Eye-Witness Testimony

  • Created by: Jess
  • Created on: 15-01-13 16:36

Devlin Committee (1976)

The Devlin Committee analysed all identity parades in England and Wales in 1973.

Of over 2000 ID parades, 45% led to identification of suspect, 82% of those identified were convicted.

In 350 cases, eye-witness testimony was only evidence, yet 74% were convicted.

The Committe recommended htat juries should not convict on the basis of eye-witness testimony alone.

1 of 19

Which factors explain why it is unreliable?

  • Reconstructive nature of memory (schemas & stereotypes)
  • Language/leading questions
  • Context
  • Emotional factors/stress
  • Age
  • Stereotyping
2 of 19

1. The reconstructive nature of memory

Allport & Postman (1947) - "when an actual perceptual fact is In conflict with our expectations, expectation may prove a stronher determinant of perception and mempry that the situation itself."

In other words, we see what we want to see, and this forms the basis of our memory.

  • When memories are adjusted to conform to our expectations, beliefs, and stereotypes - confabulation
  • We use schemas - mental representations of situations, events, etc - to organise our memories
3 of 19

1. The reconstructive nature of memory

Bartlett et al (1932) - War of the Ghosts

  • Participants listened to the story (a North American folk tale), then another, and so on.
  • On retelling, details were changed to become more familiar.
  • Memory was shaped by British culture, the more often it was recounted, the more British it became, e.g, canoe b3came boat
4 of 19

2. Language/leading questions

Loftus et al researched how a witness' memory for events could be distorted at interview and found that:

  • Leading questions can influence recall
  • Non-existent items can be inserted into memory
  • Memories can be transformed by deleting and replacing information
5 of 19

2. Language/leading questions

Loftus (1974) 

Aim: to investigate effect of leading questions on eye-witness testimony

Method: participants were shown a clip of a car accident and were asked different questions involving leading language, e.g, how fast was the car going when it hit/smashed the object. They were then asking if they saw any broken glass.

Results: participants estimated 34mph for hit, and 40.8mph for smashed. 14% of the hit participants reported seeing broken glass, compared to 32% of the smashed condition.

Conclusipn: language/leading questions can affect peoples memories for events

6 of 19

2. Language/leading questions

Loftus (1975)

Aim: to investigate the effect of leading questions on eye-witness testimony

Method: participants were shown a clip of a car accident and were asked questions involving false information, e.g, referred to a barn which wasn't there

Results: 17% of the misled group reported seeing a barn, less than 3% of the control group

Conclusion: language/leading questions can affect peoples memories or events

7 of 19

The importance of Loftus' work

Loftus' work is very important as:

  • It shows how fragile eyewitness testimony is - can easily be distorted by questions after memory was formed
  • It has provided evidence that juries should be more careful before accepting the validity of what eye-witnesses say
  • It has been useful for police - showed that they need to be very careful when questioning eyewitnesses in order to prevent memories from being distorted by their questions
8 of 19

3. Context

Lab research indicates recall is better if the personis in the same context as when the information was encoded

Study- Mapless & Devine

Method: participants were shown an act of vandalism, then were interviewed five, months later. One group were reminded of the date, time, room and immediate reactions, and a control group was given no contextual information

Results: those that were given contextual information remembered significantly more correct information

9 of 19

4. Emotional factors/stress

Contradictory evidence

  • Clifford and Scott (1978) - people who saw a film of a violent attack remembered less than a group who watched a less violent attack
  • Yuille and Cutshall (1986) - found witnesses of real life Incident had very accurate recall 5 months later. 2 misleading questions had no effect.
10 of 19

4. Emotional factors/stress - Weapon focus

Loftus (1979) - weapon focus - when a witness' attention is drawn to the weapon in a crime scene, so very few other details are recalled.

Method: participants were asked to wait outsidea laboratory for a few minutes. While they waited they heard either

  • A low key discussion about equipment failure, followed by a man comping put holdong a pen, with grease on his hands, or
  • A heated argument and sounds of breaking glass and crashing chairs, followed by a man coming out holding a paper knife covered in blood

Participants were then shown 50 photos and asked to identify the man who had left the lab

Findings - 49%  participants who saw the man with the pen identified him correctly, 33% participants who saw the knife identified him correctly.

Conclusion - the participants concentrated on the weapon which distracted them from the men. Loftus argued that anxiety caused by the weapon narrowed their focus of attention which means that their recall for peripheral details is not very accurate.

11 of 19

5. Age

Memory function is best between 25 and 65 years of age, therefore childrencould be less reliable witnesses.

Maria et al - studied memory for events with four groupd of young people. Found that recall increased slightly with age but no significant difference in accuracy.

Fivush and Shukat (1995) claimed that very young children are able to give very accurate accounts of personal experiences.

12 of 19

6. Stereotyping

Studies show constructive memory is a reflection of our individual and social beliefs.

Errors are more likely when the suspects race is different to the witness - Baddeley (1982)

Gende stereotyping also influences memory - participants in a study assumed that a bag snatcher was male when they were females - Gruneberg (1992)

13 of 19

6. Stereotyping - Allport & Postman study

Aim: to investigate effects of stereotypes on memory

Method: participants were while Americans divided into groups of seven. One participant from each group were shown a picture of an 'argument on a subway train', described it to the next participant and so on

Results: over 50% of those receiving final description reported that the black man was holding the razor. Some reported that he was holding it in a threatening manner.

Conclusions: we see what we want to see based on our existing knowledge, beliefs and expectations.

14 of 19

How can we improve the reliability of eye-witness

  • The cognitive Interview - Fisher & Geiselman (1992)
  • Hypnosis
  • Police reconstructions
15 of 19

1. The Cognitive Interview - Fisher & Geiselman (1

The cognitive interview is based on four general retrieval mnemonics:

  • Mentally reinstate environmental and personal context of the crime - imcluding sights, sound, smell, feelings, emotions
  • Encourage the repeating of every detail, no matter how trivial It seems
  • Recount the incident in a different order
  • Report from different perspectives - the view of other bystanders, or the view of the criminal
16 of 19

Geiselman et al

Aim: to compare aswer given in cognitive interview with those given in standard police interview and hypnosis

Method: participants watched short film about a violent crime. two days later, were divided into three groups, and interviewed by a police officer

Results: those interviewed in standard police interview made an average of 29.4% correct statements, 41.1% correct with a cognitive interview, and 38% correct with hypnosis

Conclusion: supports the cognitive interview as a way of obtaining more accurate eyewitness testimony

17 of 19

2. Hypnosis

Hypnosis helps witnesses to relax and, in theory, allows the hypnotist to reconstruct the details in the mind of the witness.

Supporting evidence

1977 - a bus driver, whose bus containing 26 children has been hijacked, was successfully hypnotised to recall all but one digit on the hijackser van. The three suspects were all tracked down.

Evidence against

Orme et al (1984) - analysed results of a large numer of studies in eye-witness testimony and hypnosis and found no evidence that hypnosis reliably aided recall. In fact, can lead to more confabulation and memory for false details.

18 of 19

3. Police reconstructions

Often used for crimes such as murder, robbery, and missing persons.

Aim: to obtain witnesses who have not yet come forward and to jog the memory of other witnesses - restating context

Based on the idea of cue-dependent recall, participants wear similar clothes, reconstruction takes place at a similar time of day.

These often lead to useful information being remembered, but can also lead to confabulation - adding false details to the memory of the event.

19 of 19


No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Memory resources »