4/6 & 4/8: Working memory.doc

4/8: Working memory II Capacity limits Chunking Working memory capacity isn't expanding, but it's being supported by long-term memory You can increase working memory in a more long-term matter, but it may not last for a very long time Filtering Individual differences in the ability to filter out irrelevant stimuli Can account for individual differences in learning and memory Filtering study Focus on certain stimuli, but not others Can get a measure in how many items the person can encode, bring online, and keep in mind over the delay Used scale EEG Sorted into high capacity and low capacity subjects Looked at neural response at occipital (visual representation) and parietal (spatial representation) lobes For high capacity, higher response for 4 items than 2 items. 2 items w/ 2 distractors is similar to 2 items For low capacity, 4 items = 2 items w/ 2 distractors. Weren't able to filter out irrelevant things Subjects with higher filtering efficiency had the highest working memory Why this is important Tells us where individual variability comes in w/ WM ability Gives a way to understand training data in Jaeggi study What people are being trained to do is use selective attention, such that they're doing a better job just focusing on relevant things Can transfer to problem solving, etc. Fluid intelligence is related because performance on the tasks rely on good attentional processes WM vs. LTM systems Recalling items in any order Function looks like a J Primacy effect: probability of recalling items from beginning Impaired primacy in Amnesia Because it was longer ago Still do have some primacy effect, but it's much less than people without amnesia Items at the end seem to be recalled at the same rate as regular people Factors affecting primacy effect Rate of presentation Remember fewer w/ fast presentation The more time they have, the higher the probability they'll be able to engage in processes that will make it stick List length Familiarity of materials Aging MTL lesions (amnesia) Recency effect: probability of recalling items from the end Filled delay Distracting people after they've heard the list Other things filter in and affect representation in WM Only affects recency effect Evidence for separate WM/LTM stores K.F. Impaired WM, relatively intact LTM Hardly able to remember 3 items Forms of WM Is WM Unitary? Modal model: allocate attention to sensory registers, if we continue to allocate attention, we can rehearse and maintain them BUT K.F. is still impaired in visual domain, but performs a lot better than in the auditory domain E.L.D Stroke to right fronto-temporal region Difficulty finding her way home, memory problems for unfamiliar material Severe deficit in visuospatial memory, but normal auditory WM Spatial WM Right frontal Bilateral superior parietal Maintain spatial information over a delay period Verbal WM Left inferior parietal and temporal Left inferior frontal (Broca's) & premotor Tripartite Model of WM Central executive Visuospatial sketchpad Phonological loop Phonological store: Stores active phonological representations Phonological rehearsal/control Subvocal rehearsal Accesses and maintains representations When you present auditory info, it automatically activates these areas. Phonological similiarity effect when words are similar, recall fewer of them correctly Articulatory suppression PL disrupted by overt or covert articulation of irrelevant phonological codes Blah blah blah hinders digit span PL not disrupted by motor things like rhythmic finger tapping Broca's aphasia Can understand things fine, but can only speak in one or two syllable utterances Dominant model Neurobiology of WM Review session next week Come with questions