Chapter 3 STM Flashcards
Digit span
Used in Wechsler
Reflects STM and WM
Memory span requires
Remembering what the items are
Remembering the order in which they were presented
Miller: number of chunks limit memory capacity.
Chunking takes cues from prosody, natural speech rhythms
Conrad: error likely to be similar in sound to item they replaced P/G/V
Phonological loop
Short-term store and articulatory rehearsal process
No explanation of how serial order stored
No clear specification of crucial processes involved in retrieval from phono store
Assume separate mechanisms
Order carried by context
Phonological similarity effect
Conrad
Letter span is reduced for similar sounding items
Suggests consistency with Conrad’s acoustic code
Big, high, tall easier - similar meaning
Phono similarity disappears if lists are increased in length and participants are allowed several learning trials
Phono coding not limited to STM. Without LTM we could never learn to pronounce new words
PSE Occurs at retrieval
Rehearsal blocked via articulatory suppression - say ‘the’ means you can’t refresh memory trace by subvocally pronouncing the membered material - prevents from being registered in phono store
Even while suppressing, people remember 4-5 visually presented digits - suggests although phono loop typically plays imp role in digit span, it’s not the only basis
With auditory presentation, words gain direct access to phono store despite articulatory suppression - similarity effect still occurs
Word length effect
Ppl remember sequences of 5 dissimilar one-syllable words relatively easily but not five-syllable words
Ppl remember as many words as they can say in 2 seconds
Longer words allow more decay to occur
Happens during subvocal rehearsal but recall phase too because they’re longer
If rehearsal prevented, word length effect should be lost
Articulatory suppression ‘the’ while performing memory task abolishes word length effect
Long words are more vulnerable to fragmentation and forgetting
Irrelevant sound effects
STM for visual digits impaired when required to ignore speech even in foreign language but not white noise
Irrelevant speech could gain access to phono store adding noise
Doesn’t matter how similar irrelevant speech is to remembered items
Vocal music disrupts not instrumental
Attributed to serial order mechanism, not phono store
Similarity between remembered and irrelevant items has no effect because they influence different parts of the system - no interaction
Changing state hypothesis
Jones and Macken
Retention of serial order can be disrupted by irrelevant stimuli, providing these fluctuate
Theories of verbal STM
Jones
Object oriented episodic record
To account for influence on STM of irrelevant sound
Sequences of items are represented as points of multimodal surface
Assumes both auditory and visual serial recall involve same system operating on common representation
Recall involves retrieving trajectory of points representing sequence
Irrelevant sounds create competing trajectories, disrupting recall
Not supported by subsequent experiments
STM theories
Nairne’s feature model
Single memory system
Each item assumed to be represented by set of features: modality dependent and independent
Hat has visually dependent features (mode, case) and independent (meaning)
Forgetting depends on interference - new items disrupt earlier features
Simulates wide range of results
Hard to explain how irrelevant sound disrupts recall after presentation of memory items, even when rehearsal is prevented by suppression
Word length effect disappears in mixed lists of long and short words
STM theories
SIMPLE
Brown Scale invariant memory Perception Learning Works for STM and LTM More distinctive items are more retrievable Not good for explaining serial recall
STM theories
Serial order in a box model
Order is maintained using an event-based context signal, with forgetting based on interference between events
Free recall
Serial position effect: last few items good - ‘recency effect’ ‘primacy effect’ less so
Glanzer: earlier items held in LTM, later STM
Recency challenged by demonstration that recency effects can occur under conditions in which the short-term trace ought to have been disrupted - counting backwards - Bjork
Baddeley rugby players
Primacy
Tendency to rehearse first few items as theyh come in
Crowder
Telephone posts along railway to recall list - closest most distinguishable from next
Visuo-spatial STM
Binding features
Change detection task
Luck
Participants see array of square differing in colour, followed after varying delays by pattern that is identical, or has the colour of one square changed
Colour can be remembered verbally
To prevent, ppts required to repeat numbers to occupy verbal rehearsal system
Performance declines as no of squares increases
Capacity limited to 3-4 items
In LTM, standing presented 2600 colour slides for 10 secs each then tested days later. Ppts 90% correct
Diffs between STM and LTM in visual
Phillips: as matrices more complex, performance declined
Visual LTM can store huge amt of complexity
More diffs
Rapid for visual STM
STM active rehearsal
Attempts to keep item in focus of attn
McCollough ERPs ask ppts to remember items on 1 side of visual field, observe electro activity
Amt of activation increased w no of items up to max of 4
Unsuccessful trials associated w lower level of activation
What is stored in visual STM
Vogel: ppl able to combine several features into single object with little cost
Separate features bind together - colours, shapes - not more difficult than single-feature conditions - so no more additional resources
Allen - colours easiest, shapes, bound features hardest (nonsig) - impairment no greater in binding than single-feature - binding is auto
Visual-spatial distinction
Corsi span - Simon - is spatial span
visual span - half cells are filled half blank. Ppts shown pattern and asked to reproduce it by marking filled cells in empty matrix in increasing matrices (16)
Interfering activity reduces Corsi span
Pattern span disrupted by viewing shapes
KF
Digit span of 2 items
Shallice and Warrington
Patient had specific phonological STM deficit
Performance better when digit span tested using visual presentation
Corsi block good
PV
Phonological STM affected by stroke
Intellect and language unaffected
Digit span of 2
No word length effect in STM verbal
Weak recency effect in immediate verbal free recall
Showed normal long-term recency tested by anagrams and recall
LE
Lupus Good spatial memory Impaired visual memory Bad drawing from memory Sculptures changed
MV
Stroke R frontal lobe
Spatial hindered on maze task
