Offender Profiling Notes

Offender Profiling

  • Forensic psychology
  • Dr Lu Lemanski
  • Department of Security and Crime Science

Lecture 2 Topics

  • Introduction to Offender Profiling
  • Clinical Approach
  • FBI Method (‘Criminal Investigative Analysis’)
  • Investigative Psychology
  • Geoprofiling
  • Evaluating Profiling

Introduction to Offender Profiling

  • How to solve a crime when witness identification, perpetrator confession, or inferences from physical evidence are missing?

What is Offender Profiling?

  • Interpretation of available data to indicate the type of person who may have committed the crime.
  • Profiling helps reduce the pool of suspects and focus inquiries.
  • Deriving an offender’s likely characteristics from the way they committed a particular crime (Blau, 1974).
  • Often applied to serial offenses judged to be linked.
  • Traditionally applied to serious crime types.
  • Profiles can include a wide range of information, depending on the extent/type of evidence present.

The Clinical Approach

  • Generic approach to profiling, similar to a ‘diagnostic evaluation.’
  • Expert provides opinions and assumptions about offenders.
  • Based on previous experience.
  • Role is to provide ‘insight’ about the case/offender.
  • Focuses on emotions, desires, and psychopathology.
  • Ad hoc.

A Clinical Profile: Jack the Ripper

  • ‘Jack the Ripper’ was a serial murderer targeting prostitutes in London.
  • Dr. Thomas Bond, a police surgeon, autopsied one of the Ripper victims.
  • He considered the nature of the victim’s injuries, e.g., severe pre-/post-mortem injuries of a sexual nature.
  • Believed this indicated ‘rage’ and ‘hatred of women.’
  • Dr. Bond worked up a profile based on his experience of the sort of person who would have committed such an act.
  • The murderer must have been a man of physical strength, great coolness, and daring; likely solitary and eccentric in his habits, without regular occupation, but with some income or pension, and middle-aged.

Features of a Clinical Profile

  • Drawing inferences based on clinical experience + interpretation of the crime scene.
  • Features + patterns of crime are judged to provide an insight into individual offender attributes and/or motivation, e.g.:
    • Tying up victim > indicates a need for control.
    • Targeting prostitutes > implies a hatred toward women.
  • Clinical profilers usually have a psychology background.
  • Draw on clinical experiences + expertise, e.g., gained in FMH settings.
  • Based on hands-on expertise.
  • Process is woolly.

General Process in Clinical Profiling

  1. Collate available information:
    • Victim
    • Crime scene
    • Circumstances
  2. Identify crime type at a fundamental level, e.g., ‘murder.’
  3. Identify ‘character’ of crime:
    • Weapons
    • Nature of victim injuries
    • Extent of victim injuries
  4. Understand what happened:
    • Is there evidence of emotion, desires, moods, or psychopathology in the case material?
    • Reconstruction
  5. Apply psychological theory:
    • Psychological literature is consulted to understand motive + suggest offender characteristics.

Clinical Approach Limitations

  • Subjectivity is an issue!
  • Crime scene interpretation is highly subjective.
  • Inferences are highly subjective à based on individual experiences/judgements.
  • Variation by profiler.
  • Unsystematic à difficult to replicate.

Clinical Approach: Movie Example - Silence of the Lambs

  • Clarice Starling profiles serial killer ‘Buffalo Bill’ using crime scene information and expertise/experience from forensic psychiatrist Hannibal Lecter.
  • Key takeaway: Lack of empirical evidence - profiling from experience and intuition

The FBI Method

  • AKA ‘Criminal Investigative Analysis.’
  • Developed at FBI BSU in 1970s.
  • Applied to unsolved crimes.
  • Aim à psychological, behavioral, demographic picture of the type of person likely to have committed the offense.
  • Moved away from diagnostic terminology used in the clinical approach.
  • More structured, based on real cases.
  • BSU developed an evidence-base of offenders.
  • Structured and systematic process (unlike the clinical method).
  • Rooted in evidence à inferences about the offender grounded in what we know about offenders in previous similar cases.
  • But still emphasized the application of investigative expertise gained from experience.

