Multi-Store Model of Memory

• The multi-store model of memory - key assumptions:

  this theoretical computer model was devised by researchers Atkinson and Shiffrin in 1968. There are 3 key assumptions of this model;

  • there are 3 separate unitary stores of memory
  • these 3 stores are the sensory memory (also referred to as the sensory register or the sensory store), the short-term memory and the long-term memory. They all work independently of each other
  • the model works in a serial fashion, where information has to move through each store to get to the next
• Information of each store:

  • the capacity of the store - how much information it holds
  • the duration of the store - how long it will store that information
  • the encoding of the store - in what format the information is stored
• The sensory register:

  • stimuli from the environment enters our memory via the sensory store
  • this stores information from our sense (sight, smell, sounds, taste, touch, amongst others)
  • anything that is important is passed on to the STM
  • if it is not important then it is immediately forgotten (acts as a filter for your memory)
• Sensory register capacity:

  • the capacity of this store is enormous (lots of information is currently being processed)
  • however, unless it is attended to then the information is lost
• Sensory register duration:

  • information is held very briefly in the sensory register
  • if it is important then it is paid attention to and passed to the STM. If it is not then it is forgotten
  • this store will hold information for up to 2 seconds
• Sensory register encoding:

  the sensory register can process information in any format. However, the two main stores are for sight and sound;

  • iconic encoding - the storage of images
  • echoic encoding - the storage of sounds
• Summary of the sensory register:

  • capacity - unlimited
  • duration - 2 seconds
  • encoding - all formats (mainly iconic and echoic)
• The short-term memory:

  • once information has entered the sensory register and you have paid attention to it, this information will move to the short-term memory
• Short-term memory capacity:

  • This was investigated by Jacobs, using a digit-span task
  • Jacobs gave participants a series of digits to remember, increasing in digit length by one in each trial
  • they were tested immediately on their memory of the string digits
  • they found that participants began making mistakes when they were given approximately 5-9 items (referred to as items as chunking information into categories allows you to utilise more capacity)
  • they inferred that the STM capacity was '7 plus or minus 2'
  • George Miller coined the term 'Miller's magic number' to describe the capacity of the STM
• Short-term memory duration:

  • this was investigated by Peterson and Peterson (1959) with their trigram study
  • in this repeated measures lab experiment, Peterson and Peterson gave people 'trigrams' which are nonsensical combinations of 3 letters without vowels
  • they then had to recall the trigrams after different time periods (3,6,9,12,15,18,21,24,27,30 seconds) and were tested immediately
  • to prevent rehearsal of trigrams. the participants completed a distractor task (counting backwards in 3s)
  • Peterson and Peterson found that participants started forgetting trigrams at around the 15 second mark, concluding that the duration of the STM lasts between 15-30 seconds
• Short-term memory encoding:

  • this was investigated by Baddeley (1966), who wanted to research how out short-term memories are stored
  • participants were given different lists of words to remember in this lab experiment. They were tested immediately and asked to repeat the list in order
• Short-term memory encoding:

  • Baddeley found that participants could not remember word list 1 (acoustically similar) in order, when tested immediately
  • participants muddled these words up as they all sounded the same
  • Baddeley inferred from this that the STM is encoded acoustically. This means that the words all sound the same and therefore the STM cannot separate them out into the correct order
• Summary of the short-term memory:

  • capacity - 7 +/- 2
  • duration - 15-30 seconds
  • encoding - acoustically
• The long-term memory:

  • once information has been rehearsed (elaborative rehearsal) then it can move into the long-term memory
• Long-term memory encoding:

  • this was investigated by Baddeley (1966) using the exact same procedure as his STM encoding study using the 4 word lists
  • however, this time the words were tested after a 20 minute delay
  • this time Baddeley found that participants muddled list 3 (semantically similar)
  • Baddeley inferred from this that the LTM is encoded semantically
  • if words all mean the same thing then the LTM finds it hard to separate them
• Long-term memory capacity and duration:

  • LTM capacity is unlimited and cannot be measured
  • the duration of LTM is minutes to a lifetime, as we can remember things things throughout our lives
• Long-term memory capacity and duration:

  • Bahrick (1975) did attempt to measure this with his quasi experiment looking at memories of school yearbook

  in this study, 392 American ex-high school participants aged 17-74 were tested on their memories of their classmates;

