working memory 2 Flashcards
STM capacity
Miller (1956) - magical number 7
memorise series of letters in a sequence, average = 7 +/- 2
2 ways to boost STM capacity
active rehearsal
chunking
WM capacity
Cowan’s (2001; 2010) magical number 4
only representations in focus of attention (centre of network for spreading activation) are available to conscious awareness
average capacity limit of adults focus of attention = 4 +/- 1
set-size effect
larger set = worse performance when recalling strings of letters
later items in the set interfere with the earlier ones - harder to hold into
in both simple and complex span tasks
limits of working memory (3 - list)
decay
interference
limited resource
WM decay
rapid decay with time
Thorndike’s law of disuse
time-based decay = representations get weaker with time (as a linear function)
first presented unit is strongest as it has had the longest time to be rehearsed over and over
primacy and recency effects seen because of this
does time cause WM decay
REMEMBER: time passing doesn’t cause decay, it is correlated with processes that cause forgetting
- processes cause forgetting and these are associated with time
WM decay - restoration mechanisms (2)
rehearsal - subvocal repetition of memoranda to maintain them
refreshing - think of memoranda to keep memory traces active
WM interference
mutual interference between representations limits WM
seen in set size effect - more units means they overlap and compete with each other
WM transience (2)
proactive interference = old impairs new(remember old)
retroactive interference = new impairs old (remember new)
3 types of WM interference
confusion
- present array of items verbally or visually
- need to remember position of items
- confusion when you get the position of an item wrong
superposition
- items that seem similar but are dissimilar in terms of positioning
- try to determine positioning of item
- less similar = worse performance
overwriting
- more similar = worse performance
- e.g. cat, bat, mat = easily confused
WM limited resource
capacity determined by limited quantity of resource that enables holding representations available
what is a resource
resource = limited quantity that enables cognitive function/process; probability of success increases the larger amount of resource assigned to it
- more resource given to an item = remembered better
2 resource models of WM
slot models
*egg-carton metaphor
* resources distributed in discrete units (defining the number of items you can store)
* quality of retained representations is high but not perfect
* each item gets equal share of the resource
flexible resource models
- resources are distribute flexibly, this allows for
- small number of high quality; or
- high number of low quality
2 ways WM resource can be allocated
discrete = slot model = allocation to limited number of items with no info stored about additional items
continuous = flexible resource model = equal spread of resource among all items, fewer resource per item in larger arrays
which hypothesis for WM capacity is best
none can explain all findings:
set size effect + complexity of items
set size effect when delay = 0s
domain specificity
cross-domain set-size effect
heterogeneity benefit
variation in WM capacity - who has greater capacity (4)
older children > younger children
young adults > older adults
healthy > frontal lobe damage
some people > others of same age
importance of studying variation in WM capacity
correlates with complex cog activities
predictive of things
correlations of WM with complex cognitive activities (3)
reading comprehension
reasoning
problem solving
what WM can predict (2)
cognitive development
individual differences in intellectual ability
2 tasks for measuring WM performance and reasoning
WM task:
- complex span task - storage and processing
reasoning task:
- e.g. non-verbal reasoning, finishing a pattern by figuring out the rule
proposed explanations for individual differences in WM (2)
executive attention hypothesis
binding hypothesis
executive attention hypothesis - 2 systems
Shipstead et al (2016)
attention control = executive attention
2 systems:
1 = quick and easy access to all info you know
2 = effortful, controlled processing – attention control system
executive attention in WM and reasoning tasks
single top-down executive attention system underlies WM and reasoning task performance:
EA –> maintenance or disengagement –> task performance
executive attention in a WM task
complex span task
maintenance –> of relevant info (to store the letters) and append to new info on the list
disengage –> from and suppress outdated info (equations) from previous trials
this is the ability to be selective of what info to remember or not
executive attention in a reasoning task
disengage –> from outdated rules/hypotheses between trials to prevent returning to them - new trials have new rules to determine
maintain –> info about the problem, allow systematic hypothesis testing
executive attention hypothesis - summarised
WM capacity and reasoning are linked - both arise from limited executive attention
better executive attention = better performance in WM, reasoning, and other similar tasks
issue with executive attention hypothesis
executive attention tasks don’t correlate well - difficult to directly test hypothesis
e.g. doing one task doesn’t mean you do better in another task of EA - Stroop task and flanker task don’t correlate with each other but both require EA
binding hypothesis for variation in WM
binding hypothesis = system for rapid formation of temporary bindings underlies both WM and reasoning task performance
individual differences due to differences in number of potential bindings
more bindings = more WM capacity
WM –> bound by construction and manipulation of representations and novel structures –> reasoning
binding hypothesis in WM task
complex span task
each letter is bound to its position in the list to be remembered
binding hypothesis in non-verbal reasoning task
objects are bound to semantic space
e.g. blue square and purple circle are bound by colour and shape into a grid (array) - sort by shape and by colour to find missing shape
this is more complex rules/structures than with WM task
binding hypothesis - summary
bindings = temporary links of content representations to places in a mental coordinate system
WM capacity limit = number of bindings maintained at any one time (varies between people) - arises from interference between bindings
less interference = build more complex structural representations - therefore do better in WM and reasoning tasks
issue with binding hypothesis
bindings may be built and maintained with help of executive attention
difficult to directly test hypothesis against executive attention hypothesis
both hypotheses are too entangled and interlinked
therefore need more research into it
