Describe the working-memory model and evidence that supports this theory.
Question: Describe the working-memory model and evidence that supports this theory.
Baddeley and Hitch (1974) describe the working memory model as 'a short-term memory system involved in the temporary processing and storage of information'. Over time, the model has been refined (Baddeley, 1986), and is considered to be the framework for the analysis of working memory in language processing.
The model consists of three components: The Central Executive, The Phonological Loop and the Visuo-Spatial Sketch Pad. This essay describes each one individually, giving evidence that supports the existence of each component. Research has predominantly surrounded the 'phonological loop', and as such, more information is included about this component. Some of this evidence is experimental and some neuropsychological, looking into patients who had incurred brain damage some time in their life.
The Central Executive is the primary component in the working memory model. It is believed that it coordinates activity between the Phonological loop and sketch pad and transmits information to other parts of the cognitive system. The executive retrieves information from long-term memory (Hitch, 1980) and allocates information to both components. Its capacity is limited and therefore it must process the appropriate amount of activity incurring in the sketch pad and phonological loop to provide the best resources for a task. Research into the involvement of the central executive and how it works is limited. However, advances have been made into the understanding of the component, and beliefs include that it is comprised of subprocesses including task co-ordination, planning and control of action. The main basis of research has surrounded the latter of these three, and the 'control of action' model (Shallice, 1988). For this reason it is the basis of my explanation of the central executive.
The model describes the central executive as guiding automatic action (schemas) using environmental triggers (presented as phonological and sketchpad elements). Whilst driving, for example, our driving schema takes over. The sight of a zebra crossing in front induces an environmental trigger to look, and be ready to brake in case of people crossing. The contention scheduling resolves any conflicts between schemas. However, when urgent action must be taken (such as an emergency stop) the Supervisory Attentional System (SAS) overrides the contention scheduling and controls action quickly.
Neuropsychological evidence links the Central 'controlling' Executive to the SAS. Shallice (1988) states that patients with frontal lobe damage have a malfunctional SAS that cannot intervene to reduce the effect of multiple schemas or environmental triggers. Therefore the patient is constantly distracted and finds it hard to focus on set tasks and appropriate environmental schemas. This can result in socially unacceptable, uncontrollable behaviour.
The Central executive is therefore believed to be an over-riding component, that although always in control, incurs less demand the more automated the task is. This is consistent with experimental evidence (Baddeley, 1966) whereby ...
This is a preview of the whole essay
Neuropsychological evidence links the Central 'controlling' Executive to the SAS. Shallice (1988) states that patients with frontal lobe damage have a malfunctional SAS that cannot intervene to reduce the effect of multiple schemas or environmental triggers. Therefore the patient is constantly distracted and finds it hard to focus on set tasks and appropriate environmental schemas. This can result in socially unacceptable, uncontrollable behaviour.
The Central executive is therefore believed to be an over-riding component, that although always in control, incurs less demand the more automated the task is. This is consistent with experimental evidence (Baddeley, 1966) whereby participants had to give non-stereotypical, random three-letter strings and therefore deny the use of the automatic alphabetic schema that produces stereotypical strings such as 'ABC, MNO' etc.. Results showed that participants found it harder to produce random strings when the pace increased on the task, or when other concurrent tasks had to be solved, and therefore showed that non-automated tasks increase the demand on the central executive.
As stated earlier, the central executive is the primary component of the working memory model. Alongside it are two separate 'slave systems'. Evidence for a separation of these systems can be found at a biological level. In an experiment PET scans measured the activity of two groups of participants when undergoing a tracking task. Each group of participants were given strings of squares with a letter in one of the corners. The first group of participants were asked whether the square three squares back contained a letter in the same location (spatial information task), whilst the other group were simply asked if the same letter occurred three squares back), irrelevant of location (non-speech phonological loop task). Analyses by Smith, Jonides and Koeppe (1996) showed that the first group produced a lot of brain activity in the right hemisphere, whereas the other group produced a lot in the left 'language' hemisphere, thereby, once again indicating two separate working memory components.one of which is the Phonological Loop.
In 1986, Baddeley refined the phonological loop model into two components: a phonological store, and a sub-vocal rehearsal process used in non-phonological inputs such as text and pictures that converts the information into phonological form.
Neuropsychological evidence supports this updated theory of two components - the subvocal rehearsal and the phonological short-term store. Research into the patient KF by Shallice and Warrington (1970), showed that the patient was significantly better at remembering visual than auditory tasks. This shows a problem with the short term phonological store, and defines a difference in short term memory between auditory and visual information.
