Файл: [D._Kim_Rossmo]_Geographic_Profiling(BookFi.org).pdf

© 2000 by CRC Press LLC

ination of possible connections between unsolved and solved murders of
females within the U.K.

The Policing and Reducing Crime Unit (PRC) of the British Home Office

assists the NCF in several areas, including research (e.g., theory validation,
profiling utility, rape geography, etc.), building violent criminal databases,
and determining the optimal manner of delivery for profiling services (Aitken
et al., 1995; Copson, 1993; Dale & Davies, 1994a, 1994b; Davies & Dale,
1995b; Farrington & Lambert, 1993). They are currently involved in an
assessment of all operational offender profiles within the U.K.

5.2 Organized and Disorganized Crime Scenes

The FBI profiling approach is based upon a systematic process that follows
a four-stage sequence: (1) data assimilation; (2) crime classification; (3) crime
reconstruction; and (4) profile generation (Jackson & Bekerian, 1997a). The
criminal profile generating process involves the use of decision models, based
on data inputs, to develop crime assessments and criminal profiles (Ressler
et al., 1988). Further investigation may provide feedback to the system.
Depending upon offence type, different criminal personality typologies are
employed in the profiling process. For murder cases, psychological profilers
use the categories of organized and disorganized, as determined from crime
scene behaviour and offender lifestyle. While this is an ideal dichotomy, most
real killers are mixed to some degree.

Organized offenders typically plan their crimes, likely have access to a

vehicle, and travel further from home to offend. They are usually intelligent
and sane, but may be psychopathic. Characteristics include superior intelli-
gence, social and sexual competence, skilled occupation, high birth order status,
father with stable work history, inconsistent childhood discipline, controlled
mood during crime, alcohol use, precipitating situational stress, living with a
partner, mobile, car in good condition, investigation followed in the news, and
possible change of jobs or move from town after the crime (Ressler et al., 1988).
Organized crime scene characteristics include a planned crime, stranger victim,
victim personalized, demand for submissive victim, use of restraints, conver-
sation controlled, crime scene controlled, aggression prior to death, weapon
and evidence absent, victim’s body transported, and body hidden.

Disorganized offenders typically act spontaneously and do not plan their

crimes. They tend to live or work near the crime sites. Many suffer from
some form of psychosis such as paranoid schizophrenia. Characteristics
include average to below average intelligence, social and sexual immaturity,
poor work record, low birth order status, father with unstable work history,
harsh childhood discipline, anxious mood during crime, minimal use of

© 2000 by CRC Press LLC

alcohol, living alone, crime scene near residence or work, does not follow
the investigation in the news, and minimal change in lifestyle after the crime
(Ressler et al., 1988). Disorganized crime scene characteristics include a
spontaneous crime, knowledge of victim or location, sudden violence to
victim (blitz attack), minimal use of restraints, depersonalization of victim,
minimal conversation, random and sloppy crime scene, postmortem sex,
weapon and evidence present and body left in view and not removed from
crime scene.

There is supporting evidence that crime scenes can be reliably categorized

as organized or disorganized, and that this typology is correlated with
offender characteristics (Homant & Kennedy, 1998). An FBI study of crime
scene classification (n = 220) found an interrater agreement level of 74.1%
(Ressler & Burgess, 1985). Kocsis et al. (1998) examined arson cases in New
South Wales, Australia, to determine whether the organized/disorganized
dichotomy was a subjective categorization or actually intrinsic to the crime
itself. The study found support for the latter hypothesis. A cluster analysis
of crime scene variables produced two basic groupings that matched 

a priori

organized/disorganized categories and offender characteristics. But the need
still exists for further research into this offender profiling dichotomy and the
development of objectively quantifiable scales for categorizing crime scenes
(Homant & Kennedy, 1998).