FBI Process

  1. Data assimilation
    • Time of death
    • Path reports
    • Search of ‘signatures’
    • Signatures = distinctive feature(s) of Modus Operandi (MO) carried out to satisfy the needs of the offender
  2. Crime scene classification
    • Based on ‘typologies’ e.g. ‘organised/disorganised’ ‘serial murderer’
  3. Crime scene reconstruction
    • How the crime unfolded
    • Sequence of events
    • Clarify the MO
  4. Profile generation
    • Includes hypotheses re:
      • Demographics
      • Lifestyle
      • Habits
      • Behaviours
      • Personality etc

FBI BSU Individuals

  • John Douglas (Holden Ford)
  • Robert Ressler (Bill Tench)
  • Ann Burgess (Wendy Carr)

Organised/Disorganised Typology

  • Organised crime scene:

    • Victim appears to have been targeted
    • Body is concealed
    • Murder and disposal site at different locations
    • Going equipped (weapons)
    • Neat crime scene
    • Little physical (incriminating) evidence at crime scene

  • Disorganised crime scene:

    • Seemingly random target
    • Body in plain view
    • Murder and disposal site at the same location or nearby
    • Murder committed using items available at the scene
    • Chaotic crime scene
    • Plentiful physical evidence at crime scene

Characteristics Inferred

  • Organised offender:

    • High IQ
    • Socially adept
    • Confident
    • Sexually competent
    • Educated
    • Lives with partner
    • Resourced (money, vehicle)

  • Disorganised offender:

    • Low IQ
    • Socially inadequate
    • Shy
    • Rarely dates
    • College drop out
    • Lives alone
    • Lacking resources

Issues with Organised/Disorganised Typology

  • The organised/disorganised typology was based on interviews with a sample of incarcerated offenders.
  • Interviews were ad hoc and depended on the interviewee.
  • The sample of killers interviewed was neither random nor large.
  • Based on retrospective accounts (memory).
  • Divided the sample into disorganised and organised (how?).
  • Never set out to test the discriminatory power of the dichotomy on others.
  • How accurate and reliable is the organised /disorganised offender typology?

Myth or Model? Canter et al (1994)

  • Lots of issues with the dichotomy
  • Prior lack of empirical testing
  • What if organised actions often co- occur with disorganised ones?
  • First step in testing: can crimes be reliably assigned to 1 of the 2 categories?
  • Do features of so-called ‘organised’ crime scenes form different patterns to ‘disorganised’ scenes?

Canter et al (1994) Study

  • Similar data to original FBI.
  • ID 39 elements of 100 serial sexual homicides.
  • Committed by 100 murderers (1 offence of each offender).
  • = analysis of Organised + Disorganised crime scene and offender characteristics explicitly mentioned in the FBI Crime Classification Manual.
  • To test whether org + disorg behaviours co-occur (i.e. are patterned).

The study also showed

  • ~2 x number of so-called ‘disorganised’ indicators
  • Disorganised behaviours are rarer
  • Don’t really distinguish organised vs disorganised offenders…
  • But à ‘disorganised’ indicators can be used to differentiate killers based on their interaction with the victim…
  • Empirical research can be harnessed to make profiling more scientific

Canter et al Identified 4 distinct behavioural clusters

  • Indicative of different crime scene types

Org/Disorg typology: Limitations

  • Originally derived from a small sample (n = 36)
  • Sample was opportunistic and may not be representative
  • Based on information elicited from convicted murderers
  • Assumes behavioural consistency over time and neglects situational influences

Assumptions

  • Behavioural consistency
  • Homology

Behavioural Consistency

  • Offenders must exhibit sufficient behavioral consistency across crimes.
  • Important for crime linkage and profiling.
  • For profiling, we must also be able to differentiate crimes committed by different offenders (distinctiveness).
  • Evidence suggests some degree of support for this assumption.
  • But behavioral consistency and distinctiveness does not hold for all offenders, across all situations.

