View sample Psychology of Working Memory Research Paper. Browse other research paper examples and check the list of research paper topics for more inspiration. If you need a religion research paper written according to all the academic standards, you can always turn to our experienced writers for help. This is how your paper can get an A! Feel free to contact our custom writing service for professional assistance. We offer high-quality assignments for reasonable rates.
Working memory (WM) can be thought of as the desktop of the human brain, a set of cognitive functions that allows humans to keep track of what they are thinking, what they are doing, or where they are moment to moment. It holds information long enough to make a decision, to acquire vocabulary, to mentally image the layout of a home, to support creative thinking, or remember what to do next. This article describes the theories that have been developed to explain these human abilities and some of the experimental evidence for these theories that has been derived from empirical studies of healthy adults and children and of individuals who suffer from cognitive deficits as a result of brain damage.
Academic Writing, Editing, Proofreading, And Problem Solving Services
Get 10% OFF with 24START discount code
1. Why Is The Concept Of Working Memory Useful?
The concept of working memory (WM) has figured prominently in theories and explanations for human on-line cognition, as well as offering a basis for understanding how humans accomplish a wide range of everyday tasks. It can be thought of as a mental tool that aids survival; if an organism is to interact successfully with its environment, then clearly there must be some means to record, on a temporary basis, its own most recent actions and to have some means to plan actions for the immediate future. This gives a significant advantage over organisms entirely driven by automatic responses to external stimuli that signal basic needs. Examples of how this mental tool is used in everyday interaction with the environment highlight its utility. Humans use WM every time they have to solve a problem, to perform a mental calculation, to recall where an object has just been laid down, or to repeat a strange foreign word that has just been heard.
From these examples, it is clear that the concept is much broader than the more traditional idea of short-term memory. Indeed, the concept of WM arose from the need to provide a theoretical account of a wide range of everyday tasks that require both temporary memory and on-line cognition. Until the early 1970s the focus of research on this topic was on retention of random sequences of words, letters, or digits, with the everyday example of recalling telephone numbers while dialling. However a mental faculty would not have evolved to deal with these relatively artificial tasks, and therefore WM is intended to encapsulate on-line cognition and temporary verbal and visuospatial memory in both laboratory tasks and in everyday life.
2. Theoretical Views Of Working Memory
Most memory researchers consider WM to be a useful concept that uses relatively few assumptions to explain a wide range of experimental evidence as well as important aspects of everyday cognition: it is a system with limited capacity, which loses information over periods of seconds, whose contents are subject to displacement by new input, and which interacts and is supported by other parts of the cognitive system such as stored knowledge or the products of perception. Most theories of WM include these features, however, there remains a debate as to its precise characteristics. Some theories view memory as a largely unitary system with WM as the sum total of what is currently activated from the store of knowledge, skills, expertise, and past events. Others view WM as separate from the long-term store of knowledge and events, but even among these latter theories there is some discussion as to whether the cognitive functions are supported by a single flexible system that provides both temporary storage and on-line processing, or comprises multiple components, each specialized for a particular function. A recent book edited by Miyake and Shah (1999) provides a comprehensive discussion of the different theories of WM.
The first type of theory, that WM is simply currently activated knowledge, can account for a wide range of data from experimental psychology with healthy adults, including the ability to bring rapidly to mind information related to areas of personal expertise. Theories of this kind can very successfully explain why chess masters can play several games simultaneously, expert waiters can recall details of meal orders from large groups, and why burglars can recall better than householders, or indeed the police, details of houses related to their crime. Expertise extends the apparent capacity of WM, but there is some recognition among proponents of this theory type of the need for a temporary memory system to deal with completely novel or unfamiliar material. Such theories additionally assume inhibitory processes that serve to prevent activation of information that is not strictly relevant to the task in hand. A breakdown in these inhibitory processes has been suggested as one of the corollaries of normal aging, with elderly people more prone to be misled by irrelevant information or to wander off the main topic in a conversation.
