Factors Affecting The Accuracy Of EWT

Factors Affecting The Accuracy Of EWT

1. Reconstructive Memory

   • Schemas
     • Bartlett: memory is not always a complete reproduction of an event. We reconstruct various pieces of info and use schemas to fill in the blanks
       • Schemas are small packets of information that we build up about the world that helps us make sense of it
     • Carmichael: Drawing of an object. P was asked to draw either 'a' or 'b'. Their visual representations were altered by language, so a schema affected our memory
       • Not a real life situation
     • Allport and Postman: Invesitgated stereotypes. Black man without knife but P's recalled that he did, shows that stereotypes can effect EWT
       • Not a real life situation
     • Tuckey and Brewer: video of a bank robbery. P's were better at remembering stereotypical bank robbery things "male robber, getaway driver" showing that stereotypes affect EWT
       • P's were also good at remembering counter-stereotyped information
     • Tuckey and Brewer, Cohen and Allport and Postman were all Lab Experiments which do not reflect real life. Would P's react the same way in a real-life setting
     • Allport and Postman only used white P's. Can't generalise to other cultures. Other cultures may have different schemas
     • Allport and Postman took place in 1947. This is old data, schemas may have changed over the years
     • Ethical issues within the studies. Invasion of privacy, lack of informed consent
   • Leading Questions
     • Loftus and Palmer: found that Leading Q's affect EWT as they provide 'post-event' info which is integrated with the original percepton
       • Speed is complex thing to judge, so P's more prone to be led by LQs
     • Loftus and Zanni: Did you see a/the broken headlight? 7% reported seeing 'a', 17% reported seeing 'the'. LQs can make Ps question whether they actually saw somethin
       • Video, doesn't reflect real life
     • Loftus: 98% of P's could remember the colour of the bag, and their statements were not changed, even when experienced to LQs
       • Colour is more objective to speed estimates
     • Alot of the studies are in artificial settings. Generalisability issues
     • Yullie and Cutshall: interviewed 13 people who witnessed a real life robbery. Included to LQs and found P's were not led by LQs which could show that EWT is not affected by LQs in real life situations
     • Cognitive interviews have come about a result of research into leading questions. Successful applications

2. The Role of Emotion

   • Does emotion improve or impede memory?
   • Freud would say repressed memories which are high in emotion would be blocked out if they are too painful or traumatic
   • Deffembauer: meta-analysis on studies of EWT and found that high stress had a negative impact on recall
   • MacLeod: 379 EWT reports of assualts compared to crime with no injury. No overall difference between accuracy of recall suggesting emotion does not make a difference to recall
     • Issues with privacy
   • Christianson and Hubinette: found witnesses for real bank robberies who had been threatened had better recall than onlookers
     • high ecological validity
   • Yullie and Cutshall: emotion enhances your recall of an event
   • Brown and Kulik: flashbulb memories, long lasting memories which occur during heightened emotion. Such as JFK's assassination, 9/11 etc. People are able to recall what they were doing at the time which supports the idea that emotion improves memory
     • Odiniot: interviewed 14 witnesses of a robber. 84% of information recalled was correct, and those who reported heightened emotion had better recall
       • Talarico and Rubin: 54 students recalled memories of 9/11 then of a less emotional events. Their recall (at periods of 1, 6 and 32 weeks) hardly differed for each event
         • P's reported their 9/11 recall as much more vivid
   • Loftus: Weapon Effect, showed a series of slides of a customer in a resturant. In one, customer had a cheque book, in the other a gun. P's were less likely to identify the customer if he had the gun, which shows emotional scenes can throw our EWT
     • Johnson and Scott: P's heard an argument then a man came in holding a gun/grease-covered pen. 49% in pen condition identified the man, 33% in gun identified. This supports Loftus study
       • higher ecological validity
       • More emotionally involved
   • Evaluation
     • Different people experience different emotion. How can we generalise?
     • What is emotion? Are we measuring fear, anger, anxiety
     • Yerkes Dodson Curve shows the relationship between emotional arousal and level of performance
     • Weapon Focus doesn't always have an effect. Yullie and Cutshall found weapon effect did not happen
     • Weapon effect is dependent upon the length of time of the event. Hostage situations would give the P longer to identify the criminal