We propose a five year program of investigation into problem- solving in immunohematology, leading to the development of a knowledge-based decision support system and a prototype intelligent tutoring system. Initially, the techniques and methodologies of cognitive psychology and artificial intelligence will be used in tandem to investigate the problem-solving of practitioners in immunohematology at various levels of expertise. During this period the knowledge and problem-solving used by individuals performing various interpretation tasks in immunohematology will be described from an information processing perspective. This description will be performed within the context of the generic task theory of problem-solving, which has proven successful in constructing expert systems in a variety of medical fields. In order to more fully understand the performance of immunohematologists, both correct/expert performance and error-producing behavior will be modeled. The resultant models will then be used as the basis for developing knowledge-based systems to support the education and practice of immunohematology.