The strict view that WM is no more than the sum total of currently activated experience and memories for events runs into difficulty when considering dissociations that are apparent from studies of healthy adults performing more than one task at the same time. There is substantial evidence that healthy adults can perform two quite demanding tasks concurrently (such as holding a sequence of digits and following rapid random movements of a target). There is very little impact on the performance of either task, and this can be accomplished without any particular expertise or training. Complementary studies have shown that placing a heavy demand on temporary memory places few if any restrictions on retrieval of stored knowledge about past events or that is linked to expertise. These findings suggest that the human cognitive system may comprise several components, each specialized for particular kinds of tasks and which may operate in parallel. The findings also point to the idea that WM is best considered as being separate from the system that stores knowledge of the world and of experiences.
The unitary view of memory also has difficulty explaining patterns of impairment and sparing in individuals who suffer from cognitive deficits as a result of brain damage. The cognitive impairments are often selective, for example, with very poor performance on immediate memory tasks (such as repeating back a random sequence of digits) coupled with normal performance on long-term retention (such as recalling events that occurred weeks, months, or years previously). The converse dissociation also occurs, namely normal immediate memory contrasting with very poor retrieval from long-term memory or difficulties in long-term learning. Theories assuming that working memory comprises temporary activation of relevant information from long-term memory would predict that if long-term memory access is impaired then temporary activation should also result in poor performance. Theories that assume WM to be a functionally separate part of the cognitive system can account more readily for the neuropsychological dissociations as well as the data patterns from healthy adults, and therefore may be more useful.
3. The Case For Multiple Memory Systems
During the first half of the twentieth century the prevailing view in North America was of human memory as primarily the result of observable learning of associations, and theories of learning dominated. During the same period in Europe, memory as a cognitive concept was more salient, for example in the work of Bartlett, Broadbent, and Zangwill. By the mid-1950s, the latter became the more generally agreed view, and a cognitive approach to memory gave rise to major new avenues of research. Very quickly it became clear that memory was best thought of as comprising at least two elements, with one focusing on temporary or short-term retention and the other on long-term retention. This reflected a dichotomy proposed in the seventeenth century by the British philosopher John Locke who suggested that the ‘faculty of retention’ comprised ‘contemplation’ or ‘keeping an idea in mind,’ and that this was separate from ‘memory’ or the ‘storehouse of these ideas.’ This distinction appeared again in the late nineteenth century when the North American philosopher and psychologist William James referred to primary memory which he contrasted with retention.
During the 1960s, the contrast between primary memory and what then became secondary memory was demonstrated experimentally. For example, when healthy adult volunteers are asked to retain sequences of words for immediate recall in any order, omissions tend to occur among words from the middle part of the list and much better recall is observed for items close to the beginning or close to the end of each list. If instead, the volunteers are asked to recall the items following a delay, performance on the last few items is no better than it is for items in the middle of the list. However, initial items in the list retain their advantage despite the delay. These contrasting patterns were thought to support the idea that two memory systems are reflected in recall of different parts of the list: the early list items are thought to be rehearsed and transferred into a longer term memory system. As the list progresses, rehearsal becomes more difficult and therefore few items are transferred to a long-term store. The final items are thought to be held in a temporary memory system whose contents are lost unless recall is immediate. The advantage for the final items with immediate recall is known as the ‘Recency Effect.’
Further support for the dichotomy comes from amnesic patients, that is people who, following brain damage, suffer from severe memory impairment that is characterized by an inability to learn new material. When such patients are tested for immediate free recall of word lists, their pattern of recall lacks the advantage for the initial items in the list but retains a normal recency effect. This suggests damage to the memory system responsible for retention of the early items but normal functioning of the memory system responsible for the recency effect.
