multi-store model of memory

• Multi-store Model of Memory


  The model was proposed by Atkinson and Shiffrin (1968, 1971)

  
  

  This model aims to explore memory shifts from different storage systems

  
  

  The three memory structural components are: sensory memory, short-term memory, long-term memory
• Sensory Memory


  Functions to retain information for long enough to decide if information is worthy for further processing

  
  

  There are many sensory stores, the two important examples of sensory stores are: visual sensory memory (iconic) and echoic memory.

  
  

  Information that isn't attended to is lost quickly

  
  

  After information is selected from sensory memory and is attended to, it is transferred to short-term memory
• Short-Term Memory


  Less than 0.01% of all sensory information, that is received by human senses every second, reaches consciousness

  
  

  Approx. 5% is stored permanently

  
  

  Information that isn't lost from sensory memory is transferred to short-term memory (the temporary memory store)

  
  

  Information in is no longer than its original form since it has been encoded

  
  

  When selecting information to attend to in sensory memory or to be retrieved from long-term memory, the information enters short term memory

  
  

  Duration of short-term memory is longer than sensory memory, approx. 18 - 20 seconds but can occasionally reach 30 seconds

  
  

  Decay (forgetting information) occurs at approx. 12 seconds if information is unrehearsed.

  
  

  Rehearsal keeps information in short term memory and aids the transfer of information from short term to long term memory.

  
  

  Maintenance rehearsal is maintaining information in short term memory by using mental repetition; simply repeating it over and over again without effects of interruptions.
• Short-Term Memory: Miller's Empirical Study

  Miller (1956) conducted an empirical study that showed that most people could only store around 7 unrelated independent items (numbers, letters, words etc.).

  
  

  To expand the capacity, separate bits of information can be combined into parts/chunks, which is called chunking. Think of the mobile phone number and how we read it.
• George Sperling Experiment (Short-Term & Sensory)

  George Sperling explored the properties of iconic memory. 12 letters were flashed for approximately 50 milliseconds. The participants would generally recall 4 or 5 letters but they report that they felt they saw all the letters but the memory of the other letters had faded when after writing 3-5 down.

  
  

  Sperling believed that all 9 letters are stored in the viewer's memory for a short time but only some letters could be moved into short-term memory and named as the memory fades rapidly, and the information disappears.

  
  

  
  

  George's follow-up experiment was that the participants were shown 3 rows, 4 words per row. After viewing all 12 letters, a tone would be given, indicating that the participant should report the content of 1 of the 3 lines. Participants could report lines perfectly once they are familiarized with the system. The results show that the image of all letters was stored in iconic memory after the image disappeared.
• Long-Term Memory


  Long-term memory is considered to be relatively permanent

  
  

  The information isn't the exact replica of events or facts but is stored as an outline

  
  

  Memories can be distorted when retrieved as we fill in gaps to create a more meaningful memory
• Basic Information-Processing Operations Involved in Memory

  1) Encoding: Transforming sensory inputs into a form, allowing it to be stored in memory. In other words, process of changing information into a form that could be stored in memory

  
  

  2) Storage: Holding, keeping process of memory

  
  

  3) Retrieval: Process of information being extracted/taken from memory
• Sensory Stores #1: Iconic Memory


  ☼ Temporary storage for information received visually (visual system)

  
  

  ☼ Duration of iconic memory approx. 0.2 - 0.4 seconds

  
  

  ☼ Each image is registered in iconic memory before the previous image fades, viewing the world in a continuous manner (similar to a series of frames that makes a movie seem continuous. Movies run 64 frames per second to create the illusion of movement)
• Sensory Stores #2: Echoic Memory


  ☼ Stores information received aurally (auditory system)

  
  

  ☼ Believed to be necessary for comprehension of many sounds; after hearing combination of syllables, the definition of the word is apparent.

  
  

  ☼ Duration of echoic memory is approx. between 3 - 4 seconds after the stimulus is perceived

  
  

  ☼ Similar to iconic memory, echoic memory stores lots of incoming information and these information are held in an unprocessed form
• Serial Position Effect


  This effect is when the retrieval or recall of information varies according to the relative position of the material; recall is usually better at the beginning and the end of a list than the middle.

  
  

  This effect supports the two distinct sub-systems of memory, long-term and short-term memory.
• Primacy Effect


  This effect is when the first few items in a list are recalled well, when compared with middle items, during a free recall test. This is usually attributed to more successful rehearsal of the first few items, which information enters long-term memory and can be retrieved later on.

  
  

  Items are probably rehearsed and transferred to LTM before STM's capacity is full. It is likely that items from the start of the list would be forgotten unless they are stored in LTM.

