Criminal Justice and Behavior Sept 1998 v25 n3 p319(25) Page 1

Psychological aspects of crime scene profiling.

by Robert J. Homant and Daniel B. Kennedy

Crime scene profiling is distinguished from psychological profiling and offender profiling, and the
profiling process is described. A review of the literature on offender types indicates that it may be
possible to generalize from various behavioral aspects of a crime scene to some characteristics
of the perpetrator. There is some evidence that it may be possible to type offenders and crime
scenes as organized or disorganized, and that this categorization may be related to certain
aspects of offender personality. There is also some evidence that those who are FBI trained in
profiling may perform better at the task. At this time, however, the evidence for the validity of
profiles is weak, and it is recommended that profiling not be relied on to the exclusion of other
alternatives. There is a need for more specific validity research, especially when profiling is
generalized beyond its original purpose of providing leads and focusing investigations.

© COPYRIGHT 1998 Sage Publications, Inc. and FBI-trained profilers; and to comment on the
application of profiling to various problem areas. We will
Through movies such as Silence of the Lambs and begin with a general description of the profiling process
television series such as Millennium and Profiler, crime and distinguish it from certain related attempts at
scene profiling has come to the attention of the general prediction.
public. According to Douglas, Ressler, Burgess, and
Hartman (1986), profiling may be defined as "a technique DISTINGUISHING CRIME SCENE PROFILING
for identifying the major personality and behavioral
characteristics of an individual based upon an analysis of THE PROCESS OF CRIME SCENE PROFILING
the crimes he or she has committed" (p. 405). Numerous
books and articles have described the origins of the FBI’s Profiling is referred to by various terms in the literature:
efforts to develop and implement a formal process for psychological profiling, criminal personality profiling or
crime scene profiling, which began systematically in 1978 assessment, criminal behavior profiling, offender profiling,
(Annon, 1995; Dietz, 1985; Douglas & Olshaker, 1995, criminal profiling, and investigative profiling (Annon, 1995).
1997; Geberth, 1990; Hickey, 1997; McCann, 1992; Currently, the official FBI term is criminal investigative
Pinizzotto, 1984; Ressler & Schachtman, 1992). analysis. The authors believe that, for better or worse, the
term profiling has become too well established to change.
These accounts, together with the following literature that In this article, therefore, we will use the term crime scene
we will cite, represent a mixture of journalistic, profiling to help focus on exactly what is being attempted
autobiographical, and more empirically based information and to help distinguish the area from related efforts that
on profiling. Furthermore, much of the empirical literature are also referred to as profiling.
directly relating to profiling has been published without
being subjected to peer review. For the most part, these Crime scene profiling is specifically based on the
sources give a highly positive impression of the techniques developed by the FBI’s Behavioral Science
effectiveness of crime scene profiling. Unit, which has evolved into the Profiling and Behavioral
Assessment Unit. The process was developed particularly
Although a few critics have raised cautioning voices to deal with cases of serial homicide and/or serial rape.
(Hickey, 1997; Keppel, 1995), not only does the utility of Because of the typically chance connection between
such profiling seem to be generally accepted, but it is perpetrator and victim in these kinds of crime, and
being extended into areas far beyond its original design. because a perpetrator frequently commits crimes across
The FBI currently has 12 full-time profilers who collectively various jurisdictions, such cases are especially
are involved in about 1,000 cases per year (Witkin, 1996). troublesome to local law enforcement.
In addition, many state and local police, typically
FBI-trained, apply profiling to some unknown number of The how-to of profiling is set out especially well by Dietz
cases. With some slight differences in approach, profiling (1985), Douglas et al. (1986), and Geberth (1990). The
has also gained a hold in Canada (Rossmo, 1995b), Great profiler typically begins with complete photographs and
Britain (Davies & Dale, 1995), and the Netherlands descriptions of the crime scene. This includes information
(Jackson, van den Eshof, & de Kleuver, 1994). about the general character of the location, including traffic
patterns and ease of access for various types of
The purpose of this article is to review the theoretical individuals. If the crime is a homicide, an autopsy of the
underpinning of profiling; to examine the existing studies of victim is required to assist in reconstructing the sequence
its reliability and validity, especially as practiced by the FBI of the crime; if a rape, it is hoped that the victim can
- Reprinted with permission. Additional copying is prohibited. - GALE GROUP
Information Integrity
Criminal Justice and Behavior Sept 1998 v25 n3 p319(25) Page 2

Psychological aspects of crime scene profiling.