A second experimental paradigm demonstrated that different memory codes are used in immediate and delayed recall. In one set of experiments, volunteers were given lists of words that were similar in meaning (e.g., tall, broad, wide, fat, huge, big, great) or were similar in sound, that is they were phonologically similar (e.g., man, mad, mat, map, cap, can, cat), for immediate recall or for recall after a delay. Recall of both lists was compared with memory for lists of words that were neither semantically nor phonologically similar (e.g. car, lock, spoon, book, chair, pen, dog). With immediate recall, phonologically similar words were recalled more poorly than were words from the other lists, a phenomenon known as the acoustic similarity or the phonological similarity effect. It is interesting to note that this effect appears even when the words were read by the volunteers, suggesting that the visually presented words were translated into a sound-based or phonological code in memory. In contrast, with delayed recall, the semantically similar words led to poorest recall. Because phonological similarity affected short-term but not long-term retention, while semantic similarity showed the converse, this led to the conclusion that acoustic or phonological codes were a signature of a temporary memory system (for visual or auditory presentation) and semantic codes were characteristic of the use of a longer term memory system.
4. Developing The Concept Of Working Memory
The idea of primary and secondary memory provided a useful framework to account for the accumulating body of experimental data in support of a two-component memory system. However, it was limited by the fact that virtually all of the research relied on experiments with presentation and recall of words, letters, or digits over relatively short time periods. It is clear that we can also retain information about the appearance of objects and scenes as well as their names and descriptions. Therefore, from the late 1960s, primary memory began to be seen as a system that could support temporary retention of different kinds of material as well as offering a means to process and manipulate that material. The dominant or ‘modal’ concept then became that of a short-term, or WM processing and storage system. This ‘modal model’ retained the notion that information was transferred from the senses, via WM to secondary, or long-term memory. It also incorporated the assumption that the system has a maximum capacity that could be employed by storage, by processing, or by elements of both.
However, a number of features of this model came under pressure from the 1970s onwards. The recency effect, once thought to be a major source of support for a temporary memory system, was also demonstrated in longer term recall. For example, when recalling visits to movies, where the car was parked, or attending rugby matches, individuals can often recall details of very recent events but have rather poorer recollection of experiences that are more remote. This demonstrated that recency was not unique to a temporary memory system. Also, there were demonstrations that semantic coding, thought previously to be the prerogative of long-term storage, could be used to help support short-term retention. For example, immediate free recall of real words is better than recall of nonsense words, suggesting that the additional stored information about the real words (such as their meaning) can support immediate memory. These findings did not wholly undermine the traditional view of a short-term memory, because there was other evidence that a unitary model of memory could not readily explain, such as that mentioned above from dual task performance and from brain damaged individuals. However, the findings of long-term recency effects suggested that recency might be a property of any human memory system (short or long-term), and the evidence for short-term semantic coding showed that the memory systems interact, with any one memory task probably drawing on several different aspects of the cognitive system depending on task demands.
Another core assumption of the modal model has come under increasing strain in recent years, namely the idea that information from vision or hearing could reach long-term memory only through WM. For example, one famous neuropsychological case is that of patient ‘PV’ who, following a stroke in the sylvian region of the left hemisphere, presented with a very severe deficit in retaining sequences of words or digits for immediate recall, but had no difficulty in retrieving details of autobiographical events and proved well able to learn new material. The pattern for PV and other similar patients suggests that access to long-term memory does not depend on normal function in short-term verbal retention, either for new learning or for recall of past events (e.g. Vallar and Shallice 1990).
Additional evidence comes from patients who suffer from a neuropsychological disorder known as unilateral spatial neglect. This arises most commonly from damage to the right parietal cortex of the brain, and such patients typically are unable to report details of scenes or objects to their left, despite having normal vision. A few patients show unilateral neglect only for information in their mental image or immediate memory for scenes. For example, one patient ‘NL’ (Beschin et al. 1997) could quite readily report details of everything in his immediate environment. However, if he was asked to close his eyes, he could only report from memory details that were on his right. Moreover, if asked to recall details of a familiar scene such as the square in his home town he could recall only those details that were on his imagined right. This was not a problem of accessing his long-term memory, because when asked to imagine himself standing at the opposite end of the square, he could then report details that were now on his imagined right but had been omitted before from a different imagined viewpoint. This pattern demonstrates a clear dissociation between the processes of perception, which for NL were perfectly normal, the contents of his long-term memory, which again was normal, and the operation of his visual working memory which was dramatically impaired. It suggests that WM may not act as a gateway between perception and long-term memory as assumed by the modal model and by many recent introductory textbooks on memory.
