Describe and evaluate the multistore model of memory
Describe and evaluate the multistore model of memory
The multi-store model of memory was designed by Atkinson and Shriffin (1968). They developed this model in order to explain the way our memory works; how much it can hold (capacity), and how long we can hold memories (duration). The model is based around the idea of three separate stores which hold our memories, sensory, short term and long term.
The sensory memory is collected by senses, the ears, eyes, mouth, etc. The store is constantly receiving information and so the data collected only remains in the store for a very small amount of time. The model suggests that if attention is then given to the senses then the memory transfers to the short term store. Evidence in support of the sensory store was provided by Sperling (1960); participants were shown a grid of digits and letters for 50 milliseconds and when asked to recall what they saw, the average recall was 42%. This information supports as it suggests, memory decays rapidly and therefore supports the concept of a store which cannot hold large amounts of memory. It was also centered around the use of sight supporting the sensory aspect of the store. However if this store can only hold memories for such a short amount of time should it be considered a store at all?
This is a preview of the whole essay
The short term store (STM), according to the model, is our memory for events in the immediate past (such as ordering drinks at a bar) and present. The information is said to be held in a ‘fragile state’ and the capacity is limited. Furthermore if new information enters the store it displaces memories that were there to begin with. Memories move from the short term to the long term through elaborative rehearsal. Evidence to show the existence and capacity of the STM has long been researched. Although the model was developed much later, Jacobs (1887) used the digit span technique to test the capacity of memory and in the context of this model, the short term memory. He found the average span for digits was 9.3 items and for letters 7.3, suggesting that STM capacity is affected by what we have to remember.
Miller (1956) also investigated the capacity of STM and found it to be between 5-9 or 7+/- 2 chunks. A ‘chunk’ represented a meaningful collection of items. This research supports the existence of STM as a limited store and turn suggests that memories are held in a fragile state and there must be another store for which our other memories are located. However there has also been studies that imply these ideas to be wrong, Simon (1974) suggested that the size of the chunk doesn’t affect memory and Cowan (2001) stated the number of chunks is more likely to be 4 than 7. Because of these suggestions it therefore questions the validity of the other studies and provokes further research into the capacity of the STM which can’t be disproved.
There has also been research into the duration of STM; how long a memory lasts before its no longer available. Peterson and Peterson (1959) found when rehearsal is prevented that STM lasts 20 seconds at the most. This supports the original proposals about the STM store, that memories are held in a fragile state and that information decays rapidly if not rehearsed. On the other hand Nairne et al. (1999) suggested the STM could last as long as 100 seconds emphasising the fact we still have not found an exact answer and more research has been provoked by the contradictory natures of these studies.
However some studies have suggested there is more to the multi-store model and that oversimplification is a key flaw of this model. For example, some studies have shown there is more than one part to the short term store and that instead of stores there could in fact be different layers of memory. It is also argued that the idea of separate stores could misrepresent memory; in the MSM memories move from the STM to LTM through elaborative rehearsal but Logie (1999) found evidence to suggest that the STM relies on the LTM so it cannot come ‘first’ in the model.
In MSM, the LTM is proposed as the store for events that have happened in the more distant past. The relationship between rehearsal in STM and the strength of the LTM is that more information is rehearsed in the STM the better you will remember it. Research into the duration of LTM has also been investigated. Batrick et al. (1975) designed a natural experiment when participants were shown their yearbook 48 years after leaving school and had to identify the people. It was found that they were 70% accurate, supporting the idea our LTM store can hold information for a very long time.
Much research has been done into the existence of separate stores and due to more recent technology such as brain scans; psychologists have been able to link STM and LTM to different areas of the brain. For example Beardsley (997) found the prefrontal cortex was active when individuals worked on a STM task and Squire et al. (1992) found the hippocampus to be involved when participants worked on a LTM task. This evidence suggests separate stores as they are linked to separate areas of the brain therefore strengthening the ideas about the MSM.
On the other hand some research has shown the hippocampus to be part of the formation of STMs. Scoville and Milner (1957) carried out a case study on a man called HM, who in an attempt to reduce severe epilepsy had undergone surgery which unknown to him was experimental and his hippocampus had been removed. Afterwards, HM could not form any new memories but could recall LTMs, suggesting that the hippocampus may act as a gateway that new memories must pass through. This research goes against the idea of separate stores as the hippocampus is suggested to be connected to the processing of STMs whereas the hippocampus was only linked with LTMs in Squire et al’s research. However this was a case study therefore the findings aren’t generalisable to the whole population.
In conclusion, the MSM is strong due to the testable predictions it provide and the wealth of research evidence to support it. However other evidence has contradicted it and it has been criticised as oversimplifying the memory process. So can it be the most valid and reliable model?