The most commonly used profiling methodology appears to be crime

scene analysis (the FBI approach), followed by diagnostic evaluation
(Wilson, Lincoln, & Kocsis, 1997). Investigative psychology, another profil-
ing method primarily used in England, is based upon a five-factor model
constructed from facet theory and smallest space analysis (SSA): (1)
attempted intimacy; (2) sexual gratification; (3) aggression; (4) impersonal
interaction; and (5) criminality (Canter & Heritage, 1990). The Australian
national police Criminal Profiling Research Unit developed a profiling model
from the statistical analysis of 86 sexual murders from 1960 to 1998 (Kocsis,
1999). After elimination of actions common to most crimes, four distinct
behaviour patterns emerged: (1) the predator pattern, characterized by tor-
ture, sadism and a high level of planning; (2) the perversion pattern, typified
by organized behaviour and bizarre sexual perversions; (3) the fury pattern,
distinguished by hatred, hostility, and excessively brutal attacks; and (4) the
rape pattern, in which sexual assault is the primary purpose and the murder
is accidental or committed to avoid apprehension (the offender is often
vaguely acquainted with the victim). Predator and perversion patterns are
methodical (organized), and fury and rape patterns haphazard
(disorganized).

© 2000 by CRC Press LLC

5.3 Applications of Profiling

Profiling requirements are information intensive, perhaps because of the
method’s probabilistic nature. In turn, a profile can provide a variety of
suggestions concerning offender characteristics including likely age, race, sex,
socioeconomic status, residence description, method of transportation, edu-
cation level, marital status, employment background, criminal record
particulars, psychiatric history, social and sexual development, military his-
tory, physical characteristics, habits, degree of organization, pre-offence
behaviour, post-offence behaviour, and the possibility of accomplices
(Geberth, 1996; Ressler et al., 1988).

Profiles also attempt to give a general estimate of the criminal’s residence

proximity to the crime scenes (e.g., “the offender likely lives or works in the
neighbourhood of the murders”). Background information necessary for a
profile includes, amongst other details, the addresses of the victim’s residence
and employment, where he or she was last seen, the locations of the crime
sites, and a map of the victim’s travels prior to death (Geberth, 1996).

In addition to the criminal profiling of unknown suspects and the estab-

lishment of investigative priorities, profiling techniques have been used for
a variety of other purposes. These include indirect personality assessments
(to assist in developing interviewing, undercover, or cross-examination
approaches), equivocal death analyses, development of investigative and trial
strategies, establishing grounds for search and seizure, threat assessments,
and informing linkage analyses (Ault & Hazelwood, 1995; Cavanagh &
MacKay, 1991; Jackson et al., 1994).

Offences suitable for profiling involve those incidents where the suspect

has demonstrated some form of psychopathology or the offence is of an
unusual, bizarre, violent, sexual or repetitive nature (Geberth, 1981). Profiles
have been prepared in cases of mutilation, torture, and lust murders; serial
killings; homicides involving postmortem cutting, slashing, evisceration, or
body exploration; ritualistic and Satanic or cult crimes; rapes and sadistic
sexual assaults; random arsons; child molestings; bombings; bank robberies;
and false rape allegations (Cavanagh & MacKay, 1991). It has even been
suggested that the personality traits of a murderer may be profiled based on
the characteristics of the handgun (type, model, image, calibre, and ammu-
nition) used (Bensimon, 1997).

Howard Teten (1989), a former FBI member and one of the fathers of

criminal profiling, however, cautions that the technique should only be
applied to specific types of crimes, typically those involving overt sexual
activity or loss of contact with reality by the offender. Profiling property or
drug-related crimes is also problematic; there is usually little contact with
the victim or time spent at the scene in property offences, and drug use alters

© 2000 by CRC Press LLC

the basic personality of the offender. Rape crimes where the victim was
unconscious or cannot remember the incident, or the offender was under
the influence of drugs, did not speak, used minimal force, or did not engage
in atypical sexual activity, are difficult to profile (Hazelwood, Ressler, Depue,
& Douglas, 1995). Generally, cases lacking offender behaviour or other rel-
evant information will prove difficult to profile. And because of its probabi-
listic nature, profiling is more powerful in multiple than single crime cases.
Kennedy and Homant (1997) suggest that “modern profiling in general is a
viable and intriguing investigative tool that can be useful within limits.…
[but] refinement efforts should first be directed to improving the study of
serial rape and murder before reaching too far out into other crime catego-
ries” (pp. 226, 228).

