Multi-store model of memory

• Concepts of multi-store model

  • Introduced by Atkinson and Shiffrin
  • Model contain both stores and processes
  • Memory consists of distinctly separate stores
  • Information is transferred from STM to LTM by rehearsal
  • Each store contains a certain amount of information, in a different way and for a different length of time.
  • STM and LTM are unitary stores that cannot be subdivided.
• Sensory register

  • Takes in all sensory stimuli from the environment
  • 2 main stores: iconic and echoic memory
  • Information received by sensory registers lasts only briefly, the duration is for about 1-2 seconds
  • It also has a high capacity as the sensory register is constantly receiving information
  • For information to be passed onto the STM, the persons attention need to be focuses on one of the sensory stores.
• What happens if you don’t pay attention to information in the sensory register?

  Forgetting through decay
• Short term memory

  • Limited capacity stores
  • This is because it can only hold a small amount of information before the information is forgotten.
  • Based on previous research, STM can only hold between 5-9 items.
  • Information is coded acoustically and has a duration of about 18 seconds
  • In order to keep information in STM, we need to repeat it (maintenance rehearsal)
  • Once it has been rehearsed long enough it is passed onto LTM.
• How is information forgotten in the STM?

  • Forgetting through decay or displacement
• Long term memory

  • Unlimited capacity stores
  • Information within the LTM can last for a long period of time.
  • Information is coded for semantically
  • Use the process of retrieval whereby information in the LTM is passed back to the STM which is then available to use.
• Research support for separate short-term and long-term stores

  • Baddeley’s research into the different coding in STM and LTM. He found that the STM encodes information acoustically whereas the LTM encodes information semantically.
  • The evidence suggests that the features of the STM and LTM are different in terms of their characteristics.
  • This supports the idea that there are separate memory stores which are unitary and independent from one another.
• Multi-store model only accounts for one type of rehearsal

  • Craik and Lockhart criticised the MSM as they suggested enduring memories are created by the processing you do, rather than maintenance rehearsal. They proposed information that is processes more deeply are more memorable.
  • They discovered there are 2 types of rehearsal, the first being maintenance rehearsal but also elaborative rehearsal for long-term storage. This suggests that we need to be able to link information to existing knowledge and requires you to think about the meaning of that information.
• Brain scanning support

  • Beardsley, who found that prefrontal cortex is active during STM tasks but not LTM tasks. Squire et al found that hippocampus is active when LTM is engaged in LTM task.
  • Evidence suggests we can identify distinct areas of the brain which are engaged in different memory tasks. From findings, it is apparent that depending on the task and the memory store it requires different brain regions are activated.
  • Strengthens the idea that the STM and LTM are unitary memory stores and work independently from one another
• Artificial materials

  • A lot of research studies have heavily relied upon the use of artificial stimuli in their procedures e.g recall tasks using letter or digits
  • These tasks do not reflect the way we use memory in normal everyday life which makes the MSM lack mundane realism. For the model to be reflective of real-life situations research studies need to incorporate tasks which involve everyday materials e.g peoples faces, names
  • MSM lacks external validity as it does not accurately reflect the way memory operates in the outside world.
• More than one type of STM

  • Shallice and Warrington studied an amnesic patient known as patient KF who had suffered brain damage as a result of a head injury. His digit span was tested by presenting digits to him visually (rather than reading aloud), his performance was much better, compared to verbal digit span tests.
  • Research proposes there must be one short term store which processes visual information and another that processes auditory information. The structure of the STM is far more complex than Atkinson and Shiffrin suggested and the model does not go far enough in breaking down STM