reconstruct all interactions with the rapist, especially GOALS OF PROFILING
including all the verbal and nonverbal techniques used by
the offender to gain control. The profiler also requires a Profiling was originally intended to help law enforcement
complete victim profile, including general lifestyle and a discover who the criminal is--either by narrowing an
detailed account of behaviors prior to the victimization. All overwhelming list of suspects to a small subgroup or by
physical evidence is expected to be at the disposal of the providing new avenues of inquiry. Other uses were quickly
profiler. This includes such classic clues as footprints, found, however. Thus, case histories (e.g., Douglas &
blood spatters, and tools or paraphernalia used. Such Olshaker, 1995) document the use of profiles to give police
evidence is typically given even more weight than advice on how best to interrogate a suspect and to tell
psychological speculation. In addition, any knowledge of prosecutors the approach to cross examination most likely
the perpetrator’s pre- and post-offense behavior is sought, to break down a defendant. Profiling has been used to
as well as all information from other crimes that may be help set traps to flush out the offender, for example, by
linked to the same perpetrator by physical or behavioral planting information in the media. It has been used to
evidence. predict dangerousness as a factor in sentencing or at a
parole hearing or to determine whether a threatening note
The profiler then attempts to give as complete a should be taken seriously.
description of the perpetrator as possible. This might
include gender, age, race or ethnicity, level of intelligence Proponents of profiling are normally cautious in their
or schooling, military service status, job status, living claims, stressing, for example, that the process is more of
circumstances, nature of interpersonal relationships, and an art than a science and suggesting that profiling should
even the make and color of the perpetrator’s car. The be limited to cases that show severe psychopathology
numerous descriptive statements are seen as hypotheses, (Geberth, 1990, p. 492; McCann, 1992, p. 476; Pinizzotto,
and it is not expected that all will prove correct. 1984, p. 33) and for which there is a sufficient database of
known offenders from previous similar cases. This caution
Many of the hypotheses are simply generalizations based would seem to limit profiling mostly to serial rapes and
on the profiler’s experience and training. Other hypotheses serial murders; at other times, however, profiling seems to
are arrived at much more intuitively, by mentally be extended to cases of single rape or murder, as well as
reenacting the crime and imagining what sort of person to arson, bombing, and threats of various types. Profilers
would be involved. Profilers as a group are not clear as to have also applied their efforts to distinguishing accidental,
whether they are trying to construct the personality of the autoerotic asphyxiation from suicide or homicide to
offender or merely to generate a loosely related series of hostage negotiations, stalking, and even bank robbery--not
descriptive statements. There does not seem to be any all of which necessarily involve significant
particular personality theory, psychodynamic or otherwise, psychopathology.
that guides the FBI-trained profilers. When profilers refer to
the personality of the offender, then, we take them to be THE VALIDITY ISSUE
speaking loosely, in more or less layperson’s terms, both
about the interpersonal style and the underlying motives of The issue of the validity of crime scene profiling takes
the individual. different forms depending on the context. From a law
enforcement point of view, there is no need to wait for
Dietz (1985) treats profiling as a systematic, five-step assurances that profiling in general is a valid process, as
process, in which various levels of hypotheses are long as there are not any more promising alternatives and
continuously checked against the data to arrive at a more as long as the process is used cautiously. For example, a
or less final set of attributional hypotheses or reasonably promising lead should not be abandoned just because it
specific, descriptive statements about the offender. Other does not fit a profile, and no lead should be focused on too
writers (Douglas & Olshaker, 1995; Ressler & narrowly. From a forensic point of view, when profiling
Schachtman, 1992) describe the process as being more becomes courtroom evidence that is used to link cases, for
intuitive. The final profile that emerges may range from a example, it is important that such evidence that is be seen
paragraph to several pages, depending partly on the as probabilistic, and even though conclusions are based
amount of input data and partly on how thoroughly the on an "art," it is still important that the profiler be able to
profiler explains his or her deductions through references articulate the basis for various inferences. Finally, from a
to case materials. An early example of a brief, largely social scientific point of view, there is a need to validate
unexplained profile is provided by Hazelwood, Dietz, and profiling not just for each of the individual purposes for
Burgess (1982); for an example of a more extensive which it is used (e.g., identifying suspects, predicting
profile, see Douglas et al. (1986). dangerousness) but also for each type of crime (e.g.,
homicide, rape, arson). As a research strategy, of course,

- Reprinted with permission. Additional copying is prohibited. - GALE GROUP

Information Integrity
Criminal Justice and Behavior Sept 1998 v25 n3 p319(25) Page 3

Psychological aspects of crime scene profiling.

one might begin by trying to establish the validity of the offender was more or less likely to be guilty of such a
process for its principal use and then focus later on its crime.
extended uses.
Murphy and Peters (1992) have reviewed the literature on
As far as we can determine, no one has attempted to the profiling of child sexual abusers. In general, the results
assess the validity of crime scene profiling in real-life of studies of this type of profiling were unimpressive.
situations. Such a study would present some unique Standardized personality testing could not distinguish child
problems. The main problem is the lack of an objective sexual abusers from other deviant groups (such as rapists,
criterion against which to test a sample of actual profiles. murderers, or arsonists). Penile plethysmography, a
Holmes (1989) reported that less than half of profiled behavioral method that directly measures deviant arousal,
cases had been solved. Even when the identity of the showed more promising results. The positive results,
offender is unambiguously determined, there is still a large however, were based on subjects voluntarily in treatment;
subjective element in deciding how well the person fits the there was fairly good evidence that involuntary participants
profile. Ideally, the profile should also be compared to any could suppress sexual arousal, thus appearing normal.
cleared suspects in the case to guard against taking
advantage of general statements that might fit most Crime scene profiling and psychological profiling have in
potential offenders. It may also be that if a profile is common the attempt to understand and make predictable
reasonably accurate, it is more likely to lead to an arrest, the behavior of psychologically deviant individuals. In
creating a spuriously positive impression of profiling contrast to psychological profiling, however, crime scene
accuracy for those profiles that can be evaluated. profiling starts with known behavior (the crime, as
witnessed or reconstructed from the scene) and infers
The research strategy for examining profiling has been to characteristics of the offender; in psychological profiling,
use a more piecemeal approach, for example, looking at the profiler starts with a known individual (who can be
the relationships between types of known offenders and tested and interviewed in depth) and tries to project to
their crime scenes. A major problem here is that the pool behavior.
of suitable and cooperative known serial killers, rapists,
and arsonists is relatively small and possibly not very Offender profiling. What we term offender profiling is
representative of the population of such offenders strictly an empirical procedure, in which no assumptions
(Godwin, in press). Many of the studies that we are about about the motivation or personality of the offender are
to review appear to rely on the same pool of subjects, necessary. Simply by gathering a large amount of data,
although not in any clearly defined way. Because the either systematically or by more loosely collected
intensive study of a relatively small sample is likely to take anecdotal information, law enforcement agents construct a
advantage of chance relationships among the numerous description of the type of person most commonly involved
variables, cross validation of findings is clearly called for, in a certain type of offense. For example, someone driving
although, practically speaking, seldom feasible. at a certain speed, at a certain time of day, in a certain
type of car, and of a certain general appearance may fit
OTHER TYPES OF PROFILING the profile of a drug courier and be stopped for a search.
This type of profiling has been credited with reducing some
To evaluate crime scene profiling, it is important to types of criminal behavior, but it raises legal and ethical
distinguish it from two related processes: psychological problems when ethnicity, gender, and age are part of the
profiling and offender profiling. Although the literature on profile (DeGeneste & Sullivan, 1994; Easteal & Wilson,
profiling uses all three terms more or less interchangeably, 1991).
we believe that more precise terminology is critical if we
are to fully appreciate the validity issues involved. Two Crime scene profiling shares with offender profiling the use
other procedures, geographic profiling and equivocal death of empirical data from previous similar offenses. But
analysis, are blends of the three basic types and also need although offender profiling stops with generalizations about
to be distinguished. a class of offenders, crime scene profiling is only useful
insofar as it can distinguish a particular offender from a
Psychological profiling. By psychological profiling, we general type. Also, rather than being limited to a
mean the interviewing and testing of an individual to participant’s reasonably observable features, crime scene
determine whether his personality matches the established profiling attempts to extrapolate to the entire lifestyle of the
personality characteristics of a certain class of offenders. offender.
This procedure has been most developed in the area of
child sexual abusers, where such profiles are sometimes Equivocal death analysis. Equivocal death analysis drew
called on to lend supporting evidence that a particular professional attention when the FBI was asked by the

- Reprinted with permission. Additional copying is prohibited. - GALE GROUP

Information Integrity
Criminal Justice and Behavior Sept 1998 v25 n3 p319(25) Page 4

Psychological aspects of crime scene profiling.

