A VIEW FROM MEDICINE
Guy J. Groen Vimla L. Patel McGill University
The development of theories of problem solving in educational contexts can be approached from two directions, frequently achieving different results. One approach is to directly examine the teaching of problem solving in practical situations and then to generalize from the task and domain specificity that inevitably arises. Ile second approach is to begin with a general theory and discover how it applies in specific domains. These two approaches tend to yield different results because the level of specificity tends to determine what is easy or hard to examine. Thus, in the first approach, a theory based on arithmetic may focus on quite different issues from a theory based on geometry. This issue is particularly exacerbated in medicine because the domain is organized into a complex hierarchy of specialties and subspecialties which are all connected to certain types of general knowledge that is generally assumed to be required of all practitioners. For medicine, therefore, it is useful to consider the second approach.
There exists in the literature of cognitive psychology a theory of problem solving for which a considerable amount of generality has been claimed. This is the approach developed primarily by Herbert Simon, Alan Newell, and their colleagues at Carnegie Mellon University( Ericsson & Simon, 1984; Greeno & Simon, 1988 Newell & Simon, 1972). Its most recent incarnation, in Newell's SOAR system