3. L3 working memory and executive functions Flashcards
What are the 3 assumptions of Nairne 2002 Standard Working memory model
- Short-term memory storage is based on activation of memory representations
- Active representations are easily accessible - Activation is fragile
- Information is lost due to decay
- Decay is a direct function of time and hence a constant “force”
- Decay is adaptive, no longer needed info is automatically removed - Rehearsal counteracts decay
- Refreshes activation
Empirical evidence for Nairne (2002) standard working model
Articulation rate effects
Articulatory suppression effects
- Articulation rate effects (e.g., word length)
○ Better memory for short vs. long words (e.g., ball, plum > tricycle, pineapple)
○ Shorter words allowed more rehearsal = less effect of decay - Articulatory suppression effects (“the, the, the”)
○ Memory performance declines
○ Prevents rehearsal = more effect of decay
Blocks the phonological loop
Criticisms of the standard model (2002)
Problems with activation
- activated information is immediately, easily accessible
- This ignores the fact that all retrieval is cue-dependent
- Sometimes activated information is not easily accessible
- “Release” from proactive interference (Wickens, 1970)
○ Demonstrates that in cue overload performance declines and there is proactive interference even in the most recently activated word list
○ Therefore there needs to be more than activation - A cue dependant retrieval process is not specified in the standard model
Criticisms of the standard model (2002)
problems with rehearsal
- Decay is assumed to be fixed
- so remembering hinges on one’s ability to actively maintain information in the face of constant decay = rehearsal
- so most variability in memory span should be directly attributable to rehearsal
- Recall rate can differ despite equivalent articulation (= rehearsal) rate
- Hulme et al. (1991)
○ Words and nonwords matched for articulation rate
○ e.g. holiday vs. dalihoy
○ Three different lengths: One, three and five syllable items
○ Word advantage cannot eb attributed to rehearsal rates
○ Contribution of longterm memory…better retrieval cues available for words (cue-dependence)
Criticisms of the standard model (2002)
problems with decay
articulatory suppression tasks with different suppression items
complex span tasks with more time in between items
- in an articulatory suppression task there is failure to remember the items because saying the the the prevents rehearsal.
- the standard models states that this is because with no rehearsal items decay.
- is it actually just due to an overload of items and therefore we see interference?
- Decay is assumed to be fixed
- But: The extent of forgetting depends more on the activities that occur during the retention interval than its duration
○ Articulatory suppression: “1, 2, 3, 4, 5, 6” causes more forgetting than “1, 1, 1, 1, 1, 1” (more distractors = more interference; Lewandowsky et al., 2008)
○ No cognitive activity between study and test (rest, sleep) à little forgetting even over long periods of time - This is displayed in the complex span task
○ They tested: if the notion of decay is correct, more processing time (of distractor task)must lead to more forgetting
They found that time does not cause forgetting
The attentional control model Engle et al., 1999; Engle, 2002)
Similar to Baddley but different:
- STM is inside/a part LTM
- WM capacity is defined as the “ability to control attention to maintain information in an active, quickly retrievable state”
○ only indirectly about memory (i.e., storage)
○ not about the size of a memory store instead about its abilities to control and focus
The attentional control model Engle et al., 1999; Engle, 2002)
empirical evidence
- Greater WM capacity should also mean greater ability to use attention to avoid distraction
○ cocktail party & the own-name effect?
