From: Richard M Young To: "Frank.Ritter" Subject: Re: dera report Date: Tue, 6 Jul 1999 15:31:45 +0100 Frank, I really hate doing this. Hate it with a vengeance. How did I get myself into this situation? Below is draft of sec 2.2. Do you have any ideas of references for WM in human performance, or for effects of emotion & stress? Sec on Integration will follow, I'm not sure about anything else. Is this the kind of thing you're expecting? -- Richard ----------------------------------------------------- 2.2 The role of working memory Central to all questions about human cognition and performance is the role of working memory. Working memory is implicated in almost all aspects of cognitive performance (Newell & Simon, 1972; Just & Carpenter, 1992; ** need a more human performance ref, Wickens?*** human perf handbook**? stuff on WM load and performance**?), and it is widely agreed that limitations of working memory are a major determinant of limitations of cognitive performance. Definitions of working memory are varied, but for present purposes we can take it to refer to the mechanisms that maintain and provide access to information created or retrieved during the performance of a task. Modern approaches to the psychological study of human working memory often take as their starting point the famous paper by Miller (1956) arguing that people can retain only around "7 +- 2" items in short-term memory. Later work tends to revise that estimate downwards, towards 3-4 items of unrelated information (Simon, 1974; Crowder, 1976). An influential line of work by Baddeley (1986, 1997) paints a picture of working memory as a dual system for the rehearsal of information, consisting of (1) a phonological loop, of around 2 seconds duration, for the rehearsal of phonological, acoustic or articulatory information, and (2) a visuo-spatial scratchpad, of somewhat indeterminate capacity, to play an analogous role for the maintenance of pictorial and spatial information. Other approaches within experimental psychology place more emphasis on the role of working memory in both storing and manipulating temporary information (Just & Carpenter, 1992; Daneman & Carpenter, 1980). An important recent extension to the notion of working memory comes from the study of expertise, where Ericsson & Kintsch (1995) argue that, after high levels of practice in a particular domain, people can make use of specialised "retrieval structures" in order to use long-term memory for the rapid storage of temporary information ("long-term working memory"). A recent book (Miyake & Shah, 1999) reviews a range of current approaches to the modelling of human working memory, although many of the "models" do not have the explicitness and generality needed to support the simulation of human performance in realistic tasks. Of those that do, their view of working memory varies widely. Some, such as ACT-R (Anderson & Lebiere, 1998) and CAPS (Just & Carpenter, 1992) view working memory not as a separate structural entity, but rather as an activated region of a larger, more general memory system, where the limitations of working memory derive from a limited total quantity of activation. Just & Carpenter (1992, followed by Lovett, Reder & Lebiere, 1999) extend that view to the modelling of individual differences, where different people are assumed to have different maximum quantities of available activation. A number of these ideas are put together by Byrne & Bovair (1997), who model (in CAPS) the way that a class of performance errors, in which people forget to complete subsidiary aspects of a task (such as removing the original from a photocopier), are affected by working memory load. In contrast to these resource-limited models, Soar (Newell, 1990) imposes no structural limitation on working memory. Using Soar, Young & Lewis (1999) explore the possibilities of working memory being constrained not by physical resources, but by functional limitations, and by specific kinds of similarity-based interference. In summary, the current position is that human performance is known to be highly dependent on working memory and working memory load, and to be susceptible to factors such as individual differences (Just & Carpenter, 1992), distraction (Byrne & Bovair, 1997), emotion and stress (*ref* ?), and expertise (Ericsson & Kintsch, 1995). Existing models of human performance (e.g. as reviewed in Pew & Mavor, 19**) are not strong on the modelling of working memory. Models exist (Miyake & Shah, 1999), and some approaches to cognitive modelling (ACT-R, CAPS, Soar) have potential as a basis for predicting human performance in realistic task situations. There remains a need for the investigation and development of more explicit and complete models, with broader scope, of the role of working memory in human performance. REFERENCES Anderson, J. R. & Lebiere, C. (1998) The Atomic Components of Thought. Erlbaum. Baddeley, A. D. (1986) Working Memory. Oxford University Press. Baddeley, A. D. (1997) Human Memory: Theory and Practice. Hove, UK: Psychology Press. Byrne, M. D. & Bovair, S. (1997) A working memory model of a common procedural error. Cognitive Science, 21, 31-61. Crowder, R. G. (1976) Principles of Learning and Memory. Hillsdale, NJ: Lawrence Erlbaum. Daneman, M. & Carpenter, P. A. (1980) Individual differences in working memory and reading. Journal of Verbal Learning and Verbal Behavior, 19, 450-466. Ericsson, K. A. & Kintsch, W. (1995) Long-term working memory. Psychological Review, 102, 211-245. Just, M. A. & Carpenter, P. A. (1992) A capacity theory of comprehension: Individual differences in working memory. Psychological Review, 99, 122-149. Lovett, M. C., Reder, L. M. & Lebiere, C. (1999) Modeling working memory in a unified architecture. In A. Miyake & P. Shah (Eds), Models of Working Memory: Mechanisms of Active Maintenance and Executive Control, 135-182. Cambridge University Press. Miyake, A. & Shah, P. (Eds) (1999) Models of Working Memory: Mechanisms of Active Maintenance and Executive Control. Cambridge University Press. Miller, G. A. (1956) The magic number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63, 81-97. Newell, A. (1990) Unified Theories of Cognition. Harvard University Press. Newell, A. & Simon, H. A. (1972) Human Problem Solving. Englewood Cliffs, NJ: Prentice Hall. Simon, H. A. (1974) How big is a chunk? Science, 183, 482-488. Young, R. M. & Lewis, R. L. (1999) The Soar cognitive architecture and human working memory. In A. Miyake & P. Shah (Eds), Models of Working Memory: Mechanisms of Active Maintenance and Executive Control, 224-256. Cambridge University Press. ** Wickens ? ** human perf handbook ** ? ** emotion & stress on WM ** ?