PSYCH 3CC3 Lecture 5: Criminal Profiling
19 Jun 2018
School
Department
Course
Professor
Receives the most media, tv, movie, etc. attention but is the most insignificant part of
what forensic psychologists do.
-
History of Criminal Profiling
Begins in 1841 with a publication by Edgar Allen Poe, "Murders in the Rue Morgue"
Displays French detective solving murders in the Rue Morgue
○
Shown to do things consistently with how offender profilers work now.
○
-
Sherlock Holmes by Arthur Conan Doyle
"A Study in Scarlet" - 1887
○
Many films made (e.g., Basil Rathmon, Jeremy Brett, etc.)
○
-
First real-life profile (1888): Dr. Thomas Bond
Profile of Jack the Ripper
○
Some of these inferences are just common sense, but some are pretty profound.
○
-
Purposes of Profiling
Provide offender characteristics.
Everyone agrees on this purpose.
-
1.
Help understand the crime scene.
Helps us understand the crime, how/why things were done.
-
2.
Provide leads for investigators.
What kind of person would've done this? Demographic information?
-
3.
Narrow pool of viable subjects.
4.
Prioritizes investigation of subjects.
5.
Risk of offender escalation?
E.g., "Criminals like this will often start killing at a faster rate."
-
6.
Evaluate suspect possession.
Might help understand the relevance of pieces of possessions.
-
7.
Develop interrogation strategies.
8.
Show links between crimes.
E.g., Jack the Ripper
Even today, people debate how many people Jack the Ripper actually
killed.
§
Agree on 5, but there are 3 other people that may have also been killed by
him.
§
-
9.
Supportive trial testimony. 10.
Note: not all profiles agree with ALL of these purposes (usually the first 5 are heavily
agreed upon)
Challenges of Profiling
Turning crime scene info into description of offender.
This is extremely challenging.
-
We don't have a lot of data about the associations between crime scene
information and offender characteristics.
-
It’s the small removals/changes of area at crime scene that need to be
interpreted.
-
1.
Personality tests can't do this.
We have a lot of different instruments (i.e., personality measures) that can give
you a good description of a person's general characteristics, but they can't
describe someone in as much detail as offender profiles try to provide.
-
If years of psychology research on personality cannot do this, you can imagine
how difficult it would be for one person to achieve this.
-
2.
Unscientific - intuition, experience.
"Gut feeling"
-
They cannot tell you how exactly they profile people, how exactly they use
information to develop profiles because they don't even know how they do it.
-
They rely on intuition
-
This is why television programs display profilers as having "psychic" qualities.
-
3.
Not well evaluated - utility unclear.
Criminal profilers don't reveal their methods or profiles to the public.
Especially those done by the FBI
§
-
Consider Criminal Minds - profile is given verbally to the law enforcement, no
documentation released.
-
Accuracy is questioned heavily because of this.
Therefore, the usefulness is also questioned.
§
-
4.
Note: rape, murder and arson are most commonly profiled cases; sometimes high profile
robbery.
Basic Assumptions of Profiling
Behavioural Consistency
Offenders behave similarly across offenses.
An offender will behave in the same way across all of his crimes.
§
-
Cross-offense similarities must be infrequent to be of use.
But there has to be something the individual does similarly in all of his
crimes that is different from regular crimes of the same nature.
§
-
Bennell and Carter (2002): examined 4 characteristics for 2 crimes from each of
43 serial commercial burglars.
Inter-crime distance most closely similar in pairs of burglaries.
Only thing that distinguished crimes of the same offender.□
§
Type of business, method of entry, property stolen, intra-crime behaviours,
linked but not strongly enough.
These did not differentiate. □
§
-
Barteman and Salfati (2007): examined 35 behaviours of 90 offenders in 450
serial homicide cases.
Some things did differentiate … bringing crime kit, destroying evidence
both consistent across offenses and infrequent overall.
§
Bringing weapon, restraining victim both consistent across offenses, but
very common across all offenses … doesn't distinguish.
§
-
1.
Behavioural Differentiation
Offenders differ from one another in their patterns of inter-crime similarity.
Different offenders behave differently from one another across different
crimes.
§
-
Little evidence, but plausible.
-
2.
Homology
Offenders with similar crime behaviours will have similar characteristics.
-
Something about all of the individuals behaving similarly in crimes that is the
same.
-
Little good evidence supporting this.
-
Mokros and Allison (2002):
Rapists: are those who offend similarly also similar in personal
characteristics?
No positive correlations between crime similarily and similarity in
age, employment, ethnicity, criminal records, etc.
□
§
-
Doan and Snook (2008): Used typologies to classify 87 arson cases, 175
robberies. Compared similarity of offenders with similar case types.
Homology assumption violated in 56% of arson types and 67% of robbery
types.
§
-
If this doesn't work, the profile kind of goes to shit.
-
3.
Dr. James Brussel: Profile of New York Bomber (1956)
First modern profile
-
Number of bombs sent around New York City; sent letters explaining why they were
sending the bombs.
-
Brussel provided a profile:
"Male, former employee of Consolidated Edison, injured while working there so
seeking revenge; paranoid, 50 years old, neat and meticulous persona, foreign
background, some formal education, unmarried, living with female relatives, but
not mother who probably died when he was young. Upon capture, he will be
wearing a buttoned up double-breasted jacket."
-
-
Searched ConEd's records based on profile.
When they arrested him he was wearing a buttoned up double-breasted jacket …
-
-
FBI: Behavioural Analysis (Sciences) Unit
Founded in 1972
-
Howard Tenet - 1st Director
-
Robert Ressler, Roy Hazelwood and John Douglas
John Douglas is the biggest name in criminal profiling.
-
Huge publication history for all of them.
-
-
John Douglas is the inspiration for the Agent Jack Crawford in Silence of the Lambs;
author interviewed him.
-
Agent Rossi (Criminal Minds) also based off of John Douglas.
-
-(see slide for chart)
Crime Classification Manual
Describes the process by which the FBI constructs an offender profile.
-
Also has a crime classification system.
-
Offender Profile Stages:
Profiling Inputs
Crime scene information
-
Victimology
Has become very popular□
-
Forensic information
-
Police reports (i.e., time of day, people around, neighbourhood description,
etc.)
-
Photographs
-
1.
Decision Process Model
Homicide type and style (e.g., mass murder, spree killing, serial murder,
etc.)
-
Primary intent (e.g., revenge killing)
-
Victim, offender risk (risk of being detected and caught)
-
Escalation risk
-
Time, location factors
-
2.
Crime Assessment
Crime reconstruction
-
Crime classification: organized/disorganized
-
Staging
-
Motivation
-
Crime scene dynamics (interaction between things at time of the crime)
-
3.
Criminal Profile
Demographics
-
Physical characteristics
-
Habits
-
Beliefs and values (very common in rape cases)
-
Recommendations
-
4.
Investigation5.
Apprehension6.
-
Types of Murder
Mass Murder
Double vs triple homicide
○
4 + victims killed in the same place at the same time.
○
-
Murder Spree
No cooling off period between killings
○
Murdered in different locations within a relatively short time frame.
○
-
Serial Murder
Multiple victims
○
Cooling off period between victims
○
Typically in different locations
○
-
Organized-Disorganized Distinction
Made by FBI; distinction based on interviews with convenience sample of 36 serial
murders.
-
Interviews unstructured, different for each interviewee.
-
No subsequent validation with another sample.
-
Organized: someone who has planned the crime, taken the weapons/tools with them,
cleaned up afterward.
Profile would expect that the individual is organized IRL too.
○
-
Disorganized: opposite of organized
-
In 100 US serial murder cases, no indication that organized features correlated with
each other.
-
NOTE: even though they give you characteristics of a profile, they never provide a
system for profiling.
Not a science, but an art.
○
It's a special skill people develop with experience and intuition.
○
-
Holmes and Holmes: Offender Typologies
Disorganized Asocial/ Organized Nonsocial
-
Serial Murderers (defined based on motives)
Spatial mobility - do they kill within a small geographic area or a large one?
○
Visionary - are they motivated by vision?; usually mentall disordered, usually
found not guilty because of illness.
○
Missionary - sets out to kill people of a certain type (e.g., sex workers)
○
Hedonistic - murders for personal pleasure or personal gain.
○
Power and control - self-explanatory
○
-
Rapists (older categorization system, borrowed from other people)
Power reassurance - rape to reassure themselves about their power.
○
Anger retaliation - rape to get back at someone; generally person symbolically
representing the person who wronged them.
○
Exploitative - rape to take whatever they want
○
Sadistic - often followed by murder; pleasure in inflicting pain (violent)
○
-
Child Molesters
Situational - will have sex with children if the situation presents itself, but don't
need to have sex with children to get off.
○
Preferential - prefer children sexually; pedophiles
○
-
Arsonists
Jealousy - set fires to destroy things they want, but can't have; these people
don't stay to watch fire.
○
Would-be hero - fireman, policemen, etc.; start fires and stay to watch so they're
there to help and get the fame/glory of helping.
○
Excited fire-starter - find fires arousing (re: Freud); stay at fire to watch
○
Pyromaniac - sexually aroused by starting fires; stay at fire to watch
○
-
Geography
Trying to identify the suspect based on looking at the distance between victims.
○
-
Victim Profiling
Recent focus in research and actual profiling practice
○
Importance of understanding the victim to identify the offenders.
○
-
(see slide for chart)
Keppel and Walter: Rape/Murder Offender Typologies
(see slide for chart)
Took rape classifications and applied them to rape and murder.
