Lecture 6: STS Evidence Revisited and Working Memory Flashcards
What is the duration of STS?
15-20 seconds if not rehearsed
What is encoding like in STS?
- Select from sensory memory
- Retrieve from LTS
- Reduction & Elaboration encoding
What type of code is used in STS?
A.V.L. (acoustic-verbal-linguistic)
What is the capacity of STS?
7 +/- 2 chunks of information
How does evidence with amnesic patients illustrate the existence of separate memory systems?
- “Amnesics can learn” (e.g., Puzzle WM task, teach them how to do it and they can learn. Bring them back a week later, they won’t remember learning the puzzle or meeting you but they can solve the puzzle [shows that something is getting into LTM]. Pricking them with a pin when shaking their hand, a couple weeks later they remember the aversive event [pulled away before shaking Chris’ hand]. They don’t remember you or the event but seem to remember the pin)
How did the study by Kroll et al. (1970) illustrate that coding in STS is more than just AVL?
- Visual codes in STS
- Task: shadowing + test letter (auditory vs. visual presentation)
- Retention interval 1 to 15 sec (during which time they are shadowing)
- Recall test letter
- Results: short interval recall 96%, auditory = visual and long interval: recall visual > recall auditory
- At the longer interval, the performance on visual recall was greater than auditory recall because there was no space for more audio information because they were shadowing. Suggests that the code that was used to keep that visually presented letter was different than the code that was used to keep the auditory presented letter.
- Note: Suggests visual code of test letter (presented
visually) can be maintained in STS.
How did Shulman (1971) show evidence for semantic code in STS?
- Evidence for semantic code in STS
- Task:
- List of ten words (500 ms each)
- Probe word: does it match an item from list?
- Probe type: related (boat/ship) vs. unrelated (e.g., boat / shin)
- Results: false positives when related (thinking a word like ship was on the list because it was connected to boat)
- Suggests: information in STS may include semantic code
What was the original STS/LTS distinction based on and what has been extended?
- Amnesic evidence (has been extended, people with Amnesia can learn and transfer things to LTM even though they suffered extensive damage that would prevent this)
- Code (has been extended; people can be encouraged to use a different. Generally gravitate towards using an AVL code unless otherwise inclined by the task or their abilities)
- Duration (has not been extended, still 15s without rehearsal)
- Capacity (Has not been extended, still 7+/- 2
How did Baddeley describe STS vs. Working memory?
STS: - initially viewed as a storage system - however, the same system seems to be important for active processing and manipulation of information - a new concept needed to extend STS WM: - active system for temporary store & manipulation of information
What are the components of working memory?
- central executive
- phonological loop
- visuospatial sketchpad
What is the central executive component of working memory?
- plan actions (conscious thinking), integrate information, initiate control (over what happens next, i.e., the boss or executive)/decisions
- transfer of info. to/from LTS
- rehearsal, recoding (e.g., from auditory code to visual code)
What is the phonological loop component of working memory?
- recycle information for immediate recall. Responsible for storage & rehearsal of verbal information, & phonological processing
- two parts: Phonological storage/store (passive, will fade away after 15 secs unless you reactivate them) and articulatory rehearsal (active processing)
What is the visuospatial sketchpad?
- Deals with visual and spatial (locating where things are in space) information
- visual imagery tasks
- spatial visual search tasks
- Two parts: visual cache (storage, stays for a short period of time) and inner scribe (processing)
How can arithmetic be used as an example of working memory?
- (4+5) x 2 / 3+ (12/4)
- CE - retrieves rules of computation from LTS and retrieves facts from LTS to solve: (4+5) = 9, x 2 = 18
- PL - temp stores “18”
- CE - retrieves rules and facts from LTS to solve: (12/4)=3, 3 + 3 = 6
- PL - temp store “6”
- CE - retrieves rules, retrieves items from PL to compute: 18 / 6 = 3
What are the 3 general assumptions of Working memory?
