The Multi-Store Model of Memory.

The multi-store model of memory was proposed by Atkinson and Shiffrin in 1968 and quickly became the basis for decades of research into the structure of memory.
1. Atkinson and Shiffrin proposed 3 separate memory stores or components that vary in terms of coding, duration and capacity.
2. The model is a good example of an information-processing theory, seeing cognitive processes as a sequence of stages and comparing them with the operations of a computer (Eysenck 1993).
The Model.
1. According to the multi-store model, all our sensory experiences are recorded in the brief sense organs as a very brief sensory memory, most of which is lost when the brief memory trace fades.
  1. Different types of sensory memory have been identified.
    1. Iconic memory refers to the visual sensory memory.
    2. Echoic memory refers to the auditory sensory memory.
  2. Moving from sensory memory, the information we attend to is not lost, and is transferred to our short-term memory, where it is retained for a period of 18-30 seconds.
    1. However, it is possible to maintain information in the short-term store for longer than 30 seconds by verbally rehearing information in the rehearsal loop.
      1. E.g. you might verbally rehearse a set of directions in your head as you follow a route in a car.
        Such an example illustrates the limited capacity of short-term memory: it is extremely difficult to retain a long set of directions; most people would manage to retain around 7 pieces of information in this situation.
        A process known as chunking (Miller 1956) does allow us to cope with more than 7 items at a time in short-term memory: chunking involves grouping separate items into lots or chunks, with each chunk then constituting a single item.
        E.g. Single items: 1 7 7 3 5 2 8 8 7 Chunked: 177 352 887.
2. According to Shiffrin and Atkinson, verbal rehearsal is also the means by which selected information passes from the short-term store to the more permanent long-term store.
  1. Information to be transferred to the long-term store is usually selected on the basis of its current importance to us, or its relevance in relation to what we already know about the world.
    1. E.g. A teacher might rehearse the names of students in a new class in order to transfer their names to her long-term memory, but she would be particularly motivated to retain the names of certain students, E.g. very clever or challenging students, so that she could manage future lessons more effectively.
3. Sensory Memory.
  1. Capacity:- All sensory experience.
  2. Duration:- quarter of a second visual and 4 seconds auditory.
  3. Coding (In what form is it stored):- In the form in which it is received (i.e. sense-specific).
  4. Short-term Memory.
    1. Capacity:- 7 items (can be increased by chunking.
    2. Duration:- 18-30 seconds.
    3. Coding:- Mainly phonological (i.e. auditory or sound-based.
    4. Long-term Memory.
      1. Capacity:- Unlimited.
      2. Duration:- Unlimited.
      3. Coding:- Mainly semantic (i.e. based on meaning).
4. Several studies support the multi-store model, either by demonstrating the existence of separate stores, or by illustrating the features.
  1. Research Study: Sperling (1960).
  2. The capacity of short-term memory is usually investigated in what is known as a digit-span task.
    1. Participants hear a series of numbers or letter sequences which they have to recall one at a time. The length of the sequence is increased by 1 item every time. The participant must recall each sequence in the order in which it is presented.
      1. It is usually found that most people can recall sequences of a maximum of 7 items in the correct order. Since rehearsing information in our head like this involves the use of short-term memory, these findings support the view that short-term memory has a capacity of 7, plus or minus 2 items.
        Research shows that children who are experiencing problems learning to read and write often have a very limited short-term memory capacity when tested using a digit-span task.
Primacy-Recency Effect.
1. Research into the primacy-recency effect has been used to support the existence of separate short-term and long-term stores.
  1. If you are given a series of 15-20 words and asked to recall them in any order, you will usually find that you remember several items from the start of the list and some from the end, but you are less likely to recall the items from the middle of the list.
    1. Items recalled from the start of the list (the earliest items) are due to the primacy effect, and items recalled from the end of the list (the most recent items) are due to the recency effect. The effect is also known as the serial position effect, because the likelihood of an item being remembered depends on its position in the series.
2. Supporting the Multi-store Model.
  1. When we hear the first words from the list we start to mentally rehearse them so we will be able to remember them later; this rehearsal transfers the words to our long-term memory. The rehearsal means that we tend to miss the words from the middle of the list altogether, unless they are especially unusual or relevant. The last few words in the list are retained in our short-term memory for long enough when the reader has finished speaking to enable us to recall them straight away from our short-term memory.
    1. Thus the primacy effect is a long-term memory effect and the recency effect is a short-term memory effect. Glanzer and Cunitz( (1966) varied the original primacy-recency task and provided further evidence for the existence of separate stores with different features.
      1. Research Study:- Glanzer and Cunitz (1966).
Evaluation.
1. Many different studies seem to support various aspects of the model, although a lot of the research might be criticised because it involves testing memory in an artificial way. E.g. We never normally need to recall lists of unrelated words or sequences of digits.
2. Perhaps the model is oversimplified, as it suggests that we have a single store of long-term memories. Other researchers, notably Tulving (1972), have suggested that there is more than one type of long-term store.
3. The view of short-term memory as a single store, holding up to 7 pieces of information for 18-30 seconds, is also an oversimplification. The problem can be illustrated with the case of KF, who suffered brain damage following a motorcycle accident. His short-term memory for verbal materials was very poor, but was more or less normal for visual material (Shallice and Warrington 1970). Findings such as this suggest that short-term memory is not a single store. A more sophisticated understanding of short-term memory was proposed by Baddeley and Hitch (1974), in the form of the working memory model.
4. One problem for the multi-store model is the view of rehearsal as simple rote rehearsal or verbal repetition. Studies have shown that rehearsal involving elaboration of materials, into a story for example, is more effective than simple rote rehearsal. Also, we know from experience that some material is remembered whether we rehearse it or not. Flashbulb memories of highly significant or emotional events and images, such as 9/11, seem to be readily remembered without any need to rehearse. Indeed, in everyday life we rarely rehearse information, yet we have little problem in storing huge amounts of it.
5. Recently, many researchers have questioned the idea of separate short-term and long-term memories. Nairne (1996) sees short-term memory as a tiny, highly active part of long-term memory.
6. Despite these criticisms, the multi-store model provided the basis for much research into memory and remains an important milestone in our understanding of the structures and processes involved in memory (Matlin 2002).
Coding Studies.
1. These illustrate how information in the different memory stores is coded or represented differently.
  1. In short-term tasks requiring information to be stored for less than 30 seconds, it is usually found that there is confusion with material that is sound-based (E.g. mat, man, mad, map). This confusion over same-sounding words suggests that short-term memory most involve coding of information according to how it sounds.
  2. In a long-term task with a duration of more than 30 seconds, it is typically found that participants get confused over information with the same or similar meanings (E.g. huge, wide, great, big), suggesting that long-term memory involves the use of a semantic or meaning-based code.
Clinical Studies.
1. These offer convincing evidence for the existence of separate short-term and long-term stores.
2. Clinical cases involve people who are experiencing memory impairments, perhaps because of some brain damage or disease. Such cases demonstrate how some memory functions are relatively normal, while other can be severely impaired.

Next up

The Multi-Store Model of Memory.

Description

Resource summary

Similar

	Created by Stephanie Price almost 11 years ago