The empirical decision analysis procedure'

摘要: Large numbers of primary documents are available in which decision makers explain their choices. Examples of such documents include minutes of governmental meetings, coded telegrams from embassies, reports from various governmental agencies, etc..There are also several problems in analyzing such texts in order to determine the decision makers' argumentation. First of all there is the problem of choosing the concepts one should use. Second, the problem arises whether different readers (coders) of a document using the same concepts obtain the same results (coder reliability). Third, how can one assess whether or not the instrument measures what it is intended to measure? Fourth, given a reliable representation of the argumentation, which hypotheses should be made in order to determine the choice? In a long term study, the authors of this paper have tackled these problems while developing the empirical decision analysis procedure (EDA)(Gallhofer 1978; Gallhofer and Saris 1979 a, b, c; 1982; Gallhofer and Saris 1985). The theoretical framework of the approach is decision theory as it was developed through collaboration between statisticians and experimental psychologists (Edwards 1961; Fishburn 1964; Keeney and Raiffa 1976). Although Edwards was still optimistic in 1961 about the usefulness of the Subjective Expected Utility model (SEU), the so-called rational or analytic choice model, Simon (1957) had already proposed models for bounded rationality. Later studies (eg Tversky (1969), Slovic (1975), Montgomery (1977), Huber (1982), Larichev ao (1980)) confirmed that the choices of human decision makers were not in …