Working Memory

From Psy241wiki

Jump to: navigation, search

Baddeley and Hitch's working memory model was devised as an alternative to the idea of STM proposed by Atkinson and Shiffrin's model


Components of the Working Memory Model (WM)

Central Executive(CE)

Visuo-Spatial Sketchpad (VSSP)

Phonological Loop (PL)

Episodic Buffer (EB)

Neuropsychology of the Working Memory Model

Neuropsychology of the Working Memory Model

Alternatives To The Working Memory Model

There are a number of alternatives to the Baddeley and Hitch's (1974) Working Memory Model, all of which can be used to evaluate and criticise it. However, the following are only a brief summary of the major criticisms:

1. Theories of short-term memory e.g. Nairne's (1990) Feature Model. Unlike the Working Memory Model which focuses largely on the importance of phonological coding in the working memory for the transfer from short-term to long-term memory, the Feature Model focuses on a whole range of coding e.g. semantic and episodic

2. Cowan's Embedded Processes Theory: [ Cowan considers working memory to be a part of short-term memory.According to him, WM is organized into two embedded levels. The first level consists of limitless representations from LTM that are activated. The second level is called the focus of attention and this has a limited capacity and holds up to four of the activated representations. ]

Cowan's (1999) theory, unlike Baddeley and Hitch's (1974), accounts for a limit in the capacity of attentional focus across the areas of active long term memory. Cowan argued that working memory depends on the activation of LTM, and that attentional focus controls this activation. This activation is temporary in nature and over time is subject to decay without rehearsal; however, it is multi-dimensional like the episodic buffer. Cowan also argues that unlike Miller's original seven chunk capacity, we are in fact limited to four.

Baddeley sees Cowan's model as some sort of link between Central Executive and Episodic Buffer.

Although these two points only offer a very short evaluation, they do provide a strong and alternative argument against Baddeley and Hitch's (1974) Working Memory Model and lead to a wide range of research which can be used in support of these opposing arguments.

3. Barrouillet & Bernardin's 'Time-Based Resource-Sharing Model' (2001): The authors propose that working memory tasks require attention, and once attention on a task is diverted the memory traces decay; opposed to a lack of verbalisation. An important point to note is that Barroillet and Bernardin's model does acknowledge Baddeley and Hitch's resouce-sharing characteristic of working memory, only Barrouillet and Bernardin define working memory as having a time-based component.

- Research by Towse and Hitch (1995) suggests that a temporal interval between storage and recall causes memory traces to decay, supporting Barrouillet & Bernardin's claims. Case et al. (1982) conducted a study involving a counting span task in children of varying ages. The results indicated that younger children performed worse on the task than the older children. They suggested that counting span tasks, such as the one they used, become less 'resource demanding' over age, resulting in more available resources for storage. A convincing point, but Towse and Hitch (1995) presented an alternative hypothesis termed the "memory decay" hypothesis. This hypothesis appears to be a precursor to Barrouillet and Bernardin's model as it incorporates the idea that a 'temporal interval' between storage and recall weakens memory traces. They used this idea to explain that perhaps the older children in Case et al.'s experiment could count the dots (part of the task) faster, causing a smaller temporal interval. Towse and Hitch thus conducted a test to prove their point, and eventually concluded that "counting spans depend on the duration of the counting, during which the totals may be forgotten".

Implications and links to other topics

Implications and links to other topics. This page is for any practical applications or random musings on how we can apply WM research to our lives, as well as interesting crossovers with other topics in psychology.

Working Memory Dysfunction

When the working memory system doesn't function properly this can tell us a lot about the functions of normal working memory.

Disorders of Verbal Short Term Memory

When the working memory system doesn’t function properly this can tell us a lot about the functions of normal working memory. There are several case studies that can tell us a lot of disorders of working memory. The first is patient PV (Baddeley et al., 1988). This patient had a left hemisphere stroke which resulted in brain damage in the left hemisphere, which resulted in normal LTM (paired associate learning, story recall) but impaired verbal STM (auditory digit span was only 2-3 items). As a consequence of having an impaired verbal STM they were unable to decide if long sentence were true or false (for example, the world divides the equator into two hemispheres; the northern and southern) because they couldn’t remember what was at the start of the sentence. This suggests that the phonological loop plays a role in comprehending material. They were also unable to learn pairs of novel words with words they already knew (van-jalka). These experiments helped reach the conclusion that PV is unable to hold items in a short term phonological store. This results in inability to transfer information into the long-term phonological storage.