Often rape and murder overlap
○
-
Note: don't need to remember all of the specific profiles; they're unclear and the
research doesn't show that they're accurate
-
Power Assertive:
○(see slide for description)
-
Power Reassurance:
○(see slide for description)
-
Anger Retaliatory:
○(see slide for description)
-
Anger and Excitation:
○(see slide for description)
-
Keppel and Walter (1999):
Power assertive is the most common
○
Next is Anger Retaliatory, then Power Reassurance
○
Anger Excitation is the least common
○
-
Brent Turvey
Publishes texts on criminal profiling
-
Academy of Behavioural Profiling
Pretty dead organization right now.
○
His company is called Forensic Solutions
Designed to train profilers
○
○
-
Started the Journal of Behavioural Profiling
Stopped publishing ~ 10 years ago
○
-
There just isn't enough literature on criminal profiling.
-
Turvey: Inductive vs Deductive Profiling
Inductive Profiling:
Starts with statistical generalizations ( e.g., most rapists are single males aged 18
to 30).
Based off of a bunch of data collected from things like FBI interviews, etc.
-
○
We then apply these generalizations to individual cases.
○
Problem: these general conclusions do not logically flow.
○
-
Deductive Profiling:
Starts with specific crime scene evidence (e.g., there are motorcycle tracks at the
crime scene.)
○
Conclusion: if they belong to the rapist, then he has access to a motorcycle.
○
These conclusions are often more logical because you start with the crime and
trace evidence back to the criminal.
○
-
Turvey wants to replace the common inductive profiling with deductive profiling
because it has more of a logical basis.
But you still cannot be certain that conclusions made in deductive profiling are
true.
○
-
Turvey Deductive Profiling Model: Four Components of Profiling
Forensic and Behavioural Evidence1.
Every investigation collects information about the crime.
-
E.g., witness and victim statements, crime photos, wound pattern and blood spatter
analysis, ballistics evidence, other forensic evidence.
-
Victimology2.
Turvey pushed this new idea a lot
-
E.g., physical characteristics, habits, lifestyle, relationships, risk level (e.g., do they do
things that would make them more susceptible to criminal activity)
-
Can these things tell us anything about the offender?
-
Crime Scene Characteristics3.
We consider what happened and in what order.
-
g., method of attack, nature of sexual or violent acts, verbal behaivour, precautionary
acts (i.e., did they clean up the crime scene? Did they clean up the body?)
-
Deduction of offender characteristics 4.
There is no system of offender profiling that tells you how to do this; intuitive/artistic.
-
Turvey's Behaviour-Motivational Typology
Applied this to all crimes; argues that these motives are not specific to certain crimes,
but are widely applicable to all crimes.
-
Description more of the crime characteristics
-
Includes:
Power reassurance
○
Power assertive
○
Anger retaliatory
○
Anger excitation
○
Profit (i.e., like hedonist)
○
-
Hickes and Sales (2006): Criticism of Non-scientific Systems of Profiling
Lack of goals and standards.1.
A lot of these systems do not outline the purpose/goals of the profile.
-
No agreement on goals of profiling
-
No agreement on crimes suitable for profiling.
Everyone agrees that you can profile murderers.
○
Most agree that you can profile rapists.
○
Many, but not all, agree that you can profile arsonists.
○
Others are very mixed.
○
-
No standards to evaluate the success of profiling.
Not validation of the accuracy or usefulness of the profiling.
○
Few attempts to evaluate.
○
-
Use of unclear terms and definitions.2.
Serial killers defined differently.
Some systems defines serial killers differenty than the FBI.
○
-
Modus operandi (way of operating) sometimes fixed, sometimes changes from crime to
crime - depends on the system whether it can change or not.
-
No description of how to determine a criminal "signature."
One area where this is clear defined is in bombing - bombers construct bombs in
their own, specific way.
○
-
Many other unclear terms …
-
Misuse of typologies3.
Inappropriate use of typologies
E.g., Offender may fall into several categories; fail to meet full criteria for any,
but these systems for them into a box.
○
-
Inconsistency within typologies
E.g., H&H: use pedophile and child molester both as synonyms and as different
There is a distinction to be made between these two terms.
-
○
E.g., in one category, an offender my be both sloppy and cunning?
○
E.g., in one category, an offender may be described has lacking empathy, but
later their behaviour is described has being the result of the loss of a loved one.
○
-
Overlapping categories
Many categories, same characteristics (e.g., single male, use of weapons, age,
etc.)
○
Might be identical in several different categories, so how do you discriminate?
You can't.
○
-
Limited value of typologies
Don't help police narrow suspect list; more like a horoscope than a useful
description.
○
-
Rely on intuition, professional knowledge4.
Not based on scientific information
-
Huge concern
-
Lack of clear procedures5.
NONE of the systems or authors provide a clear list of instructions to construct a
profile.
-
They can't tell you because it's based on a feeling or intuition.
-
Little evidence of investigative value6.
Very few attempts have been made to consider the usefulness of profiles.
-
Studies that have been done show that they're not that useful at all.
-
Canter's Scientific Model (wanted to put profiling on scientific footing; said that there were
questions that needed to be answered in order to do this.)
Behavioural Salience1.
What are the behavioural features of the crime that might help identify the offender?
-
Distinguishing between offenders2.
How do we indiciate differences between offedners and crimes?
-
Inferring characteristics3.
What inferences can we make about offenders that would help identify him?
-
What information will lead us to characteristics of a particular offender?
-
Linking offenses4.
How can we determine whether same offender committed a series of crimes?
-
E.g., goes all the way back to Jack the Ripper; 5 murders that have been confirmed as
being performed by Jack the Ripper, but there were 2-3 others that we're still debating
on.
-
Theoretical Links: Action -> Offender (Canter)
How do we see the relationship between crime scene characteristics and the offender?
-
Psychodynamic Typologies
Certain psychodynamic characteristics, Freudian types will lead to certain types
of offense behaviours.
○
Maybe we can use that differentiation of freudian psychological types to identify
a crime scene behaviours as being part of a certain individuals work.
○
-
Personality Differences
E.g., introverts vs extroverts
○
People with different personality characteristics might perform crimes
differently.
○
-
Career Routes
Most criminals do not engage in the same types of crimes for their entire criminal
careers.
○
Very often murders and rapists have committed other crimes first.
○
Maybe we can distinguish individuals based on the route they took on their way
to the current set of crimes.
E.g., did they come from assaultive crimes? Victimless crimes? Robbery?
Etc.
-
○
-
Socioeconomic Subgroups
E.g., blue collar crime vs white collar crime
○
-
Interpersonal Narratives
The internal stories that the offender is working, that they're trying to realize in
life.
○
How can we identify this and link it to crimes and crime scene characteristics?
○
This is the only possibility accepted by Canter.
Thinks we should be looking at the role the victim plays within the
offender's interpersonal narrative.
-
Applies to rape and murder typically because they have identifiable
victims.
-
○
-
Interpersonal Narratives: Role of the Victim
Victim as Object:
Offender has no feeling or connection with victim, who does not play any active
role in the assault.
○
Victim is just a piece of furniture, doesn't stand for anything.
○
High desire for control:
May involve mutilation
-
Low intelligence, poor contact with reality
-
Eccentric
-
From dysfunctional background
-
○
Low desire for control:
Sexual component prominent
-
Murder not major goal
-
Will not understand seriousness of crime
-
Probably little contact with people, undemanding job
-
○
-
Victim as Vehicle:
Offender is tragic hero, angry about self and his lot in life.
○
Assault regain him his rightful place in life/world.
○
High desire for control:
Like FBI's spree murderer
-
May kill many in single bout
-
May also kill self "Samson syndrome"
-
○
Low desire for control:
FBI's organized offender
-
Killings more deliberate
-
Motivated by failed/lost relationship
-
Offender intelligent, good social skills
-
○
-
Victim as Person:
Offender sees victim as individual, tries to understand their experience, but has
no true empathy for them.
○
No distinction between high and low desire for control here.
○
-
Canter's Smallest Space Analysis
Canter looks at many crimes and looks at specific aspects of the behaviours involved in
these crimes.
E.g., whether trophies were taken, whether weapons used were kept or
destroyed, was the victim tortured, etc.
○
-
Each of these words describes a type of crime scene activities.
-
Looked at the extent to which these events occurred together - intercorrelations.
-
Uses SSA to display these characteristics in a meaningful way.
Displays on a graph the distance between different behaviours.
○
Graphical depiction of the commonalities that occur in the crime.
○
-
Canter looks at these and says what type of interpersonal narratives these clusters
better apply to.
-
Argument goes - if you know the different characteristics of the crime scene, you'll be
able to determine their interpersonal narratives.
But what does the narrative even tell you?
○
-
Hicks and Sales Critique (of Canter)
Lack of conceptual clarity:
Victim's role in themes unclear
○
Inconsistency within themes
○
-
Unverifiable assertions re: inner narratives:
We cannot see into the minds of offenders to see what there intropsychic
narrative is.
○
… so what is the point of using them if a) we cannot know them and b) cannot
use them to narrow down the pool of suspects.
○
-
SSA inappropriate:
No factors indicated by analysis
○
Themes not linked to offender characteristics
No evidence, not even by Canter
-
Therefore, they can't be used to identify the offender from a pool of
suspects.
-
○
Can themes even predict offender characteristics?
Can't predict
-
Might be able to confirm
-
○
-
Hickes and Sales (2006):
What evidence should be gathered? 1.
Focuses on the pieces of evidence that would lean into a profile.
-
How should we interpret evidence?2.
None of the current books tell you how to do this because they don't believe that you
can teach people that.
-
What are relationships between evidence and offender characteristics? 3.
Which offender characteristics important to investigating crime? 4.
What is it that we can tell about the offender from the evidence?
-
None of the current systems can do this except for the non-scientific approaches?
-
D.C. Beltway Sniper (2002)
People got out of the car to get gas and they'd be shot.
-
After 5 or 6 people were murdered, FBI released a profile.
