSONALITY PARADOX IN OFFENDER PROFILING
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
strate with sufficient strength any claim that he or she can reliably and consistently
identify character traits from crime scenes. The evidence would lack a reliable
foundation.
In Regina v. Stagg (1994), the judge also questioned whether profile evidence
was sufficiently well accepted as a scientific method according to strict definitions
in Daubert v. Merrell Dow Pharmaceuticals, Inc. (1993). Although the English
courts have not been overly concerned with this issue, the concern has been more
prevalent in U.S. courts, particularly in light of recent developments. In broad
terms, federal standards on the admissibility of expert testimony require that to be
considered expert evidence, the proffered witness (a) must qualify as an expert by
knowledge, skill, experience, or training; (b) must testify to scientific, technical,
or other specialized knowledge; and (c) his or her testimony must assist the trier
of fact. In cases following the Daubert criteria for admitting scientific evidence,
the following facts must be considered: whether the method consists of a testable
hypothesis, whether the method has been subject to peer review, the known
potential rate of error, the existence and maintenance of standards controlling the
technique’s operation, whether the method is generally accepted, the relationship
of the technique to methods that have been established to be reliable, the
qualifications of the expert witness testifying on the basis of the methodology, and
the nonjudicial uses of the method.
If the foregoing criteria were applied to the many opinions about likely
offender characteristics on the basis of types of classification systems endemic in
some profiling methods, the “reliability” of profile evidence would be extremely
controversial. As outlined earlier, the few studies that have assessed success rates
have been limited and have relied heavily on subjective opinions of the investi-
gating officers. Moreover, rates of “success” are difficult to evaluate because it is
extremely difficult to assess the extent to which the profile aided in the identifi-
cation of the offender. Where efforts have been made to extract such information,
results are not encouraging. For example, in Copson’s (1995) study, profiles
assisted in less than 25% of cases in solving the case, and in only 5 cases out of
184 (2.7%) did the profile lead to the identification of the offender. Studies have
also indicated that profilers do not process crime scene information differently
from any other group (e.g., students, police officers) in which they are untrained
in profiling methods (Pinizzotto & Finkel, 1990).
Courts in the United Kingdom, at least, are aware of the potential prejudicial
value of statements that appear to come from authoritative sources but are not
necessarily any more helpful than jurors’own opinions (Ormerod, 1999). A
particular danger may lie in Bem and Allen’s (1974) demonstration that individ-
uals readily attribute behavior to global traits and that much profiling advice
appears to reinforce this inappropriate heuristic. Therefore, a profiler proffering
advice on types of offenders, and how these classifications relate directly to sets
of characteristics, may be particularly prone to prejudicial thinking because it
exploits inaccurate laypersons’perspectives of behavior.
However, is it possible that arguments could be made to admit expert
testimony on offender profiling on the basis of a more flexible application of
standards, as outlined in Kumho Tire Co. v. Carmichael (1999)? The plaintiffs in
this case argued that the court’s application of the Daubert criteria was too
inflexible, resulting in the exclusion of testimony regarding visual observations
130 ALISON, BENNELL, MOKROS, AND ORMEROD
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
made by an expert with extensive experience in tire failure analysis. On the basis
of patterns of wear and tear on tires, the plaintiffs sought to use the expert to
support the view that the tire that blew out on a minivan driven by Carmichael,
thereby leading to a fatal accident, was defective. The issue regarding the
admission of the expert testimony centered on whether observations based on
extensive visual and tactile practical experience represent a standard of knowl-
edge sufficient to comprise technical or other specialized knowledge beyond the
knowledge of the jurors.
Profilers with extensive practical case experience may argue that this standard
of knowledge applies to them and thus affords them the privileged position of
commenting on offenses that the layperson would have little or no opportunity to
experience. This tacit knowledge may indeed lead to a greater appreciation of the
multivariate factors that underpin various forms of offending behavior. However,
if an offender profiler were to present profiling evidence in court, it would be
essential for them to make clear that their testimony was based on personal
opinion and experience rather than scientific evidence. As O’Connor and Krauss
(2001) made clear, the burden would be on profilers to know what their limits are
and to effectively articulate these limits to the court.
