the multistore model Atkinson & Shiffrin

the multistore model Atkinson & Shiffrin

1. capacity

   • short term memory
     • Joseph Jacobs
       • found average recall to be 7+-2 items
       • presented numbers and asked for immediate recall
       • lacks mundane realism - lab study
       • this would mean we are unable to remember words longer than 9 letters
         • so miller said out stm can hold 7+-2 chunks of information, and personal experience would determine what was a meaningful chunk
     • the recency effect
       • you remember the most recent chunks more than others
       • this is not always the case
   • long term memory
     • the capacity is hard to measure
     • it seems to be infinite
     • some people seem to be able to recall lots of detail for many years, other people not so much

2. duration

   • short term memory
     • Peterson & Peterson
       • used trigrams to test stm duration
       • found worse recall after longer delay
       • memory in stm is fragile and easily forgotten
       • lacks mundane realism
       • doesn't account for the duration of pictures, sounds or smells
   • long term memory
     • Bahrick et al
       • found 90% accuracy for recalling faces and names 34 years after leaving school, and 80% accuracy for names and 40% for faces 48 years after leaving school
     • ltm duration is possibly infinite
     • hard to test = just because you can't think of something at the time doesn't mean its been forgotten

3. encoding

   • Baddeley
     • stm is acoustic coding
       • worse recall if the words sound similar
     • ltm is semantic coding
       • worse recall if words have similar meanings
     • did not take account of visual cues
     • Posner
       • said that visual information is important. it is easier to identify A and A as being the same as A and a

4. evaluation

   • strengths
     • defines stm and ltm
     • evidence from different encoding
     • definite differences in capacity between stm and ltm
     • different durations in stm and ltm been proven
   • weaknesses
     • overemphasis on rehearsal. maybe in lab but too consciously in real life
     • lacks description. how does info go from acoustic to semantic coding?
     • info probably contacts ltm before rehearsal in stm, this is not explained by model
     • reductionist and oversimplified