Studies of healthy adults are pointing in a similar direction. Research over the last decade has explored the possible constraints on creative thinking or mental discovery. In one form of these experiments, volunteers are asked to imagine a set of shapes such as a triangle, a horizontal line, and a circle. They are then to manipulate and combine these shapes in their mental imagery to form a recognizable object which they then name and draw. There is no correct answer except that all the shapes must be included, and they must retain their original shape (e.g., a circle is not to be used as an oval). One possible response generated by a volunteer in these experiments is shown in the upper part of Fig. 1. This demonstrates that working memory can be used for mental discovery. However, if instead, volunteers are asked to combine the shapes of real objects rather than generic shapes, they have much greater difficulty in doing so. One successful drawing from a volunteer is shown in the lower half of Fig. 1 which is an attempt to combine the shapes of an apple, a banana, and a comb. The problem with object shapes was not their complexity, but the fact that they have a meaning associated with them. This made it difficult, for example, to think of an apple as anything other than an apple, because this requires reinterpretation of the shape. This suggests that the contents of working memory have meaning, and they can only have meaning if the perceived shape is interpreted in the context of stored knowledge. That is, it appears that information from the senses goes through and activates the contents of long-term memory before it becomes available for retention or manipulation in working memory, quite the reverse of the traditional flow of information in the modal model of memory.
A final, rather convincing demonstration that the contents of working memory are interpreted comes from experiments with ambiguous figures. Figure 2 can be interpreted in one of two ways, as a duck or as a rabbit, and if we continue to inspect the figure then our interpretation ‘flips’ between one and the other. However, in some ingenious experiments, Chambers and Reisberg (1985, see Cornoldi et al. 1996) showed this figure for only a brief time to a large number of volunteers who were asked to report from their memory what the figure depicted. All of the volunteers reported it as either a duck or a rabbit, but none of them could then reinterpret their image in memory and see the figure as anything other than their first interpretation. Nevertheless, they could do so when asked to draw their image on paper moments later. In other words, again, the contents of working memory are interpreted and under some circumstances are very difficult to reinterpret unless we can use the processes of perception with a physical drawing. This offers quite convincing evidence that perception feeds into long-term memory which activates relevant knowledge for manipulation with working memory (for discussions see Cornoldi et al. 1996, Helstrup and Logie 1999).
A second assumption of the modal model was that of WM as a single, flexible, limited capacity system that could undertake both temporary storage and some processing of the material being stored. This theme remains in one contemporary strand of WM research that has focused on measures of individual differences in cognitive capacity. In a typical task, healthy adult volunteers are presented with a series of sentences. The requirement is to recall the last word of each sentence after all the sentences have been presented. The number of sentences is gradually increased until the participant can no longer accurately recall all of the final words, and the maximum number of words that can be recalled is taken as a measure of the individual processing and storage capacity. The measure in various forms has been shown to correlate well with other measures of cognitive ability such as reading comprehension or mental arithmetic performance. These correlations are then used to argue that WM is a mental capacity for processing and storage that can be applied to a wide range of both laboratory and everyday cognitive tasks. In this sense, the sentence ‘span’ (or WM capacity) measure has been quite useful, and offers the basis for a standardized test of cognitive ability. Set against its pragmatic value, the underlying theoretical assumptions are that there is a single flexible cognitive system supporting performance. One possible alternative view is that the task taps several aspects of cognition, and that this same range of cognitive functions is employed for other complex cognitive tasks such as reading comprehension. A further limitation is that the test provides only a measure of memory in the context of processing (retention of the final words while reading). In a recent test of the above assumptions, a separate measure was taken of processing capacity by asking volunteers to verify each of the sentences for their semantic content without a memory requirement. Memory ability was measured independently by asking volunteers to recall lists of individual words. Participants were then asked to combine verification with recall of all of the final words, and the task demands were set at the maximum that had been determined for each task component separately. If WM were a single flexible, limited capacity system, we might expect performance to be very poor when these two demanding task elements were combined. Results showed that despite the high processing demand and high memory demand, performance on both elements of the task remained very close to those obtained when the subjects were performing the tasks independently.
