The use of "process-tracing" techniques for examining individuals' decision strategies has been criticized widely for yielding results that are uninformative and often confounded. The present paper describes a new technique based on a "structural" approach that allows efficient assessment of finely discriminated decision strategies via a HyperCard program called DECOUT. The program presents individuals with an information matrix (e.g., describing five alternative pesticides on three attributes). Individuals are asked to choose the best or the worst alternative in the matrix for 12 problems. Information in the matrix is constructed so that choice of an alternative by an individual is uniquely associated with use of one particular decision strategy. The strategies include information integration, maximin, maximax, pessimism-optimism, and minimax-regret processes. An experiment using the program is described, and possible directions for future research and development are discussed.A long-standing problem for researchers in decision making has been the lack of an efficient and accurate technique for assessing decision strategies. We have developed a new technique whereby minimal data can be used in order to distinguish among specific decision strategies. Our technique revitalizes the seemingly forsaken "structural" approach in decision-making research, insofar as it estimates an individual's decision strategy on the basis of his/her decision outcome. The main feature of our technique is that we can provide individuals with decision problem information in such a way that their preferred option (alternative) in the problem is uniquely associated with one of five specific decision strategies.Our technique has both methodological and theoretical value. In this paper, we focus on the methodological value first. We then describe an example of a theoretical question that can be examined by using the technique. We report an experiment in which the technique was used, and finally, we discuss possible directions for future research and development.
METHODOLOGICAL VALUEOur technique is an alternative to the recently popular "process-tracing" techniques. Process-tracing typically focuses on various parameters of individuals' information search and acquisition strategies, such as the order in which information is retrieved and the duration ofaccess to information (for reviews, see Ford, Schmitt, Schechtman, Hults, & Doherty, 1989;Harte, Westenberg, & van The authors gratefully acknowledge Sylvie Noel for her generous help with programming the experimental stimuli. The DECOUT program will run on Apple Macintosh computers with the application HyperCard; it can be obtained by sending a blank 3.5-in. Apple-formatted disk to M. Finucane,