-
Profile:
White male, 24 to 40 years old
○
Not a sniper; not military
○
Lives in or near community; no children
○
Firefighter or construction worer
○
Possible terrorist links
○
-
Offenders: John Allen Mohammed; Lee Malvo
African-American male, 42 and 17
○
Mohammed spent 16 years in the National Guard, Army expert marksman - just
as good as a sniper
○
No terrorist links, but did admire terrorist works
○
-
Profile was inaccurate and didn't lead to the capture.
-
Baton Rouge Serial Killer (2003)
FBI Profile:
White male, 25 to 35 years old.
○
Unsophisticated social outcast
○
Awkward around women; doesn’t get along with them
○
Impulsive, quick-tempereed
○
-
Offender: Derrick Todd Lee
African-American male, 34 years old
○
Smooth, charming, "casanova"
○
Out with different women each night
○
-
Atlanta Olympic Bomber (1996)
FBI Profile:
Single, white middle-class male
○
Intense interest in police work
○
-
Arrested Richard Jewell
Because he fit the profile and was a security guard at the venue
○
Sued the government for ruining his potential police career.
○
-
Offender: Eric Rudolph
Anti-abortion activist
○
Lower-class single-parent family
○
Strong hatred for government and authority - opposite of the interest in police
work
○
-
Unabomber (1978-1995)
Sent a variety of bombs to universities and airlines
-
2 different profiles developed - one wildly inaccurate
-
FBI Profile (Tafoya):
White male, early 50s
○
University graduate with advanced degree
○
Background in science, math or engineering
○
Strongly anti-technology
○
-
Offender: Ted Koczynski
Not found because of the profile.
○
He wrote a manifesto in handwriting - his own brother recognized the
handwriting as his and turned him in.
○
White male, early 50s
○
M.S. and Ph.D. in math
○
Living alone in Montana cabin
○
Strongly anti-technology (correct, but not a great leap because he admitted it
this in the manifesto)
○
-
Paul Bernardo (1987-1992)
FBI profile (for the rapes):
Single white male, 18-25 years old
○
Lives in Scarborough area
○
High school education only
○
Angry, disparaging toward women
○
Spotty work record
○
-
Offender: Paul Bernardo
Single, white male
○
18 at first rape
○
Living in Scarborough area
○
University graduate
○
Gainfully employed (Price Waterhouse, Amway)
○
-
Again, he was not caught on the basis of the profile.
-
Bernardo Likeness
Rare case where the facial reconstruction of the offender was strikingly accurate
○
-
Criticisms of Profiling
Personality models lack empirical support. 1.
Assumes primacy of personality over situation.2.
Big debate in psychology for a long time - to what extent is behaviour determined of
personality vs situation.
-
Behaviour is usually the result of the demands (i.e., the demands of the crime)
-
Profiles are vague and ambiguous - fit many people. 3.
E.g., usually white males
-
Many aspects of profile are common sense4.
E.g., "Male, 18-30" - well, duh … most criminals are males between the ages of 18 and
30.
-
Kocsis et al (2000):
Studied 6 groups of Australian profilers:
35 police officers
○
30 psychologists
○
30 2nd year university students
○
20 psychics
○
5 profilers
○
23 university economics students (control)
○
-
Given details of two solved homicides:
Crime scene report, forensic biologist's report, forensic entomologist's report,
pathologist's postmortem report, info on victim, plan and photos of crime scene
and victim.
○
-
Then asks them a serious of questions about the offender's characteristics.
Given a score (# of items correct about offender)
○
Psychics and control groups don't do as well as others for physical characteristics,
but everyone else does equally as well.
Profilers BARELY better.
-
○
For behavioural characteristics, police do the worst and psychologists do the
best.
○
For offender cognitive characteristics, profilers significantly better (but only
about 1.5 pts higher)
○
For personality characteristics, psychologists do the best, but again, it's barely
better.
○
-
Point is … even though there might be a significant difference between profilers and
other groups, they're meaningless because they're small - the experts should be MUCH
better, but they're only slightly better.
-
-(see graphs)
Alison et al (2003):
Examined 880 statements in 21 profiles from US, UK (including FBI)
Majority of profile statements about offense and already known to investigators.
○
55% of offender descriptions not verifiable (i.e., emotions, motives, thoughts,
etc.)
Don't help you because they don't identify people …
-
○
24% of offender descriptions ambiguous (e.g., poor heterosexual social skills).
What do they even mean?
-
Don't help you …
-
○
6% of offender descriptions contain incompatible alterntives (e.g., unemployed
or manual job)
○
-
Kocsis (2004):
Studied 6 groups of Australian profilers:
3 profilers
○
13 police detectives
○
12 fire investigators
○
21 2nd year B.S. students (chemistry)
○
47 community college students (control)
○
-
Given details re: serial arsonist (13 fires)
Crime scene schematics, incident reports, forensic reports, 12 captioned photos.
○
-
Asked about the characteristics of the offender.
Again, marked # correct.
○
-
-(see graphs)
Consistent with previous study … profilers should be way better, but they're not.
-
So if profiling is a reputable career and something you can be an expert in, why are they not
much better than everyone else at profiling?
Copson (1995)
Surveyed 184 detectives who had used profiles regarding the usefulness of profiles.
-
83% said profiles "useful"
61% - said profiles helped them understand the case
○
52% - said profiles supported their own opinions
○
Basically, they helped us validate our own personal views, that's probably why
law enforcement continues to rely on profiles.
○
-
What percentage of profiles actually helped you identify the offender?
2.7%
○
-
Recall: Dr. Thomas Bond (1888)
In 1988, John Douglas produced his own profile of Jack the Ripper, 100 years later.
-
Couldn't come up with anything better than Dr. Bond's,
Even though 100 years have passed and John Douglas' expertise.
○
-
Where are we now?
" … this research confirms the perceptions of those who have concluded the criminal
profiling relies on weak standards of proof and that profilers do not decisively
outperform other groups when predicting the characteristics of an unknown criminal …
profiling appears at this juncture to be an extraneous and redundant technique for use
in criminal investigations. Criminal profiling will persist as a pseudoscientific technique
until such time as empirical and reproducible studies are conducted on the abilities of
large groups of active profilers to predict, with more precision and greater magnitude,
the characteristics of offenders."
Snook et al (2008)
-
-
The time is not coming soon when this will be a set science …
-
Note: Hicks and Sales, based on this
critique would argue that this is not
the basis on which we should turn
profiling into a scientific enterprise.
Note: Basically, Hicks and Sales don't
have a system, but they think this is what
a good system should have, what it
should indicate
Criminal Profiling
Tuesday, March 6, 2018
9:56 AM
Receives the most media, tv, movie, etc. attention but is the most insignificant part of
what forensic psychologists do.
-
History of Criminal Profiling
Begins in 1841 with a publication by Edgar Allen Poe, "Murders in the Rue Morgue"
Displays French detective solving murders in the Rue Morgue
○
Shown to do things consistently with how offender profilers work now.
○
-
Sherlock Holmes by Arthur Conan Doyle
"A Study in Scarlet" - 1887
○
Many films made (e.g., Basil Rathmon, Jeremy Brett, etc.)
○
-
First real-life profile (1888): Dr. Thomas Bond
Profile of Jack the Ripper
○
Some of these inferences are just common sense, but some are pretty profound.
○
-
Purposes of Profiling
Provide offender characteristics.
Everyone agrees on this purpose.
-
1.
Help understand the crime scene.
Helps us understand the crime, how/why things were done.
-
2.
Provide leads for investigators.
What kind of person would've done this? Demographic information?
-
3.
Narrow pool of viable subjects.
4.
Prioritizes investigation of subjects.
5.
Risk of offender escalation?
E.g., "Criminals like this will often start killing at a faster rate."
-
6.
Evaluate suspect possession.
Might help understand the relevance of pieces of possessions.
-
7.
Develop interrogation strategies.
8.
Show links between crimes.
E.g., Jack the Ripper
Even today, people debate how many people Jack the Ripper actually
killed.
§
Agree on 5, but there are 3 other people that may have also been killed by
him.
§
-
9.
Supportive trial testimony. 10.
Note: not all profiles agree with ALL of these purposes (usually the first 5 are heavily
agreed upon)
Challenges of Profiling
Turning crime scene info into description of offender.
This is extremely challenging.
-
We don't have a lot of data about the associations between crime scene
information and offender characteristics.
-
It’s the small removals/changes of area at crime scene that need to be
interpreted.
-
1.
Personality tests can't do this.
We have a lot of different instruments (i.e., personality measures) that can give
you a good description of a person's general characteristics, but they can't
describe someone in as much detail as offender profiles try to provide.
-
If years of psychology research on personality cannot do this, you can imagine
how difficult it would be for one person to achieve this.
-
2.
Unscientific - intuition, experience.
"Gut feeling"
-
They cannot tell you how exactly they profile people, how exactly they use
information to develop profiles because they don't even know how they do it.
-
They rely on intuition
-
This is why television programs display profilers as having "psychic" qualities.
-
3.
Not well evaluated - utility unclear.
Criminal profilers don't reveal their methods or profiles to the public.
Especially those done by the FBI
§
-
Consider Criminal Minds - profile is given verbally to the law enforcement, no
documentation released.
-
Accuracy is questioned heavily because of this.
Therefore, the usefulness is also questioned.
§
-
4.
Note: rape, murder and arson are most commonly profiled cases; sometimes high profile
robbery.
Basic Assumptions of Profiling
Behavioural Consistency
Offenders behave similarly across offenses.
An offender will behave in the same way across all of his crimes.
§
-
Cross-offense similarities must be infrequent to be of use.
But there has to be something the individual does similarly in all of his
crimes that is different from regular crimes of the same nature.