Navy to investigate the cause of an explosion aboard the geographic profile can then be used to further refine the
USS Iowa that killed 47 sailors in 1989. Briefly put, the FBI crime scene profile. Rossmo sees geographic and crime
concluded that the explosion was the result of a scene profiling as enhancing each other in more than a
murder-suicide, committed by a spurned homosexual merely additive way. Regarding validity, Rossmo (1995a)
gunnery officer. Discontent with this conclusion led to a found that with serial killers, the use of geographic profiling
congressional investigation, which commissioned a panel was able to narrow the search for the residence, on
of 14 prominent psychologists and psychiatrists to review average, to about 6% of the potential area of
the FBI procedures (Jeffers, 1991; Poythress, Otto, interest--depending primarily on the number of crime sites
Darkes, & Starr, 1993). A clear majority of the panel were available to the profiler.
critical of the FBI procedures and conclusions, mainly on
the grounds that there was no scientific basis for them. To summarize, the five types of profiling that we have
differentiated all have in common the goal of making some
The FBI agents involved in the process (and testifying sort of inference about an individual’s personality or
before Congress) viewed equivocal death analysis as an behavior. All are, to some extent, dependent on a model
extension of crime scene profiling. Poythress et al. (1993) that portrays human behavior as having a good deal of
specifically noted the lack of concern on the part of one of consistency. The inferential burden seems to us to be by
these agents for the issue of validity and also noted the far the greatest in crime scene profiling: Not only must the
total absence of any meaningful studies of the validity of personality of an unknown individual be assembled from
equivocal death analysis. However, Poythress et al. view behavioral clues, but that hypothesized personality must
equivocal death analysis as much more closely related to then be used to generate further attributions about the
psychological autopsy (which attempts to clarify the individual. Because trait theory and the concept of offender
motivation of a known individual who engaged in a types is central to this process, we will briefly review some
particular act--suicide or accidental death) than to crime of the key studies in these areas before turning our
scene profiling. We agree, and we consider equivocal attention to studies directly involved with the validity of
death analysis as more akin to psychological than to crime crime scene profiling.
scene profiling--except that the known individual’s
personality must be reconstructed without his or her THE THEORETICAL BASIS FOR PROFILING
cooperation. The lack of concern for validity and the
unwarranted degree of certainty displayed by the FBI TRAIT THEORY AND PROFILING
agents in the USS Iowa incident certainly raise a caution
flag over other work done by FBI profilers. Crime scene profiling rests on the assumption that at least
certain offenders have consistent behavioral traits. This
Geographic profiling. Developed as part of his doctoral consistency is thought to persist from crime to crime and
dissertation at Simon Fraser University by Detective also to affect various non-criminal aspects of their
Inspector Kim Rossmo of the Vancouver Police personality and lifestyle, thus making them, to some
Department (1995a, 1995b), geographic profiling blends extent, identifiable.
the insights of modem geography with the contributions of
environmental criminology (Brantingham & Brantingham, In the field of psychology, trait theory has had a somewhat
1981; see also Kennedy, 1990). It is clearly allied to crime checkered past, with many researchers finding little
scene profiling, in that it attempts to generalize from linked consistency of particular traits across time and situations,
crime scene locations to the probable residence or base of let alone any pattern of personality across different traits
operations of an unknown offender. Although primarily (Kendrick & Funder, 1988). Trait theorists have responded
empirical (taking into account such variables as bus routes by shifting to measures that combine data from a large
and travel time), geographic profiling does employ the number of observations across situations. Hartup and van
concept of a mental map and tries to reconstruct a Lieshout (1995) also suggest that antisocial behavior might
psychological representation of the crime-relevant areas in be an especially stable behavior trait for aggressive
which the offender feels comfortable. A related approach persons--whether because their personalities do not
used by Godwin and Canter (1997) also attempts to change much, the genetic contribution remains constant,
construct a social psychological portrait of the offender. or they tend to remain in aggression-fostering
environments.
Rossmo (personal communication, June 19, 1997)
considers geographic profiling as a distinct step in a Caspi and Moffitt (1995) cite a number of studies that,
criminal investigation, one that would normally follow after taken together, support the theory that a relatively small
the construction of a crime scene profile, which might be group of adolescents engage in life-course persistent
used in helping to formulate the mental map. The antisocial behavior. This form of antisocial behavior has

- Reprinted with permission. Additional copying is prohibited. - GALE GROUP

Information Integrity
Criminal Justice and Behavior Sept 1998 v25 n3 p319(25) Page 5

Psychological aspects of crime scene profiling.

