5 - Profiling and Classification

why use classification?

• theoretical development

• guides assessment practices

• guides treatment practices

why not use classification?

• important individual differences may be obscured

• categorizations often seen as pejorative

humans have a strong tendency to _____

categorize

• better to use this natural inclination than ignore it

issues where classification may be relevant

• level of risk

• institutional housing decisions

• degree of supervision required

• offender needs

• treatment inclusion

are between-group differences theoretically informative?

yes

• developmental trajectories and risk factors can help identify later problems

• circumstantial risk management post-release

Chaiken and Chaiken study

8 dimensions based on past crimes yields 256 possible combinations, but 10 of those accounted for 59% of inmates

• leaves some people unclassified

• doesn’t guarantee clinical differences between groups

• lots of combos

Megargee’s good system criteria (7)

• comprehensive

• unambiguous

• statistically reliable

• type distinctions are valid

• sensitive to individual changes

• treatment relevant

• economical to apply

systems based on ________ rather than impressions fair better

cohesive personality / criminogenic theory

most attention now is on ____ _____ systems

empirically derived

• most testable and accurate

impressionistic system

an individual, or group, think they recognize a pattern or type among some members of a clinical/institutional sample

• no theory or method of measurement

• essential features identified after the fact

• prognostic is poor

• ex., DSM-5

problems with DSM-5 as an impressionistic system

• many NOS categories

• systematic disagreements (ex., schizoaffective vs. schizophrenic)

• many personality disorders go hand in hand

  • predictions are circular

• assumes discontinuity among personality disorders (which is untrue)

  • discrete taxons

theoretically derived systems

stem from a developmental or personality theory

• relates specific attributes or points of maturation to antisocial behaviour

• measure instruments draw heavily on concepts central to theory

  • introduces significant bias

conceptual level (example of theoretically derived system)

based on ability to adopt other perspectives

• attempts to tap reasoning process more so than outcome

• incomplete sentences task

responsivity principle

delivering treatment in a way that is tailored to the individual and meaningful to them to maximize effectiveness

• used in both impressionistic and theoretically derived systems

• ex., lower conceptual level = more behaviourally oriented interventions

empirically derived classifications

internet data used to profile you

based on two statistical techniques

• cluster analysis

• factor analysis

cluster analysis

numerical taxonomy

groups subjects according to level of similarity as measured by variables selected by the researcher

• ex., MMPI based scheme

factor analysis

correlational

good for reducing a large number of variables to a smaller number

• may help identify “underlying dimensions” like when 1-2 variables are responsible for the others

• ex., Hewitt and Jenkins 4-factor solution

Minnesota Multiphasic Personality Inventory

most commonly used clinical personality measure (cluster analysis)

• 1200 convict profiles yielded 10 types

• classifies almost all profiles (96%)

• results replicated

characteristic criminal based on MMPI

able, 17%

elevated psychopathy (4) and mania (9)

problems with MMPI

• predictive accuracy in question, poor correspondence between MMPI and MMPI-2 profiles

• high scale overlap (characteristics are not orthogonal)

• are 10 groups really needed?

• profile derivation was atheoretical, results may not be unique to instrument

Hewitt & Jenkins 4-factor solution

• example of factor analysis

• 4 factors

  • unsocialized aggressive behaviour (UA)

    • poor prognosis

    • LCP

  • socialized delinquency (SD)

    • better prognosis

    • AL

  • over-inhibited (OI)

    • more likely mentally ill

  • phuscial deficiency

    • less interpretable

profiling

used to describe serial offenders, high premium to catch as they are likely to offend again

collection of related techniques:

• psychological profiling

• criminal personality profiling

• criminal behaviour profiling

• investigative profiling

• criminal investigative analysis

goal of profiling

describe perpetrators in order to narrow the investigative field

profiling sources of information

• autopsy results

• victim verbal reports

• examination of crime scenes / photos

• test and interview data (psychological profiling only)

criminal profiling

relies heavily on intuition

• first attempts in WWII to predict military beahviour of enemy officers

• directs attention of investigators to most likely subjects

• akin to interpreting personality profile

• makes general assertions

• works best with frankly ill and highly sadistic individuals

_____ of police psychologists express discomfort with profiling

70%

two reasons

• constitutional concerns, is it inappropriate to profile based on race etc.

