Multi-Store Memory Model

Long term memory (LTM) is where information is kept beyond that of sensory or STM. Long term memory has a seemingly unlimited capacity, with a duration of anything from a few seconds to a lifetime. It is where all of our knowledge and skills are kept, every skill and technique is stored here. Without LTM, we would not be able to perform even the most simple of tasks. In LTM, information is encoded in terms of its meaning rather then its sound, for instance if the word ‘ barn ‘ is to be remembered, it is more likely we would remember words with a similar meaning, i.e. shed, hut, then a word with a similar spelling or sound, i.e. born, because that would be a sound based error, which is an unusual mistake in LTM. Semantic is another word for ‘ meaning ‘, and many researchers refer to LTM as semantic memory.

Support for this model is vast, Jacobs investigated the serial digit span. He used letters and numbers (apart from W and 7 as they have 2 syllables). Using a metronome an item would be presented every half second. The average digit span was just over nine whereas letters was just over seven. He also found that as we age we improve at being able to remember both letters and digits. Jacobs research backs up Millers ‘ magic number ‘ theory also, but it also shows that how much we remember can be influenced by what is actually is we are trying to remember.

Further support comes from Bahrick in 1975. He wanted to find out about the duration of long-term memory. So he carried out an experiment and used participants from the US because of their tradition of high school yearbooks having a statement and picture of each person in the year.

They carried out 3 tasks. Remembering names of ex classmates, recognised pictures which had been mixed up and to recognise names of people from their school. They found that verbal and visual recognition was highest in those who had left school less than 15 years previously and 48 years  after school, accuracy had dropped off to 70% verbal names and 80% visual faces. This shows that LTM is not permanent, and also that we remember things better with recognition then just recall.

There are general strengths and weaknesses of the model for example, Murdock and the serial position curve. He found that when people try to remember a list of words, they tend to remember the first few words, the primacy effect, and the last few words, the recency effect, and the middle words are usually forgotten. This can be explained by the idea that the first few words where probably rehearsed alot, so they would be remember, and the last few words would be remembered and they would still be in the approximate 30 second time slot of the STM.

The distinction between STM and LTM has been supported by cases of amnesia, for example Clive Wearing. He suffers from a 7 second memory. He cannot learn new skills or remember new things as information cannot be transferred into his LTM. But before his memory was damaged, the things he remembered such as playing the piano were unharmed, supporting two different memory stores. Also a distinction can be seen in people with Korsakoff’s syndrome. This is a memory disorder caused by alcohol abuse and people with this have problems with their long term memory but their short term memory is still intact. So they can have normal conversations but have no memory of them later. This shows that brain damaged people clearly suggest that short term memory and long term memory are completely separate.

Some may say that the multi store model of memory is too simplistic, for instance the studies which support it, namely Conrad, Jacobs and Peterson and Peterson all only used letters numbers and words, which only tests one part of the memory, therefore it could be said these findings are being unfarily generalized for the whole of the memory. It is also argued that the idea that there is only one single STM memory store is innacurate. It is assumed that STM operates using sound based information, but Shallice and Warrington (1974) show this is not the case. They found their patient’s STM problem was restriced to verbal material, such as words and letters. His in STM tests for non verbal sounds, such as telephone ringing were normal. This suggests STM must have at least 2 different sound based mechaninsms, for verbal and non-verbal sounds.The model suggests also only one LTM store. However, Tulving (1989) distinguishes between Episodic (early childhood episodes and events), Procedural (how to do things), and Semantic (Meanings and tasks), the multi store model does not discuss these types of LTM.

Also STM is more complex then previously thought, Baddeley and Hitch propose a working model of memory, which is active and composed of several sub systems, controlled by one central executive. These sub systems may work together or alone depending on nature of the task. Also the MSM takes no account of the nature of information to be recalled and treats all information as the same. Not all information needs rehearsal. Psychologists have studies the nature of flash bulb memories when people have a strong or very detailed memory of a particular event, such memories are associated with events, which are unexpected, seen as important and create strong emotional arrousal.

The MSM is an interesting interpretation of how the memory could work, and many of its ideas are backed up by some researches, but many of these supporters could be argued to be too simplistic, and the same could be said of the whole model. The MSM does explain many parts of the day, but it is still just a model and no definites can be proved.