It seems improbable that any individual would have experienced a sufficient
number of cases to enable him or her to offer the level of detailed advice that
exists in many profiling reports. Many profilers comment on crimes that are
exceptionally rare, such as sexual homicide. Therefore, by definition, they would
not have had the opportunity to work on hundreds of such cases. Moreover, for
reliable conclusions, each case used to proffer an opinion would have to be similar
in nature to the profiled case. Due to the behavioral complexity involved in such
cases, it is unlikely that any individual expert would have encountered enough
cases that are sufficiently similar to be considered “extensive practical experi-
ence.” There is a qualitative and quantitative lack of experience in the field. Given
the research reviewed in this article, one could reasonably argue that these limits
are so overwhelming as to make any evidence based on offender profiles inap-
propriate for use in court, under federal approach to expert evidence, even with
the flexibility of the approach outlined in Kumho Tire Co. v. Carmichael (1999).
Conclusion
In terms of its evidential value, profiling cannot circumvent the problems of
legal relevance if it is adduced to seek to show that, on the basis of a profile
constructed from the crime scene, it is more probable that a particular defendant
committed the crime. Few profilers would or could claim, on the basis of
experience (or, indeed, any method), that they could single out the actual perpe-
trator. As Ormerod (1996b) stated, “[T]his above all other hurdles appears to be
the one almost guaranteed to trip up the psychological profile” (p. 350). Ormerod
(1996b) concluded with a quote from Smith (1993), which warned that psycho-
logical profiling “does not of course provide the identity of the offender, but it can
indicate the type of person being sought, i.e., the type most likely to have
committed a crime possessing certain unique characteristics” (p. 245). However,
the arguments outlined in this article assert that, because many contemporary
profiling methods rely on a naı¨ ve trait approach, even Smith’s more humble claim
131THE PERSONALITY PARADOX IN OFFENDER PROFILING
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
is overly optimistic. The notion that particular configurations of demographic
features can be predicted from an assessment of particular configurations of
specific behaviors occurring in short-term, highly traumatic situations seems an
overly ambitious and unlikely possibility. Thus, until such inferential processes
can be reliably verified, such claims should be treated with great caution in
investigations and should be entirely excluded from consideration in court.
References
Alison, L. J., & Canter, D. V. (1999). Professional, legal and ethical issues in offender
profiling. In D. V. Canter & L. J. Alison (Eds.), Profiling in policy and practice (pp.
21–54). Aldershot, England: Ashgate.
Alison, L. J., Smith, M., & Eastman, O. (2001, January). Toulmin’s philosophy of
argument and its relevance to offender profiling. Paper presented at the 6th Interna-
tional Investigative Psychology Conference, University of Liverpool, Liverpool,
England.
Annon, J. S. (1995). Investigative profiling: A behavioral analysis of the crime scene.
American Journal of Forensic Psychology, 13, 67–75.
Åsgard, U. (1998). Swedish experiences in offender profiling and evaluation of some
aspects of a case of murder and abduction in Germany. In Case Analysis Unit (BKA)
(Eds.), Method of case analysis: An international symposium (pp. 125–130). Weis-
baden, Germany: Bundeskriminalamt Kriminalistisches Institut.
Ault, R. L., & Reese, J. T. (1980). A psychological assessment of crime profiling. FBI Law
Enforcement Bulletin, 49, 22–25.
Badcock, R. (1997). Developmental and clinical issues in relation to offending in the
individual. In J. L. Jackson & D. A. Bekarian (Eds.), Offender profiling: Theory,
research and practice (pp. 9–41). Chichester, England: Wiley.
Bandura, A. (1969). Principles of behavior modification. New York: Holt, Rinehart &
Winston.
Bem, D. J., & Allen, A. (1974). On predicting some of the people some of the time: The
search for cross-situational consistencies in behavior. Psychological Review, 81,
506–520.
Bennell, C., Alison, L. J., Stein, K. L., Alison, E. K., & Canter, D. V. (2001). Sexual
offences against children as the abusive exploitation of conventional adult–child
relationships. Journal of Social and Personal Relationships, 18, 149–165.
Blau, T. H. (1994). Psychological profiling. In T. H. Blau (Ed.), Psychological services for
law enforcement (pp. 261–274). New York: Wiley.
