Prof. Greg Francis 1/2/13 PSY 200: Intro. to Cognitive Psychology 1 Purdue University Memory discrimination PSY 200 Greg Francis Lecture 19 How to take a test. Purdue University Discrimination Many cognitive tasks require you to discriminate between events/stimuli Is this a real smile? Is this fruit ripe? Is the stapler on the desk? The same kind of discrimination is required for memory Purdue University Discrimination Discrimination is difficult because memories can come from lots of different sources Consider so-called “False memory” studies 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 Purdue University False memory 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 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 in the list some words were not seen » including the special target Purdue University False memory 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 CogLab data (193 participants) Type of selected items ! !Percentage of recalls! In original list ! !!75.5! Normal distractor (not in list) ! 7.4! Special distractor (not in list) !70.3!Purdue University False memory These types of findings suggest that our memories are not necessarily accurate, we can remember things that never occurred able to be manipulated, to a certain extent, I can make you have certain memories Why does the false memory effect happen?Prof. Greg Francis 1/2/13 PSY 200: Intro. to Cognitive Psychology 2 Purdue University False memory With every to-be-remembered item you store some information, but not only information about the item Other information is automatically generated as well smooth, bumpy, road, tough, sandpaper, jagged, ready, coarse, uneven, riders, rugged, sand, boards, ground, gravel Memory: Item 1 Memory: Item 2 Memory: Item 3 smooth baby rough smooth baby rough bumpy road smooth baby rough bumpy road tough nails Purdue University False memory At the end of the trial, you have a lot of items in memory that are related to the list Some of them are items that were actually on the list and some of them are items that were “generated” but not actually on the list Reporting all items from memory is not going to lead to good performance Both types of memory items are real, but only one type matches the physical stimuli Good performance on this task requires discrimination between memories generated by physical stimuli and memories generated by internal processes Source monitoring Memory smooth baby rough bumpy road tough nails jagged peaks sandpaper gravel stones uneven Purdue University Discrimination Good memory recall usually requires not only recall of an item from memory You also must identify the correct item relative to the appropriate context or time frame The current trial The context of the experiment Relative to an earlier event At a particular moment in time Purdue University Interference Retroactive interference (RI) new information prevents recall of previous information e.g., Overwriting a computer file. Proactive interference (PI) prior learning prohibits new learning e.g., Learning new cultural customs. Purdue University Proactive interference May be due to a variety of effects One is that memory involves discriminating new from old Visual memory See a set of photos Then see a test photo and decide if new or old Memory: Trial 1 Memory: Trial 2 Memory: Trial 3 Purdue University Proactive interference Suppose you see this building as the test on the third trial You have a match in memory, but is from trial 1, not trial 3 You may report it having been shown in trial 3 In general, previous trials make memory discrimination more challenging Memory: Trial 1 Memory: Trial 2 Memory: Trial 3Prof. Greg Francis 1/2/13 PSY 200: Intro. to Cognitive Psychology 3 Purdue University Proactive interference One finds proactive interference for lots of memory tasks CogLab serial position experiment I looked at recall of the first letter in each list, averaged across all students 0.60.650.70.750.80.850.90.9511 2 3 4 5 6 7 8 9 10 11 12 13 14 15TrialProportion of times first letter is recalledPurdue University Release from PI Proactive interference weakens for different stimulus types Run two Brown-Peterson type experiments XJF WRM DBL NRX Control XJF WRM DBL 942 Experimental Trial 1 Trial 2 Trial 3 Trial 4 Purdue University Release from PI Trials 1-3 show build up of PI Experimental group shows release of PI on Trial 4 0204060801001 2 3 4ControlExperimentTrial Percentage correct Purdue University Release from PI Works for many kinds of memory tasks Many kinds of stimuli 5, 7, 9, 1 vs 5, 7, 9, HAND NICE, SUNNY, ENJOY, PUPPY vs NICE, SUNNY, ENJOY, KILL D, D, D, D vs D, D, D, F News stories Purdue University Memory system Every memory system must have at least two components/processes Storage retrieval We have described proactive interference as being due to difficulty discriminating new items from previous items But there is an alternative explanation Proactive interference might prevent items from being stored and thereby make them unrecallable Purdue University Working memory For example, working memory has a storage interference hypothesis for the phonological loop Working memory suggests that interference can occur by blocking ACP rehearsal (articulatory suppression, Brown-Peterson task, word length effect) within the PS when items sound similar both of these block the storage of items (items fall out of the loop)Prof. Greg
View Full Document