| Summary: | This thesis describes a study of novice problem solving in the domain of medicine. The study attempts to answer questions pertaining to the diagnostic accuracy, the generation and change of diagnostic hypotheses, and the use of clinical findings in the course of solving clinical cases with similar presenting complaints. Two specific issues are addressed: (1) how does the initial case presentation suggesting a common disease schema affect the diagnostic problem solving process of novices and intermediate subjects? (2) what are the processes the subjects used in coordinating hypothesis and evidence during diagnostic problem solving? === Medical trainees (students and a resident) were given four clinical cases to solve and think-aloud protocols were collected. The verbal protocols were analyzed using methods of protocol analysis. The results show that second year medical students interpreted clinical cases in terms of the more common disease schema, regardless of the initial presentation of the case. More advanced students, although unable to make a correct diagnosis in most instances, were less susceptible to such confusions. Only the resident was able to interpret the cases in terms of different disease schemata, reflecting knowledge of the underlying disease process. The semantic analysis of the protocols revealed that most students, especially at lower levels of training, misinterpreted or ignored the evidence that contradicted their initial hypotheses and made use of a mixture of forward and backward reasoning; a finding consistent with previous research. Implications for educational training and for a theory of novice problem solving in medicine are presented.
|
|---|