Boon, J. (1997). Contribution of personality theories to psychological profiling. In J. L.
Jackson & D. A. Bekarian (Eds.), Offender profiling: Theory, research and practice
(pp. 43–59). Chichester, England: Wiley.
Bowers, K. (1973). Situationism in psychology: An analysis and a critique. Psychological
Review, 80, 307–336.
Britton, P. (1992). Review of offender profiling. London: Home Office.
Britton, P. (1997). The jigsaw man. London: Bantam Press.
Burgess, A., & Hazelwood, R. (Eds.). (1995). Practical aspects of rape investigation: A
multidisciplinary approach (2nd ed.). New York: CRC Press.
Canter, D. V. (1995). Psychology of offender profiling. In R. Bull & D. Carson (Eds.),
Handbook of psychology in legal contexts (pp. 343–355). Chichester, England: Wiley.
Canter, D. V. (2000). Offender profiling and criminal differentiation. Legal and Crimi-
nological Psychology, 5, 23–46.
Canter, D. V., & Fritzon, K. (1998). Differentiating arsonists: A model of firesetting
actions and characteristics. Legal and Criminological Psychology, 3, 73–96.
132 ALISON, BENNELL, MOKROS, AND ORMEROD
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
Cervone, D., & Shoda, Y. (1999a). Beyond traits in the study of personality coherence.
Current Directions in Psychological Science, 8, 27–32.
Cervone, D., & Shoda, Y. (1999b). The coherence of personality: Social–cognitive bases
of consistency, variability and organization. New York: Guilford Press.
Copson, G. (1995). Coals to Newcastle: Part 1. A study of offender profiling. London:
Home Office, Police Research Group.
Craik, M., & Patrick, A. (1994). Linking serial offences. Policing, 10, 181–187.
Cronbach, L. J., & Gleser, G. C. (1957). Psychological tests and personnel decisions.
Urbana: University of Illinois.
Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S. Ct. 2786 (1993).
Davies, A., Wittebrood, K., & Jackson, J. L. (1998). Predicting the criminal record of a
stranger rapist. London: Home Office, Policing and Reducing Crime Unit.
Delaware v. Pennell, 584 A. 2d 513 (Del. Super. Ct. 1989).
Douglas, J. E. (1981). Evaluation of the (FBI) psychological profiling program. Quantico,
VA: FBI Academy, Institutional Research and Development Unit.
Douglas, J. E., Burgess, A. W., Burgess, A. G., & Ressler, R. K. (1992). Crime
classification manual: A standard system for investigating and classifying violent
crime. New York: Simon & Schuster.
Douglas, J. E., & Munn, C. (1992). Violent crime scene analysis: Modus operandi,
signature, and staging. FBI Law Enforcement Bulletin, 62,1–20.
Douglas, J. E., Ressler, R. K., Burgess, A. W., & Hartman, C. R. (1986). Criminal
profiling from crime scene analysis. Behavioral Sciences and the Law, 4, 401–421.
Dudycha, G. J. (1936). An objective study of punctuality in relation to personality and
achievement. Archives of Psychology, 204, 1–319.
Epstein, S. (1979). The stability of behavior: I. On predicting most of the people most of
the time. Journal of Personality and Social Psychology, 37, 1097–1126.
Farrington, D. P. (1997). Human development and criminal careers. In M. Maguire, R.
Morgan, & R. Reiner (Eds.), The Oxford handbook of criminology (2nd ed., pp.
361–408). Oxford: Oxford University Press.
Golding, S. (1975). Flies in the ointment: Methodological problems in the analysis of the
percentage variance due to persons and situations. Psychological Bulletin, 82, 278–
288.
Goldblatt, P. (1992). Psychological offender profiles: How psychologists can help the
police with their enquiries. Surrey, England: University of Surrey.
Green, E. J., Booth, C. E., & Biderman, M. D. (1976). Cluster analysis of burglary M/Os.
Journal of Police Science and Administration, 4, 382–388.
Groth, A. N. (1979). Men who rape: The psychology of the offender. New York: Plenum.
Grubin, D. (1995). Offender profiling. Journal of Forensic Psychiatry, 6, 259–263.