been found to correlate with various genetic, sample of 92 rape victims. The rapes were found to be
environmental, and personality variables that help account classifiable as either power-assertive (44%),
for its persistence. Presumably, individuals who display power-reassurance (21%), anger-retaliation (30%), or
life-course persistent antisocial behavior make up the bulk anger-excitement (5%). Although many descriptive data
of the adult incarcerated offender population. A much were given for the four types of rapists, no data were given
smaller subset of this group, in turn, would be as to how reliably a given rape or rapist could be classified
representative of the type of offender for whom profiling is as one of the four types (or whether a person might
thought to be appropriate. Various research studies have change from one type of rapist to another).
attempted to identify these offender subtypes.
The theorizing of Groth et al. (1977), as modified
OFFENDER TYPES somewhat by Hazelwood and Burgess (1987), became
central to the FBI’s efforts to profile serial rapists. In-depth
Crime scene profiling is based on the assumption that a interviews were conducted with 41 incarcerated serial
portrait of an individual offender can be drawn that will rapists (Hazelwood & Warren, 1989a, 1989b, 1990). In
distinguish that person from what is known about a class one study of these rapists, Hazelwood, Reboussin, and
of offenders in general. A particular serial rapist, for Warren (1989) identified a subgroup of serial rapists whom
example, needs to be distinguishable from the generalized they termed "increasers." Unlike the majority (n = 31) of
picture that could be obtained by simply using modal the serial rapists, the increasers (n = 10) escalated their
values of known serial rapists. Although not absolutely use of force with succeeding rapes. They also differed in
necessary in a logical sense, the existence of reliably that they raped more frequently and engaged in more
identifiable subtypes within any class of offenders at least sadistic acts. However, no developmental differences were
supports the claim that such narrowing of the field is found between the two subgroups, so it is not clear how a
possible. Furthermore, much of the research work on profiler might use the information that a particular series of
profiling has involved the attempt to identify such subtypes linked rapes was probably committed by an increaser.
of offenders. Furthermore, in this particular study, all coding
disagreements between interviewers were resolved
One of the most frequent areas for developing offender "through an assessment of all available information" (p.
typologies concerns sex offenders, particularly rapists. 69). Thus, we do not know how reliable any attempt would
This research has been reviewed by Prentky and Knight be to classify someone as an increaser, even based on
(1991; see also Knight & Prentky, 1990) in the context of direct interviews with the individual.
developing their own nine-category typology of rapists for
the Massachusetts Treatment Center. Prentky and Knight A second study using these same 41 serial rapists dealt
identified several variables that reliably distinguished with the accuracy of using various scales to classify serial
certain aspects of rapists’ behavior. Perhaps most relevant rapists and rape crime scenes (Warren, Reboussin,
to the area of crime scene profiling is the finding that Hazelwood, & Wright, 1991). Although this study produced
lifestyle impulsivity and the presence of sexual fantasies some seemingly impressive agreements in classifying
are important predictors of recidivism (and therefore are rapists and rape incidents in terms of Groth et al.’s (1997)
relevant to serial rape). Other variables, such as irrational four-category system, as well as distinguishing increasers
attitudes, dominance, sadism, history of child sexual and nonincreasers, there was a serious methodological
abuse, social competence, aggression, and alcohol use problem. The same information that was used by the
have also been found to be of varying usefulness in researchers to classify a rape as, say, power-reassurance,
distinguishing among rapists. Whether these separate was also used to score the various scales (e.g., did the
variables can be combined into an empirically valid, rapist reassure the victim of no intent to harm?). About all
practically useful typology, such as that proposed by that the results mean, then, is that a set of scales can be
Knight and Prentky, remains to be demonstrated. applied consistently to both crime scenes and rapists. If a
technique can be developed to independently rate both
Groth, Burgess, and Holmstrom (1977) proposed a simple crime scenes and the rapists involved (without relying on
typology of rapists that was based on two main types, information common to both), and if a good fit is found
each divided into two subtypes. According to Groth et al., between the two, then a lot will have been done to
power, anger, and sexual motives are present in all rapes, establish some theoretical underpinning for profiling.
with either power or anger predominating. Rape is seen as
a pseudosexual act, meaning that the sexuality involved is Other studies of offender types were more exploratory,
merely instrumental to gratifying the dominant motives of looking for patterns of behavior among distinctive samples
power or anger. The authors examined 225 rapes, based of offenders. Dietz, Hazelwood, and Warren (1990)
on reports of 133 convicted rapists and an unrelated examined the case histories of 30 "sexually sadistic

- Reprinted with permission. Additional copying is prohibited. - GALE GROUP

Information Integrity
Criminal Justice and Behavior Sept 1998 v25 n3 p319(25) Page 6

Psychological aspects of crime scene profiling.

criminals" that had been submitted to the National Center The classification is complicated somewhat by the use of a
for the Analysis of Violent Crime. A representative finding mixed type. Exactly how much overlap of elements is
was that 53% of the offenders kept detailed records of needed before a crime scene should be called mixed is not
their crimes, and 93% showed careful planning of those clear; probably most crime scenes could be classified as
crimes. This was seen as indicating the sadist’s mixed. Also, because the FBI profiler is more likely to be
tremendous need for control and dominance. (See also called in on the tougher cases, profiling is more likely to be
Warren, Hazelwood, & Dietz, 1996, for an overlapping attempted on relatively organized crime scenes, so that
study.) any statement about the distribution of the crime scene
classification variable would be misleading. Some points
Ressler, Burgess, Douglas, Hartman, and McCormack that are not clear in the literature include whether the
(1986) studied the case histories of 28 sexual murderers. mixed category should be applied only to the crime scene
The murderers were categorized into two groups based on or also to the offender, and whether and how offenders
presence or absence of self-reported childhood sexual might change or evolve during their career from organized
abuse. Numerous differences between the two groups to disorganized and vice versa.
were found, although most were only marginally significant
because of the small sample size. As expected, all of the The difference between organized and disorganized
dependent variable differences (e.g., in nightmares, cruelty offenders is further complicated by three distinctions that
to animals, assaultiveness) were associated with higher profilers make about the crime scene: the MO (modus
levels of pathology in the abused subgroup. More operandi), the signature, and (possible) staging (Douglas
important for the area of profiling, at least one crime scene & Munn, 1992; Geberth, 1995). The MO refers to the
difference did emerge: 78% of the abused group mutilated method used by the offender to accomplish the crime; it is
their victims, compared to 42% of the nonabused group (p essentially learned and changeable behavior. The
= .07). signature refers to behaviors that are related to the
offender’s personality--specifically, the unique
To summarize this section on offender types, several fantasies--and that go beyond what is needed to
studies of extreme groups of rapists and murderers have accomplish the crime. Although it is described as "never
found that some personality and behavioral distinctions changing" (Douglas & Munn, 1992, p. 3), it is also possible
can be made within these groups and that there is some for the signature to "evolve" from crime to crime. Profilers
reliability in making these distinctions. There is also some typically argue that the signature is more important than
evidence linking these distinctions to differences in how the MO, both for linking crimes and for deriving an offender
offenders carry out their crimes. In many respects, profile. Finally, staging refers to deliberate efforts by the
however, these studies are closer to psychological rather offender (or others) to alter the crime scene so as to
than crime scene profiling, in that they typically begin with mislead investigators. The presence of staging may be
a sample of known offenders. In the next section, we will detected by "inconsistencies in the forensic findings and in
look at those studies that bear even more directly on the the overall `big picture’ of the crime scene" (Douglas &
reliability and validity of crime scene profiling. Munn, 1992, p. 7). Douglas and Olshaker (1995) see
staging as an aspect of MO in that it is part of the
DIRECT STUDIES OF CRIME SCENE PROFILING criminal’s basic plan for getting away with the crime.
Sometimes, however, the alteration in the crime scene can
CRIME SCENE CLASSIFICATION VARIABLES represent posing--for example, using the victim as a sort of
prop to communicate a symbolic message-and thus can
Before looking at specific studies of the profiling process, it be a part of the signature (Douglas & Olshaker, 1995).
will be necessary to clarify the distinction between Exactly how the profiler can determine whether some
organized and disorganized crime scenes and offenders. aspect of a particular crime scene represents posing as
This distinction is fairly well described in the profiling opposed to staging is not clear.
literature. Ressler and Burgess (1985) list some 25
variables that distinguish the personality, socioeconomic It would seem from these definitions that the categories
background, and crime scene behavior involved in the two organized/ disorganized apply partly to the signature
types of offender. Compared to disorganized offenders, aspect and partly to the MO. On one hand, the degree of
organized offenders are described as more intelligent, organization is seen as reflective of the criminal’s
more socially competent, more likely to be responding to personality (Ressler, Burgess, Douglas, Hartman, &
some precipitating situational stressor, and more likely to D’Agostino, 1986) and therefore related to signature. On
show care, planning, and control in the criminal act. the other hand, the fact that some offenders may learn
Because of these factors, organized offenders are viewed from their early crimes and become more careful offenders
as more difficult to catch. seems more characteristic of the MO aspect of the crime.