A second case that can tell us a lot is the case of SR (Baddeley, 1993). SR was a 23 year old graduate student who had no history of neurological damage but an impaired verbal STM (poor spelling, difficulty learning second language). Like patient PV he struggled to learn word-nonword pairs but could learn word-word pairs. This supports the idea that an inability to hold items in a short term phonological store means that they cannot be learnt as they can’t be transferred to the long term phonological store.

The third case is that of patient SC (Trojano et al., 1992). Like patient PV he had a left hemisphere stroke, and a poor verbal STM (although his visual STM was normal). He took longer to remember all words than matched controls but eventually reached their level in learning bisyllabic concrete nouns, trisyllabic concrete nouns and abstract nouns but never reached control's level in function words (why, what). This suggests that the phonological loop is important in remembering function words which aren’t richly encoded.

Overall, these results suggest that important functions of verbal working memory are: comprehension of difficult material; learning novel words; remembering familiar words which aren’t richly encoded in semantic memory. This has implications for language development in children as it suggests that children with a poor working memory might struggle learning words.

Gathercole and Baddeley (1989) investigated language development in children. They tested them with a non-word repetition test (where 40 nonwords were read out by an English female and the children were required to repeat each one). A significant correlation was found between non-word repetition and vocabulary ability when children started school, even when the effects of age and non-verbal IQ was removed, and this relationship was maintained a year later. Non-word repetition score at aged four predicted degree of vocabulary development during 1st year at school. This provides further evidence for the fact that the phonological loop aids language learning. The new word will go into the phonological loop and will then be transferred to the long term store, if the word cannot be comprehended by the phonological loop then it will never be transferred. Gathercole and Baddeley (1990) suggest that ‘the contribution of phonological memory to learning a new word seems likely to be located in the processes of achieving stable representation of the initially novel sound sequence in long-term memory’.

Disorders of Visual Short Term Memory

Disorders of visual short term memory provide evidence for the fact that the visuospatial sketchpad is split into two components. Patient LH (Farah et al., 1988) was poor at remembering colour and shapes but had a good memory for spatial information, suggesting that the inner scribe was intact but non the visual cache. Patient MV (Carlesimo et al., 2001) had a good visual memory for colours and shapes but performed poorly on spatial tasks, implying that the visual cache was intact but that the inner scribe was impaired.

The case study conducted on patient ELD by Hanley et al. (1991) can also inform us about normal working memory. Patient ELD had an aneurysm in the right hemisphere and presented with a number of complaints, mainly that they couldn’t remember new routes or faces (however they could recognise new voices so it wasn’t amnesia). During a formal assessment ELD was found to have: above average IQ; average performance on a standard memory test; good at remembering new verbal info; normal at identification of old famous faces, but was impaired on: remembering new visual information; identification of new famous faces. Hanley et al. (1991) wanted to see if these problems, as well as those in LTM, could be explained by a deficit in visual STM; a deficit in the visuospatial sketchpad. They concluded that as ELD has difficulty in remembering short sequences of visuospatial information but her phonological loop seemed to function normally, her deficit can be interpreted as a specific impairment in the visuospatial sketchpad. ELD shows poor imagery for material she has been exposed to since her illness but not for information required before her illness. Thus her deficit seems to be in generating new images rather than in achieving old representations. Finally, it is possible that ELD's visuospatial sketchpad impairment may underlie her impairment is visual LTM, if the visuospatial sketchpad plays a role in transferring information into LTM as well as maintaining it in STM.

Disorders of the Central Executive

Alzheimer’s disease is a progressive dementia which results in impairments in memory, perception and language. Patients often have a very poor visual STM and a low span for verbal STM. Baddeley et al. (1986) attempted to find out of these STM deficits can be interpreted as an impairment in the central executive. They did this by using a task which made heavy demands on the central executives to co-ordinate information from different sources (a dual task). Normal elderly were no more disrupted than young subjects on the dual task, but Alzheimer’s patients were very seriously disrupted by the requirement to coordinate the two tasks and the extent of disruption the in Alzheimer’s patients increased with development of their dementia. This suggests that those with Alzheimer’s do have an impairment in central executive function as the central executive is believed to be responsible for dual task performance.

Patient RC had a left frontal lobectomy and was also believed to suffer from impaired central executive function (Allain et al., 2001). Like those with Alzheimer’s he too performed poorly on dual tasks, and his performance decreased as the tasks became harder. This suggests that the frontal lobe is important in functioning of the central executive. It also implies that the central executive has a limited capacity, as his performance (as well as that of controls) decreased as the tasks became harder.

Personal tools