  • free recall of as many names from their high school as possible
  • a photo recognition test where they were asked to identify former classmates in a set of 50 photos, only some of which were their classmates
  • a name recognition test
  • a photo and name matching test
• Long-term capacity and duration:

  Bahrick found the following;

  • after 48 years of leaving high school, name recognition had 80% accuracy
  • after 48 years of leaving high school, photo recognition had 40% recall
  • after 34 years of leaving high school, photo and name matching had 90% accuracy
  • after 30 years of leaving high school, free recall had 30% accuracy
  • after 15 years of leaving high school, free recall had 60% accuracy

  Bahrick inferred from his quasi experiment;

  • LTM can last up to a lifetime
  • LTM has an unlimited capacity
  • providing cues such as photos and names is better than free recall for LTM
• Summary of long-term memory:

  • capacity - unlimited
  • duration - minutes to a lifetime
  • encoding - semantically
• Evaluation of memory studies:

  Jacobs, Peterson and Peterson, Baddeley;

  S - high control

  W - low ecological validity

  Bahrick;

  S - high ecological validity

  W - low control
• Evaluation of the multi-store model:

  S - support that the stores are separate - Murdock's serial position curve

  S - support that the stores are separate - the H.M case study

  S - support that the stores are separate - Baddeley's encoding studies

  W - evidence that the STM is more complex - the K.F case study
• Murdock's serial position curve;

  • Murdock read participants 20 words out and asked them to recall them in any order
  • they recorded how many participants remembered each word, depending on the word's position in the word list
  • they then plotted this on a graph, showing that participants could remember the first 7 or so words (where they had time to rehearse) and the last 7 words (the words they heard most recently). The words in the middle were forgotten
• Support that the stores are separate - Murdock's serial position curve (identify):

  • there is support for the multi-store model's idea that there are separate STM and LTM stores of memory
• Support that the stores are separate - Murdock's serial position curve (explain):

  • the serial position curve shows that words at the start of a word list are remembered well as they are rehearsed and placed into the LTM (primary effect) whilst words at the end of the list are remembered well as they are still in the STM (recency effect)
  • when there is delay in testing the words, the recency effect disappears, supporting the idea that these words are in the STM
• Support that the stores are separate - Murdock's serial position curve (conclusion):

  • this adds validity to the model as it supports the idea that they are separate unitary stores of STM and LTM
  • this is used in marketing and advertising (companies often show their logo or catchphrase at the end)
• Support that the stores are separate - the H.M case study (identify):

  • there is support for the multi-store model's idea that there are separate STM and LTM stores of memory
• Support that the stores are separate - the H.M case study (explain):

  • the case study of H.M found that the removal of the hippocampus led to him being unable to form new long-term memories
  • his memories from before the surgery were intact and he could store information for up to 30 seconds in his STM, but he couldn't place these into his LTM
• Support that the stores are separate - the H.M case study (conclusion):

  • this adds validity to the model as it supports the idea that there are separate unitary stores of STM and LTM
  • this lacks population validity (it is not generalisable)
• Support that the stores are separate - Baddeley's encoding studies (identify):

  • there is evidence to support the model from Baddeley's encoding studies
• Support that the stores are separate - Baddeley's encoding studies (explain):

  • the Baddeley encoding studies found that when tested immediately (testing our STM) the acoustically similar words get muddled, suggesting how the STM is encoded acoustically and we cannot untangle them
  • when tested after delay (testing the LTM) the semantically similar words get muddled, suggesting LTM is encoded semantically and we cannot untangle them
• Support that the stores are separate - Baddeley's encoding studies (conclusion):

  • this adds validity to the model as it supports the idea that there are separate unitary stores of STM and LTM
  • however, it is a lab experiment so lacks ecological validity. This task doesn't apply to everyday life
• Evidence that the STM is more complex - the K.F case study (identify):

  • the model can be criticised as being overly simple
• Evidence that the STM is more complex - the K.F case study (explain):

  • the case study of K.F found that his memory was damaged after a motorcycle accident and that he could only remember 3-4 digits at a time
  • however, his memory for images was left untouched
  • this suggests that there are separate stores for visual and verbal short-term memory
• Evidence that the STM is more complex - the K.F case study (conclusion):

  • this lowers the validity of the model as it suggests that its explanation of STM is overly simple and needs expanding to include separate visual and verbal stores