- Reprinted with permission. Additional copying is prohibited. - GALE GROUP

Information Integrity
Criminal Justice and Behavior Sept 1998 v25 n3 p319(25) Page 7

Psychological aspects of crime scene profiling.

In any event, the interrelationships of these concepts need based strictly on observable crime scene variables, the
more clarification. overall number of significant findings does represent an
important beginning for showing a link between crime
ORGANIZED VERSUS DISORGANIZED KILLERS scene and personality/background variables.

Ressler, Burgess, Douglas, Hartman, and D’ Agostino RELIABILITY OF CRIME SCENE CLASSIFICATION
(1986) compared 24 organized with 12 disorganized
sexual killers on a wide variety of variables. The authors The extent to which crime scenes could be reliably
assume that the 36 sexual killers have been accurately classified as organized/disorganized was the subject of a
classified as organized or disorganized. Although a citation study by Ressler and Burgess (1985). Sixty-four cases
is given to research that describes the basis for this were taken from FBI files. An agent familiar with the case
classification, that source (Ressler & Burgess, 1985) only presented all relevant crime scene information to five other
states that the interviewers made such a distinction. The agents of varying experience. The measure of reliability
troubling aspect of this is that it seems highly likely that the was the percentage of agreement by the five listeners with
classification was made on the basis of information both the classification of the presenting agent. The bulk of the
about the offender and about the crime scenes involved. classifications was into the categories organized or
Thus, a subsequent discovery that the classification leads disorganized, although the categories mixed and unknown
to significant differences on a number of crime scene were also used (making agreement by chance somewhat
variables would be essentially circular. less likely). The obtained agreement was 74.1%; that is,
about three fourths of the time, the agents listening to the
In addition to data on the 36 offenders themselves, data presenter’s details agreed with the presenter’s
were available for 118 of their crime scenes and victims. classification of the crime scene. Between any two agents,
These data were scored according to 357 variables, across all 64 crime scenes, the agreements ranged from
divided into four categories: background of offender, 45% to 89%, with experienced agents achieving at least
nature of the offense (primarily offender behavior), nature 62% agreement.
of the victim, and crime scene (primarily observable
physical evidence). Numerous statistically significant Significance levels for the percentages are not given, and
differences were found in these variables when comparing kappa coefficient would have been the appropriate statistic
organized and disorganized offenders. Besides finding for measuring reliability here. Although Ressler and
consistent (though overlapping) differences in the crime Burgess (1985) report the results as a positive finding, the
scenes, the researchers found numerous variables that results also show a good deal of disagreement, especially
could be used in a profile to "identify the subject as an for colleagues listening to the same presenter’s summary
individual" (Ressler, Burgess, Douglas, Hartman & of the various crime scenes. For example, even though
D’Agostino, 1986, p. 297). 48.4% of the time the right answer was "organized," one
agent only got 51.7% of his answers correct, and even two
Twenty-four of these variables are identified as statistically trained agents agreed with each other only 62% of the
significant. However, as many as 15 of these significant time--not much better than guessing.
findings can be attributed to chance because it appears
that as many as 300 of the original 357 variables may THE ROLE OF FANTASY
have qualified as potential profile variables. Furthermore,
some of the significant findings do not seem particularly Prentky et al. (1989) based a study of fantasy in serial
compelling. For example, the organized offender is sexual homicide directly on the research on crime scene
described as "likely to change jobs or leave town" classification and provide some evidence relevant to the
(Ressler, Burgess, Douglas, Hartman & D’Agostino, 1986, validity of that process. Prentky et al. compared 25 serial
p. 300). An inspection of the data, however, shows that sexual murderers with 17 single sexual murderers (32 of
although no disorganized offender did these things (out of the cases taken from the same 36 FBI cases studied by
a possible 21 homicides), organized offenders left town Ressler, Burgess, Douglas, Hartman & D’Agostino, 1986).
only 11 out of 97 possible times and changed jobs 8 times They hypothesized that the serial murderers would
(and probably these 8 are included in the 11 who left evidence a higher degree of fantasy in their motivation and
town). A profiler who deduced that an organized offender that this fantasy, via rehearsal, would result in more
was likely to change jobs or leave town would be wrong organized crime scenes. Crime scenes were categorized
89% of the time--even presuming the offender was independently by FBI profilers. As hypothesized, the first
accurately classified as organized based on the crime murder committed by the serial killers was more likely to
scene. Nevertheless, to the extent that the offenders could be organized: 68% versus only 24% for the single
have been classified as organized versus disorganized murderers. Also, as hypothesized, the serial murderers

- Reprinted with permission. Additional copying is prohibited. - GALE GROUP

Information Integrity
Criminal Justice and Behavior Sept 1998 v25 n3 p319(25) Page 8

Psychological aspects of crime scene profiling.

were much more likely to have been acting out conscious case materials from two actual closed cases: a homicide
fantasies (86% vs. 23%) in their homicides. The and a rape. Participants were instructed to write detailed
relationship between fantasy and crime scene organization profiles about the probable offender based on the case
is not reported directly, but given the respective materials and were asked a series of objective questions
percentages, there would have to be a strong relationship. about the offender (with the correct answers based on the
One null result was that the serial killers were no more actual convicted offenders). Results are complicated both
likely to have exhibited planning prior to the crime (42% vs. by the number of variables on which the groups were
41%), although the authors report some problems with the compared and by the fact that the FBI experts did not
measurement of the planning variable. participate in all phases of the study.