Homology

  • Offenders who exhibit similarity in their crime scene behaviors must also share similar characteristics.
  • No or little evidence in support of homology assumption.
  • Underlying assumption is that crime scene behavior is an expression of psychological attributes…
  • BUT behavior is also situationally determined. e.g. Victim resistance.
  • Much offender profiling downplays situational influences

Situational Factors

  • Any similarity between offences can be because of situational factors.
  • The degree of violence exerted by the attacker during sexual assaults, for example, has tentatively been linked to both personality disorder, such as paranoid and antisocial tendencies (Proulx, Aubut, Perron, & McKibben, 1994) as well as to the disinhibitory effect of alcohol (Seto & Barbaree, 1995).
  • This illustrates that it may be problematic to infer an antisocial personality structure within an offender profile if excessive violence was present in a given case, since the perpetrator could in fact have been intoxicated. (Mokros & Alison, 2002)

Investigative Psychology

  • Pioneered by David Canter in the UK.
  • Similar aims to FBI approach – inferring offender characteristics from crime scene behaviour.
  • BUT more evidence-based.
  • Uses more data from more crime scenes to increase reliability.
  • Crime scenes broken down to à behaviours.
  • Statistical analysis determines whether those behaviours are related to certain offender characteristics.
  • Moves away from the ‘why’ – motive is unimportant.

Classifying Behaviours

  • Co-occurrence of crime scene behaviours can lead to classification.
  • Solved cases show that certain crime scene behaviours are linked to different offender characteristics.
  • Allowing for inferences about that person based on their actions
  • E.g. Salfati and Canter (1999) found:
    • Certain types of crime scene behaviours co-occur in ‘themes.’
    • These themes are associated with certain offender types…

Salfati + Canter 1999 SSA

  • Aggression studies suggest differences between instrumental vs expressive crimes
  • Are there such stylistic distinctions in aggressive actions in murder scenes?
  • Do crime scene behaviours link to offender characteristics?
  • Frenzied, Impulsive, Opportunistic, Thoughtful elements

Salfati + Canter Study Results

  • IF these themes are linked to certain types of offender characteristics…

  • Then those characteristics should ALSO co-occur with behavioural themes

  • The offender characteristics DID co-occur with similar behavioural themes

  • So:

    • It is possible to classify crime scene behaviours and infer related characteristics…
    • Independently of ‘expertise’

Location

  • Canter also noted that previous efforts lacked consideration of offender LOCATION
  • Residential location was included in Canter’s early successful profiles
  • Postulated à distribution of attacks arose from an offenders daily activities
  • I.e. we offend in known locations

Barnum Effect

  • Cognitive bias: tendency to find sense in vague and ambiguous statements.
  • Given enough ambiguous predictions, some will inevitably be correct.
  • Human tendency to focus attention on items that match our “story” and disregard items that do not (related to confirmation bias)
  • Officers in Alison et al (2003) attended to the info in the profiles that was correct, but ignored that which ran contrary to expectations.
  • The Railway Rapist

The Railway Rapist

  • One of the earliest profiles where psychologists identified an offender based on geographical relationship to a victim.
  • Environmental psychology à how space and surroundings influence people and behaviour
  • Relevant to how offenders interact with potential target areas
  • Geography gradually becomes important in offender profiling…

Revisiting the Ripper

  • Inference of personality from crime scene behaviours
  • The importance of the killer’s interaction with situational factors
  • The importance of crime location in relation to likely residence (e.g. all crimes were walkable)

Geographic Profiling

  • an investigative methodology that uses locations of a connected series of crimes to determine the most probable area of offender residence.
  • Most offenders operate in predictable locations
  • Mathematical formulae à likely places of a repeat offender’s residence or work
  • Helps to prioritise a search area for investigators (patrols, surveillance, canvassing, missing bodies)
  • Helps to manage information (e.g. prioritising suspects for interrogation)

Crime Pattern Theory (Brantingham + Brantingham)

  • We are creatures of habit
  • Offenders operate within awareness spaces – places they visit frequently and are familiar with
  • Individuals tend to offend close to their home or other significant personal nodes (e.g. work)
  • Crime sites might tell us something about an offender’s routine activities and their anchor points

Journey to Crime

  • The distance between an offender’s home and crime location
  • Most crime trips are short (highest % of incidents occur nearest home)
  • Offenders prefer not to travel far