Detective John Baeza (1999) of the New York Police Department outlines

criteria for what a police task force should expect from a psychological or
geographic profile. Profilers need to operate within their area of expertise,
integrate physical with behavioural evidence, and be prepared to visit the
crime sites. The report itself should be written, comprehensive, and provide
investigative suggestions. Profilers have to be able to explain their logic, and
investigators must be willing to ask questions about methods and conclu-
sions. He observes that while profiles are only a tool and their results are not
set in stone, they can play a useful role in task force management through
suspect and tip prioritization. FBI and Fellowship profilers often follow a
case consultation model where investigators present the details of their crimes
to a panel of profilers using photographs, maps, charts, and reports.  After
questions and answers, a brainstorming session leads to a collaborative profile
and specific case strategies, all documented by the investigator.

5.4 Critiques

Profiling has generated its share of interest, controversy, and censure (Jeffers,
1991; Jackson et al., 1994; Jenkins, 1994; Levin & Fox, 1985; Masters, 1994;
Poythress, Otto, Darkes, & Starr, 1993; Rosenbaum, 1993; Turco, 1990). It
has been suggested that profiles provide information that is general and
ambiguous at best, and misleading at worst. In particular, profiling has been
criticized as wanting programmatic validity and reliability research, and for
lacking a proper theoretical basis. Bartol (1996) surveyed 152 full and part-
time police psychologists and found that 70% were uncomfortable with the
validity and usefulness of profiling; many thought more research was needed.
The investigative psychology approach has been questioned in Britain
because of a lack of empirical verification and the mysterious methodology
underlying smallest space analysis (Copson & Holloway, 1996). Meanwhile,

© 2000 by CRC Press LLC

an analysis of profiling results from a small number of cases in Australia
produced mixed results (Wilson & Soothill, 1996).

Perhaps the most serious censure to date of profiling, at least within

North America, originated from a U.S. Congressional review (Jeffers, 1991).
The U.S. House of Representatives Armed Services Committee (HASC), while
reviewing the Navy investigation of the 1989 explosion on the USS 

Iowa

,

became concerned over the manner in which a conclusion of “intentional
action” had been reached (Poythress et al., 1993). Gunners Mate Clayton
Hartwig was singly blamed for the tragic incident that resulted in the deaths
of 47 sailors. He was alleged to have set off an explosion of a 16-inch turret
gun in order to commit suicide (Nelson, 1990).

The Naval Investigative Service (NIS) had requested that the National

Center for the Analysis of Violent Crime conduct a reconstructive psycho-
logical evaluation. This procedure, part of the FBI’s profiling service, is
termed equivocal death analysis (EDA), and is conceptually similar to a
psychological autopsy (Ebert, 1987; Poythress et al., 1993; see Fowler, 1986,
for example). In suspect cases, EDA attempts to determine manner of death
— accidental, suicidal, homicidal, or indeterminate — through the use of
retrospective psychological techniques.

The HASC was concerned over the validity and reliability of EDA, and

these worries were only exacerbated by the disinterest displayed by NCAVC
profilers for issues of validity and data integrity (Poythress et al., 1993). They
asked for assistance from the American Psychological Association (APA),
whose officials responded by forming a review panel comprising 12 psychol-
ogists and 2 psychiatrists. The majority of the APA experts subsequently
found the FBI’s analysis invalid, and the panel was unanimous in its criticism
of EDA procedures, methodology, and unarticulated limitations (

U.S.S. Iowa

Tragedy

, 1990). The lack of concern expressed by BSU profilers over validity

and reliability was found to be unacceptable (

Review of Navy investigation

,

1990). In short, the FBI did not pass the APA peer review.

Calling the process an investigative failure, the HASC noted that: “The

major problem with the Navy investigation is that it fell into the trap of an
excess of certitude. Thin gruel became red meat. Valid theories and hypotheses
were converted into hard fact” (Poythress et al., 1993, p. 10). The APA expert
panel cautioned: “The conclusions and inferences drawn in psychological
reconstructions are, at best, informed speculations or theoretical formulations
and should be labeled as such ... The kinds of unequivocal, bottom-line state-
ments offered by the FBI in its EDA report, regarding Clayton Hartwig, are
not defensible within the technical limitations of our science” (Poythress et
al., p. 12). In 1990, Sandia National Laboratories in New Mexico was able to
replicate the turret gun explosion through accidental means, further under-
mining the Navy’s investigative conclusions (Jeffers, 1991; Nelson, 1990).