- There is only one WM System and it operates on many tasks (math, reading,
driving) - WM has a limited storage and processing capacity
- Demands of task #1 can limit performance on task #2
What are the assumptions regarding the two subsystems?
- Involve simple low-level processing
- rehearsal / maintenance of information - Subsystems are domain specific
- phonological vs. visuospatial - Each subsystem has own limited (small) pool of
attentional resources
- may also drain (rely on) resources from the CE
What is the phonological store?
- passive store that holds verbal information
- forgotten unless rehearsed & refreshed
What is the articulatory loop?
- active refreshing of verbal information
What is the articulatory suppression effect?
- memory for words poorer when concurrently try to
say something responsible for storage & rehearsal - “act of speaking” uses up resources in the
articulatory loop à so words from list cannot be
refreshed and are lost - Real world effect – it’s hard to read while in room
with other people talking
What is the phonological similarity effect?
- memory poor when try to remember words that are
similar - “boat, bowl, bone , bore” harder than “stick, pear,
friend, cake” - why? Similar sound codes get confused in the
phonological store
What is the focus generally on when discussing the phonological loop? what evidence is there for other codes?
- focus is generally on verbal aspects of phonological loop, but also some evidence for other codes:
- Musical pitch: find get confused by similar pitches just like the phonological similarity effect
- Nonverbal communication like ASL
How did the study on ASL demonstrate the existence of other codes?
- congenitally deaf, skilled at ASL
- presented word lists written or in ASL sign:
- word lists were either phonological similar (shoe,
through, new) or cherologically similar (hand
movements) - found memory confusions based on cherological
relatedness! - deaf people coded (& recoded) in ASL, & used ASL
code in phonological loop
What evidence exists for the visuospatial sketchpad?
- Evidence: mental rotation (of 3D objects)
- people shown first figure and must indicate whether second figure is same or different (the second figure has been rotated)
- requires mental rotation
- Picture plane (e.g., left to right; first shape in panel A), depth plane (e.g., back to front; first shape in B panel)
- results: mental (image) rotation can be done. Time to decide same/different varies with how far need to be rotated. The farther people had to mentally rotate the image (i.e., degree of rotation 20 degrees, 60 degrees et.), the more time it took to decide if it was the shape (i.e., linear relationship)
- visual codes can be/ are used to temporarily store
and manipulate information in STS
What is the focus of dual task method?
Focus is on evidence showing separate phonological
loop vs. visuo-spatial sketchpad subsystems
How do we assess working memory?
- Typically based on dual-task studies
- perform two tasks at same time
– if there are not separate WM systems, we would see interference on all combinations of tasks but if there are separate systems we would expect to see selective interference - if tasks do not affect each other, assume utilizing
different WM systems - if task mutually interfere, assume tasks utilizing
same WM system
What was the dual task method study by Logie, Zucco, and Baddeley (1990) do?
Primary tasks: Visual Memory Span: - grid of squares on display - ½ (random) filled in - 3 sec presentation, blank display, new display - new display one square location empty - subjects to indicate (point) which changed - requires visuo-spatial sketchpad Letter Memory Span: - remember 5 letters - requires phonological loop Secondary tasks: Mental Imagery Task: - hold mental image - requires visuo-spatial sketchpad Simple Mental Addition Task: - 25 + 4 - requires phonological loop
What were the results of he dual task method study by Logie, Zucco, and Baddeley (1990)?
- data represents drop in performance in the dual-task
conditions versus original baseline performance
– higher up on graph more drop from baseline to dual-task - graph shows interference was greatest for:
1. two phono tasks (Letter span + Adding)
2. two visual tasks (Visual span + Imaging) - Conclude: supports notion of separate phonological vs. visuo-spatial subsystems
- If there is only one system in working memory then all the tasks should interfere with each other but if we have selective interference (i.e., cant do two phonological tasks at once but can do a visuospatial and phonological at once) then it indicates that there are separate systems that can work simultaneously
What is the neuropsychological evidence for the phonological loop?