Case Study (Rossmo + Rombouts, 2008)

  • 42 burglaries over 24 months
  • Mostly targeting single-family homes backing onto parks
  • Targeted properties all observable from public places
  • Point of exit was always at rear
  • Offender brought a chair to put against back fence to gain entry (MO)
  • Property stolen was generally limited to cash and jewellery…
  • Offender was likely a professional burglar
  • “The underlying premise here is that the nature of crime- related locations tells us something about the victim, the offender, and how each interacted with the environment during the course of the crime.” (Rossmo + Rombouts, 2008)

Evaluating Profiling

  • Officer satisfaction
  • Profile accuracy
  • Assessing geoprofiling

Officer Satisfaction (Copson, 1995)

  • Interviewed police officers in the UK
  • Profiling advice reportedly helped in 14% of investigations
  • 83% of respondents deemed profiles to be “operationally useful” (but what does this mean?)
  • Identified offender in 3% of investigations
  • Profiles found to provide comfort to investigators through reinforcing their own judgement
  • 92% said they would use a profiler again
  • Problems with ‘consumer’ satisfaction surveys? (small samples of officers – no clear picture)

Profile Accuracy (Alison et al, 2003)

  • 46 police officers / 2 groups:
    • Group A à genuine description of actual offender
    • Group B à fabricated description of offender
  • The fabricated descriptions were deliberately different (i.e. opposed) to the real ones
  • Both groups à same crime details + same offender profile
  • PS rated how accurately the profile was perceived to have portrayed the offender
  • Most in both samples rated the profiles as accurate
  • Mean accuracy ratings of genuine offender description did not differ from fabricated offender

Assessing Geoprofiling

  • Probabilistic process: X doesn’t mark the spot (Harries + LeBeau 2007)
  • Less empirical research on accuracy of geoprofiling strategies
  • Support mostly comes from anecdotal evidence and investigations
  • Relies on accurate data and identification of linked offences
  • Townsley + Sidebottom (2010) not all offenders exhibit distance decay in the journey to crime
  • Dark figure: models are based on offenders who are caught
  • Data likely to be skewed towards offenders who commit crimes near home (easier to catch?)

Dangers of Profiling Example: Rachel Nickell Case

  • Model Rachel Nickell stabbed 49 times on Wimbledon common with 2 year old son
  • The police lacked physical evidence à profiler was consulted
  • Profiler Dr Paul Britton created a profile of the murderer, indicating:
    • Violent sexual fantasies
    • An expectation to kill again
  • Although many suspects were interviewed, Britton caused police to focus on Colin Stagg, as he closely met the offender profile provided
  • Was seen walking on the common (but no physical evidence)
  • Had written a letter to another woman about a sexual fantasy

Rachel Nickell Case Outcome

  • Undercover officer wrote sexually explicit letters, in attempt to entice Stagg to confess; instead…
  • Blinkered focus on Stagg lead to missed links with the real offender (tunnel vision, confirmation bias)
  • Stagg falsely confessed to a different murder (the police established this was fabricated)
  • Stagg (an innocent man) was held for a year in prison, due to his attempts to please the officer trying to entrap him
  • The case was deemed inadmissible in court, due to the methods used to gather ‘evidence’ against him
  • Serial rapist Robert Knapper later admitted to the murder

Summary

  • Growing belief that profilers can accurately + consistently predict offender characteristics from crime scene examination
  • Snook et al. warn that this belief is an illusion:
    • Critical analysis of profiling shows that the field lacks theoretical grounding + empirical supports
    • The belief is due to biases in the way people receive and process the information they are given
  • This is important, as profilers have the ability to mislead investigators
    • Hindering the apprehension of guilty criminals
    • Leading to wrongful convictions
  • Be cautious of anecdotal evidence pertaining to the utility of profilers

Takeaway Message

  • More scientific evaluation of criminal profiling methods is needed…
  • Criminal Profiling might actually work! (but the burden is on those making the claims to prove this)
  • People deserve to have an accurate view of profiling (and the use of the scientific method is the way to achieve this)
  • The influence of profiling on investigations overall, is unknown (research can help to determine this, positive or negative) (Snook et al., 2008)