Grubin, D., Kelly, P., & Brunsdon, C. (2001). Linking serious sexual assaults through
behavior. London: Home Office, Research, Development and Statistics Directorate.
Hampson, S. E. (1995). The construction of personality: An introduction (2nd ed.).
London: Routledge.
Hartshorne, H., & May, M. A. (1928). Studies in the nature of character: Vol. 1: Studies
in deceit. New York: Macmillan.
Hickey, E. W. (1997). Serial murderers and their victims (2nd ed.). New York: Wads-
worth.
Holmes, R. M., & DeBurger, J. (1988). Serial murder. Newbury Park, CA: Sage.
Homant, R. J., & Kennedy, D. B. (1998). Psychological aspects of crime scene profiling:
Validity research. Criminal Justice and Behavior, 25, 319–343.
House, J. C. (1997). Towards a practical application of offender profiling: The RNC’s
criminal suspect prioritization system. In J. L. Jackson & D. A. Bekerian (Eds.),
133THE PERSONALITY PARADOX IN OFFENDER PROFILING
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
Offender profiling: Theory, research and practice (pp. 177–190). Chichester, En-
gland: Wiley.
Jackson, J. L., Van Koppen, P. J., & Herbrink, J. C. M. (1993). Does the service meet the
needs? An evaluation of consumer satisfaction with specific profile analysis and
investigative advice as offered by the Scientific Research Advisory Unit of the
National Criminal Intelligence Division (CRI), The Netherlands. Leiden: NSCR.
Knight, R. A., & Prentky, R. A. (1990). Classifying sexual offenders: The development
and corroboration of taxonomic models. In W. L. Marshall, D. R. Laws, & H. E.
Barbaree (Eds.), The handbook of sexual assault: Issues, theories, and treatment of
the offender (pp. 27–52). New York: Plenum.
Knight, R. A., Warren, J. I., Reboussin, R., & Soley, B. J. (1998). Predicting rapist type
from crime-scene variables. Criminal Justice and Behavior, 25, 46–80.
Kumho Tire Co. v. Carmichael, 119 S. Ct. 1167 (1999).
Langevin, R., Paitich, D., & Russon, A. E. (1985). Are rapists sexually anomalous,
aggressive, or both? In R. Langevin (Ed.), Erotic preference, gender identity, and
aggression in men: New research studies (pp. 17–38). Hillsdale, NJ: Erlbaum.
Lord, C. G. (1982). Predicting behavioral consistency from an individual’s perception of
situational similarities. Journal of Personality and Social Psychology, 42, 1076–
1088.
Louisiana v. Code, 627 So. 2d 1373 (La. 1993).
Magnusson, D., & Endler, N. S. (1977). Interactional psychology: Present status and
future prospects. In D. Magnusson & N. S. Endler (Eds.), Personality at the cross-
roads: Current issues in interactional psychology (pp. 3–35). Hillsdale, NJ: Erlbaum.
Mischel, W. (1968). Personality and assessment. New York: Wiley.
Mischel, W. (1990). Personality dispositions revisited and revised: A view after three
decades. In L. Pervin (Ed.), Handbook of personality: Theory and research (2nd ed.;
pp. 111–134). New York: Guilford Press.
Mischel, W. (1999). Introduction to personality (6th ed.). Fort Worth, TX: Harcourt
Brace.
Mischel, W., & Peake, P. K. (1982). Beyond de´ja` vu in the search for cross-situational
consistency. Psychological Review, 89, 730–755.
Mokros, A., & Alison, L. J. (in press). Is profiling possible? Legal and Criminological
Psychology.
Newcomb, T. M. (1929). Consistency of certain extrovert–introvert behavior patterns in
51 problem boys. New York: Columbia University, Teachers College, Bureau of
Publications.
New Jersey v. Fortin, 162 N.J. 517 (2000).
North Carolina v. Wallace, 351 N.C. 481 (2000).
O’Connor, M., & Krauss, D. (2001). Legal update: New developments in Rule 702.
American Psychology–Law Society News, 21, 1–18.
Olweus, D. (1977). A critical analysis of the modern interactionist position. In D.
Magnusson & N. S. Endler (Eds.), Personality at the crossroads: Current issues in
interactional psychology (pp. 221–233). Hillsdale, NJ: Erlbaum.
