working memory 1 Flashcards
working memory definition
storage and processing of information in the present moment
The ability to hold goal-relevant information in mind (active state) for ongoing task in the physical absence of this information
flexible workspace - used to guide behaviour
WM = an ability
WM - flexibility
can hold any info - make arbitrary relationships between items
e.g. an authentication code is arbitrary but we use it in relation to getting access to what we want
stages of working memory experiments
encoding
retention interval
retrieval
WM vs LTM (4)
WM:
active (easy access)
relevant to goal/task
immediate use
limited capacity
LTM:
remote (needs to be cued)
everything learned/remembered
permanent (ish)
unlimited capacity
multicomponent model of WM - components
Baddeley and Hitch (1974)
central executive
visuospatial sketchpad
episodic buffer
phonological loop
LTM
VSSP, EB, and PL all transfer info between LTM and WM
- both directions
multicomponent model of memory - features (3)
hierarchical organisation
multiple components with functional responsibilities
interaction of attention, LTM with present stimulus
central executive function
coordination of storage systems and control of attention to stimuli
storage systems: visuospatial sketchpad and phonological loop function
processing of senses
senses:
visuospatial and auditory information
taste (olfaction) and touch (haptics) are not formal components –> go from CE to sense to episodic buffer (not in VSSP or PL)
episodic buffer function
binding of multimodal information to form episodic memories
assumptions of components of multicomponent model (3)
central executive = flexible allocation of attention
storage systems (VSSP and PL) = domain-specific short-term storage, only deal with info specific to their domain
episodic buffer = binding of information from different sources
phonological loop - more detailed
auditory/linguistic input –>
phonological short-term store <–> subvocal rehearsal (between these 2 is the articulatory loop, goes round between them)
rehearsal is needed to remember information in WM through this articulatory loop
subvocal rehearsal
saying things in your head
word length effect
recall in WM is a function of time
2 seconds => we can remember the number of words we can articulate in 2 secs, if we cannot rehearse them in this time they decay
can recall more short words than long words - as long words take longer to articulate (say or sub-vocalise)
[think of experiment in lecture with memorising country names]
word length effect - Welsh
Ellis and Hennelley (1980)
- bilingual english and welsh speakers
- can recall more english than welsh digits
- as welsh digits have longer spoken duration
word length effect - chinese
stigler et al (1986)
- better chinese digit span
- chinese digits have short spoken duration
conclusions from word length effect studies - welsh and chinese
language can have profound impact on memory
phonological similarity effect
tendency for recall to be depressed where the items “sound” similar in WM - semantic similarity
similar meaning does not effect working memory
suggests that coding is phonological
semantic similarity study
assessed recall for semantically related vs unrelated words
with interference - backwards counting task between encoding and recall - or without
results:
relatedness protects against interference
sematic links strengthen ability to recall information
these are inconsistent with previous findings for multi-component model
articulatory suppression
the uttering of an irrelevant word (“the, the, the..”) whilst being presented with words to remember
word length effect and articulatory suppression
articulatory suppression abolishes the word length effect with visual presentation – participants can’t transform words into phonological codes – because phonological loop is used up by saying other words so cannot be used to memorise words
word length effect is NOT abolished with auditory presentation - presumably as words enter straight into phonological store so speech doesn’t interfere
suggests suppression occupies the articulatory control processes (for visual presentation) but does not prevent direct access to phonological store (for auditory presentation)
is there a phonological loop in deaf signers
(in core reading too)
evidence shows PL exists in deaf signers
4 signature effects on deaf signers immediate recall of signs:
- phonological similarity effect - similar motions of signs
- word length effect - how long a word takes to sign
- articulatory suppression effect
- irrelevant speech effect
proposed a sign based phonological loop - the sign loop
2 elements of the sign loop (deaf)
PL in deaf signers
sign-based phonological store - codes e.g. hand shape, orientation, location, movement
manual articulatory rehearsal mechanism - refreshes info in the phonological store
visuo-spatial sketchpad - two types of info
what type of study used to show this
visual imagery = what
spatial information = where
can study with mental rotation task
mental rotation task
Shepherd and Melzer (1971)
for VSSP
- shown two images of 3d objects
- asked is the object the same object, different, or are they mirrored (mirrored = enantiomorphs)
- involves visual and spatial memory
results:
3 types of trial:
- picture-plane rotations = image rotates = visual - viewing images as they are now
- depth rotations = 3D rotation of shape - new perspective = spatial - more related to 3D modelling
- control
- the more an object has been rotated from the original or more spatially changed, the longer it takes them to determine if the two images are of the same object or mirrored
- linear correlation between reaction time for same pairs and angle of rotation
blind and sighted peoples VSSP as evidence for functional distinction of visual and spatial
blind people generate spatial representations of environment almost as accurately as sighted (Schmidt et al., 2013)
this uses same mechanisms as sighted people but without visual input - shows distinct systems
maintain sense of where you are in environment without visual stimulus - spatial ability without visual imagery
