week 8 - models of memory - working memory Flashcards
working memory
proposed by?
baddeley and hitch 19974
working memory
what is STM for?
not just passive retention
neccessary for lanuage understanding, mental arithmetric, reasoning, problem solving
—> emphasis on working
essential to performing tasks that are not just memory
working memory
STM - mulitcomponent
different systems
the single STM store in the multicompenemt model is replaced by 4 working memory components
working memory
4 working memory components
central executive (attentionally limited “control” system)
2 “slave” systems - phonological loop and visuo-spatial sketchpad
later also added: episodic buffer
working memory
4 working memory components - summary
central executive
modality free
central pool of mental resources
control and decision processes
cf. attention
no real storage
working memory
4 working memory components - summary
phonological loop
inner ear and voice
verbal rehearsal
info=speech based
working memory
4 working memory components - summary
episodic buffer
holds anf integrates diverse information
working memory
4 working memory components - summary
visuospatial sketchpad
inner eye
spatial and/or visual coding
working memory
components relatively ……
independent and of limited capacity
working memory
testable predictions:
if two tasks use the same component …
… they cannot be performed successfully together
logic of dual task experiment
working memory
testable predictions:
if two tasks use different components…
… it should be possible to perform them as well together as separately
logic of dual task experiment
working memory
phonological loop
what
tempory storage of sppech like info
(verbal STM)
working memory
phonological loop
2 primary structures
phonological structure
articulary loop
working memory
phonological loop
- phonological structure
(inner ear)
temory storehouse
passive
limited in time (+-2secs) and capacity
code = speech based
working memory
phonological loop
- articulary loop
(inner voice)
active rehearsal component
linked to speech
(it is refresh of the phonological structure)
working memory
phonological loop
purpose
holding onto information
language acquisition device
working memory
phonological loop
diagram
see notes
circular arrow —> phonological store ——————-
indirect access (visual presentation) needs subvocal articulation
direct access (auditory presentation)
working memory
phonological loop
evidence
phonological similarity effect
word length effect
unattended speech effect
articulatory supression
working memory
phonological loop
evidence: phonological similarity effect
errors more likely to be phonologically similar to correct item
- F for S, B for G
more likely to misremember if items in list sound similar
- DBCTGP harder than KWTQLR
- mad cap man map harder than pen day cow bar
(badeley 1966)
—> items in phonological store based on phonological code
—-> reduces discriminability of items in store
working memory
phonological loop
evidence: word length effect
memory span for short words is greater than memory span for long words
(able to recall more shorter words)
working memory
phonological loop
evidence: word length effect
- what causes it?
is the word length effect caused by articulation duration or related to the no. of syllables in the word?
- longer words usually have both (longer articulation as well as more syllables) the two are correlated
the word length effect is due to articulation duration
–> therefore it is evidence for the phonological loop
- span for words with shorter articulation is greater
- the span is as many words as the articulatory loop can actively rehearse within 2 seconds (which is how long the phonological loop lasts)
working memory
phonological loop
evidence: word length effect
- real example impact
people who have a language with faster speaking rate (chinese) seem to have a larger span
simply because they can actively rehearse more words in the articulatory loop
working memory
phonological loop
evidence: unattended speech effect
performance impaired if other verbal material needs to be ignored
- even with nonsense syllables
- even in a different language
- even with vocal music
- less with instramental music
- not with (very loud) white noise
(salame and baddeley 1982, 1989)
—> irrelevant spoken material can gain access to phonological store
—> filter to distinguish between noise and speech
working memory
phonological loop
evidence: articulatory suppression
articulatory suppression is preventing rehearsal by overt or covert articulation
- say “thethethethe” while learning words (out loud or in head)
the consequences are different whether the to-be-learned items are presented visually or auditory
- this relates to the two ways info enter the phonological loop
1) auditory presented info has direct access
2) visually presented info needs to be sub-vocally rehearsed first
working memory
phonological loop
evidence: articulatory suppression
- while learning items presented visually
result in the speaker not being able to use sub-vocal articulation
that is evident by the word length effect disappearing
(this effect relies on sub-vocal articulation)
working memory
phonological loop
evidence: articulatory suppression
- while learning words presented auditory
have no effect on sub-vocal articulation (items have direct access to phonological store)
that is evident by word length effect not disappearing
working memory
phonological loop
evidence: articulatory suppression
what happens to performance overall
overall worse performance
as usual when doing two things at the same time
working memory
visuospatial sketchpad
what
a system for setting up and manipulating images and spatial movement
limited capacity for processing spatial, visual and kinaesthetic information
working memory
visuospatial sketchpad:
two components
visual cache
inner scribe
working memory
visuospatial sketchpad:
two components - visual cache
visual involvement about shape and colour
(the what)
working memory
visuospatial sketchpad:
two components - inner scribe
spatial and movement information
(the where)
working memory
visuospatial sketchpad:
two distinct components…
.. interference depends on secondary task
(dual task experiments)
working memory
visuospatial sketchpad:
function
construction, maintenance and manipulation of mental images
—> isomorphic (similar attributes) relation to perceptual images
imagine a rabbit next to an elephant/fly:
does it have whiskers?