The study suffers from a serious flaw in its sampling, in Six findings seem especially relevant here. (a) The
that the serial killers all came from FBI files, whereas most FBI-trained detectives wrote much longer, more detailed
of the single killers were taken from Massachusetts profiles (the FBI experts were not included in this
Treatment Center records. If the thoroughness or focus of comparison). (b) Five police detectives (not part of the five
the records varied by source, which is not unlikely, then participating groups) rated the FBI-expert or FBI-trained
the difference in the presence of fantasy as well as the profiles as the most helpful. (c) FBI-trained detectives
categorization of the offenders as organized or scored more objectively correct responses on the sex
disorganized may be simply an artifact of the data source. offender case than other groups (FBI experts not given).
It should also be pointed out that it is the very However, the FBI-trained group had the poorest score on
disorganization of the crime scene that may have resulted the homicide case. Group differences were generally
in the single killers being caught before they could small--one or two points on a 15-point scale. (d) Subjects
progress to become serial killers. At most, then, we can were asked to identify the correct offender from a lineup
say that the findings of this study are supportive of some consisting of five written descriptions of possible suspects.
aspects of profiling but are in need of more In the sex offense case, all six FBI experts and five of six
methodologically rigorous replication. FBI-trained detectives were correct, trailed by experienced
detectives (four of six) and psychologists (three of six).
THE PROCESS OF PROFILING Only one of six students was correct. Results are not given
for the homicide case, except to report that the two profiler
Although many methodological weaknesses need to be groups were no longer superior, and all groups did poorer
addressed, the studies reviewed to this point offer some than with the rape case (which must have been difficult for
support for the claim that experienced profilers can classify the student group). (e) The FBI-trained detectives recalled
crime scenes with some reliability and that this more details of the homicide case but were edged out by
classification can be related to identifiable features of the the experienced detectives on the sex offense case (FBI
offender. What remains to be determined is whether experts not given). (f) FBI-trained detectives cited more
profilers do anything more than would experienced details as important for both cases. Groups did not differ in
investigators applying old-fashioned deduction. In the how they processed the case details (i.e., in whether they
Prentky et al. (1989) study, for example, 58% of the serial used specific vs. global details to make attributions).
killers had IQs above 110, compared to 29% of the single
killers. This intelligence variable might account for both the The authors felt that the generally higher accuracy of the
organization in the crime scene and the greater use of profiles in the rape case may have occurred because that
conscious fantasy rehearsal. Perhaps any experienced case had more details available (thanks to the victim’s
investigator would be able to distinguish a relatively report). Furthermore, too many details had to be
planned, organized crime from an impulsive, disorganized eliminated from the cases because of privacy
one and make a generalization about the offender’s issues--which hurt the homicide case more because of the
probable intelligence (and, therefore, school achievement, generally lower level of information. In any event, the
job level, and social adaptation). Pinizzotto and Finkel (1990) study gives only limited
support to the claim that profiling adds a significant degree
To test the effectiveness of trained crime scene profilers of expertise to what is normally done by experienced
as opposed to other criminal investigators, Pinizzotto and detectives. However, insofar as profiling originated as an
Finkel (1990) compared five groups of participants on two attempt to deal with serial offenses, where much more
profiling tasks. The five groups consisted of four FBI information would be available across crime scenes, this
profiling experts, six police detectives specially trained in study may not be a fair test of profiling.
profiling by the FBI, six experienced but not trained police
detectives, six clinical psychologists, and six inexperienced SATISFACTION WITH PROFILING
undergraduate students. Each group received extensive

- Reprinted with permission. Additional copying is prohibited. - GALE GROUP

Information Integrity
Criminal Justice and Behavior Sept 1998 v25 n3 p319(25) Page 9

Psychological aspects of crime scene profiling.

Two other studies of profiling may be termed user tends to encourage pushing the envelope to new
satisfaction studies. The first of these was conducted by applications. Keppel (1995) has observed that many FBI
the FBI itself in 1981, when profiling had barely begun as profiles are generalizations about what is known of serial
an organized process (Holmes, 1989; Pinizzotto, 1984). killers and are not helpful for narrowing suspect lists.
Beginning with FBI files on 192 profiled cases, it was
determined that 88 of them had been solved. Of these 88, IMPLICATIONS FOR THE USE OF PROFILING
in 15 cases (17%), a profile helped in the identification of a
suspect. In a number of other cases, the responding Our take on the evidence at this point is that it is important
agencies reported that profiling helped to focus the to expect that a significant number of mistakes will occur
investigation or to locate or prosecute a suspect. In only with profiling. Where these mistakes can be guarded
17% of the cases were the profiles deemed to be of no against, there is no reason not to use it. For example, in
assistance. Thus, the data can be read in two radically the area of criminal investigation, certainly no significant
different ways. In 62% of the 192 cases where profiling leads should be overlooked simply because someone
was tried, it was of no known assistance. On the other does not fit a profile, and no particular suspect should be
hand, the profiles may have been highly accurate in the focused on without other supporting evidence.
unsolved cases. If we limit our attention to cases where
the outcome was known, profiling was at least of some In other situations, advice is harder to give. An interviewer
help 83% of the time. may only get one chance to break a suspect in an
interrogation or, even more so, during a courtroom
More recently, in 1994, Bartol (1996) conducted a survey cross-examination. In such situations, if a particular profiler
of 152 police psychologists. Overall, 70% of the police has developed a track record of credibility and has specific
psychologists did not feel comfortable with profiling and advice as to how to proceed, one would need to have
seriously questioned its validity and usefulness. Before strong reasons to ignore the advice. Or, in dealing with a
giving too much weight to this study, it would be important hostage taker where lives are at stake and one or another
to have more details on how much experience the strategy must be followed, again the experienced profiler’s
respondents actually had with professionally done crime advice, backed by soundly articulated reasoning, should
scene profiles. Nevertheless, the Bartol study does raise a be given some weight.
cautionary note.
SUGGESTED RESEARCH
SUMMARY OF RELEVANT FINDINGS
Research on the validity of crime scene profiling suffers
In summary, there is enough research to suggest that especially from two limitations. One is the lack of access
crime scene profiling may have sufficient reliability and for neutral researchers to a representative sample of
validity to be useful for some purposes. The literature actual profiles. The second problem concerns the criterion:
suggests that the concept of behavioral traits and The accuracy of a profile can only be determined in cases
consistency across situations is respectable, if measured where the criminal has been clearly identified. This second
in broad contexts. Some antisocial behavior, especially if problem is compounded by the possibility that accurate
based on underlying psychopathology, may have a high profiles may be more likely to result in solved cases; thus,
degree of consistency. Within the narrow category of inaccurate profiles may be less likely to come to light.
sexual offenders and murderers, some theoretically
reasonable and reliable distinctions can be made. Crime The practical issue with profiling is whether it leads to an
scenes can be categorized with some degree of reliability increase in successful police investigations. In this sense,
and have been found to correlate with some offender even an inaccurate profile may be useful, for example, by
characteristics. Those trained in profiling have been found stimulating a line of inquiry. Probably the only way that this
to produce longer, more detailed reports, possibly with practical validity (or utility) could be determined would be
increased accuracy, and field agencies have been by a truly randomized experiment in which profiling was
generally positive in their feedback. withheld from some otherwise suitable cases. We do not
seriously propose that such a study be done, of course, on
At the same time, all of the supporting data seem the grounds that no method that has promise should be
somewhat tentative, and much of the research is in-house withheld from the types of cases with which profilers
and done on the same common core of offenders. typically deal.
Inaccurate profiles often seem to be ignored or forgotten
(Jenkins, 1994; Porter, 1983; Rosenbaum, 1993). A more limited study that might be possible would be to
Although all profilers caution that profiling is an art and that make available all profiles from solved cases. This would
mistakes can be made, the occasional dramatic success at least permit researchers to make more or less objective