separate visual and spatial systems study
Klauer and Zhao (2004)
participants did one of two tasks:
- memory of dot location on a grid (spatial)
- memory of Chinese characters - which was presented (visual)
sometimes complete task with colour discrimination task (visual interference)
sometimes complete task with with movement discrimination (spatial interference)
predictions
if there are separate visual and spatial systems:
- spatial interference task should disrupt performance more on spatial than visual task
- visual interference task should disrupt performance more on visual than spatial
results
- supported predictions
- movement interfered with both task - more with dots than characters
- colour interfered with characters but not dots location
idea that domain specificity causes conflict –> visual-visual task doesn’t work well due to being overworked
issue with CE assumption of flexible allocation of attention
CE is a homunculus - a critical part of the model that is not explained further
homunculus = “small human like thing” - like a small person in your head making decisions and controlling things
doesn’t say how it does that despite WM relying on that
issue with the assumption of storage systems as domain-specific short-term storage
are two separate domain-specific storage systems needed e.g. VSSP and PL
the fact that both are split down further too
also other senses are needed - haptics, taste, smell
idea of less is more - why split it up so much
WM study with storage and processing
complex span task to assess storage and processing of information
- present a letter to remember (storage)
- present a maths equation to say if it makes sense (processing)
- another letter to remember
- maths equation again and say if it makes sense
- letter to remember
- maths equation
etc
at the end, recall all stored letters
accuracy of recall = DV
this requires switching between using working memory to store and manipulate information
testing domain-specificity of storage systems - method of testing
use complex span task to see if tasks conflict with each other
use verbal and/or visuospatial info for storage and/or processing to see if performance varies - due to conflict
e.g. verbal - verbal = same domain
e.g. verbal - visuospatial = different domain
storage
verbal = letters, numbers, words
visuospatial = location on a grid, shapes, patterns
processing
verbal = decision about equations, lexicality, rhyme
visuospatial = decisions about symmetry, orientation, identity
domain-specificity testing - predictions
domain-specificity predicts lower recall performance for same domain than different domain
due to overloading of component of WM - resource is already used up
Vergauwe et al (2022) - domain-specificity improved study - rationale
rationale = previous studies used task combinations that vary on more aspects than just the representational domain - response modality required –> this confound might have affected the results
previous experimental design issues
Vergauwe et al (2022) - domain-specificity improved study - method
complex span task
verbal storage = auditory nonwords (e.g. lorro) - remember these - at the end click and select which words were presented to them
verbal processing = make judgement of if two letters are rhymes e.g. B+D are, A+J are not
visuospatial storage = remember locations in a grid - click where this was at the end
visuospatial processing = asked whether two letters share axis of symmetry e.g. CD do (horizontal) but AC do not (vertical and horizontal)
Vergauwe et al (2022) - domain-specificity improved study - results
no difference between same-domain and different-domain combinations
therefore there may be no need for two separate storage systems
however, lacks validity as it is the only study on this - needs more support
this is just the start of challenging the assumptions of the model
issue with the assumption that the EB is the binding of info from different sources
are two separate memory systems needed (WM and LTM)
e.g. no effort needed to remember some LTM things so is this actually being drawn out of LTM into WM or is there more overlap
Occam’s razor = explanations with fewer assumptions are preferred
LTM as a spread of activation
representations in LTM are activated by similar words which are linked in a network
e.g. “red” activates concepts such as: colour, roses, apples, fire etc.
these other activations then spread further across linked network
Cowan’s embedded-process model (1995)
2 embedded components
- focus of attention = current contents of working memory
- activated part of long term memory = nearest terms in network to it
WM holds a limited amount of info temporarily in a heightened state of availability for use in ongoing info processing
idea of WM as a subset of representations in LTM, not a separate system
activate specific parts of LTM when we need it - more focussed - becomes more easily accessible when activated
Cowan’s embedded-process model (1995) - with a complex span task
idea that letters to remember (storage) are not new concepts - often used symbols and so will be in the “focus of attention” and easily accessible - already have these concepts that just needed activating
but this would then spread and activate related things - how would a single representation be selected for processing and recall? - flaw of this model
Oberauers (2009) - three-embedded components model
picture as a bullseye over a network of spreading activation
added layer to cowan’s model which allows for activation of a single representation
narrow focus of attention = one currently selected representation - can be recalled and processed
broad focus of attention = region of direct access to representations bound to current context - readily activated for when it is needed
activated part of long-term memory = wider spreading activations
Oberauers three-embedded components model - with a complex span task
+ unanswered questions (3)
the specific letters needed for storage and recall are in the narrow focus of attention
- unclear how many representations are in broad focus of attention
- capacity of processing
- variation across individuals