These results point to the notion that temporary memory might be served by different aspects of the cognitive system from those used to support on-line processing. This argument in turn suggests that WM might be best thought of as the orchestrated operation of multiple cognitive instruments, each one with a specific role for temporary memory or for processing.
5. Working Memory As A Multiple Component Workspace
The most influential model of multicomponent WM was proposed in 1974 by two British scientists, Alan Baddeley and Graham Hitch. The model included some elements of the modal model by incorporating processing and memory, however, these were viewed as distinct. Three components were originally proposed to provide, respectively, temporary verbal storage, temporary visuospatial storage, and a coordinating or ‘executive’ function. The verbal storage component was originally named the articulatory loop, although subsequently it has been referred to as the phonological loop, and to consist of two components, a passive phonologically based temporary store and a mental verbal rehearsal process. The visuospatial component was originally referred to as the visuospatial scratch pad, but has subsequently been thought to comprise a specialist temporary memory store (or visual ‘cache’) for retaining the visual appearance of objects and scenes linked to a system (an ‘inner scribe’) for retaining, and possibly rehearsing, sequences of movements and more dynamic properties of scenes. The coordinating mechanism is referred to as the central executive, but in more recent years has been thought of as a collection of executive functions, including the coordination of multiple task performance, the allocation of attention, and the implementation of strategies and heuristics used in problem solving and mental discovery or when learning and retrieval are demanding of attention.
A schematic representation of a recent version of the multiple component model is shown in Fig. 3. The diagram illustrates the various specialist subcomponents in addition to the suggested flow of information from perception through a long-term knowledge base into working memory. In the latter sense the contents of WM comprise some of the products of perception and of subsequent mental processing. WM can then be considered as a multicomponent workspace that stores and manipulates information which has been activated from the long-term store, rather than as a primary route into long-term memory from the environment. Detailed discussion of the evidence for the features of the model can be found in Logie (1995), Miyake and Shah (1999) or Richardson et al. (1996).
This conception of WM fits neatly with the notion that people are strategic in how they use their cognitive functions. For example, recalling a list of words can be accomplished by rehearsing the sounds of the words, thinking about their meaning, organizing them into groups, or imagining the objects or concepts to which they refer. As a result, a phenomenon such as the phonological similarity effect (see above) with visually presented words only appears if the participants rely on mentally rehearsing the sounds of words prior to recall. There is evidence that some participants in these experiments spontaneously use other strategies that might rely on visual or semantic codes, and for them the phonological similarity effect does not appear (e.g. Logie et al. 1996). The same argument holds for other well-established phenomena such as the recency effect. Such strategic use of different aspects of WM would not arise if the material did not activate knowledge stored in long-term memory prior to it being held in WM for the period until recall is required.
