THE EFFECTS OF MANAGEMENT SCIENCE INTERVENTIONS ON THE SOLVING OF UNSTRUCTURED PROBLEMS (PROBLEM-SOLVING, DECISION ANALYSIS)

This dissertation reports research into the nature of unstructured real world problems, the means by which humans attempt to solve such problems, and the effects which management science interventions can have on such problem solving activity. Both conceptual and theoretical research was conducted. As regards the former, the dissertation includes a detailed conceptual analysis of problem and related concepts; it analyzes the notion of problem structure; and it develops a detailed taxonomy or problem classification scheme. A review of major theories of problem solving leads to a descriptive language proposed as appropriate for this domain. The notion of an intervention is developed, and an interventional assessment is made of the practise of decision analysis. Empirical research involved an experiment in which subjects were asked to make an investment decision, after reading a scenario and selecting information cards from a search board. Treatment subjects received a scenario including an admittedly partial decision tree which did not include the ultimately decisive decisional contingency. Subjects were encouraged to verbalize throughout the experimental sessions, which were conducted one-on-one with almost two hundred student subjects. All verbalizations were recorded and subsequently analyzed. Chi-square tests were used to identify possible interactions among variables. A significant treatment-performance interaction was identified: subjects using the scenario-with-tree were less able to solve the problem than nontreatment subjects. Further analyses identified three sources of this overall effect, involving the differential abilities of treatment vs. nontreatment subjects to select critical information cards and to properly interpret the critical information when accessed. Further qualitative analyses of empirical data informs a model of unstructured problem solving in which such notions as strategy, perspective, understanding, recognition, inference, and judgment are especially important. The research supports the conclusions that management science interventions can have adverse effects on problem solving performance, and that the processes whereby humans solve unstructured problems is more complex, and requires a far richer descriptive language, than prevailing accounts would suggest