Construction of group systems of expert advantages using the a-technology of applying classical decision-making criteria

Abstract

Decision making is a more important system-forming characteristic of expert activity. Therefore, studying the specifics of the relevant choices and their optimization, especially from the perspective of the influence of the human factor, is an urgent scientific and practical task. Among the components of this influence, which simultaneously determine the attitude of specialists to the indicators and characteristics of the objects of examination, in particular the features of investment attractiveness, systems of advantages are identified, by which we mean an ordered series of these features: from the most significant, acceptable, weighty, etc. — to less significant. The qualimetry of the significance of the features of the investment attractiveness of objects of examination in the ordering scale is linear, therefore it makes the corresponding measurements “rough” and can even provoke statistical errors of the І-ІІ kind when moving from individual systems of preferences to group ones. The research involved m = 90 specialists who are constantly involved by SSI “UkrISTEI” in conducting various examinations and who, using our methodology, built individual systems of advantages on a spectrum of n = 18 characteristic features of the investment attractiveness of the objects of examination. Using a multi-step technology for identifying and filtering out marginal thoughts, as well as eliminating “survivor bias”, four subgroups were identified from the original sample, mC = 30 people, mH = 12 people, mM = 11 people, mT = 6 people, with consistency group opinions satisfies the range of system-information consistency criteria we introduced at a high level of significance α = 1 %. It is substantiated that the mC subgroup is the basic. A decision matrix has been constructed, where the ranks of investment attractiveness features are defuzzified by the corresponding normalized weight coefficients determined by the method of prioritization. To solve this matrix, classical decision-making criteria (Wald, Savage, Bayes-Laplace, Hurwitz) were applied and group systems of advantages were obtained, characterized by the features of these criteria. A high statistically probable coincidence of the advantages of group systems obtained by this method has been established and ways for further development of α-technology have been outlined.