The working memory model was proposed by Baddeley and Hitch in 1974. This was an alternative to the multi-store model because they believed it was too simplistic and did not explain how memories are transferred into the long term memory without rehearsal which can be supported by flashbulb memories e. g. the death of princess Diana. The working memory model focuses on the short term memory and suggests that unlike the multi-store model, there is more than one store. Baddeley and Hitch used the term ‘working memory’ to refer to the part of the memory that is active or working.
This could be as simple as playing a game, calculating sums or reading a sentence all of which are collecting data to be stored. The working memory model consists of three parts; these are the central executive, phonological loop and visuo-spatial sketchpad. The central executive controls attention and draws on the other two systems known as the ‘slave systems’. The central executive has a limited capacity; in other words it cannot attend to too many things at once. This is supported by the dual task technique in 1976 where participants were given two tasks to do simultaneously.
The first task used the central executive which was a simple sentence verification task e. g. participants were asked to answer true or false when shown the sentence B is followed by A. The second task involved the central executive and the phonological loop where participants had to repeat the word ‘the’ over and over again whilst working out the sentence verification task. The third task used the central executive and the central executive and phonological loop. In this task participants had to say random digits between 1 and 9 whilst completing the sentence.
Hitch and Baddeley found that the time taken on task 3 was significantly longer because the task involved using the same component, the central executive making completing the task more difficult. They also found that when different components were used, such as in task 2, the performance was not affected. The problems with the central executive are that little is known about what it actually does therefore it is difficult to prove its existence since there is no specific information linked to this area like visual or auditory information.
Recent research has found that the central executive differs between individuals and it can be easily selectively impaired by brain damage. Baddeley coined the term dysexecutive syndrome (DES) to describe dysfunctions of the central executive As well as this, many psychologists believe that more research is needed to clarify the precise nature, capacity, and functioning of the central executive as it is not clear whether it forms a single unitary system or consists of separate components.
If various functions are used it is suggested that the central executive is responsible for tasks such as timesharing and selective attention. The second part to the working memory is the phonological loop. This can be further divided into the phonological store and articulatory process. The phonological loop has a limited capacity and deals with auditory information as well as preserving the order of information. The information goes round in a loop-hence the name, and simply holds words you hear, sometimes referred to as the inner ear.
It appears that this store is used for learning new words and simply holds auditory data. The information is said to enter this store automatically and the information that is rehearsed prevents decay. There are many findings to support the phonological loop; one is the effect of phonological similarity. Lists of words that are acoustically similar are more difficult to remember than those that are acoustically dissimilar. Semantically similar words seem to have no effect which supports the assumption that verbal information is encoded largely phonologically in working memory.
Another piece of evidence that supports this store is the word-length effect. This describes that people can cope with shorter words better than they can with longer ones. This is because the phonological loops holds information you can say in approximately 2 seconds before it is passed on for more processing, this makes it harder remembering a list of longer words because it takes more time to say them therefore the individual is unable to store them. Longer words cannot be rehearsed because there simply isn’t enough time and they don’t fit.
The word-length effect disappears when the individual is given an articulatory suppression task such as the task 2 one. By repeating the word ‘the’ continuously the shorter words cannot be rehearsed because the articulatory process is tied up. It cannot do two things at once. The words can’t be rehearsed more quickly than the longer ones therefore it does not matter how long the words are when another task that involves the same component is being used. In this case the articulatory process is assumed to be blocked consequently the memory trace is more susceptible to decay.
The articulatory process is used for words that are heard or seen, these words are silently repeated, which acts like an inner voice. The third part to the working memory model is the visuo-spatial sketchpad. This store temporarily stores visual/spatial data. Visual information is what things look like such as colour, texture and shape and spatial information is the relationship between two things such as locations and movement in space. These two tasks link together such as a visual task of working out how many windows are in your house, and the spatial task of counting them using the relationship between the rooms in the house.
Although this seems that this store many be unitary, a notion has been supported that there may be another subdivide within the working memory model where the visuo-spatial sketchpad splits up into the visual cache and inner scribe. Support for the visuo-spatial sketchpad and its distinction from the phonological loop comes from Quinn and McConnell in 1996 where dynamic visual noise was contrasted with static visual noise. This was introduced to the individual and then a visual task was presented.
They found that in the first experiment the results showed that no interference has been caused when the visual task was given. The second experiment was carried out to find out why this was in which they concluded that the static field is susceptible to decay and so fails to cause interference. In the final experiment the dynamic visual noise was further researched where they discovered that by manipulating the number of changes in the noise field, this played an importance in the interference. This research supports the theory that the phonological loop and visuo-spatial sketchpad are separate stores.
A new piece of research by Boutla et al researched the puzzle of working memory for sign language. Since sign language is visual information then we would expect that this would be stored in the visuo-spatial sketchpad however evidence shows that the working memory for speech and sign is very similar. For both speech and sign the information is maintained in the phonological loop and they both use articulatory process to rehearse the information sub-vocally or sub-manually. For sign language, phonological coding is based on manual rather than oral features e. g. hand/arm orientation.
Despite this information, storage capacity has been found to differ significantly for speech and sign, with speakers consistently exhibiting a longer span than signers. Further support comes from physiological studies where different areas of the brain have been active during the different STM memory processing tasks. This diagram shows the areas in the brain in correspondence with the processing of information. In locating where these parts of the brain are situated we have a deeper understanding of what they do. Each part of the brain governs part of the memory that is part of the cortex therefore the working memory making more sense.
This can sometimes explain brain-damaged patients who often suffer damage to part of the STM such as KF’s motorcycle accident where his mainly his verbal information was affected supporting the working model of separate stores within the STM. Strength’s of the working model include the fact that we don’t have to verbally rehearse information in order for it to be transferred to LTM like the multi-store model suggests. The working model says that verbal rehearsal is optional but not necessary for the information to be stored.
Another strength is that it proves that we can perform tasks together when using different components possibly because they are located in different areas of the brain therefore there is less chance of a mix up between neurons in the brain. The model is praised for its validity because it is able to incorporate a large amount of information on STM and working memory. It has also generated a lot of research into the two ‘slave systems’ in particular the phonological loop which has inspired greater research into experimental psychology, neuropsychology, and cognitive neuroscience.
Some weaknesses to the model appear in the central executive where very little information is known about it, as well the model still being too simplistic and fails to recognise any ‘real’ cognitions, it is difficult since the brain and mind is complex. (Andrade 2001). There is some evidence to suggest that intelligence may play a part in working memory but yet again this evidence has little validity or support from other psychologists. Furthermore the model does not acknowledge how the brain processes change over time, or how much practice we put in to remembering things.
Since we know that over time memory decreases more information is needed to say what happens to the three systems within the working memory. A final criticism to the model is that much of the research done to support the working model was done in laboratory settings therefore it lacks ecological validity because it lacks being able to explain memory in real life as well as a lack of understanding about how this may affect other people e. g. individual differences such as gender, age, ethnicity therefore lacks external validity as well.