§
-
Bennell and Carter (2002): examined 4 characteristics for 2 crimes from each of
43 serial commercial burglars.
Inter-crime distance most closely similar in pairs of burglaries.
Only thing that distinguished crimes of the same offender.□
§
Type of business, method of entry, property stolen, intra-crime behaviours,
linked but not strongly enough.
These did not differentiate. □
§
-
Barteman and Salfati (2007): examined 35 behaviours of 90 offenders in 450
serial homicide cases.
Some things did differentiate … bringing crime kit, destroying evidence
both consistent across offenses and infrequent overall.
§
Bringing weapon, restraining victim both consistent across offenses, but
very common across all offenses … doesn't distinguish.
§
-
1.
Behavioural Differentiation
Offenders differ from one another in their patterns of inter-crime similarity.
Different offenders behave differently from one another across different
crimes.
§
-
Little evidence, but plausible.
-
2.
Homology
Offenders with similar crime behaviours will have similar characteristics.
-
Something about all of the individuals behaving similarly in crimes that is the
same.
-
Little good evidence supporting this.
-
Mokros and Allison (2002):
Rapists: are those who offend similarly also similar in personal
characteristics?
No positive correlations between crime similarily and similarity in
age, employment, ethnicity, criminal records, etc.
□
§
-
Doan and Snook (2008): Used typologies to classify 87 arson cases, 175
robberies. Compared similarity of offenders with similar case types.
Homology assumption violated in 56% of arson types and 67% of robbery
types.
§
-
If this doesn't work, the profile kind of goes to shit.
-
3.
Dr. James Brussel: Profile of New York Bomber (1956)
First modern profile
-
Number of bombs sent around New York City; sent letters explaining why they were
sending the bombs.
-
Brussel provided a profile:
"Male, former employee of Consolidated Edison, injured while working there so
seeking revenge; paranoid, 50 years old, neat and meticulous persona, foreign
background, some formal education, unmarried, living with female relatives, but
not mother who probably died when he was young. Upon capture, he will be
wearing a buttoned up double-breasted jacket."
-
-
Searched ConEd's records based on profile.
When they arrested him he was wearing a buttoned up double-breasted jacket …
-
-
FBI: Behavioural Analysis (Sciences) Unit
Founded in 1972
-
Howard Tenet - 1st Director
-
Robert Ressler, Roy Hazelwood and John Douglas
John Douglas is the biggest name in criminal profiling.
-
Huge publication history for all of them.
-
-
John Douglas is the inspiration for the Agent Jack Crawford in Silence of the Lambs;
author interviewed him.
-
Agent Rossi (Criminal Minds) also based off of John Douglas.
-
-(see slide for chart)
Crime Classification Manual
Describes the process by which the FBI constructs an offender profile.
-
Also has a crime classification system.
-
Offender Profile Stages:
Profiling Inputs
Crime scene information
-
Victimology
Has become very popular□
-
Forensic information
-
Police reports (i.e., time of day, people around, neighbourhood description,
etc.)
-
Photographs
-
1.
Decision Process Model
Homicide type and style (e.g., mass murder, spree killing, serial murder,
etc.)
-
Primary intent (e.g., revenge killing)
-
Victim, offender risk (risk of being detected and caught)
-
Escalation risk
-
Time, location factors
-
2.
Crime Assessment
Crime reconstruction
-
Crime classification: organized/disorganized
-
Staging
-
Motivation
-
Crime scene dynamics (interaction between things at time of the crime)
-
3.
Criminal Profile
Demographics
-
Physical characteristics
-
Habits
-
Beliefs and values (very common in rape cases)
-
Recommendations
-
4.
Investigation5.
Apprehension6.
-
Types of Murder
Mass Murder
Double vs triple homicide
○
4 + victims killed in the same place at the same time.
○
-
Murder Spree
No cooling off period between killings
○
Murdered in different locations within a relatively short time frame.
○
-
Serial Murder
Multiple victims
○
Cooling off period between victims
○
Typically in different locations
○
-
Organized-Disorganized Distinction
Made by FBI; distinction based on interviews with convenience sample of 36 serial
murders.
-
Interviews unstructured, different for each interviewee.
-
No subsequent validation with another sample.
-
Organized: someone who has planned the crime, taken the weapons/tools with them,
cleaned up afterward.
Profile would expect that the individual is organized IRL too.
○
-
Disorganized: opposite of organized
-
In 100 US serial murder cases, no indication that organized features correlated with
each other.
-
NOTE: even though they give you characteristics of a profile, they never provide a
system for profiling.
Not a science, but an art.
○
It's a special skill people develop with experience and intuition.
○
-
Holmes and Holmes: Offender Typologies
Disorganized Asocial/ Organized Nonsocial
-
Serial Murderers (defined based on motives)
Spatial mobility - do they kill within a small geographic area or a large one?
○
Visionary - are they motivated by vision?; usually mentall disordered, usually
found not guilty because of illness.
○
Missionary - sets out to kill people of a certain type (e.g., sex workers)
○
Hedonistic - murders for personal pleasure or personal gain.
○
Power and control - self-explanatory
○
-
Rapists (older categorization system, borrowed from other people)
Power reassurance - rape to reassure themselves about their power.
○
Anger retaliation - rape to get back at someone; generally person symbolically
representing the person who wronged them.
○
Exploitative - rape to take whatever they want
○
Sadistic - often followed by murder; pleasure in inflicting pain (violent)
○
-
Child Molesters
Situational - will have sex with children if the situation presents itself, but don't
need to have sex with children to get off.
○
Preferential - prefer children sexually; pedophiles
○
-
Arsonists
Jealousy - set fires to destroy things they want, but can't have; these people
don't stay to watch fire.
○
Would-be hero - fireman, policemen, etc.; start fires and stay to watch so they're
there to help and get the fame/glory of helping.
○
Excited fire-starter - find fires arousing (re: Freud); stay at fire to watch
○
Pyromaniac - sexually aroused by starting fires; stay at fire to watch
○
-
Geography
Trying to identify the suspect based on looking at the distance between victims.
○
-
Victim Profiling
Recent focus in research and actual profiling practice
○
Importance of understanding the victim to identify the offenders.
○
-
(see slide for chart)
Keppel and Walter: Rape/Murder Offender Typologies
(see slide for chart)
Took rape classifications and applied them to rape and murder.
Often rape and murder overlap
○
-
Note: don't need to remember all of the specific profiles; they're unclear and the
research doesn't show that they're accurate
-
Power Assertive:
○(see slide for description)
-
Power Reassurance:
○(see slide for description)
-
Anger Retaliatory:
○(see slide for description)
-
Anger and Excitation:
○(see slide for description)
-
Keppel and Walter (1999):
Power assertive is the most common
○
Next is Anger Retaliatory, then Power Reassurance
○
Anger Excitation is the least common
○
-
Brent Turvey
Publishes texts on criminal profiling
-
Academy of Behavioural Profiling
Pretty dead organization right now.
○
His company is called Forensic Solutions
Designed to train profilers
○
○
-
Started the Journal of Behavioural Profiling
Stopped publishing ~ 10 years ago
○
-
There just isn't enough literature on criminal profiling.
-
Turvey: Inductive vs Deductive Profiling
Inductive Profiling:
Starts with statistical generalizations ( e.g., most rapists are single males aged 18
to 30).
Based off of a bunch of data collected from things like FBI interviews, etc.
-
○
We then apply these generalizations to individual cases.
○
Problem: these general conclusions do not logically flow.
○
-
Deductive Profiling:
Starts with specific crime scene evidence (e.g., there are motorcycle tracks at the
crime scene.)
○
Conclusion: if they belong to the rapist, then he has access to a motorcycle.
○
These conclusions are often more logical because you start with the crime and
trace evidence back to the criminal.
○
-
Turvey wants to replace the common inductive profiling with deductive profiling
because it has more of a logical basis.
But you still cannot be certain that conclusions made in deductive profiling are
true.
○
-
Turvey Deductive Profiling Model: Four Components of Profiling
Forensic and Behavioural Evidence1.
Every investigation collects information about the crime.
-
E.g., witness and victim statements, crime photos, wound pattern and blood spatter
analysis, ballistics evidence, other forensic evidence.
-
Victimology2.
Turvey pushed this new idea a lot
-
E.g., physical characteristics, habits, lifestyle, relationships, risk level (e.g., do they do
things that would make them more susceptible to criminal activity)
-
Can these things tell us anything about the offender?
-
Crime Scene Characteristics3.
We consider what happened and in what order.
-
g., method of attack, nature of sexual or violent acts, verbal behaivour, precautionary
acts (i.e., did they clean up the crime scene? Did they clean up the body?)
-
Deduction of offender characteristics 4.
There is no system of offender profiling that tells you how to do this; intuitive/artistic.
-
Turvey's Behaviour-Motivational Typology
Applied this to all crimes; argues that these motives are not specific to certain crimes,
but are widely applicable to all crimes.
-
Description more of the crime characteristics
-
Includes:
Power reassurance
○
Power assertive
○
Anger retaliatory
○
Anger excitation
○
Profit (i.e., like hedonist)
○
-
Hickes and Sales (2006): Criticism of Non-scientific Systems of Profiling
Lack of goals and standards.1.
A lot of these systems do not outline the purpose/goals of the profile.
-
No agreement on goals of profiling
-
No agreement on crimes suitable for profiling.
Everyone agrees that you can profile murderers.
○
Most agree that you can profile rapists.
○
Many, but not all, agree that you can profile arsonists.
○
Others are very mixed.
○
-
No standards to evaluate the success of profiling.
Not validation of the accuracy or usefulness of the profiling.
○
Few attempts to evaluate.
○
-
Use of unclear terms and definitions.2.
Serial killers defined differently.
Some systems defines serial killers differenty than the FBI.