- better at identifying the components of an image if the mental image in large
working memory
visuospatial sketchpad:
mental scanning
manipulation of info in the VS sketchpad (kosslyn 1978)
- learn map of island, keep in mind, now go from well to tree
- mental scanning between two imagined landmarks increases linearly as the distance between them increases
working memory
visuospatial sketchpad:
neurophysiological findings
eg, difference brain areas active during visual (occipital) and spatial (parietal) working memory tasks
but:
extensive interaction with other cognitive systems and the central executive (really independent?)
working memory
central executive:
what?
attention system
- maintain task goals and goal-related info. and use this to direct / bias your processing
working memory
central executive:
attentional capacity is …
… limited
working memory
central executive:
most important?
most important and active component
(but not well understood)
working memory
central executive:
tasks:
Directing attention to task, updating task at hand
switching / shifting between strategies
selective attention and inhibition
working memory
central executive:
where?
probably in pre frontal cortex
(not exclusively)
patients with frontal lobe damage: problems of attentional control
working memory
central executive:
evidence
dysexecutive syndrome
alzheimer’s patients
working memory
central executive:
evidence - dysexecutive syndrome
disruption of CE due to frontal lobe damage
1) perseverance: sort deck of cards by suit, ok: cannot then sort by value (cannot interrupt ongoing schema)
2) utilization behaviour: automatic responding to cue in environment (fails to focus attention)
3) catatonia: remain motionless and speechless for hours (unable to initiate schemas)
patients do not share a single pattern of impairment
- suggests different parts of PFC are responsible for different aspects of executive functioning
working memory
central executive:
evidence - alzheimer’s patients
problems with distributing attention between 2 tasks (CE function)
- 2 tasks together: disproportionate decline in performance
- can follow conversation with one person, much more difficult if more
working memory
central executive:
one unitary system?
probably not one unitary CE system but multiple executive mechanisms
working memory
episodic buffer:
what?
phonological loop and VS sketchpad are modality specific (verbal vs visual/spatial)
episodic buffer can integrate info into single complex structure or episode
working memory
episodic buffer:
how much info can it hold?
about 4 pieces / chunks of info in multidimensional code
working memory
episodic buffer:
assists in …
… binding
integrating info about location, size, colour, smell, size, feels of objects, scenes
working memory
episodic buffer:
interacts with both…
… perception and LTM
working memory
evaluation
have recast STM (multistore model) as working memory
this is progress
- rehearsal mandatory process –> optional rehearsal in the phonological loop
- forgetting only due to decay –> interference effects
BUT: cannot forget that we are all individuals and differ from eachother –> different working memories (individual differences research)
working memory
capacity
many different ways to capture WM capacity
1) operation span (solve maths probs/remember word)
2) reading span
- read sentences (processing) and remember last word (storage) of each sentence
- related to language comprehension across the life span
3) corsi block task (for visuospatial memory)
working memory
capacity correlates with…
… fluid intelligence
working memory
capacity related to…
… attention control
high capacity individuals less likely to be distracted by external (eg. outside noise) and internal (eg. mind-wondering) stimuli
