Eyewitness Testimony

1. Effects of misleading info on accuracy of EWT

Loftus (1975) (CAR ACCIDENT. BARN)
Aim: EWT affected by what happens after event as memories are subject to distortion by post even info. This is called misinformation acceptance. She tested this.
Procedure:  shows p's films of events leading up to a car accident. They were divided into 2 groups.
Control group- asked questions that actually happened.
Experimental group- asked question with misleading info (barn)
Results: 17% of exp group reported seeing barn, 3% control group made this error.
Conclusion: p's given misleading post event info absorbed this as actual event.

Loftus (1978)
Aim: there is no real change to original memory, but p's alter what they say as a result of demand c's.
Procedure: offered reward money if they could correctly recall details from film of accident.
group1- was pedestrian knocked over by a car, stopped at stop sign.
group 2- stopped at yield sign.
2 days later, p's given q's about accident including misleading info of reference to 'stop and yield sign (vise versa)
p's then looked at slides pointing out what was the real film. They were divided into 4 groups.
1. no reward
2. $1 each correct answer
3. $5 each correct answer
4. $25 person in group scoring most points.
Results: 70% p's made error on misleading info despite the offer of reward showing their original memory had been altered as a result of misleading post event info.
Loftus and Palmer (1974):

Aim: memories can become distorted by info after event.

Procedure: Independent measures design. Independent variable used is the verb. The dependent variable in p's first exp is speed estimate. In second its whether they believed they saw glass or not.
45 students from Washington uni.
Watched a car crash and had 5 conditions of which the verb changed.
Results: The verb effected speed estimates. The mean average of smashed was 40mph in contrast to the 'contacted' of 31mph.
Evaluation: results could be due to response bias factors, the p could have been unsure therefore adjusts their estimation to fit with the expectations of the questioner.
2nd experiment
Procedure: 150 students viewed 4 second car accident on film. The independent variable was manipulated by the wording of the question.
3 conditions of changing the verb e.g. smashed
One week later the p's answered qs about it and were asked if they saw any broken glass (there was none)
Results: the majority reported seeing glass.
Conclusion: the verb had an effect on the mis-perception of glass.
Two kinds of info go into memory of event. First is the info obtained from actual event, second is info supplied  after event.

Other factors affecting accuracy of EWT

-reconstructive memory which is storing replica of events. We blend in elements of our knowledge and experience to make it memorable= a schema.
List (1986) (SHOPLIFTING)
Aim: to investigate the above schema idea.
Procedure: drew list of elements that may occur in shoplifting scenario. P's had to rate in order how likely these were to occur in a shoplifting accident. She then compiled a video of 8 shoplifting incidents that included  elements that people rated as high probability and low and showed them to new p's. A week later they had to recall them.
Result: p's more likely to recall high probability than low and reported seeing high elements that weren't in the video.
Conclusion: due to the gap between seeing the videos and being asked questions about it, p's lost some info of the exact replica therefore blended in elements of their own knowledge and experience to make it more memorable.
Tuckey and Brewer (2003)
Aim: Further investigated reconstructive memory and schema. 
Procedure: found people think schema of robbers = dark clothes, male etc. They showed video of a bank robbery
Result: found p's had better recall for elements of film that conformed to their schema than to elements that didn't.
Conclusion: schema affects memory and recall of an event as this idea has already stored itself in the individuals memory.

3. How to improve accuracy of EWT

The cognitive interview technique
Fisher (1987)
-studied real interviews by detectives in Florida over 4 months. Found witnesses were bombarded with brief, direct, close ended questions, were interrupted and not allowed to talk freely, broke concentration and encourage short answers with less detail.
-On the basis of this,
Geiselman (1985) 
developed the cognitive interview technique as a more effective tool for police investigations.
1. Context reinstatement
2. report everything
3. recall from changed perspective
4. recall in reverse order
-this provided cues, open ended qs etc.
Fisher (1990)
-demonstrated effects of CIT in police in Miami.
Trained detectives to use enhances CIT with genuine crime witness and found it increased recall.

Factors which effect eye witness testimony

1. Consequentiality:

Studies are too controlled therefore p's are usually aware that they are in an artificial situation and their responses will not have any consequences. 

Foster (1994)- (BANK ROBBERY)

Aim: to see if witnesses more likely to be accurate if they believed their evidence would influence a conviction.

Procedure: p's watched a video of a robbery and had to pick out the robbers from an identity parade. They were told it was genuine whereas the others knew it was a simulation.

Result: p's more accurate in condition where testimony had consequences. 

Evaluation: artificial. 

It is an indication that factors operating in real life situations are different than those in experimental situations. 

2. Previous experience:

the witness may combine what they had seen in the past and what they had recently seen together and create a whole new scene.

3. Individual differences:

Some people more susceptible to misinformation than others.
Tomes and Katz (1997) 'share the following characteristics'
- Poor recall for event
-score high on measures of imagery vividness
-score high on scores on measures of empathy
People also resist misleading information if it is blatently incorrect.
Loftus (1979) (RED PURSE)
Aim: To test misleading information
Procedure: gave p's set of slides of a red purse being stolen. They were then given an account of event including errors e.g. 'the purse was brown'
Result: In a recall test all but two of p's resisted the misinfo.
Conclusion: memory for info particularly striking at the time is less susceptible to effects of misinfo than memory for peripheral details.

4. Age of witness:

Children are susceptible than adults to absorbing post event info into original memory.
Poole and Lindsay (2001) (SCIENCE DEMONSTRATION)
Aim: To find out if the same factors that affect accuracy in adults also operate in children and to investigate whether children are more susceptible to absorbing post event info.
Procedure: engaged 3-8 yr old into science demonstration. The parents read them a story which contained elements from the demonstration.  They were then questioned about the demonstration. In a second phase of the exp the children were asked to recall where they got the info from.
Result: they incorporated new info from story in original memory.
Phase 2 the older kids revised their account of the demonstration however younger children couldn't do it.

Gordon (2001): (KIDS ACCURATE BUT SUSCEPTIBLE TO SUGGESTION)
'children provide detailed and accurate witness statements but susceptible to suggestion and account should be reviewed with caution.'
Davies (1994): (KIDS VALUABLE TESTIMONY)
'differences between kids and adults show kids provide valuable testimony provided care is taken in interviewing process.'

http://www.youtube.com/watch?v=f41JjnYkruk