○
-
Modus operandi (way of operating) sometimes fixed, sometimes changes from crime to
crime - depends on the system whether it can change or not.
-
No description of how to determine a criminal "signature."
One area where this is clear defined is in bombing - bombers construct bombs in
their own, specific way.
○
-
Many other unclear terms …
-
Misuse of typologies3.
Inappropriate use of typologies
E.g., Offender may fall into several categories; fail to meet full criteria for any,
but these systems for them into a box.
○
-
Inconsistency within typologies
E.g., H&H: use pedophile and child molester both as synonyms and as different
There is a distinction to be made between these two terms.
-
○
E.g., in one category, an offender my be both sloppy and cunning?
○
E.g., in one category, an offender may be described has lacking empathy, but
later their behaviour is described has being the result of the loss of a loved one.
○
-
Overlapping categories
Many categories, same characteristics (e.g., single male, use of weapons, age,
etc.)
○
Might be identical in several different categories, so how do you discriminate?
You can't.
○
-
Limited value of typologies
Don't help police narrow suspect list; more like a horoscope than a useful
description.
○
-
Rely on intuition, professional knowledge4.
Not based on scientific information
-
Huge concern
-
Lack of clear procedures5.
NONE of the systems or authors provide a clear list of instructions to construct a
profile.
-
They can't tell you because it's based on a feeling or intuition.
-
Little evidence of investigative value6.
Very few attempts have been made to consider the usefulness of profiles.
-
Studies that have been done show that they're not that useful at all.
-
Canter's Scientific Model (wanted to put profiling on scientific footing; said that there were
questions that needed to be answered in order to do this.)
Behavioural Salience1.
What are the behavioural features of the crime that might help identify the offender?
-
Distinguishing between offenders2.
How do we indiciate differences between offedners and crimes?
-
Inferring characteristics3.
What inferences can we make about offenders that would help identify him?
-
What information will lead us to characteristics of a particular offender?
-
Linking offenses4.
How can we determine whether same offender committed a series of crimes?
-
E.g., goes all the way back to Jack the Ripper; 5 murders that have been confirmed as
being performed by Jack the Ripper, but there were 2-3 others that we're still debating
on.
-
Theoretical Links: Action -> Offender (Canter)
How do we see the relationship between crime scene characteristics and the offender?
-
Psychodynamic Typologies
Certain psychodynamic characteristics, Freudian types will lead to certain types
of offense behaviours.
○
Maybe we can use that differentiation of freudian psychological types to identify
a crime scene behaviours as being part of a certain individuals work.
○
-
Personality Differences
E.g., introverts vs extroverts
○
People with different personality characteristics might perform crimes
differently.
○
-
Career Routes
Most criminals do not engage in the same types of crimes for their entire criminal
careers.
○
Very often murders and rapists have committed other crimes first.
○
Maybe we can distinguish individuals based on the route they took on their way
to the current set of crimes.
E.g., did they come from assaultive crimes? Victimless crimes? Robbery?
Etc.
-
○
-
Socioeconomic Subgroups
E.g., blue collar crime vs white collar crime
○
-
Interpersonal Narratives
The internal stories that the offender is working, that they're trying to realize in
life.
○
How can we identify this and link it to crimes and crime scene characteristics?
○
This is the only possibility accepted by Canter.
Thinks we should be looking at the role the victim plays within the
offender's interpersonal narrative.
-
Applies to rape and murder typically because they have identifiable
victims.
-
○
-
Interpersonal Narratives: Role of the Victim
Victim as Object:
Offender has no feeling or connection with victim, who does not play any active
role in the assault.
○
Victim is just a piece of furniture, doesn't stand for anything.
○
High desire for control:
May involve mutilation
-
Low intelligence, poor contact with reality
-
Eccentric
-
From dysfunctional background
-
○
Low desire for control:
Sexual component prominent
-
Murder not major goal
-
Will not understand seriousness of crime
-
Probably little contact with people, undemanding job
-
○
-
Victim as Vehicle:
Offender is tragic hero, angry about self and his lot in life.
○
Assault regain him his rightful place in life/world.
○
High desire for control:
Like FBI's spree murderer
-
May kill many in single bout
-
May also kill self "Samson syndrome"
-
○
Low desire for control:
FBI's organized offender
-
Killings more deliberate
-
Motivated by failed/lost relationship
-
Offender intelligent, good social skills
-
○
-
Victim as Person:
Offender sees victim as individual, tries to understand their experience, but has
no true empathy for them.
○
No distinction between high and low desire for control here.
○
-
Canter's Smallest Space Analysis
Canter looks at many crimes and looks at specific aspects of the behaviours involved in
these crimes.
E.g., whether trophies were taken, whether weapons used were kept or
destroyed, was the victim tortured, etc.
○
-
Each of these words describes a type of crime scene activities.
-
Looked at the extent to which these events occurred together - intercorrelations.
-
Uses SSA to display these characteristics in a meaningful way.
Displays on a graph the distance between different behaviours.
○
Graphical depiction of the commonalities that occur in the crime.
○
-
Canter looks at these and says what type of interpersonal narratives these clusters
better apply to.
-
Argument goes - if you know the different characteristics of the crime scene, you'll be
able to determine their interpersonal narratives.
But what does the narrative even tell you?
○
-
Hicks and Sales Critique (of Canter)
Lack of conceptual clarity:
Victim's role in themes unclear
○
Inconsistency within themes
○
-
Unverifiable assertions re: inner narratives:
We cannot see into the minds of offenders to see what there intropsychic
narrative is.
○
… so what is the point of using them if a) we cannot know them and b) cannot
use them to narrow down the pool of suspects.
○
-
SSA inappropriate:
No factors indicated by analysis
○
Themes not linked to offender characteristics
No evidence, not even by Canter
-
Therefore, they can't be used to identify the offender from a pool of
suspects.
-
○
Can themes even predict offender characteristics?
Can't predict
-
Might be able to confirm
-
○
-
Hickes and Sales (2006):
What evidence should be gathered? 1.
Focuses on the pieces of evidence that would lean into a profile.
-
How should we interpret evidence?2.
None of the current books tell you how to do this because they don't believe that you
can teach people that.
-
What are relationships between evidence and offender characteristics? 3.
Which offender characteristics important to investigating crime? 4.
What is it that we can tell about the offender from the evidence?
-
None of the current systems can do this except for the non-scientific approaches?
-
D.C. Beltway Sniper (2002)
People got out of the car to get gas and they'd be shot.
-
After 5 or 6 people were murdered, FBI released a profile.
-
Profile:
White male, 24 to 40 years old
○
Not a sniper; not military
○
Lives in or near community; no children
○
Firefighter or construction worer
○
Possible terrorist links
○
-
Offenders: John Allen Mohammed; Lee Malvo
African-American male, 42 and 17
○
Mohammed spent 16 years in the National Guard, Army expert marksman - just
as good as a sniper
○
No terrorist links, but did admire terrorist works
○
-
Profile was inaccurate and didn't lead to the capture.
-
Baton Rouge Serial Killer (2003)
FBI Profile:
White male, 25 to 35 years old.
○
Unsophisticated social outcast
○
Awkward around women; doesn’t get along with them
○
Impulsive, quick-tempereed
○
-
Offender: Derrick Todd Lee
African-American male, 34 years old
○
Smooth, charming, "casanova"
○
Out with different women each night
○
-
Atlanta Olympic Bomber (1996)
FBI Profile:
Single, white middle-class male
○
Intense interest in police work
○
-
Arrested Richard Jewell
Because he fit the profile and was a security guard at the venue
○
Sued the government for ruining his potential police career.
○
-
Offender: Eric Rudolph
Anti-abortion activist
○
Lower-class single-parent family
○
Strong hatred for government and authority - opposite of the interest in police
work
○
-
Unabomber (1978-1995)
Sent a variety of bombs to universities and airlines
-
2 different profiles developed - one wildly inaccurate
-
FBI Profile (Tafoya):
White male, early 50s
○
University graduate with advanced degree
○
Background in science, math or engineering
○
Strongly anti-technology
○
-
Offender: Ted Koczynski
Not found because of the profile.
○
He wrote a manifesto in handwriting - his own brother recognized the
handwriting as his and turned him in.
○
White male, early 50s
○
M.S. and Ph.D. in math
○
Living alone in Montana cabin
○
Strongly anti-technology (correct, but not a great leap because he admitted it
this in the manifesto)
○
-
Paul Bernardo (1987-1992)
FBI profile (for the rapes):
Single white male, 18-25 years old
○
Lives in Scarborough area
○
High school education only
○
Angry, disparaging toward women
○
Spotty work record
○
-
Offender: Paul Bernardo
Single, white male
○
18 at first rape
○
Living in Scarborough area
○
University graduate
○
Gainfully employed (Price Waterhouse, Amway)
○
-
Again, he was not caught on the basis of the profile.
-
Bernardo Likeness
Rare case where the facial reconstruction of the offender was strikingly accurate
○
-
Criticisms of Profiling
Personality models lack empirical support. 1.
Assumes primacy of personality over situation.2.
Big debate in psychology for a long time - to what extent is behaviour determined of
personality vs situation.
-
Behaviour is usually the result of the demands (i.e., the demands of the crime)
-
Profiles are vague and ambiguous - fit many people. 3.
E.g., usually white males
-
Many aspects of profile are common sense4.
E.g., "Male, 18-30" - well, duh … most criminals are males between the ages of 18 and
30.
-
Kocsis et al (2000):
Studied 6 groups of Australian profilers:
35 police officers
○
30 psychologists
○
30 2nd year university students
○
20 psychics
○
5 profilers
○
23 university economics students (control)
○
-
Given details of two solved homicides:
Crime scene report, forensic biologist's report, forensic entomologist's report,
pathologist's postmortem report, info on victim, plan and photos of crime scene
and victim.
