Lecture 2 - STM & WM

Question 1

Verbal info in the STM

Answer 1

1. STM stores info phonologically
   > Conrad & Hull (1964) - serial recall of consonants phonologically similar or distinct. letters easier to remember if distinct as not as acoustically confusing.
2. STM has limited capacity
   > Miller (1956) - capacity is 7+-2. chunking inc amount of info retained.
3. STM limited duration
   > Glanzer & Cunitz (1966) - free recall of one-syllable nouns either immediate, 10s after or 30s after. first (primacy) and last (recency) better recalled. delaying recall eliminares recency effects. STM traces decay between 10-30s and primacy effects do not reflect STM rather LTM performance
   - for items earlier in lsit which have not been affected there is more time for rehearsal

Question 2

the phonological loop

Answer 2

• contains 2 subcomponents:
  > phonological st store - limited capacity & duration
  > articulatory control process: subvocal/vocal rehearsal to prevent delay & translates visual info into speech based code for storage.
• PL explains 3 effects of verbal STM:
  1. phonological similarity effect - immediate serial recall of verbal material to be reduced when items sound similar. Baddeley 1996 - impaired recall for phonologically similar words. not affected by semantic similarity
  2. word length effect - verbal mem span dec when longer words used. Baddeley 1975 - serial recall of visually presented words. either contains 1, 2, 3, 4, or 5 syllables. recall drops with word length. reading speed also drops. optimum reading speed is 2.3 words per second.
  3. irrelevant sound/unattended speech effect - verbal STM disrupted by concurrent fluctuating sounds inc both speech and music. Colle & Welsh (1976) - serial recall of auditory presented 8 letter list either with or without foreign word noise. higher errors in nosie condition = spoken words impair recall of verbal material as phonological material is automatically encoded in PL. first words less prone to errors.

Question 3

visual info in STM

Answer 3

• non-namable visual material that cannot be vocalised:
  > phillips (1974) change detection task of checkerboard. accuracy dec as function of retention time - longer retention - more decay & lower accuracy. espec for larger matrices (8x8)
• visual material with prevented vocalisation
  > articulatory suppression: secondary verbal task (e.g. count back) occupies articulatory control processes & visual material cannot be fed into the phonological store of the loop.
  > Luck & Vogel (1977) change detection task of coloured squares retained with or without load (counting to 3). accuracy dec as function of set size to same extent for both conditions.
• STM 2 stores: one for auditory and vocalised visual info and one for non-vocalised visual info.

Question 4

visuo-spatial STM

Answer 4

• visual STM has 2 paths: visual object memory (what - ventral/temporal: V1, 2, 4 & IT), & spatial location memory (where - dorsal/parietal: V1, 2, 3, MT).

Question 5

STM has 2 separate stores

Answer 5

• evidence of double dissociations of STM:
  > shallice & warrington (1970) - KF deficits in verbal info but intact memory for visuospatial info.
  > Carlesimo et al (2001) - MV deficits in spatial info but intact visual and verbal info.
• not in line with atkinson & shiffrin idea of a single STM component
• updated model is WMM

Question 6

WMM

Answer 6

• STM holds small amount of info & serves as mental workspace for cog tasks.
  > phonological loop: helps language acquisition - Baddeley et al. 1988 patient PV has digit span of 2 but preserved visual STM, LTM & language = pure PL deficit. PV could not learn russain words but good for native words.
  > VSS: aids visual & spatial imagery - Shepard & Metzler (1971) - mental rotation with 3 conditions: 2d match, 3d match, non-match. needed to assess whether the images are same or different and measured how long. the greater the angle of rotation the longer it takes to decide if objects the same. similar results bt Kosslyn et al. (1978) - location distance and time to mental travel inc linearly = mental image of map created.
  > CE: attentional controller which distributes resources to subsystems. Baddeley et al. (1991) - either single tasks (visuopursuit tracking of spotlight OR digit span) OR dual task with alzheimers and control pp. performance in dual tasks dec espec for alzheimers.

Question 7

problems of the multicomponent model

Answer 7

• fails to explain chunking
  > chunking can be independnet of LTM (Baddeley & wilson 2002) - amnesic patients show pose recall absent of LTM = chunking still takes place so can recall right after display
• fails to explain visual similarity effects in verbal recall
  > Logie et al 2000 - small adv for visually dissimilar words more than similar words in verbal recall.

Question 8

episodic buffer

Answer 8

• added in 2012
• a storage component that assumes a multidimensional code, allowing the various subcomponents of WM to interact with LTM
• holds approx 4 chunks of multi-dimensional info based on range of dif dimensions (visual, verbal, semantic) via connections to VSS, PL & LTM
• pulls together separate streams of info from senses and binds them in CE before splitting to VS or PL
• how does it interact with LTM? loops of rehearsal & retrieval generate & activate LT info in buffer which links to subsystems