Prescrição ou descrição da decisão humana: uma discussão dos modelos subjacentes aos SADS

Abstract

The decision process is an important aspect in the lives of individuals and, because of the complexity that it takes in certain situations, the application of multi-criteria methods have been presented as aid in making decision (GOMES, and CARIGNANO ARAYA, 2004). Thus, it arises the concern in relation to the way of decisions are made or how they should be taken in the theory (Dillon, 1998), which motivated us to conduct this study, i.e, to check if a Multi-criteria Decision Support system fits the process of human decision. For this, we took as basis the fundamental elements of AHP - Analytic Hierarchy Process (Saaty, 1981), which allowed up to five categories of analysis in this study, namely: OEC - (Order of Choice Criteria), OCA (Order Comparison of Alternatives), CPP (pairwise comparison), VPA (Provision of a weight vector for each attribute) and HAA (the hierarchy of attributes for evaluation). Accordingly, the experiment was conducted in the laboratory, so that the individual perform the task of choosing notebook independently, with the aid of a Multicriteria Decision Support system AHP MAKH-ER system and without the aid of the DSS, i.e, only the individual's mind to guide the decision on which it was used a protocol verbal think aloud. The presentation of the results showed significant differences for four from five categories of analysis under the AHP. Accordingly, through hypothesis testing, the only category that had no significant difference in both tasks was the order of choice criteria (OEC). Regarding to the other categories, OCA, CPP, VPA and HAA, these significant differences in both tasks, and it can be seen that the makers did not follow the prescriptive model of the system. The discussion of the results reveals important peculiarities in the process of decision making, since the decision-making process within a normative perspective, using a DSS is similar only to the Order of choice criteria in relation to human decision making. Regarding the other categories of analysis, it was not found statistical evidence to claim that the process is similar, although the final decision has been the same, as confirmed by the principle of invariance of the theory of rational decision (Tversky and Kahneman, 1986). For this reason, the hypothesis of this study cannot be confirmed, ie, it is not possible to say, in light of scientific observation that a Multi-criteria Decision Support system fits the human decision making. With regard to the conduct of the decider on options for finding information, it can be said that the process is similar in both tasks, since both the tasks with the help of DSS as in without the help of DSS, the decision maker sought the information criterion, which verifies the value of the three alternatives for a certain criteria before making the weights to the attribute in question.