| Exemplar Similarity and the Development of Automaticity |
[Nov. 19th, 2007|01:29 pm] |
One of the foundations for the work on expertise comes from the studies of chess masters by de Groot(1965) and Chase and Simon (1973). These seminal studies suggested that oni important key to achieving chess mastery seemed to lie in improved perceptual processing of the layout of chess pieces, rather than more rapid evaluation of legal chess moves. This perceptual skill results from years of practice. Whereas novices seem to rely on slow, conscious, deductive reasoning, experts seem to rely on fast, relatively unconscious processing - the chess master "sees" the right moves. Chase and Simon explained chess mastery in terms of the size of perceptual structures ( or chunks) that experts use relative to novices. Exemplar Similarity and the Development of Automaticity. Thomas J. Palmeri. Journal of Experimental Psychology: Learning, Memory, and Cognition
Volume 23, Issue 2, March 1997, Pages 324-354 |
|
|
| Comments: |
In the present context, these chunks could have their basis within a rich exemplar memory system. Furthermore, Chase and Simon ( see also de Groot, 1965) foudn that when confronted with random patterns of chess pieces, chess masters' performanc eon memory and perception tasks using such configurations was reduced to that of novices (see, however, Gobet & Simon, 196).
Recent research on the development of expertise in medical diagnosis has also begun to recognize shifts from highly analytic, conscious reasoning process to nonanalytic, perceptual, and memory-based processes ( e.g. Groen & Patel, 1988). For example, in a study using procedures similar to those of Chase and Simon ( 1973), Coughlin and Patel (1987) gave normal cases and randomly ordered cases to both physicians and medical students. Mirroring the results with chess masters, memory for the randomly presented case studies was as bad for the physicians as it was for the medical students. Physicians and medical students formulate information from clinical cases differently - the physicians recognize pattern of familiar problems but the students generally not.
The present research is consistent with some emerging notions of the development of perceptual expertise and automaticity in classification. In a variety of domains, such as diagnostic radiology, biological taxonomy, or paleontology, experts spend many years learning to classify objects on the basis of subtle perceptual cues. Novices in these domains are often presented with detailed sets of analytic rules that can be used to classify objects. These rules are often distilled from experts asked to formalize their thought processes - the implicit assumption has been that experts develop and use intricate sets of classficiation rules. Expert performance in these domains, however, may be more than merely committing more and more rules to memory. Rather, certain aspects of expertise may reflect the vast memory for examples the expert has experienced.
Experts clearly have ready access to this structured knowledge when called upon to justify a given diagnosis or when asked to pass knowledge along to physicians in training. The main issue here is that the initial ( automatic?) diagnosis may reflect the use of nonanalytic, seeming unconscious, presumably memory-based processes. Just as the chess master may "see" the right move, the expert medical practitioner may "see" the right diagnosis. Current research suggests that the role of such highly developed perceptual classification skills has been underappreciated. | |
![[info]](http://lj.rossia.org/img/imported-profile.gif){kind=small}