6. Applying The Concept Of Working Memory
The idea of a multiple component WM has proved useful in understanding some of the cognitive deficits suffered by individuals with brain damage as well as accounting for important aspects of healthy cognition. Discussed above are some of the insights into cognitive deficits arising from the use of the model, and how studies of patients with such deficits help the understanding of normal cognition. The application of WM theory to some aspects of mental discovery and creative thinking has also been mentioned. There have been several other areas of application, including the use of the phonological loop in explaining important aspects of vocabulary learning and mental arithmetic. For example, one feature of the phonological loop is that it can retain a recently heard word, mentally rehearse the word, and then generate the word as articulate speech. A number of studies have examined the ability of young children to repeat aloud new or nonsense words that they have just heard. Their ability to do so correlates well with their general competence in language and predicts their language ability several years later. Patients with severe verbal short-term memory deficits also appear to have difficulty acquiring new foreign vocabulary. These findings have led to the suggestion that the phonological loop is a cognitive function that has evolved to support the acquisition of language, and that has the emergent property that it can support temporary verbal memory.
Among the acquired skills that appear to depend on the phonological loop are counting and mental arithmetic. Keeping track of a running total in counting relies on subvocal rehearsal of the counting sequence, while keeping track of partial solutions in mental arithmetic again appears to rely on subvocalization of the subtotals involved, unless the solution is already well learned. For example, for most people the answer to 4 + 5 = ? would be available without the need for mental calculation, because it would be recalled as an arithmetic fact. However, the answer to 23 + 19 + 48 + 34 = ? would require a series of operations with temporary memory for the subtotals (held in the phonological loop) as each part of the sum is completed. Evidence for these conclusions arises from studies in which volunteers perform counting or mental arithmetic while suppressing articulation through repeating an irrelevant word. Accuracy in counting and in arithmetic is disrupted by this procedure but is not disrupted by performing other irrelevant tasks such as tapping the table in a regular pattern or even when being presented with irrelevant speech sounds. The phonological loop therefore seems to support accuracy in counting and arithmetic.
The visuospatial resources of WM can support a wide range of other everyday cognitive tasks such as remembering where we are on the page while reading, maintaining and updating a representation of objects in our immediate environment, mental scanning to plan a route around a familiar town, or imagining the configuration of furniture in an empty room.
7. Conclusion
WM is an evolving concept which may be subject to change in the light of new evidence or questions as to its detailed characteristics. Nevertheless, it allows the interpretation of a great deal of evidence and phenomena from both laboratory tasks and from everyday experience.
Bibliography:
- Baddeley A, Della Sala S 1996 Working memory and executive control. Philosophical Transactions of the Royal Society of London B 351: 1397–403
- Beschin N, Cocchini G, Della Sala S, Logie R H 1997 What the eyes perceive, the brain ignores: A case of pure unilateral representational neglect. Cortex 33: 3–26
- Cornoldi C, Logie R H, Brandimonte M A, Kaufmann K, Reisberg D 1996 Stretching the Imagination: Representation and Transformation in Mental Oxford University Press, New York
- Della Sala S, Logie R H 1993 When working memory does not work. The role of working memory in neuropsychology. In: Boller F, Spinnler H (eds.) Handbook of Elsevier Science Publishers BV, Amsterdam, Vol. 8, pp. 1–63
- Helstrup T, Logie R H (eds.) 1999 Working memory and mental discovery. Special issue of the European Journal of Cognitive Psychology. The Psychology Press, Hove, UK
- Just M A, Carpenter P A, Keller TA 1996 The capacity theory of comprehension: New frontiers of evidence and Psychological Review 103: 773–80
- Logie R H 1995 Visuo-Spatial Working Lawrence Erlbaum Associates, Hove, UK
- Logie R H, Della Sala S, Laiacona M, Chalmers P, Wynn V 1996 Group aggregates and individual reliability: The case of verbal short-term memory. Memory and Cognition 24(3): 305–321
- Miyake A, Shah P (eds.) 1999 Models of Working Cambridge University Press, New York
- Richardson J T E, Engle R W, Hasher L, Logie R H, Stoltzfus E R, Zacks R T 1996 Working Memory and Human Cognition. Oxford University Press, New York
- Vallar G, Shallice T (eds.) 1990 Neuropsychological Impairments of Short-term Memory. Cambridge University Press, Cambridge, UK