  
  

  Primacy effect still occurs if there is a delay of more than 12-30 seconds, between learning and recalling items.
• Recency Effect


  This effect is when the final few items in a list are recalled well, when compared with middle items, during a free recall test. This is usually attributed to the final words being retained in short-term memory.

  
  

  Recency effect would still occur even if the list of items is increased. Maintenance rehearsal is probably used to retain the information. However, recency effect wouldn't occur when there is a delay of more than 12-30 seconds between learning and recalling items (delayed free recall). Delaying recall would consist of having the participants do something distracting before recalling.

  
  

  Middle items are presented too late to be adequately rehearsed and transferred into long-term memory. Furthermore, the information is too early to be held in short-term memory without rehearsal. (Seen as U-shape graph)
• Study Supporting Short-Term Memory

  Peterson & Peterson (1959): Experiment 1

  Conducted an experiment showing that short-term memory has duration of approx. 12-30 seconds unless the information is rehearsed. This study also demonstrates the role of maintenance rehearsal in short-term memory.
• Peterson & Peterson (1959): Method


  Participants were asked to remember a single nonsense trigram of 3 consonants such as 'dkt'. Recalls made by participants were recorded after various time delays, which are: 3, 6, 9, 12, 15, or 18 seconds.
• Peterson & Peterson (1959): Results

  80% recall (very good) after 3 seconds but only 10% recall after 18 seconds
• Peterson & Peterson (1959): Experiment 2

  Method

  Participants were given a task, which prevented them from rehearsing the trigram. They were asked to count backwards by threes from the time each trigram was presented to the time a light signaled to then recall it. This is filled delay.
• Peterson & Peterson (1959): Experiment 2

  Results

  Recall after a filled delay was poor even after 15 seconds. The study demonstrates the role of displacement in short-term memory
• Evaluation of the Multi-store Model

  ☼ The model doesn't adequately explain the interaction between different memory stores. An example would be LTM must interact with STM because chunking is successful when meaning is given to the material that is chunked. Furthermore, we also probably rely on information that is already stored in LTM to help determine the information we would pay attention to in sensory memory.

  
  

  ☼ The model doesn't show why coding of information changes between the different memory stores.

  
  

  ☼ Other explanations of memory such as levels of processing (Craik & Lockhart 1972) don't set a clear distinction between STM and LTM. Instead, it views memory as a continuum rather than separate/distinct systems.
• Who created the multi-store model?


  Atkinson and Shiffrin (1968)
• What does the multi-store model show?


  How information flows through the memory system, suggests memory is made up of three stores linked by processing
• What are the 3 stores?


  Sensory register, Short term memory store (STM), long term memory store (LTM)
• What does the model look like?


  Stimulus from the environment —> sensory register (e.g. iconic or echoic) —> STM

  • STM (prolonged) maintenance rehearsal —> LTM

  • LTM retrieval —> STM

  • for any store, if no rehearsal leads to forgetting
• What is the sensory register?


  A stimulus from the environment e.g. the sound of someone's name, will pass into the sensory registers.
• How many stores are in the sensory register?


  5 (one for reach of our senses)
• What are the main two stores in the sensory register?


  Iconic memory and echoic memory
• What is iconic memory store in the sensory register?


  Where visual information is coded visually
• What is echoic memory in the sensory register?


  Where sound or auditory information is coded acoustically.
• What is the duration for material in the sensory registers?


  Very brief, less than half a second
• What is the capacity of sensory registers like?


  They have a high capacity (possibly unlimited)
• What is coding like in the sensory register?


  Coded in the same way it's received from the senses
• What is the capacity of the STM?


  7 +/- 2 (5-9)
• What is the coding of the STM?


  Acoustically
• What's the duration of the STM?


  15-30 seconds (if not rehearsed)
• What is Maintenance rehearsal?


  When we repeat material to ourselves over and over again, we can keep the information in our STMs as long as we rehearse it, if rehearsed for long enough, it passes into the LTM
• What is the capacity of the LTM


  Unlimited
• What is the duration of the LTM?


  Up to a lifetime
• What is the coding of the LTM?


  Mainly semantic
• Why does Retrieval occur?


  When we want to recall memories stored in the LTM, it has to be transferred back into the STM (by retrieval)
• What are 4 evaluation points for the MSM?

  1) supporting evidence research that the STM and LTM are qualitatively different

  2) there might actually be more than one type of STM

  3) there might be more than 1 type of rehearsal

  4) there is more than 1 type of LTM
• Why is supporting evidence a strength for the MSM?

  Studies show that the STM and LTM are qualitatively different, e.g. baddely showed people get acoustically similar words mixed up when using the STM and semantically similar words when using the LTM, telling us it's because they code differently. Different kinds of coding tells us that they are different stores, therefore supporting the MSM idea of them being two separate stores.