○ They found High WM-span people less likely than low-span people to notice their own name in “unattended ear”
○ unless instructed to, then pattern reverses (Colflesh & Conway, 2007)
Kane, Bleckley, Conway, & Engle (2001)
- Examined whether individual differences in WM capacity are related to individual differences in attentional control
- Individuals classified as high or low WM span
- Tested attentional control using a prosaccade/antisaccade task (saccade = eye movement)
- Found that low-span participants were slower to identify targets in antisaccade conditions eg low span found it more difficult to control attention
- The better you are able to ignore distracting information, the less irrelevant information enters your WM and interferes with the critical information in WM
Hierarchical model(s) of WM (Cowan,1988, 1999; Oberauer, 2002, 2009)
- WM has evolved to serve higher cognition
- Short-term storage is just a by-product
- No sharp distinction between LTM and STM
- Instead, focus on hierarchical levels of activation
- LTM
- Activated subset of LTM – not in consciousness
The focus of attention – highly activated info in consciousness (~ 4 items)
Empirical evidence for Hierarchical model(s) of WM (Oberauer, 2002, 2009)
triangles and squares counting task
- Some evidence that the single item being processed at any time has a special status
Garavan (1998)
○ Task: Maintain two counts of squares and triangles, press key when updated
○ WM updating 300-500 ms quicker if same count is updated (i.e., 1 triangle…2 triangles)
○ Some evidence that the single item being processed at any time has a special status
Garavan (1998)
○ Task: Maintain two counts of squares and triangles, press key when updated
○ WM updating 300-500 ms quicker if same count is updated (i.e., 1 triangle…2 triangles)
○ Only one count can be manipulated in the focus-of attention, the other maintained outside—moving FoA takes time
- LTM (basis for WM)
- An associative network of representations that activate each other automatically along their association
- LTM as basis for WM easily explains effects of cue-dependence that trouble the “standard model
components of Oberauer’s (2002, 2009) hierarchical WM model
Activated LTM
Direct-Access Region
Focus-of-Attention
Cognitive coordinate system CCS
Executive control
- Activated LTM
○ Active set of LTM representations
○ Mostly cue dependancy
○ Allows for quick access to relevant information (task-, goal-dependent) - Direct-Access Region
○ Retrieval into DA-region makes a small number of elements immediately accessible
○ Contains what you are “holding in STM”
○ Forgetting from WM is not due to decay but interference and resulting failure to access information
○ Holds 3 to 4 items - Focus-of-Attention
○ Selects single representation for cognitive action
○ Can select only from the DA-region - Cognitive coordinate system CCS
○ Where you arrange and integrate items of information in a flexible or novel manner - Executive control
○ Executive processes eg: attempting to retrieve elements into te direct access region from activated memory
○ Deciding to update the contents of the DA region
○ Shifting the Focus-of-Attention
Executive processes serve executive functions, Not a single “CE(O)”, More of an “executive committee”
3 main executive functions of executive control in the hierarchical model Miyake et al. (2000) as well as some higher level cognitive tasks which are also evidence for this
- Mental set (task, attention) shifting
○ e.g., plus/minus task: 5+3=… 7+2= vs. 5-3=… 7+2= - Memory updating
○ e.g., running memory task: F…J…H…S…Y… - Inhibition of prepotent responses
e.g., Stroop task: BLUE vs. RED
as well as higher-level cognitive tasks
Miyake et al. (2000) – Individual differences study
- Three tasks for each executive functions shifting, updating and inhibition
- Found a weak/moderate correlation between all 3 executive functions so if someone is good at shifting there’s some idea that they will be good at updating
- Meaning they’re all doing similar but different things
- moderately correlated with each other (r’s > 0) but clearly separable (r’s < 1)
Wisconsin Card Sorting Test (WSCT) test for shifting abilities
- cards differ in terms of colour, shape, & number
- participant required to sort cards according to “rule” that test administrator only knows
- rule changes periodically
- response perseveration = no. of cards before participant changes their sorting strategy when rule changes
Tower of Hanoi test for inhibition
- 3 discs on a peg with 3 pegs,
- Move tower from one side to the other
- Move only one disc at a time -
- Larger disk must never go on top of smaller disk
- “Illegal” moves must be inhibited
- Random number generation task tests for updating
executive functioning deficits in the hierarchical WM model
ADHD
- impaired inhibition à hyperactivity, distractibility & impulsivity *
Schizophrenia
- impaired shifting and updating
Frontal lobe dementia (Morbus Pick)
- executive functions are “frontal lobe functions” * etc. use Wisconsin card sorting test to diagnose
What does working memory capacity predict
Complex span task performance predicts higher cognitive abilities
- Complex span task performance (WM capacity) is an extremely good predictor of higher-level cognitive abilities
- Better predictor than simple span tasks (STM capacity)
WM capacity predicts:
- Attentional control (Kane et al., 2001)
- Reading performance (Daneman & Carpenter, 1980)
- Math skills (Hitch, Towse, & Hutton, 2001) * Reasoning (Kyllonen & Christal, 1990)
- Skill acquisition (Woltz, 1988)
- Intelligence (Oberauer et al., 2005)
- WM the “engine of cognition” (Jonides, 1995)
WM and intelligence – Latent variable approach (Engle et al., 1999)
- 9Working memory but not STM predics fluid intelligence