○
-
Then asks them a serious of questions about the offender's characteristics.
Given a score (# of items correct about offender)
○
Psychics and control groups don't do as well as others for physical characteristics,
but everyone else does equally as well.
Profilers BARELY better.
-
○
For behavioural characteristics, police do the worst and psychologists do the
best.
○
For offender cognitive characteristics, profilers significantly better (but only
about 1.5 pts higher)
○
For personality characteristics, psychologists do the best, but again, it's barely
better.
○
-
Point is … even though there might be a significant difference between profilers and
other groups, they're meaningless because they're small - the experts should be MUCH
better, but they're only slightly better.
-
-(see graphs)
Alison et al (2003):
Examined 880 statements in 21 profiles from US, UK (including FBI)
Majority of profile statements about offense and already known to investigators.
○
55% of offender descriptions not verifiable (i.e., emotions, motives, thoughts,
etc.)
Don't help you because they don't identify people …
-
○
24% of offender descriptions ambiguous (e.g., poor heterosexual social skills).
What do they even mean?
-
Don't help you …
-
○
6% of offender descriptions contain incompatible alterntives (e.g., unemployed
or manual job)
○
-
Kocsis (2004):
Studied 6 groups of Australian profilers:
3 profilers
○
13 police detectives
○
12 fire investigators
○
21 2nd year B.S. students (chemistry)
○
47 community college students (control)
○
-
Given details re: serial arsonist (13 fires)
Crime scene schematics, incident reports, forensic reports, 12 captioned photos.
○
-
Asked about the characteristics of the offender.
Again, marked # correct.
○
-
-(see graphs)
Consistent with previous study … profilers should be way better, but they're not.
-
So if profiling is a reputable career and something you can be an expert in, why are they not
much better than everyone else at profiling?
Copson (1995)
Surveyed 184 detectives who had used profiles regarding the usefulness of profiles.
-
83% said profiles "useful"
61% - said profiles helped them understand the case
○
52% - said profiles supported their own opinions
○
Basically, they helped us validate our own personal views, that's probably why
law enforcement continues to rely on profiles.
○
-
What percentage of profiles actually helped you identify the offender?
2.7%
○
-
Recall: Dr. Thomas Bond (1888)
In 1988, John Douglas produced his own profile of Jack the Ripper, 100 years later.
-
Couldn't come up with anything better than Dr. Bond's,
Even though 100 years have passed and John Douglas' expertise.
○
-
Where are we now?
" … this research confirms the perceptions of those who have concluded the criminal
profiling relies on weak standards of proof and that profilers do not decisively
outperform other groups when predicting the characteristics of an unknown criminal …
profiling appears at this juncture to be an extraneous and redundant technique for use
in criminal investigations. Criminal profiling will persist as a pseudoscientific technique
until such time as empirical and reproducible studies are conducted on the abilities of
large groups of active profilers to predict, with more precision and greater magnitude,
the characteristics of offenders."
Snook et al (2008)
-
-
The time is not coming soon when this will be a set science …
-
Note: Hicks and Sales, based on this
critique would argue that this is not
the basis on which we should turn
profiling into a scientific enterprise.
Note: Basically, Hicks and Sales don't
have a system, but they think this is what
a good system should have, what it
should indicate
Criminal Profiling
Tuesday, March 6, 2018
9:56 AM
Receives the most media, tv, movie, etc. attention but is the most insignificant part of
what forensic psychologists do.
-
History of Criminal Profiling
Begins in 1841 with a publication by Edgar Allen Poe, "Murders in the Rue Morgue"
Displays French detective solving murders in the Rue Morgue
○
Shown to do things consistently with how offender profilers work now.
○
-
Sherlock Holmes by Arthur Conan Doyle
"A Study in Scarlet" - 1887
○
Many films made (e.g., Basil Rathmon, Jeremy Brett, etc.)
○
-
First real-life profile (1888): Dr. Thomas Bond
Profile of Jack the Ripper
○
Some of these inferences are just common sense, but some are pretty profound.
○
-
Purposes of Profiling
Provide offender characteristics.
Everyone agrees on this purpose.
-
1.
Help understand the crime scene.
Helps us understand the crime, how/why things were done.
-
2.
Provide leads for investigators.
What kind of person would've done this? Demographic information?
-
3.
Narrow pool of viable subjects.4.
Prioritizes investigation of subjects. 5.
Risk of offender escalation?
E.g., "Criminals like this will often start killing at a faster rate."
-
6.
Evaluate suspect possession.
Might help understand the relevance of pieces of possessions.
-
7.
Develop interrogation strategies. 8.
Show links between crimes.
E.g., Jack the Ripper
Even today, people debate how many people Jack the Ripper actually
killed.
§
Agree on 5, but there are 3 other people that may have also been killed by
him.
§
-
9.
Supportive trial testimony.
10.
Note: not all profiles agree with ALL of these purposes (usually the first 5 are heavily
agreed upon)
Challenges of Profiling
Turning crime scene info into description of offender.
This is extremely challenging.
-
We don't have a lot of data about the associations between crime scene
information and offender characteristics.
-
It’s the small removals/changes of area at crime scene that need to be
interpreted.
-
1.
Personality tests can't do this.
We have a lot of different instruments (i.e., personality measures) that can give
you a good description of a person's general characteristics, but they can't
describe someone in as much detail as offender profiles try to provide.
-
If years of psychology research on personality cannot do this, you can imagine
how difficult it would be for one person to achieve this.
-
2.
Unscientific - intuition, experience.
"Gut feeling"
-
They cannot tell you how exactly they profile people, how exactly they use
information to develop profiles because they don't even know how they do it.
-
They rely on intuition
-
This is why television programs display profilers as having "psychic" qualities.
-
3.
Not well evaluated - utility unclear.
Criminal profilers don't reveal their methods or profiles to the public.
Especially those done by the FBI
§
-
Consider Criminal Minds - profile is given verbally to the law enforcement, no
documentation released.
-
Accuracy is questioned heavily because of this.
Therefore, the usefulness is also questioned.
§
-
4.
Note: rape, murder and arson are most commonly profiled cases; sometimes high profile
robbery.
Basic Assumptions of Profiling
Behavioural Consistency
Offenders behave similarly across offenses.
An offender will behave in the same way across all of his crimes.
§
-
Cross-offense similarities must be infrequent to be of use.
But there has to be something the individual does similarly in all of his
crimes that is different from regular crimes of the same nature.
§
-
Bennell and Carter (2002): examined 4 characteristics for 2 crimes from each of
43 serial commercial burglars.
Inter-crime distance most closely similar in pairs of burglaries.
Only thing that distinguished crimes of the same offender.□
§
Type of business, method of entry, property stolen, intra-crime behaviours,
linked but not strongly enough.
These did not differentiate. □
§
-
Barteman and Salfati (2007): examined 35 behaviours of 90 offenders in 450
serial homicide cases.
Some things did differentiate … bringing crime kit, destroying evidence
both consistent across offenses and infrequent overall.
§
Bringing weapon, restraining victim both consistent across offenses, but
very common across all offenses … doesn't distinguish.
§
-
1.
Behavioural Differentiation
Offenders differ from one another in their patterns of inter-crime similarity.
Different offenders behave differently from one another across different
crimes.
§
-
Little evidence, but plausible.
-
2.
Homology
Offenders with similar crime behaviours will have similar characteristics.
-
Something about all of the individuals behaving similarly in crimes that is the
same.
-
Little good evidence supporting this.
-
Mokros and Allison (2002):
Rapists: are those who offend similarly also similar in personal
characteristics?
No positive correlations between crime similarily and similarity in
age, employment, ethnicity, criminal records, etc.
□
§
-
Doan and Snook (2008): Used typologies to classify 87 arson cases, 175
robberies. Compared similarity of offenders with similar case types.
Homology assumption violated in 56% of arson types and 67% of robbery
types.
§
-
If this doesn't work, the profile kind of goes to shit.
-
3.
Dr. James Brussel: Profile of New York Bomber (1956)
First modern profile
-
Number of bombs sent around New York City; sent letters explaining why they were
sending the bombs.
-
Brussel provided a profile:
"Male, former employee of Consolidated Edison, injured while working there so
seeking revenge; paranoid, 50 years old, neat and meticulous persona, foreign
background, some formal education, unmarried, living with female relatives, but
not mother who probably died when he was young. Upon capture, he will be
wearing a buttoned up double-breasted jacket."
-
-
Searched ConEd's records based on profile.
When they arrested him he was wearing a buttoned up double-breasted jacket …
-
-
FBI: Behavioural Analysis (Sciences) Unit
Founded in 1972
-
Howard Tenet - 1st Director
-
Robert Ressler, Roy Hazelwood and John Douglas
John Douglas is the biggest name in criminal profiling.
-
Huge publication history for all of them.
-
-
John Douglas is the inspiration for the Agent Jack Crawford in Silence of the Lambs;
author interviewed him.
-
Agent Rossi (Criminal Minds) also based off of John Douglas.
-
-(see slide for chart)
Crime Classification Manual
Describes the process by which the FBI constructs an offender profile.
-
Also has a crime classification system.
-
Offender Profile Stages:
Profiling Inputs
Crime scene information
-
Victimology
Has become very popular□
-
Forensic information
-
Police reports (i.e., time of day, people around, neighbourhood description,
etc.)
-
Photographs
-
1.
Decision Process Model
Homicide type and style (e.g., mass murder, spree killing, serial murder,
etc.)
-
Primary intent (e.g., revenge killing)
-
Victim, offender risk (risk of being detected and caught)
-
Escalation risk
-
Time, location factors
-
2.
Crime Assessment
Crime reconstruction
-
Crime classification: organized/disorganized
-
Staging
-
Motivation
-
Crime scene dynamics (interaction between things at time of the crime)
-
3.