Phonological Loop:
- scans taken while rehearsing digits/letters. Mostly left-hemisphere parietal & frontal lobe activation.
- Broca’s area 44, 40 (articulation)
- Premotor and supplementary Motor Area (6)
___________ and _______________ evidence
supports assumption of that there are 3 core WM systems
Experimental and neuropsychological
How did Dobbs and Rule (1987) study central executive and aging?
- Paced Auditory Naming Task (PANT)
- PANT task utilizes the Central Executive
- For older people, performance started to drop around 2 back, on average (old/indepedent and healthy the best, old second best, old not independent/less healthy did the worst). Shows a difference in executive functioning in younger people than older people
- young people can do 2 back and 3 back with very little drop in performance but start to experience drop around 4 back (still much better performance, on average, than older people)
How did Working Memory and Driving Johannsdottir & Herdman (2010) study Working Memory and Driving?
- Simulator driving study: Series of 60 sec trials
- Primary task: Drive and report relative locations of surrounding cars (front v. Back) at end of each trial
- Secondary tasks: verbal (verbal rhyming task) vs. spatial memory (tap a pattern on keyboard while driving, a pre learned pattern)
Results: - Verbal task didn’t interfere with being able to locate cars in front of them however locating cars in front was impaired when doing the spatial tapping task
The performance results were the opposite when locating cars behind - Shows: Verbal WM (phonological loop) required for awareness of vehicles behind. Spatial WM (visuospatial sketchpad) required for awareness of vehicles in front
Why is visuospatial not very useful (and thus doesn’t interfere with driving performance) for information behind us?
- Its really hard to keep a visual representation of information shown to us in the mirror, because its mirror image and you only get to sample it every 7-10 seconds (cant keep staring at it while driving). Thus, visuospatial sketchpad is not very useful for keeping track of info behind you because its confusing and not sampled as often as it needs to be, so people perform well on the tasks. Thus because visual spatial processing is not helpful for the stuff behind you, what do we do instead? Our system uses a different code, what we need to know is someone is close or really close so I cant change lanes (i.e., threat assessment rather than knowing where they are visually. We verbally code the information (“close” or “really close” so I cant slam on my breaks quickly) because we cant easily form a visual representation or sample it often. That’s why the phonological task interferes with locating cars behind us because that’s the system we are using to keep that location information active.
How did Heenan and Herdman further illustrate that visual WM is associated with cars in front of us and phonological WM is associated with cars behind us?
- Simulator driving study with naturalistic conversation task
- Primary task: Drive and report relative locations of
surrounding cars (front vs. back) at end of each trial. - Secondary task: 20 questions conversation task - the driver asks a series of 20 questions to figure out experimenter’s target word (e.g., whale)
- Results: performance did not vary as a results of conversation when reporting relative locations of cars in front but when locating cars behind, having a conversation decreased performance
conclude:
– Conversation utilizes WM and this lowerd driver awareness of other cars
– Conversation especially utilizes phonological WM (the phonological loop)
– The P-loop is required to keep track of vehicles behind them
What implications do these studies have for the real world, like texting and driving?
One of the reasons people believe that Talking on the phone while driving is so dangerous is because not only are you ahving a conversation (phonological loop) which impairs processing of things behind you btu you are also visualizing the person you are speaking with since they are not physically present (visuospatial processing) so you are impaired in keeping track of things in front of you
What is the neuropsychological evidence for the central executive?
Central Executive:
- task switching & dual-tasks
- Brodmann’s area 46 (LH, DLPFC)
What is the neuropsychological evidence for the visuospatial sketchpad?
Visuospatial sketchpad:
- mainly right hemisphere areas including:occipital cortex, posterior parietal lobe, premotor and DLPFC regions
- note: object memory = mainly DLPFC and spatial location = mainly premotor region