- Reprinted with permission. Additional copying is prohibited. - GALE GROUP

Information Integrity
Criminal Justice and Behavior Sept 1998 v25 n3 p319(25) Page 10

Psychological aspects of crime scene profiling.

judgments about the accuracy of the various elements of Brantingham, P. J., & Brantingham, P. L. (Eds.). (1981).
the profile. One limitation that we foresee in such a study Environmental criminology. Beverly Hills, CA: Sage.
is that profilers often do not clarify the basis for their
predictions. Thus, a piece of physical evidence that leads Caspi, A., & Moffitt, T. E. (1995). The continuity of
to a straightforward conclusion may make the profile maladaptive behavior: From description to understanding
sound much more intuitive than it really is. There may also in the study of antisocial behavior. In D. Cicchetti & D. J.
be legitimate policy reasons to deny neutral researchers Cohen (Eds.), Developmental psychology (Vol. 2, pp.
access to FBI files. One obvious concern has to do with 472-511). New York: John Wiley.
the privacy needs of victims and their families. A less
obvious problem might be the publication of too much Davies, A., & Dale, A. (1995). Locating the stranger rapist
information that could be useful to offenders who stage (Special Interest Series: Paper 3). London: Police
crime scenes. Nevertheless, until neutral researchers have Research Group, Home Office Police Department.
some way of determining the accuracy of the various
elements of a representative sample of profiles, compared DeGeneste, H. I., & Sullivan, J. P. (1994). Policing
against some meaningful baseline data, the validity of the transportation facilities. Springfield, IL: Charles C Thomas.
entire process has to remain in a great deal of doubt.
Dietz, P. E. (1985). Sex offender profiling by the FBI: A
In the meantime, we recommend that more attention be preliminary cconceptual model. In M. H Ben-Aron, S. J.
paid to the specific concepts used by profilers. For Hucker, & C. D. Webster (Eds.), Clinical criminology: The
example, objective scales for categorizing crime scenes assessment and treatment of criminal behavior (pp.
could be developed. MO, signature, staging, and posing 207-219). Toronto: Clarke Institute of Psychiatry.
need to be more carefully operationalized and
distinguished. Relationships between crime scenes and Dietz, P. E., Hazelwood, R., & Warren, J. (1990). The
offender characteristics, when known, should be sexually sadistic criminal and his offenses. Bulletin of the
cross-validated on new samples of offenders. American Academy of Psychiatry and Law, 18, 163-178.

On a larger scale, more objectively gathered qualitative Douglas, J. E., & Munn, C. (1992, February). Violent crime
data on experiences with profiles should be obtained. This scene analysis: Modus operandi, signature, and staging.
would involve a more careful tracking of profiles that are FBI Law Enforcement Bulletin, pp. 1-20.
generated and a comparison with actual offenders when
outcomes are known. How exactly did the profile lead or Douglas, J., & Olshaker, M. (1995). Mind hunter: Inside the
not lead to the offender? How many of the elements of the FBI’s elite serial crime unit. New York: Star.
profile fit the offender? Although some of this information is
available anecdotally in biographical material (Douglas & Douglas, J., & Olshaker, M. (1997). Journey into darkness.
Olshaker, 1995, 1997; Ressler & Schachtman, 1992), it New York: Scribner.
does not seem to have been gathered systematically (e.g.,
to permit failures to be more carefully scrutinized). Douglas, J. E., Ressler, R. K., Burgess, A. W., & Hartman,
C. R. (1986). Criminal profiling from crime scene analysis.
Finally, more attention should be paid to the various uses Behavioral Sciences and the Law, 4, 401-421.
to which profiling is being put--from various types of
crimes, to interrogation advice, to predictions in negligent Easteal, P. W., & Wilson, P. (1991). Preventing crime on
security cases (Kennedy & Homant, 1997). It is important transport: Rail, buses, taxis, planes. Canberra: Australian
that a halo effect not be created, whereby a finding of Institute of Criminology.
profiling success in one area might be taken as indicative
of all possible uses. Geberth, V. J. (1990). Practical homicide investigation:
Tactics, procedures, and forensic techniques (2nd ed.).
REFERENCES Boca Raton, FL: CRC Press.

Annon, J. S. (1995). Investigative profiling: A behavioral Geberth, V. J. (1995). Criminal personality profiling: The
analysis of the crime scene. American Journal of Forensic signature aspect in criminal investigation. Law and Order,
Psychology, 13, 67-75. 43, 45-49.

Bartol, C. (1996). Police psychology: Then, now, and Godwin, M. (in press). Profiling serial killers: A theoretical
beyond. Criminal Justice and Behavior, 23, 70-89. analysis of motivational classifications. Liverpool, UK:
University of Liverpool.

- Reprinted with permission. Additional copying is prohibited. - GALE GROUP

Information Integrity
Criminal Justice and Behavior Sept 1998 v25 n3 p319(25) Page 11

Psychological aspects of crime scene profiling.

