On the importance of identifying the correct ‘problem space’

Abstract Recent research on problem solving (Newell and Simon, 1972; Pollard, 1982) and decision making (Einhorn and Hogarth, 1981; Tversky and Kahneman, 1981) emphasized the importance of representation and framing in predicting and explaining subjects' responses. Representation in the present study has been investigated via the concept of ‘problem space’. A set of three experiments is reported in which it is shown that correct identification of the problem space is a major determinant in predicting subjects' responses. It is argued that knowledge of subjects frame and problem space is a necessary requirement to understand their behavior. The complexity involved in using the concepts of ‘problem space’ and ‘frames’ is briefly discussed.