Evidence for the phonological loop has been extensive over the past 30 years and is found in four distinct areas:
* Articulatory suppression
* Word Length
* Phonological Similarity
* Irrelevant speech
Peterson and Johnson (1971) proved the articulatory suppression technique inhibits the use of the sub-vocal rehearsal process. In their experiment participants were shown a list of written words to remember whilst engaging in the activity of vocally repeating irrelevant words. The constant verbal rehearsal of the irrelevant words inhibited the use of the subvocal pathway, making it very hard for participants to use the rehearsal process to transfer the listed words into the phonological short-term store. This resulted in decay of the listed words and shows that the recall of letters and digits is based on the phonological loop.
The spoken length of words has been shown to affect performance on memory tasks (Baddeley, Thomsen, Buchanan, 1975). Words that take longer to say (eg, 'harpoon') have been proven to be significantly harder to recall than those that take less time (eg, paper), and no longer do we determine short term memory size in term of items (Miller, 1958). This is compounded by evidence from Ellis and Hennelly (1980) who used bilinguals in Welsh and English to perform a digit span test. The results showed that they had a better span in English numbers, than in Welsh numbers, which invariably have longer utterances. This evidence can be used to show the sub-vocal rehearsal element of the phonological loop, whereby longer words take longer to be rehearsed than smaller ones, and therefore are rehearsed less. This in turn means the sub vocal rehearsal element is less likely to be able to maintain them in the phonological short-term store compared to shorter words.
Evidence of phonological coding is further enhanced by early research from Conrad and Hull (1964). In experiments by them, it was found that phonetically similar groups of letters/words (eg, PGV) were harder to recall than those that were not (eg, KWH). This is because the loop finds it hard to discriminate between the letters, due to their phonetic similarity, so therefore the loss of one phoneme can make the word / letter indistinguishable from another:
P = / p i: / - pea
B = / b i: / - bee - Phonetically similar, as shown by /i:/ which represents 'ee' sound.
K = / k ?I / - kay
Y = / w aI / - why - Not phonetically similar as shown by /?I/ which represents 'a' sound and /aI/ which represents 'i' sound.
Further investigation by Baddeley (1966) showed that semantic or orthographically similar words were not harder to recall, and that therefore there is a definite part in short-term memory that concentrates on the phonological sounds of words.
Irrelevant speech has also helped prove the existence of the phonological loop. An experiment by Salamé and Baddeley (1982) consisted of participants having to remember digit lists whilst hearing irrelevant speech. The results showed that participants were most affected by the irrelevant speech that was phonologically similar to the memory items, and therefore produced overload and confusion in the subvocal rehearsal.
The second 'slave' system to the central executive is the visuo-spatial sketchpad. The system has not been investigated in as much depth as the phonological loop, and as such evidence concerning its function is limited. However, research so far indicates that the sketchpad is a component which generates stores visual and spatial information, and therefore encodes verbal material in the form of imagery, and not phonologically. Like, the phonological loop, only concurrent tasks affecting the central executive and itself are believed to have a significant negative impact on visuo-spatial task results (Baddeley, 1992).
Evidence has been found using an experiment involving the 'pegword' strategy, whereby it is believed the sketchpad is used to generate information into an image. In the experiment (Baddeley and Lieberman, 1980), one group of participants were asked to remember a list of words using the 'pegword' strategy, whereas the other group were asked to use a simple 'rote' strategy. These results were then compared with a repeat of the experiment, but this time participants were subjected to a visuo-spatial tracking task. The results showed a clear negative difference in the performance of the 'pegword' participants, believed to be due to an overload in the visuo-spatial component. The 'rote' strategy participants were unaffected.
Evidence of such a component can also be found at a child's developmental stage. Hitch, Halliday, Schaafstal, and Schraagen (1988) discovered that children aged five find it easier to remember phonologically similar pictures than ten year old children, but find it harder to remember visually similar pictures than older children. This shows that there are two obvious components, and that the visuo-spatial sketchpad is primarily used when young, until the phonological loop begins to have more of an effect.
In conclusion, the Baddeley and Hitch (1974) model for working memory has extensive supporting evidence for the model. Over time, the model has been refined (Baddeley, 1986), and other models have been used to help explain how a component works (Shallice, 1988, 'control of action' model). The main belief surrounding the model is that working memory contains three components, of which one (the central executive) can affect, and have overall control over both other components. Encoded memory is believed to be divided into phonological (including speech and written speech) and visual - spatial components. Further evidence has then shown a higher-order central executive component that controls the amount of information received from each so as to benefit us the most in a task.
