What type of experiments primarily determined what different components there were in the Working Memory model? Give an example study.
- Dual-task experiments
- eg. Robbins et al (1996)
- Chess move selection and effect of several concurrent tasks
- Key pressing task (visuo-spatial sketchpad) and random number generation (central executive) affected quality of chess moves.
- However, word repetition task (involve phonological loop) had no effect.
- Chess move selection involves central executive and visuospatial sketchpad, but not phonological loop.
[Phonological loop] List the two key empirical findings that provide insights into the structure of the phonological loop.
- Phonological similarity effect
- Word length effect
Describe the phonological similarity effect and the study associated with it.
- Baddeley (1966)
- Serial recall of list of phonologically similar words (eg. Fee, He, Knee) significantly worse than from list of phonologically dissimilar words (Bay, Hoe etc)
- whereas semantic similarity had little effect on recall
- Suggests that speech-based representations are used in storing the words (+ recall requires discrimination between memory traces)
Describe the word length effect and study associated with it.
- Recall of list of long words worse than recall of short words.
- Cultural differences in mean digit span also illustrates this effect - Chinese digits are relatively quick to say, whereas in Welsh it is not. Digit span higher in China than in Wales (also could be correlated to maths performance?)
- Baddeley (1975)
- If word length effect due to phonological loop, disrupting subvocal rehearsal should reduce effect.
- Participants silently mouth irrelevant digits (articulatory suppression) during presentation and recall of words
- This eliminated word length effect - suggesting phonological storage capactiy determined by rate of rehearsal.
So, from these 2 empirical findings, what did Baddeley conclude about the structure of the phonological loop?
- Distinction between 2 components...
- phonological store: speech perception
- articulatory control process: speech production (giving access to phonological store for visual stimuli)
- So... explanation for the findings
- Phonoglocal similarity effect: confusion between similar representations in phonological store
- Word length effect: time taken to rehearse longer words via articulatory control process
[Visuospatial sketchpad] There have been less study on this one. Most have focused on what type of study? Give example.
- The effect on different tasks of concurrent visual visuospatial processing
- Baddeley et al (1975)
- encode material either by rote verbal learning or imagery-based strategy
- when task combines with pursuit rotor tracking (tracking a moving light) performance using imagery-based strategy was disrupted.
However, Baddeley's use of rotor tracking involves both visual perception and spatial localisation. Which are more important for disrupting imgaery-based learning? Study please.
- Baddeley & Lieberman (1980)
- Repeated Baddeley (1975) experiment, but contrasting specifcally visual (making brightness judgments) and specifically spatial (pointing at moving pendulum while blindefolded, guided by auditory tone) concurrent tasks
- Learning using imagery strategy disrupted most by spatial concurrent task
From Badderley and Lieberman's study, it was shown that a distinction needed to be made between visual and spatial information processing. Who proposed this and what did it entail?
- Logie (1995)
- Visuospatial working memory can be divided into 2 components...
- Visual cache: passively stores info about visual form and colour and decays and gets interfered by new info
- Inner scribe: process spatial info and allows active rehearsal of info in visual cache
- (However, lecturer thinks it is weak model - ASK!)
How is Logie's model of visuospatial sketchpad supported by neuropsychological evidence?
- Patient NL: preserved perceptual skills but could not describe details of scene from memory (Beschin et al)
- Patient LH: perfrom better on spatial processing tasks than visual imagery (Farah et al)
[Episodic buffer]. So, these teo systems allow temporary storage of modality-specific info, but various findings are difficult to explain with this distinct-systmes approach. Outline some of these findings.
- Baddeley (1984): articulatory suppression does reduce memory span for visually-presented material too (from 7-->5) but doesn't eliminate it like phonological loop model predicts
- Patients: with impaired ST phonological memory (with auditory span of 1 digit) can recall more digits with visual presentation (ASK WHY THIS IS RELEVANT!)
- Chincotta et al (1999): dual-task method used to study memory span for Arabic numerals (1,2,3) and digit words (one, two, three). Participants used both verbal and visual representations in performing task.
- These suggest verbal and visual info must be combined and stored somewhere in working memory.
What was a further issue/phenomena that couldn't be explained simply by 2 modality-specific stores
- Memory span for meaningful sentences can be as much as 15-16 words, much more than normal phonological loop capacity (Baddeley et al, 1987)
- Traditional explanation is that info from LTM is used to integrate words into meaningful 'chunks'
- BUT in densely amnesic patients (with grossly impaired LTM) can exhibit normal immediate sentence span (Baddeley & Wilson)
- Also densely amnesic patient able to continue playing bridge, keeping track of all cards that have been played (Baddeley 2000)
- Working memory model needed new temporary storage system allowing verbal and visual codes to be combined and linked into multi-modal representation.
Outline the evidence from neuroimaging for a distinct, multi-modal short-term store.
- Prabhakaran et al (2000)
- Working memory task requiring retention of integrated verbal and spatial info
- Result: activation in right frontal cortex greater for retention of integraded info - consistent with episodic buffer
- Posterior regions exhibited material-specific working memory effects - consistent with phonological loop and visuospatial sketchpad
- Overall, episodic buffer useful addition, but lacks detailed account of how it integrates info from other components and from LTM is lacking.
[Central executive] Most important but least known about because of its complexity and somewhat vague specification. However, certain functions have been identified. What are they? (4)
- Smith & Jonides (1999)
- Identified major functions of central executive:
- 1. switching attention between tasks
- 2. planning sub-tasks to achieve a specified goal
- 3. selective attention to certain stimuli while ignoring others
- 4. updating and checking contents of other working memory stores
Describe a study which demonstrated the existence/mechanism of central executive.
- Baddeley (1996)
- Random generation
- Participants hold 1-8 digits in mind while trying to generate random sequence of key presses.
- Randomness decreases as digit memory load increases
- Suggesting greater demands on general-purpose, limited capacity central executive
What are other things Baddeley found about the central executive with the random generation paradigm?
- Concurrently reciting alphabet or counting didn't affect performance but alternating between letters and numbers decreased randomness
- suggesting that rapid switching of attention is another function of central executive
What is the impairment of the central executive called? What are the symptoms?
- Dysexecutive syndrome
- Rylander: disturbed attention, increased distractibility, difficulty grasping the whole of a complicated state of affairs, cannot master new types of task in new situations
What brain regions are associated with dysexecutive syndrome?
- D'Esposito: dorsolateral regions of frontal lobe show greater activity in dual-task than single-task conditions
- Duncan et al: lateral frontal cortex is neural basis of 'general intelligence' characterised as specific system involve in control of diverse forms of behaviour.
NOTE: Read summary slide on PP and ask lecturer questions! Bit hard to understand some of the bits - especially episodic buffer and central executive.
Good job. Keep going Ron.