TRANSFER OF TRAINING IN THE GENERATION OF DIAGNOSTIC HYPOTHESES: THE EFFECT OF LOWERING FIDELITY OF SIMULATION

Abstract

Summary. Practice on simulations of diagnostic problem solving can provide the knowledge base required for generating hypotheses for new problems. However, this kind of learning may be inhibited if the simulation requires the learner to practise the convergent type of hypothesis testing typical of most diagnostic situations. This prediction was tested in an experiment in which 40 adults practised on a simulation of a diagnostic task, and were then tested on their ability to generate hypotheses for a new problem. The degree of convergence/divergence of the hypothesis testing in the practice task was varied experimentally. It was found that subjects who practised the more convergent form of hypothesis testing acquired less hypothesis generating ability (P < 0.05) than those who practised the more divergent form of hypothesis testing. These findings imply that, in certain circumstances, transfer performance may be improved by lowering fidelity of simulation in the practice task. This contradicts the widespread acceptance in theories of transfer that maximum similarity of the practice and criterion situations always produces maximum transfer.