EYEWITNESS TESIMONY

1
EYEWITNESS TESIMONY
2
RELIABILITY OF EYEWITNESSES

• Brown stated that whilst judges believe EWT to be
  the least trustworthy evidence, jurors find it
  more persuasive than any other evidence
• Wells carried out an experiment where Ps
  believed they had witnessed a woman stealing a
  calculator.
• 58 could identify her from a set of 6 photos

3
DEVLIN REPORT

• This report found that there were 850 cases where
  EWT was the only evidence.
• In 74 of these cases, the accused was found
  guilty by the jury
• The Devlin Committee considered the reliability
  of such testimony and advised that no jury should
  convict on EWT alone
• In the USA many prisoners who were convicted by
  EWT have been released since DNA testing has been
  used, and they were discovered innocent of the
  crime

4
UNRELIABLE EVIDENCE

• Filling in the gaps of our memeory is called
  RECONSTRUCTIVE MEMORY
• It relies on stereotypes
• The effects of language are a further issue

5
RECONSTRUCTIVE MEMORY

• A schema is an organised package of information
  containing our knowledge of the world, e.g. what
  to do when you go to a restaurant
• IF we encounter stimuli which conflict with our
  schemas, we reconstruct the memory to fit in with
  our schema
• Allport and Postman demonstrated this with an
  experiment p. 60/75

6
SCHEMA THEORY

• Bartlett suggested that the process of
  remembering is an ACTIVE reconstruction of the
  bits that are stored
• Schema theory indicates that prior expectations
  will influence our perceptions
• So our stereotypes will lead to changes in how we
  subsequently remember information
• Stereotypes are schemas which summarise large
  amounts of data.

7
APFCC RECONSTRUCTIVE MEMORY

• AIM to see if people reconstruct information
  from another culture to fit into their own
  schemas
• PROCEDURE The war of the ghosts was read to
  Ps. This is a native american indian folk tale.
  They were asked to recall the story days months
  and a year after

8
APFCC RECONSTRUCTIVE MEMORY

• FINDINGS Bartlett found that recall was
  distorted. Changes made were in line with Ps
  own culture.
• Flattening (unfamiliar details left out.
• Simplifying (leaving details out)
• Sharpening (elaborating some content)
• All these made the story easier to remember

9
APFCC RECONSTRUCTIVE MEMORY

• CONCLUSIONS Accuracy of memory is low. People
  reconstruct data to fit in with their existing
  schema. This has implications for EWT
• CRITICISMS
• High external validity because it tests
  everyday memory
• - Bartlett assumed that the distortion happened
  at retrieval, whereas the shemas may have been
  altered at encoding

10
STRENGTHS OF BARTLETTS RESEARCH

• Studied the effect of conflicting information on
  memory
• He demonstrated how memory is affected by past
  experiences, e.g. cultural factors
• He demonstrated that memory is prone to
  distortion and error
• His research enabled us to understand how
  information is distorted in memory
• It has been influential in EWT

11
COHENS STUDY

• PROCEDURE Ps were shown a video of a man and
  woman eating a meal.
• They were told she was waitress/librarian
• They were asked to describe her personality
• FINDINGS Ps were more likely to recall
  information which was consistent with the
  stereotype
• Furthermore, telling Ps her occupation BEFORE
  they watched the video increased stereotyping
• Cohens study shows that schemas and stereotypes
  are important both at the initial storage and
  retrieval

12
INFLUENCE OF SCHEMAS ON EWT

• OKLAHOMA BOMBING p.62. information received
  subsequent to a crime influences what people
  think they saw. One EW says he saw 1 man visit a
  garage to hire the van to put the bomb in,
  another says he saw 2 men. The first EW came to
  remember a second man

13
TUCKEY AND BREWERP.78

• Collected peoples schema for a bank robbery
• EWs to a bank robbery used their schema to
  assist in their recall of the robbery
• Thus, ambiguous detail is altered to fit in with
  a persons robbery schema

14
EVALUATION OF RECONSTRUCTIVE MEMORY

• Schemas influence the way data is stored and can
  explain how data is retrieved
• However, some psychologists believe that schema
  theory exaggerates inaccuracy of memories.
• We are capable of remembering things word for
  word, e.g. actors and actresses. Schema theory
  does not account for this
• Schema theory cannot predict what and how people
  remember information, because we cannot know
  which schemas are being used

15
FACTORS AFFECTING RECONSTRUCTIVE MEMORY

• HIGH EMOTION Johnson and Scotts study
• Ps overheard a (rigged) violent fight in a
  neighbouring room
• Condition 1 Ps saw a confederate come into
  their room with a bloody knife (high emotion)
• Condition 2 Ps saw a confederate come into
  their room with a greasy pen (low emotion)
• Ps were asked to identify culprit.
• Evidence that high emotion condition was better
  (flashbulb) was not strong

16
HYPNOSIS

• Some police forces have used hypnosis in the hope
  of gaining information such as car number plates
  and the physical features of wanted criminals
• HYPERMNESIA means the improvement of memory
  using hypnosis
• The problem with hypnosis is knowing which
  memories are reliable, because hypnotised
  individuals make more errors and are more
  suggestible

17
PUTNAMS STUDY

• Showed Ps a video of an accident between a car
  and a bicycle
• They were then asked questions, some were
  misleading
• Half the Ps were hypnotised for the question
  session
• Hypnotised Ps made more errors, especially with
  the misleading questions

18
LEADING QUESTIONSAPFCC

• The language used in questioning eyewitnesses may
  alter what they remember. Loftuss study shows
  that leading questions affect post-event
  information
• AIMS to see if leading questions have an effect
  on the recall of a car accident

19
APFCC

• PROCEDURES 45 American students were shown a
  film of a car accident
• They were asked how fast the cars were going when
  they
• crashed/ bumped/ collided/ contacted
• 1 week later they were asked
• Did you see any/the broken glass

20
APFCC

• FINDINGS Estimated speed was influenced by the
  verb used
• Smashed mean estimate 41mph
• Hit - mean estimate 34mph
• 32 of smash group said they had seen broken
  glass
• 14 of hit group said they had seen broken glass

21
APFCC

• CONCLUSION Language can affect the memory of
  eyewitnesses. Possibly the original memory was
  reconstructed
• CRITICISMS
• Lacks mundane realism, because it was conducted
  in a lab
• Other studies show that central details are not
  distorted, only peripheral ones are
• Has informed police on the nature of
  questioning EWs

22
HOW CAN EVIDENCE BE MADE MORE RELIABLE?

• YUILLE AND CUTSHALL Gross p.356
• Real life study of 21 witnesses to a robbery
• Eyewitness accounts were very accurate, and
  accuracy did not diminish over time
• Accounts were not distorted by leading questions
• This study has high external validity

23
GEISELMAN

• He argued that interview techniques would take
  into account that
• A retrieval cue is more effective if it overlaps
  with a memory trace.
• This is the encoding specificity principle
• If information cannot be retrieved with one cue,
  then another should be used, e.g. forming an
  image, reconstructing the event

24
THE COGNITIVE INTERVIEW

• This interview led to 41 correct statements as
  opposed to 29 in the standard police interview.
  It differed by
• Minimising distractions
• Allowed eyewitness to speak slowly
• Pausing between response and next question
• Reducing eyewitness anxiety
• Reviewing the eyewitnesss description

25
EXPERT WITNESSES

• Loftus was called in to help the jury decide.
• They were told that
• People forget over time
• High stress can lead to impaired memory
• People are not good at cross-race identification
• Memory is influenced by stereotypes