Criminal Profile
Demographics
-
Physical characteristics
-
Habits
-
Beliefs and values (very common in rape cases)
-
Recommendations
-
4.
Investigation5.
Apprehension6.
-
Types of Murder
Mass Murder
Double vs triple homicide
○
4 + victims killed in the same place at the same time.
○
-
Murder Spree
No cooling off period between killings
○
Murdered in different locations within a relatively short time frame.
○
-
Serial Murder
Multiple victims
○
Cooling off period between victims
○
Typically in different locations
○
-
Organized-Disorganized Distinction
Made by FBI; distinction based on interviews with convenience sample of 36 serial
murders.
-
Interviews unstructured, different for each interviewee.
-
No subsequent validation with another sample.
-
Organized: someone who has planned the crime, taken the weapons/tools with them,
cleaned up afterward.
Profile would expect that the individual is organized IRL too.
○
-
Disorganized: opposite of organized
-
In 100 US serial murder cases, no indication that organized features correlated with
each other.
-
NOTE: even though they give you characteristics of a profile, they never provide a
system for profiling.
Not a science, but an art.
○
It's a special skill people develop with experience and intuition.
○
-
Holmes and Holmes: Offender Typologies
Disorganized Asocial/ Organized Nonsocial
-
Serial Murderers (defined based on motives)
Spatial mobility - do they kill within a small geographic area or a large one?
○
Visionary - are they motivated by vision?; usually mentall disordered, usually
found not guilty because of illness.
○
Missionary - sets out to kill people of a certain type (e.g., sex workers)
○
Hedonistic - murders for personal pleasure or personal gain.
○
Power and control - self-explanatory
○
-
Rapists (older categorization system, borrowed from other people)
Power reassurance - rape to reassure themselves about their power.
○
Anger retaliation - rape to get back at someone; generally person symbolically
representing the person who wronged them.
○
Exploitative - rape to take whatever they want
○
Sadistic - often followed by murder; pleasure in inflicting pain (violent)
○
-
Child Molesters
Situational - will have sex with children if the situation presents itself, but don't
need to have sex with children to get off.
○
Preferential - prefer children sexually; pedophiles
○
-
Arsonists
Jealousy - set fires to destroy things they want, but can't have; these people
don't stay to watch fire.
○
Would-be hero - fireman, policemen, etc.; start fires and stay to watch so they're
there to help and get the fame/glory of helping.
○
Excited fire-starter - find fires arousing (re: Freud); stay at fire to watch
○
Pyromaniac - sexually aroused by starting fires; stay at fire to watch
○
-
Geography
Trying to identify the suspect based on looking at the distance between victims.
○
-
Victim Profiling
Recent focus in research and actual profiling practice
○
Importance of understanding the victim to identify the offenders.
○
-
(see slide for chart)
Keppel and Walter: Rape/Murder Offender Typologies
(see slide for chart)
Took rape classifications and applied them to rape and murder.
Often rape and murder overlap
○
-
Note: don't need to remember all of the specific profiles; they're unclear and the
research doesn't show that they're accurate
-
Power Assertive:
○(see slide for description)
-
Power Reassurance:
○(see slide for description)
-
Anger Retaliatory:
○(see slide for description)
-
Anger and Excitation:
○(see slide for description)
-
Keppel and Walter (1999):
Power assertive is the most common
○
Next is Anger Retaliatory, then Power Reassurance
○
Anger Excitation is the least common
○
-
Brent Turvey
Publishes texts on criminal profiling
-
Academy of Behavioural Profiling
Pretty dead organization right now.
○
His company is called Forensic Solutions
Designed to train profilers
○
○
-
Started the Journal of Behavioural Profiling
Stopped publishing ~ 10 years ago
○
-
There just isn't enough literature on criminal profiling.
-
Turvey: Inductive vs Deductive Profiling
Inductive Profiling:
Starts with statistical generalizations ( e.g., most rapists are single males aged 18
to 30).
Based off of a bunch of data collected from things like FBI interviews, etc.
-
○
We then apply these generalizations to individual cases.
○
Problem: these general conclusions do not logically flow.
○
-
Deductive Profiling:
Starts with specific crime scene evidence (e.g., there are motorcycle tracks at the
crime scene.)
○
Conclusion: if they belong to the rapist, then he has access to a motorcycle.
○
These conclusions are often more logical because you start with the crime and
trace evidence back to the criminal.
○
-
Turvey wants to replace the common inductive profiling with deductive profiling
because it has more of a logical basis.
But you still cannot be certain that conclusions made in deductive profiling are
true.
○
-
Turvey Deductive Profiling Model: Four Components of Profiling
Forensic and Behavioural Evidence1.
Every investigation collects information about the crime.
-
E.g., witness and victim statements, crime photos, wound pattern and blood spatter
analysis, ballistics evidence, other forensic evidence.
-
Victimology2.
Turvey pushed this new idea a lot
-
E.g., physical characteristics, habits, lifestyle, relationships, risk level (e.g., do they do
things that would make them more susceptible to criminal activity)
-
Can these things tell us anything about the offender?
-
Crime Scene Characteristics3.
We consider what happened and in what order.
-
g., method of attack, nature of sexual or violent acts, verbal behaivour, precautionary
acts (i.e., did they clean up the crime scene? Did they clean up the body?)
-
Deduction of offender characteristics 4.
There is no system of offender profiling that tells you how to do this; intuitive/artistic.
-
Turvey's Behaviour-Motivational Typology
Applied this to all crimes; argues that these motives are not specific to certain crimes,
but are widely applicable to all crimes.
-
Description more of the crime characteristics
-
Includes:
Power reassurance
○
Power assertive
○
Anger retaliatory
○
Anger excitation
○
Profit (i.e., like hedonist)
○
-
Hickes and Sales (2006): Criticism of Non-scientific Systems of Profiling
Lack of goals and standards.1.
A lot of these systems do not outline the purpose/goals of the profile.
-
No agreement on goals of profiling
-
No agreement on crimes suitable for profiling.
Everyone agrees that you can profile murderers.
○
Most agree that you can profile rapists.
○
Many, but not all, agree that you can profile arsonists.
○
Others are very mixed.
○
-
No standards to evaluate the success of profiling.
Not validation of the accuracy or usefulness of the profiling.
○
Few attempts to evaluate.
○
-
Use of unclear terms and definitions.2.
Serial killers defined differently.
Some systems defines serial killers differenty than the FBI.
○
-
Modus operandi (way of operating) sometimes fixed, sometimes changes from crime to
crime - depends on the system whether it can change or not.
-
No description of how to determine a criminal "signature."
One area where this is clear defined is in bombing - bombers construct bombs in
their own, specific way.
○
-
Many other unclear terms …
-
Misuse of typologies3.
Inappropriate use of typologies
E.g., Offender may fall into several categories; fail to meet full criteria for any,
but these systems for them into a box.
○
-
Inconsistency within typologies
E.g., H&H: use pedophile and child molester both as synonyms and as different
There is a distinction to be made between these two terms.
-
○
E.g., in one category, an offender my be both sloppy and cunning?
○
E.g., in one category, an offender may be described has lacking empathy, but
later their behaviour is described has being the result of the loss of a loved one.
○
-
Overlapping categories
Many categories, same characteristics (e.g., single male, use of weapons, age,
etc.)
○
Might be identical in several different categories, so how do you discriminate?
You can't.
○
-
Limited value of typologies
Don't help police narrow suspect list; more like a horoscope than a useful
description.
○
-
Rely on intuition, professional knowledge4.
Not based on scientific information
-
Huge concern
-
Lack of clear procedures5.
NONE of the systems or authors provide a clear list of instructions to construct a
profile.
-
They can't tell you because it's based on a feeling or intuition.
-
Little evidence of investigative value6.
Very few attempts have been made to consider the usefulness of profiles.
-
Studies that have been done show that they're not that useful at all.
-
Canter's Scientific Model (wanted to put profiling on scientific footing; said that there were
questions that needed to be answered in order to do this.)
Behavioural Salience1.
What are the behavioural features of the crime that might help identify the offender?
-
Distinguishing between offenders2.
How do we indiciate differences between offedners and crimes?
-
Inferring characteristics3.
What inferences can we make about offenders that would help identify him?
-
What information will lead us to characteristics of a particular offender?
-
Linking offenses4.
How can we determine whether same offender committed a series of crimes?
-
E.g., goes all the way back to Jack the Ripper; 5 murders that have been confirmed as
being performed by Jack the Ripper, but there were 2-3 others that we're still debating
on.
-
Theoretical Links: Action -> Offender (Canter)
How do we see the relationship between crime scene characteristics and the offender?
-
Psychodynamic Typologies
Certain psychodynamic characteristics, Freudian types will lead to certain types
of offense behaviours.
○
Maybe we can use that differentiation of freudian psychological types to identify
a crime scene behaviours as being part of a certain individuals work.
○
-
Personality Differences
E.g., introverts vs extroverts
○
People with different personality characteristics might perform crimes
differently.
○
-
Career Routes
Most criminals do not engage in the same types of crimes for their entire criminal
careers.
○
Very often murders and rapists have committed other crimes first.
○
Maybe we can distinguish individuals based on the route they took on their way
to the current set of crimes.
E.g., did they come from assaultive crimes? Victimless crimes? Robbery?
Etc.
-
○
-
Socioeconomic Subgroups
E.g., blue collar crime vs white collar crime
○
-
Interpersonal Narratives
The internal stories that the offender is working, that they're trying to realize in
life.
○
How can we identify this and link it to crimes and crime scene characteristics?
○
This is the only possibility accepted by Canter.
Thinks we should be looking at the role the victim plays within the
offender's interpersonal narrative.
