MSM, coding, capacity & duration

• Multi-store model of memory: Atkinson & Shiffrin (1968) -

  • SR info from the environment is stored, duration is less than 1 second, Split into 5 senses: Iconic - visually encoded, Echoic - auditory encoded. Info must be attended to for it to move to STM.
  • STM duration up to 30 seconds, for it to move to LTM maintenance rehearsal has to occur. If not info decays.
  • LTM permanent store, capacity & duration is unlimited. Info must be recalled for it to move from LTM to STM.
• Multi-store model -

  
• AO3 MSM -

  • Serial position effect: Glanzer + Cunitz, pps saw 20 words for 3 secs each, Recall for middle words was poor. Beginning words (primary effect) due to the capacity of 7+/-2 items & end words (recency effect) supports idea of STM & LTM separate stores, rehearsal & limited capacity & duration of STM. However, it was a lab study, so lacks mundane realism as it may not explain memory irl. Lacks mundane realism.
• AO3 MSM -

  • case of HM, brain surgery to relieve his epilepsy, his hippocampus was removed & he lost his LTM but STM was still intact & working which supports the idea of 2 separate stores.
  • STM is not a unitary store, Shallice + Warrington (1970) studied KF, a patient with amnesia. He could recall digits he read, but not those read to him. Must be a visual & auditory store, inadequate explanation, idiographic approach.
  • Craik + Watkins, MSM ignored other types of rehearsal e.g. elaborative - linking new knowledge to pre-existing knowledge/memory via schemas.
• Coding - how we store memory

  • Baddeley: Coding in STM

  Pps given different lists of words to recall in correct order. 4 lists:

  1. acoustically similar
  2. acoustically dissimilar
  3. semantically similar
  4. semantically dissimilar

  Recall was worst with list 1, suggests we encode acoustically.
• Coding in LTM: Baddeley -

  • Same as STM study, Pps given 4 lists & had to recall after 20 minutes. Found worst recall with semantically similar words suggest LTM codes semantically as we get meanings confused.
• AO3 Coding: Baddeley -

  • lacks ecological validity, used artificial material lacks mundane realism.
  • findings can be replicated, increases reliability standardisation used.
  • machine reductionism e.g. as we recall things we want to remember.
• Capacity - max amount of info that can be stored.

  Jacobs (1887) Coding in STM: pps presented with digits which they had to recall. Digits increased till pps failed to recall back in correct sequence. Approx. 9 digits, 7 letters. Capacity increased with age. STM has limited storage, capacity of about 5-9 items.
• AO3 Jacobs & Miller Coding in STM:

  • Individual differences e.g. age, chunking.
  • Lacks ecological validity
  • Machine reductionism
  • Miller may have overestimated capacity, more recent research - Cowan (2001) found STM more likely to be around 4 chunks.
• Miller published 'The Magical Number Seven, Plus or Minus Two' - an article explained how we chunk in the STM. Coding is around 5-9 items.
• Duration - amount of time info can be held in memory.

  • Peterson & Peterson Duration in STM:
  • Experimenter said a nonsense trigram followed by a 3 digit number. Nonsense trigram had no meaning. Pps was asked to count backwards in 3's/4's - retention. Interval varied between 3-18s. Pps remembered about 90% at 3 secs, recall dropped 2% when there was an 18 second interval.
• AO3 Duration in STM Peterson & Peterson -

  • Sample cannot be generalised, sample isn't representative as only university students were part of the sample does not represent memory in all individuals as other confounding variables e.g. age may have an effect.
  • Lacks ecological validity, an attempt to reduce EV's via retention interval which is a strength.
  • Replicable & follows the scientific method.
• Duration in LTM: Bahrick et al -

  • Pps (17-74) tested on their memory of their classmates.
  • Photo recognition: pps shown 50 yearbook photos some from their own, to see if they could recognise & name them.
  • Free-recall: pps asked to list names they remember from their graduating class.
  • PR: pps within 15yrs of graduating recalled 90% accurate in identifying faces dropped to 70% after 48yrs.
  • FR: pps within 15yrs were about 60% accurate dropped to 30% after 48yrs.
• AO3 Duration in LTM: Bahrick et al -

  • Sample is representative & can be generalised to more age groups (17-74). More representative & has population validity.
  • Hard to control all variables making the findings less reliable. May mean meaningful info is stored as they were recalling their own high school classmates, but not all other types can't generalise.
• Duration Research - Peterson and Peterson - Trigrams study (1959), Bahrick - Yearbook study (1974).

  AO3 -

  • use of artificial material, e.g. recall of trigrams, lists of unconnected words etc
  • use of artificial laboratory setting, low ecological validity
  • discussion of issues of validity (higher in Bahrick study)
  • alternative explanations – Peterson and Peterson’s findings may be more to do with interference than duration.
• Capacity of STM - participants are read a sequence of letters/numbers and asked to repeat the same sequence back immediately. An additional digit is added on each subsequent trial to measure the capacity of STM (the digit span technique).
• Duration of STM - participants are given a trigram (three-letter nonsense syllable) and then asked to count backwards from a certain number for a specified time. They are then asked to recall the original trigram.