Objectives: Decentralizing medical education to community settings has raised issues of instructional quality. The need to evaluate community-based instruction accents the need to adopt a systems perspective, moving beyond factors known to comprise general clinical teaching effectiveness to include factors that focus on instruction as a process. Application of evaluation models using traditional input-output analysis can be flawed. This approach--dubbed the 'black box'--typically examines inputs and outputs, but often ignores throughputs.
Design: In this article we open the black box, using theory to examine the underlying processes that define community-based medical education. We first describe the components and processes of an instructional model that is framed by the philosophy of quality and grounded in experiential learning theory. Without examining the critical processes at work inside the black box - i.e. how students come to acquire clinical knowledge and how behaviours are influenced - it is difficult to assess which programme features contribute to success. Tensions created by the absence of consensus on the outcomes of instruction and the challenge of developing adequate measures are highlighted.
Setting: State University of New York at Buffalo.
Subjects: Clinician-teachers, learners and patients in the environment.
Results and conclusions: We conclude with describing a tool for evaluating community-based instruction that is guided by the context of our model.