-
Applies to rape and murder typically because they have identifiable
victims.
-
○
-
Interpersonal Narratives: Role of the Victim
Victim as Object:
Offender has no feeling or connection with victim, who does not play any active
role in the assault.
○
Victim is just a piece of furniture, doesn't stand for anything.
○
High desire for control:
May involve mutilation
-
Low intelligence, poor contact with reality
-
Eccentric
-
From dysfunctional background
-
○
Low desire for control:
Sexual component prominent
-
Murder not major goal
-
Will not understand seriousness of crime
-
Probably little contact with people, undemanding job
-
○
-
Victim as Vehicle:
Offender is tragic hero, angry about self and his lot in life.
○
Assault regain him his rightful place in life/world.
○
High desire for control:
Like FBI's spree murderer
-
May kill many in single bout
-
May also kill self "Samson syndrome"
-
○
Low desire for control:
FBI's organized offender
-
Killings more deliberate
-
Motivated by failed/lost relationship
-
Offender intelligent, good social skills
-
○
-
Victim as Person:
Offender sees victim as individual, tries to understand their experience, but has
no true empathy for them.
○
No distinction between high and low desire for control here.
○
-
Canter's Smallest Space Analysis
Canter looks at many crimes and looks at specific aspects of the behaviours involved in
these crimes.
E.g., whether trophies were taken, whether weapons used were kept or
destroyed, was the victim tortured, etc.
○
-
Each of these words describes a type of crime scene activities.
-
Looked at the extent to which these events occurred together - intercorrelations.
-
Uses SSA to display these characteristics in a meaningful way.
Displays on a graph the distance between different behaviours.
○
Graphical depiction of the commonalities that occur in the crime.
○
-
Canter looks at these and says what type of interpersonal narratives these clusters
better apply to.
-
Argument goes - if you know the different characteristics of the crime scene, you'll be
able to determine their interpersonal narratives.
But what does the narrative even tell you?
○
-
Hicks and Sales Critique (of Canter)
Lack of conceptual clarity:
Victim's role in themes unclear
○
Inconsistency within themes
○
-
Unverifiable assertions re: inner narratives:
We cannot see into the minds of offenders to see what there intropsychic
narrative is.
○
… so what is the point of using them if a) we cannot know them and b) cannot
use them to narrow down the pool of suspects.
○
-
SSA inappropriate:
No factors indicated by analysis
○
Themes not linked to offender characteristics
No evidence, not even by Canter
-
Therefore, they can't be used to identify the offender from a pool of
suspects.
-
○
Can themes even predict offender characteristics?
Can't predict
-
Might be able to confirm
-
○
-
Hickes and Sales (2006):
What evidence should be gathered? 1.
Focuses on the pieces of evidence that would lean into a profile.
-
How should we interpret evidence?2.
None of the current books tell you how to do this because they don't believe that you
can teach people that.
-
What are relationships between evidence and offender characteristics? 3.
Which offender characteristics important to investigating crime? 4.
What is it that we can tell about the offender from the evidence?
-
None of the current systems can do this except for the non-scientific approaches?
-
D.C. Beltway Sniper (2002)
People got out of the car to get gas and they'd be shot.
-
After 5 or 6 people were murdered, FBI released a profile.
-
Profile:
White male, 24 to 40 years old
○
Not a sniper; not military
○
Lives in or near community; no children
○
Firefighter or construction worer
○
Possible terrorist links
○
-
Offenders: John Allen Mohammed; Lee Malvo
African-American male, 42 and 17
○
Mohammed spent 16 years in the National Guard, Army expert marksman - just
as good as a sniper
○
No terrorist links, but did admire terrorist works
○
-
Profile was inaccurate and didn't lead to the capture.
-
Baton Rouge Serial Killer (2003)
FBI Profile:
White male, 25 to 35 years old.
○
Unsophisticated social outcast
○
Awkward around women; doesn’t get along with them
○
Impulsive, quick-tempereed
○
-
Offender: Derrick Todd Lee
African-American male, 34 years old
○
Smooth, charming, "casanova"
○
Out with different women each night
○
-
Atlanta Olympic Bomber (1996)
FBI Profile:
Single, white middle-class male
○
Intense interest in police work
○
-
Arrested Richard Jewell
Because he fit the profile and was a security guard at the venue
○
Sued the government for ruining his potential police career.
○
-
Offender: Eric Rudolph
Anti-abortion activist
○
Lower-class single-parent family
○
Strong hatred for government and authority - opposite of the interest in police
work
○
-
Unabomber (1978-1995)
Sent a variety of bombs to universities and airlines
-
2 different profiles developed - one wildly inaccurate
-
FBI Profile (Tafoya):
White male, early 50s
○
University graduate with advanced degree
○
Background in science, math or engineering
○
Strongly anti-technology
○
-
Offender: Ted Koczynski
Not found because of the profile.
○
He wrote a manifesto in handwriting - his own brother recognized the
handwriting as his and turned him in.
○
White male, early 50s
○
M.S. and Ph.D. in math
○
Living alone in Montana cabin
○
Strongly anti-technology (correct, but not a great leap because he admitted it
this in the manifesto)
○
-
Paul Bernardo (1987-1992)
FBI profile (for the rapes):
Single white male, 18-25 years old
○
Lives in Scarborough area
○
High school education only
○
Angry, disparaging toward women
○
Spotty work record
○
-
Offender: Paul Bernardo
Single, white male
○
18 at first rape
○
Living in Scarborough area
○
University graduate
○
Gainfully employed (Price Waterhouse, Amway)
○
-
Again, he was not caught on the basis of the profile.
-
Bernardo Likeness
Rare case where the facial reconstruction of the offender was strikingly accurate
○
-
Criticisms of Profiling
Personality models lack empirical support. 1.
Assumes primacy of personality over situation.2.
Big debate in psychology for a long time - to what extent is behaviour determined of
personality vs situation.
-
Behaviour is usually the result of the demands (i.e., the demands of the crime)
-
Profiles are vague and ambiguous - fit many people. 3.
E.g., usually white males
-
Many aspects of profile are common sense4.
E.g., "Male, 18-30" - well, duh … most criminals are males between the ages of 18 and
30.
-
Kocsis et al (2000):
Studied 6 groups of Australian profilers:
35 police officers
○
30 psychologists
○
30 2nd year university students
○
20 psychics
○
5 profilers
○
23 university economics students (control)
○
-
Given details of two solved homicides:
Crime scene report, forensic biologist's report, forensic entomologist's report,
pathologist's postmortem report, info on victim, plan and photos of crime scene
and victim.
○
-
Then asks them a serious of questions about the offender's characteristics.
Given a score (# of items correct about offender)
○
Psychics and control groups don't do as well as others for physical characteristics,
but everyone else does equally as well.
Profilers BARELY better.
-
○
For behavioural characteristics, police do the worst and psychologists do the
best.
○
For offender cognitive characteristics, profilers significantly better (but only
about 1.5 pts higher)
○
For personality characteristics, psychologists do the best, but again, it's barely
better.
○
-
Point is … even though there might be a significant difference between profilers and
other groups, they're meaningless because they're small - the experts should be MUCH
better, but they're only slightly better.
-
-(see graphs)
Alison et al (2003):
Examined 880 statements in 21 profiles from US, UK (including FBI)
Majority of profile statements about offense and already known to investigators.
○
55% of offender descriptions not verifiable (i.e., emotions, motives, thoughts,
etc.)
Don't help you because they don't identify people …
-
○
24% of offender descriptions ambiguous (e.g., poor heterosexual social skills).
What do they even mean?
-
Don't help you …
-
○
6% of offender descriptions contain incompatible alterntives (e.g., unemployed
or manual job)
○
-
Kocsis (2004):
Studied 6 groups of Australian profilers:
3 profilers
○
13 police detectives
○
12 fire investigators
○
21 2nd year B.S. students (chemistry)
○
47 community college students (control)
○
-
Given details re: serial arsonist (13 fires)
Crime scene schematics, incident reports, forensic reports, 12 captioned photos.
○
-
Asked about the characteristics of the offender.
Again, marked # correct.
○
-
-(see graphs)
Consistent with previous study … profilers should be way better, but they're not.
-
So if profiling is a reputable career and something you can be an expert in, why are they not
much better than everyone else at profiling?
Copson (1995)
Surveyed 184 detectives who had used profiles regarding the usefulness of profiles.
-
83% said profiles "useful"
61% - said profiles helped them understand the case
○
52% - said profiles supported their own opinions
○
Basically, they helped us validate our own personal views, that's probably why
law enforcement continues to rely on profiles.
○
-
What percentage of profiles actually helped you identify the offender?
2.7%
○
-
Recall: Dr. Thomas Bond (1888)
In 1988, John Douglas produced his own profile of Jack the Ripper, 100 years later.
-
Couldn't come up with anything better than Dr. Bond's,
Even though 100 years have passed and John Douglas' expertise.
○
-
Where are we now?
" … this research confirms the perceptions of those who have concluded the criminal
profiling relies on weak standards of proof and that profilers do not decisively
outperform other groups when predicting the characteristics of an unknown criminal …
profiling appears at this juncture to be an extraneous and redundant technique for use
in criminal investigations. Criminal profiling will persist as a pseudoscientific technique
until such time as empirical and reproducible studies are conducted on the abilities of
large groups of active profilers to predict, with more precision and greater magnitude,
the characteristics of offenders."
Snook et al (2008)
-
-
The time is not coming soon when this will be a set science …
-
Note: Hicks and Sales, based on this
critique would argue that this is not
the basis on which we should turn
profiling into a scientific enterprise.
Note: Basically, Hicks and Sales don't
have a system, but they think this is what
a good system should have, what it
should indicate
Criminal Profiling
Tuesday, March 6, 2018 9:56 AM
