Aims and Context
Cognitive psychology can be linked to memory. What is memory? - the retention and use of prior learning. After an incident eyewitness memory (e.m) goes through 3 stages - enconding (info of people and event in long term mem), storage, and retrieval. But - Probs with mem = EWT - descript by witness - the incident. EWT - imp area of study in cog psy + mem studies + used in trials as evidence, but the validity of the evidence - affected by many psycholog errors e.g. mem of the event - altered by info gained after -> concerns with reliability. Also - EWT can be inf by leading questions (Q that prompts an answer - colour of the scarf? Vs. Was wearing scarf?) - 1 = sugg was wearing one. Contributes - concerns with EWT - L and P partic interest in this. 1973- 74% covicted relied soley on EWT
Marshall - found that when airforce personell knew in advance they'd estimate the speed of a vehicle (12mph)(answers ranged from 10 - 50) - wide range - speeds - susceptable to leading Qs. Filmore - use diff words to see if it effects how fast we think the vehicles moved - smashed/hit etc.
Aims - investig. accuracy of mem. in particular wanted to investig. effect of leading Qs on est. - speed. EXP1 - see if lead Qs distort accuracy os EW recall. EXP2 - see if lead Qs created response bias or if actually alter persons mem.
EXP1 - 45 ppts - all of which - students. split into - 5 grps - 9 ppts in each - they were shown 7 short film clips of different accidents from police films. segments ranged 5 - 30 secs. for each group the order of the film clips differed. after each clip ppts asked to complete Qaire. among th Qs was 1 crit Q= how fast were the cars going when they _____ eachother. the word in the space varied from group to group, and when all put together used the words: hit, smashed, contacted, bumped and collided. ppts estimates of speed recorded - mph.
EXP2 - involved a new set of 150 sudent ppts - part 1 - ppts shown a film of a mutliple car crash. the actual accident lasted 4secs. then asked a set of Qs including one criticle Q about speed. ppts divide into 3 groups (50). group 1 = smashed. group 2 = hit. group 3 = control - no crit Q. Part 2 - one week later ppts asked to return 2 lab and answered more Qs on the filmed accident. inc. 1 crit Q "did you see any broken glass. no broken glass but those who predicted higher speed presumably might answer yes.
Findings and Conclusions
- F - EXP1 - highest mean est speed - Q included - smashed - 40.8 mph. the lowest mean est speed - Q included - contacted - 31.8mph. collided = 39.3, bumped = 38.1, hit = 34.0
- C - EXP1 - smashed - (most violent verb) - reflect in highest mean est speed. contacted = least violent - reflect - lowest mean speed. relation= stronger verb - higher est of speed. so wording of Q does effect mem of event - more severe words - higher est. L+P sugg 2 poss explans for the results of exp1 - response bias - diff speed ests occurred b/c crit word - influ/biases pers response. also said - mem could of been altered - crit word changes pers mem + how they perceive the acc. i.e. more or less severe. to prove 2nd point - EXP2.
- F - EXP2 - smashed condition - 16 report - said broken glass - 34 - not broken glass. hit - 7 broken glass - 43 - not broken glass. control - 6 broken glass - 44 not broken glass. results show - verb used - orig Q inf whether ppts thouught seen broken glass.
- C - EXP2 - sizable minority in smashed cond saw broken glass - perhaps word - smashed - associated with broken glass. fewer in hit - not broken glass - b/c perhaps - hit not associated with broken glass. although sim 2 the no.s for hit cond the least amount of ppts tht reported seeing broken glass was in the control group. sugg even without post event info being sugg to us our mems of events may not be 100% acc. Res showed tht the way a Q is asked - inf answer - not due resp bias - all ppts asked if seen brken glass - sugg leading Q altered mem of event.
method - lab exp - strength - precise control of extran variables - (location, video, Qs asked ,IV and DV - cause and effect) - s b/c - IV tht effect DV. strength -allowed L+P to vary posit of crit Q (EXP2) and the ord of film clips (EXP1) - s b/c it helps reduce order effects and increase internal validity. weakness - took place - lab - sit - artificial - not real life - ppts didnt have any pers invol in the events and so the judgement of speed may be diff b/c not how usually witness events - no stresses arise like with real life sit of a car crash - w b/c lack of ecol valid and low mundane realism -makes findings diff to gen to real life. weakness - ppts aware in study - dem char - effect reduced - embed crit Q among other Qs.
PPTS - all students - weakness - may be younger than the gen pop - diff educ experiences - used to remembering info and so better at mem tasks. may have also - diff driving exp. - diff est.s of speed. w b/c may not gen to population. Strength - independant grouping used - s b/c reduces demand characteristics as ppts only experience one cond. and so are less likely to guess the researchers aim. weakness - only 9 ppts used in each group - small sample, w b/c not generalise to entire population - as group s small - may have certain characteristics in common - may perform diff to another group.
Ethics - L+P - not informed consent - deception - may cause distress. However - not tell true aim - make study meaningless - any distress caused - relatively harmless - no injuries caused.
- Support -
- Loftus and Zanni - in a lab exp showed ppts a film of a multiple car accident. Asked - A/THE broken headlight? - no broken headlight really - however, people were more than twice as likely to say that they had seen THE broken headlight (17% as opposed to 7%) - support L+P's - b/c it shows that misleading post event info has an impact on memory. it is also frequently asked, that if the ppts could be mis-lead on something as simple as this - how much more could they be mis-lead on if they were given more complex info. Both of the methods used in the studies = similar b/c same strengths - controlled, precision and replicability.
- Loftus - showed 150 ppts a film of a car crash - afterwards - put into 2 groupsand each group asked 10 Qs about what they had seen. Group 1 - asked Qs directly related to the film. Group 2 - all the same Qs apart from one that said, "how fast was the white sports car going when it passed the barn when travelling along the country road?" - misleading - no barn. After a week the ppts were asked a further 10 Qs - including, "did you see a barn?" 17.3% of those in group 2 said yes and only 2.7% did in group 1. The resultsa imply that the barn had been falsely remembered because of post event info was mistaken for being part of the actual accident. So similar to L+P b/c in both studies the original representation of the event had been altered.
Alternative research cont.
- Refute -
- Christianson & Hubinette - questioned 110 witnesses, which between them had witnessed 22 bank robberies. some of them had been onlookers in the bank, whereas others were bank employees who had been directly threatened. Victims were more accurate in their memories and remembered more detail about what the robbers wore and their behaviour as opposed to the bystanders. this means that people are better at remembering events that cause them direct stress and in real life situations. So it contradicts L+P's findings b/c it shows how in real life, witnesses can recall info accuratly when in a real life sit - in L+P's study the ppts had no real involvement and so it cannot bee generalised to fit every situation. But it doesnt show us whether post event information or leading Qs had anything to do with the answers given.
- Loftus - in a lab exp - showed ppts a set of slides that showed the theft of a large red purse from a handbag. they were then asked to read an account of the theft from an alleged professor of psychology - the story gave the errror that the purse was brown. it was found that 98% of the ppts correctly remembered that the purse was red. this contardicts L+P's findings b/c it found rthat eyewitness recollection is not as easily corrupted as L+P's research would suggest. But it could be argued that therre is more accuracy in this piece of research b/c the colour of the purse is quite a basic feature, whereas speed is a more complex estimation.