Psychology G543 - Making a Case

Key Study - Bruce et al (2002) - Recognising faces

• study whether better likenesses could be produced if multiple witness composites of the same person were generated and used in combination

Results:

• If 4 composites were created using evoFIT and an average produced, it was rated a better likeness than the individual composit
• In a line up task, more correct identifications were made from the average image

Evaluation:

• Useful - average of lots of evoFITS are a better likeness
• High ecological validity - line up task would happen to a witness
• Low reliability - pps may not try hard knowing a crime hadn't been committed

Key Study - Loftus, Loftus and Messo (1987) - Factors Influencing Identification

• Lab experiment about whether weapon focus influences identification
• Participants were presented a series of slides depicting an event in a fast food restaurant
• In one half, pps saw customer hand over a dollar bill
• Other half, pps saw customer draw a gun
• They recorded eye movements of participants which showed those in the weapon condition fixated on the gun, and those in the dollar condition didn't fixate on the dollar

Results:

• Memory in the weapon condition was significantly worse than in the dollar condition - supports hypothesis that presence of a weapon adversely affects perception and recollection

Evaluation:

• Useful - for defendant lawyers if eyewitness describes the suspect with the gun - know they are not reliable
• Low generalisability - gun laws different in America - somewhere like the middle-east where gun crimes are regular, witness may not focus on weapon
• Low ecological validity - carried out in a lab and pps were watching slides - not like a crime in a real situation

Key Point: The Cognitive Interview - Background

Fisher (1989) - 4 basic principles underlying the cognitive interview technique:

• Interview Similarity - recall is improved by taking place in a similar environment to the event
• Focused Retrieval - environment in which recall takes place should facilitate concentration
• Witness-Compatible Questioning - interviewers will encourage recall more successfully if they allow witnesses to follow their own schema's
• Extensive Retrieval - witnesses should be encouraged to remember events from a different perspective in the hope this may trigger recall

Key Study - Geiselman and Fisher (1985) - the Cognitive Interview

• Compared the effectiveness of the cognitive interview with standard police interviews + interviews under hypnosis
• After watching police training films of simulated violent crimes, pps were asked 48 hours later to recall as much as possible

Results:

• Both CI and hypnosis techniques significantly recalled more correct items than the standard interview group

Evaluation:

• Useful - for police to get as much info as possible
• 

Key Study - Vrij and Mann (2001) - Detecting Lies

• Aimed to test police effectiveness in distinguishing truth and lies in a real-life high-stake situation
• Case used in which suspect who was later convicted of murder was videotaped during interviews
• The viedeotapes were analysed and edited into 3 truthful and 3 deceptive fragments which were shown to serving police officers
• Officers asked on a 7-point Likert scale after each fragment: 1) Is he lying? 2) Is he tense? 3) Is he controlling his behaviour? 4) Is he having to think hard?
• After responding, pps were asked to write down behavioural cues which prompted their decision

Results:

• Overall accuracy rate of PO's was significantly above the level of chance
• Mean scores on truth/lying revealed that PO's are better at detecting truths

• Those poor at judging relied on sterotypical behaviours
• Those good used more subtle cues

Evaluation:

• Useful - in improving police interview technique
• High in ecological vaildity - case used was real with real interviews
• Low in generaliseability - only one suspect - behaviour may not have been typical

Key Study: Granhag and Stromwall (2001) - Interrogation Techniques

• Analysed the value of multiple interrogations in assessing whether the suspect is lying or not

Results:

• Both interrogators and observers performed poorly
• No significant difference in accuracy between observers who watched interrogation once and those who watched it 3 times
• Improved performance in observers who watched it once then assessed the truthfulness, then watched video twice more before reassessing truthfulness

Conclusion : Interrogators showed truth bias

Evaluation:

• Useful - there is truth bias present and suggests a different way to do interrogations
• Low in ecological validity - participants were undergrad students, not PO's - not as experienced

Key Point - False Confessions - Background

Gudjonsson (1992) - there are 2 kinds of false confessions:
1) coerced compliant (manipulated and you go along with it)
2) coerced internalised (manipualted and you end up believing what police are saying)

Key Study - Gudjonsson (2006) - False Confessions

• Studied 10,472 Icelandic students
• National study into the background, behaviour and mental health of adolescents

Results:

• A quarter of pps said they had been interrogated by police
• Half of that quarter said they had made a false confession - this rate was highest amongst those interrogated more than once

• False confessors reported more psychological problems and a higher delinquency rate than other youths

Evaluation:

• Useful - warns police that youths interrogated more than once are at risk of making a false confession
• Reliable - large sample used
• Low generaliseability - Icelandic students - crime rate may differ elsewhere
• Self report method - students may not have been accurate or honest and liable to social desireability

Key Point: Top-Down Typology

Nic Daeid - behavioural profiling depends on 3 basic psychological principles:

1) that people influence each others actions

2) most human behaviour develops and changes over time and in relation to direct           experiences of the consequence of actions taken

3) types of behaviour reflect personality and most behaviours satisfy a wish, desire or need

Key Study - Canter et al (2004) - Top-Down Typology

• analysed Missen Corpus of Serial Killers
• data concerned cases of serial killings committed in the US and sources were published
• content analysis of case report material in the Missen Corpus enabled 39 criteria to be identified that corressponded to crime scene characteristics as either organised or disorganised

Results: 

• found twice as many disorganised crime scene behaviours than organised - shows disorganised behaviour are more fully recognised and reported
• found examples of co-occuring characteristics - both organised and disorganised

Researchers used alternate scientific method of data analysis and found:

• no clear division between organised and disorganised variables
• suggests instead that all serial killers are organised and that what distinguishes one from another is the nature of the factors described as disorganised

Canter et al believed there is no scientific basis for the top-down approach.

Evaluation:

• Low reliability - data taken from newspapers
• Supports psychology as a science - shows value of applying scientific method rather than formulation theories

Key Point: Bottom-Up Approach

Canter and Heritage (1990) - found 5 key variables that identified a behavioural pattern:

1) vaginal intercourse

2) no reaction to victim

3) impersonal language

4) surprise attack

5) victims clothing is disturbed

The 5 factor theory proved helpful in establishing whether 2 or more offences are committed by the same person.

Key Study: Snook, Canter and Bennell (2002) - Bottom-Up Approach

• investigated whether people could be trained to predict likely home location of serial offenders using heuristics based on similar principles those used by Dragnet (computer profiling system Canter developed)
• distance-decay and circle hypothesis were heuristics they were told to follow

Results: 

• After given the 2 geographical profiling, pps in general improved their predictions and some were able to make predictions as accurate as Dragnet

Evaluation:

• Useful - helps police locate serial offenders without relying on Dragnet - has a complex input of data

Key Study: John Duffy: The Railway ****** - Case Studies

• between 1982-1986 he committed 24 sexual assults and 3 murders near railways in North London
• Canter analysed assults and created a profile
• when Duffy was found, his profile was very accurate
• Duffy was already on police database - one of nearly 2,000 suspects linked to crimes by blood group
• Canter's geographic profiling determined his arrest

Dangers of offending profiling:

• If you're following the wrong person, the real killer can get away
• profiling doesn't catch criminals - it narrows down suspects
• if suspect is not on any police database they won't be caught!

Evaluation:

• Useful - for catching criminals if they are already in a database
• Dangerous - only narrows your search