Godwin, M., & Canter, D. (1997). Encounter and death: Kendrick, D. T., & Funder, D. C. (1988). Profiting from
The spatial behavior of U.S. serial killers. Policing: An controversy: Lessons from the person-situation debate.
International Journal of Police Strategy and Management, American Psychologist, 43, 23-34.
20(1), 24-38.
Kennedy, D. B. (1990). Facility site selection and analysis
Groth, N., Burgess, A. W., & Holmstrom, L. L. (1977). through environmental criminology. Journal of Criminal
Rape: Power, anger, and sexuality. American Journal of Justice, 18, 239-252.
Psychiatry, 134, 1239-1243.
Kennedy, D. B., & Homant, R. J. (1997). Problems with the
Hartup, W. W., & van Lieshout, C.F.M. (1995). Personality use of criminal profiling in negligent security tort litigation.
development in social context. Annual Review of Trial Diplomacy Journal, 20, 223-229.
Psychology, 46, 655-687.
Keppel, R. D. (1995). The riverman. New York: Pocket.
Hazelwood, R., & Burgess, A. (Eds.). (1987). Practical
aspects of rape investigation: A multidisciplinary approach. Knight, R. A., & Prentky, R. A. (1990). Classifying sexual
New York: Elsevier North-Holland. offenders: The development and corroboration of
taxonomic models. In W. L. Marshall, D. R. Laws, & H. E.
Hazelwood, R. R., Dietz, P. E., & Burgess, R. N. (1982). Barbaree (Eds.), The handbook of sexual assault: Issues,
Sexual fatalities: Behavioral reconstruction in equivocal theories, and treatment of the offender (pp. 23-52). New
cases. Journal of Forensic Science, 27, 763-773. York: Plenum.

Hazelwood, R. P., Reboussin, R., & Warren, J. I. (1989). McCann, J. T. (1992). Criminal personality profiling in the
Serial rape: Correlates of increased aggression and the investigation of violent crime: Recent advances and future
relationship of offender pleasure to victim resistance. directions. Behavioral Sciences and the Law, 10, 475-481.
Journal of Interpersonal Violence, 4, 65-78.
Murphy, D. W., & Peters, J. M. (1992). Profiling child
Hazelwood, R. R., & Warren, J. W. (1989a, January). The sexual abusers: Psychological considerations. Criminal
serial rapist: His characteristics and victims (Part 1). FBI Justice and Behavior, 19, 24-37. Pinizzotto, A. J. (1984).
Law Enforcement Bulletin, pp. 10-17. Forensic psychology: Criminal personality profiling. Journal
of Police Science and Administration, 12, 32-40.
Hazelwood, R. R., & Warren, J. W. (1989b, February). The
serial rapist: His characteristics and victims (Conclusion). Pinizzotto, A. J., & Finkel, N. J. (1990). Criminal
FBI Law Enforcement Bulletin, pp. 18-25. personality profiling: An outcome and process study. Law
and Human Behavior, 14, pp. 215-233.
Hazelwood, R. R., & Warren, J. W. (1990, February). The
criminal behavior of the serial rapist. FBI Law Enforcement Porter, B. (1983, April). Mind hunters. Psychology Today,
Bulletin, pp. 11-16. pp. 44-52.

Hickey, E. W. (1997). Serial murderers and their victims Poythress, N., Otto, R. K., Darkes, J., & Start, L. (1993).
(2nd ed.). Belmont, CA: Wadsworth. APA’s expert panel in the congressional review of the USS
Iowa incident. American Psychologist, 48, 8-15.
Holmes, R. (1989). Profiling violent crimes: An
investigative tool. Newbury Park, CA: Sage. Prentky, A. P., Burgess, A. W., Rokous, B. A., Lee, A.,
Hartman, C.,
Jackson, J. L., van den Eshof, P., & de Kleuver, E. E.
(1994). Offender profiling in the Netherlands (Report Ressler, R., & Douglas, J. (1989). The presumptive role of
NSCR WD94-03). Leiden, the Netherlands: The fantasy in serial sexual homicide. American Journal of
Netherlands Institute for the Study of Criminality and Law Psychiatry, 146, 887-891.
Enforcement.
Prentky, R. A., & Knight, R. A. (1991). Identifying critical
Jeffers, H. P. (1991). Who killed precious? Chicago: dimensions for discriminating among rapists. Journal of
Congdon and Weed. Consulting and Clinical Psychology, 59, 643-661.

Jenkins, P. (1994). Using murder: The social construction Ressler, R. K., & Burgess, A. W. (1985, August). Violent
of homicide. New York: Aldine. crime. FBI Law Enforcement Bulletin, pp. 1-32.

- Reprinted with permission. Additional copying is prohibited. - GALE GROUP

Information Integrity
Criminal Justice and Behavior Sept 1998 v25 n3 p319(25) Page 12

Psychological aspects of crime scene profiling.

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, I, 288-308.

Ressler, R. K., Burgess, A. W., Douglas, J. E., Hartman,

C. R., & McCormack, A. (1986). Murderers who rape and
mutilate. Journal of Interpersonal Violence, I, 273-287.

Ressler, R. K., & Schachtman, T. (1992). Whoever fights

monsters. New York: Simon & Schuster.

Rosenbaum, R. (1993, April). The FBI’s agent

provocateur. Vanity Fair, pp. 122-136.

Rossmo, D. K. (1995a). Geographic profiling: Target

patterns of serial murderers. Doctoral dissertation, Simon
Fraser University, Burnaby, British Columbia, Canada.

Rossmo, D. K. (1995b). Place, space, and police

investigations: Hunting serial violent criminals. In J. E. Eck
& D. Weisburd (Ers.), Crime and place: Crime prevention
studies (Vol. 4, pp. 217235). Monsey, NY: Criminal Justice
Press.

Warren, J. I., Hazelwood, R. R., & Dietz, P. E. (1996). The

sexually sadistic serial killer. Journal of Forensic Sciences,
41,970-974.

Warren, J. I., Reboussin, R., Hazelwood, R. R, & Wright, J.

A. (1991). Prediction of rapist type and violence from
verbal, physical and sexual scales. Journal of
Interpersonal Violence, 6, 55 -67.

Wikin, G. (1996, April 22). How the FBI paints portraits of

the nation’s most wanted. U.S. News & World Report, 32.

AUTHORS’ NOTE: Correspondence concerning this article

should be addressed to Robert J. Homant, Department of
Criminal Justice, University of Detroit Mercy, P.O. Box
19900, Detroit, MI 48219-0900; e-mail:
homantr@[Link].

- Reprinted with permission. Additional copying is prohibited. - GALE GROUP

Information Integrity