Clinical— experimental analysis of design problem solving

Abstract Two studies of desgin problem solving are'reported. Experiment 1 presents an observational study of an actual client-designer worksession. Analysis of the session transcript reveals a systematically structured interaction. The client and the designer decompose the overall design problem into subproblems, each of which is smaller and somewhat more well structured that the overall problem. Experiment 2 is a laboratory study. The ‘client’ role is simulated by an instruction booklet; subjects play the ‘designer’ role. Again, it is found that subjects spontaneously structure the elements of a design problem into subproblems the nature of which is systematically related to aspects of problem structure. There is high intersubject agreement as to how the decomposition into subproblems should proceed.