Forgetting IIE 269: Cognitive Psychology Greg Francis Lecture 20 How accurate is eye-witness testimony? Fundamental fact There is no method other than object physical evidence to verify the accuracy of a memory Memory is a cognitive experience Confidence in the memory is another cognitive experience You can be very confident and still be wrong Of course, we must be correct fairly often, or our lives would be a total mess! Long-long term memory Subjects from one high school tracked down Using yearbook asked to match names with faces or recall names for faces Percentage correct Long-long term memory Not much forgetting! 120 100 80 60 40 20 ~25 years 0 ~50 years 0 200 400 600 Time (months) Name matching Picture cueing Forgetting Penfield (1959) This suggests that if you learn something well, you have a good chance of remembering it forever! Brain surgeon stimulate brain regions before operating want to know what is being removed Most of the list-type experiments deal with Conscious patients report vivid memories memories that are not of this semi- unable to recall normally permanent type stimulation of temporal lobes even so, there is room for lots of errors These results are fairly well-known, even Let s look at this issue... outside of memory research they are usually misunderstood 1
Penfield (1959) Suggests that memories are stored but normally unreachable (context things again) But the memories are nearly impossible to verify the few attempts find that the memories are not true» people describe places they have never visited, It is more likely that stimulation feels like a memory, even though it is not your awareness of remembrance is a product of your brain it can be stimulated, even without a real memory Highly emotional events tend to produce strong memories e.g. JFK assassination Challenger explosion Oklahoma City bombing Earthquakes September 11, 2001... People vividly recall details surrounding event where they were when they heard what people said clothing worn time of day,.. People are confident about their reports however... Subjects are often incorrect about specific details it is often hard to verify accuracy Subjects have lots of confidence in memory, but that can be misleading retelling of the story ( I remember vividly when Kennedy was shot. I was ) probably reinforced the story, if not the true memory The significance of flashbulb memories is still in question they are a real phenomenon about the experience of memory, but probably not super-memory Hypnotism Conflicting evidence clinical studies find hypnotism helps recover lost memories experimental studies find no effect Primary difference seems to be that experimental studies can measure accuracy of memories, while clinical studies cannot Hypnotism makes subjects more confident about their recall accuracy Hypnotism e.g., subjects view slides demonstrating a crime hypnotize some subjects test recall of slide details test confidence of subjects Accuracy Confidence Hypnotized 11.9 9.0 Normal 12.8 7.3 2
Familiarity Confidence in memories is generally based on a sense of familiarity which is a state of the brain, not a characteristic of the memory Consider so-called studies An example list is smooth, bumpy, road, tough, sandpaper, jagged, ready, coarse, uneven, riders, rugged, sand, boards, ground, gravel the special target is rough, which is not shown to the subject as in CogLab subject views a list of words the list of words have something in common» they are all related to a target word After viewing the list, the subject must go through a set of words and identify which ones were in the just seen list some words were just seen some words were not seen» including the special target The main finding is that the special target is often identified as part of the just seen list even though it was not Sometimes people will even report that they recall seeing the special target but this is impossible because it was never shown Discuss CogLab data (144 participants) Type of selected items In original list 79.7 Normal distractor (not in list) 2.9 Special distractor (not in list) 67.1 Percentage of recalls These types of findings suggest that our memories are not necessarily accurate, we can remember things that never occurred constructive, we build memories from the events we experience, the special target must be related to the words in the list able to be manipulated, to a certain extent, I can make you have certain memories» more easily than you might like to believe! More significantly... Donald Thomson was accused of rape and picked out of a lineup by the victim He was on live TV at the time of the rape The victim had the TV on at the time of rape misattributed the face on TV for the face of her attacker very accurate report of the crime, otherwise There are important situations where it matters that the memory be precise Unfortunately, memory of events are easily influenced Loftus & Palmer (1974) subjects view a series of slides showing a Datsun driving down a a street, stopping at a stop sign, turning right, and knocking down a pedestrian 3
Subjects then answer a series of questions about the slides (two groups, my emphasis added) Did you see the car when it stopped by the stop sign? or Later, show slides and ask subjects if they were part of the original set real slide contains STOP sign fake slide contains YIELD sign Did you see the car when it stopped by the yield sign? Embedded in lots of other questions Misinformation effect Compare accuracy according to pre-test questions Subjects without a misleading question--90% accurate Subjects with a misleading question -- 20% accurate In a follow-up, the experimenters asked those with misleading questions if they thought they were misled» 90% say no Paying money for correctness had no effect! Loftus, Miller & Burns (1978) The misinformation effect gets stronger with a week delay before memory test Loftus & Palmer (1974) Subjects view a movie of a car wreck Answer questions like (my emphasis) How fast were the cars going when they smashed each other? How fast were the cars going when they contacted each other? Follow-up questions Was there any broken glass?» 33% say yes» 10% say yes There was no broken glass. Loftus & Zanni (1975) Filling-in the gaps subjects view a film of an automobile accident answer questions like (my empahsis) Did you see a broken headlight? Did you see the broken headlight? 4
Significance Loftus has a procedure that implants a memory of being lost in a mall Basically just have subject read a plausible story (with some details that could be true) Later the subject remembers the story as something that happened to him/her It is very easy for a therapist to implant false memories into patients Conclusions Evidence of memories over very long periods of time sometimes no forgetting of well-learned events Many techniques believed to provide accurate memories, do not flashbulb memories hypnotism brain stimulation Memories can be easily influenced by questions, interpretation, and context Next time Amnesia Anterograde amnesia Retrograde amnesia Unusual characteristics Repression What s wrong with my wife? 5