Ormerod, D. (1996a). The evidential implications of psychological profiling. Criminal
Law Review, 92, 863–877.
Ormerod, D. (1996b). Psychological profiling. Journal of Forensic Psychiatry, 7, 341–
352.
Ormerod, D. (1999). Criminal profiling: Trial by judge and jury, not criminal psycholo-
gist. In D. V. Canter & L. J. Alison (Eds.), Profiling in policy and practice (pp.
207–261). Aldershot, England: Ashgate.
Peterson, D. R. (1968). The clinical study of social behavior. New York: Appleton-
Century-Crofts.
134 ALISON, BENNELL, MOKROS, AND ORMEROD
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
Pinizzotto, A. J., & Finkel, N. J. (1990). Criminal personality profiling: An outcome and
process study. Law and Human Behavior, 14, 215–233.
Prentky, R. A., & Burgess, A. W. (2000). Forensic management of sexual offenders. New
York: Kluwer Academic/Plenum.
Proulx, J., Aubut, J., Perron, L., & McKibben, A. (1994). Troubles de la personnalite´ et
viol: Implications the´oriques et cliniques [Personality disorders and violence: Theo-
retical and clinical implications]. Criminologie, 27, 33–53.
Proulx, J., St-Yves, M., Guay, J. P., & Ouimet, M. (1999). Les aggresseurs sexuels de
femmes: Sce´narios de´lictuels et troubles de la personnalite´ [Sexual aggressors of
women: Offence scenarios and personality disorders]. In J. Proulx, M. Cusson, & M.
Ouimet (Eds.), Les violences criminelles (pp. 157–185). Sainte-Foy, Quebec, Canada:
Les Presses de l’Universite´ Laval.
Regina v. Stagg. (1994). CCC 14th September.
Ressler, R. K., Burgess, A. W., Douglas, J. E., Hartman, C. R., & D’Agostino, R. B.
(1986). Sexual killers and their victims: Identifying patterns through crime scene
analysis. Journal of Interpersonal Violence, 1, 288–308.
Rossmo, D. K. (2000). Geographic profiling. Boca Raton, FL: CRC Press.
Shoda, Y. (1999). Behavioral expressions of a personality system: Generation and
perception of behavioral signatures. In D. Cervone & Y. Shoda (Eds.), The coherence
of personality: Social–cognitive bases of consistency, variability, and organization
(pp. 155–181). New York: Guilford Press.
Shoda, Y., Mischel, W., & Wright, J. C. (1994). Intraindividual stability in the organi-
zation and patterning of behavior: Incorporating psychological situations into the
idiographic analysis of personality. Journal of Personality and Social Psychology, 67,
674–687.
Smith, C. (1993). Psychological offender profiling. The Criminologist, 244, 224–250.
Turvey, B. (Ed.). (1999). Criminal profiling: An introduction to behavioral evidence
analysis. New York: Academic Press.
Vernon, P. E. (1964). Personality assessment: A critical survey. New York: Wiley.
Wilson, M., Jack, K., & Butterworth, D. (1996). The psychology of rape investigations:
A study in police decision making. Liverpool, United Kingdom: University of Liv-
erpool, Department of Psychology.
Witkin, G. (1996). How the FBI paints portraits of the nation’s most wanted. U. S. News
& World Report, 32.
Wright, J., & Mischel, W. (1987a). A conditional analysis of dispositional constructs: The
local predictability of social behavior. Journal of Personality and Social Psychology,
53, 1159–1171.
Wright, J., & Mischel, W. (1987b). Conditional hedges and the intuitive psychology of
traits. Journal of Personality and Social Psychology, 55, 454–469.
Zelli, A., & Dodge, K. A. (1999). Personality development from the bottom up. In D.
Cervone & Y. Shoda (Eds.), The coherence of personality: Social-cognitive bases of
consistency, variability, and organization (pp. 94–126). New York: Guilford Press.
135THE PERSONALITY PARADOX IN OFFENDER PROFILING
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.
A preview of this full-text is provided by American Psychological Association.
Content available from Psychology Public Policy and Law
This content is subject to copyright. Terms and conditions apply.