• not good enough of science basis

profiling is more of an ____ than a science

art

<___ of profile cases are solved

50%

• partly because profiling only for the most challenging cases

Pinizotto and Finkel findings of profiling accuracy study

found that trained profilers did only slightly better with sex offender profiling, but no better with violent offenders than police detectives, non-forensic psychologists, and university undergrads

reconstructive psychological autopsy (psychological autopsy)

post mortem descriptions based on a known offender

• intended to add to profiling database

• attempts to address historically dynamic features, not just static

psychological profiling

begins with known perp

• findings added to database

• test selection is critical

  • one relevant test is of greater value than many irrelevant tests

geographic profiling

based on analysis and prediction of movement patterns, highly statistical

• uses Rigel Analyst software or Dragnet software

• uses distance decay functions based on the journey to crime model

• yields jeopardy surface (level of risk in certain location)

• probability based on minimizing the mean distance from all known crime sites attributed to that offender

paradigm

way of doing something, viewpoint, perspective

distance decay function

graph of distance from home compared to probability of offending, peaks on all sides of home a bit away

rasterization

used to determine the lowest distance that their home base is likely to be based on all the crime locations

jeopardy surface

elevated points on the map have highest probabilities of home base

Rossmo’s minimal criteria for profiling

• minimum 3 years experience investigating interpersonal crime

• “superior” investigative skill

case linkage analysis

joint probability, multiplying aspects of crime to see if they were committed by same person

canter’s circle hypothesis

serial offender’s home base most likely to be located within a circle defined by the two most distance crime scenes

• works best for small areas

architectural psychologist

Canter

• give people advice on designing businesses based on flow of human traffic

• a prediction, will this Starbucks stay in business here?

Snook’s circle hypothesis study

undergrads given circle hypothesis instructions saw a significant decrease in their error

• found no advantage to using Dragnet over the circle hypothesis experimental group

• conclusion: neither extensive experience or a computer is necessary to profile

three classes of crime scene features

• modus operandi

• signature

• staging

modus operandi

the way the violent act was committed, follows same pattern of human behaviour

• time, place, weapon, victim characteristics, type of crime, location, method of entry

• refined over time and with practice, MO becomes stable eventually as processes become automated

• can also de-evolve to become sloppy and less patterned

signature (calling card, personation)

features extra to the crime itself

• may become highly ritualized

• pattern of mutilation or property destruction, taking trophies, leaving something behind, notes/writing, positon of body, burning body

• may provide more info than an unstable MO

staging

alteration to the crime scene after the crime was committed

• may not have been done by assailant

  • ex., family making suicide look like murder

• offender may attempt to psychologically undo, or reverse the act

  • particularly in acts of passion

classifications of crime scenes

• organized

• disorganzed

• mixed

organized crime scene

• evidence of planning to eliminate obstacles

  • reduces chances of detection

  • allows greater specification of victim characteristics

  • may become cocky

• abduction, murder, disposal usually not in same locations

disorganized crime scene

• reflects work of more impuslive, less intelligent criminal

• crimes in angry, hihgly aroused state

• no movement of evidence (including body) from crime scene

• victims more random

mixed crime scene

features of both organized and disorganized

usually means poorly planned crime fell apart, encountering unexpected resistance

Turvey said most crime scenes exist along a _____ of classification, not a dichotomoy

continuum

what features Turvey thinks should be considered for crime scenes

• point of contact

• primary scene

• secondary scene

point of contact

where the perp first had contact with the victim

primary scene

place where something criminal took place for the first time

secondary scene

place where the actual large crime took place

• there may also be intermediate scenes and the dumpsite scene

method of approach

how the assailant will approach the victim

• mainly for rapists

• may reveal information about relationship to victim

1. blitz

2. surpise

3. con

blitz

(method of approach)

limits opportunity to work out a defense resulting in injurious physical assault

surprise

(method of approach)

laying in wait for the victim, having knowledge of their schedule

can turn a 2nd degree into 1st degree murder

con

(method of approach)

employs a ruse, may involve repeated contacts over time

hunting strategies

involve looking for a victim, greater movement for the perp than just method of approach

1. hunter

2. poacher

3. troller

4. trapper

hunter

(hunting strategy)

goes out and gets a victim fairly close to home (31%)

• geoprofiling most effective for them

• distance decay function

poacher

(hunting strategy)

goes into another town or city, doesn’t operate from home (54%)

• fewer victims

• still have a base of operation

troller

(hunting strategy)

crimes of opportunity, victims are encountered while other routine tasks are carried out (11%)

trapper

(hunting strategy)

lure/trick victims to a location where the perp has control (1%)

attack strategies (Rossmo)

1. raptor

2. stalker

3. ambusher

raptor

(attack strategies)

attack is immediate upon location of victim (79%)

• serial predators

stalker

(attack strategies)

watches, follows, waits to catch victim in a vulnerable, off-guard state (<1%)

ambusher

(attack strategies)

always works close to home or another prepared and familiar location (21%)

types of motives

revenge, financial gain, jealousy, power, non-sadistic or sadistic sexual gratification, insanity, self-defense