On the Consistency of Expert Assessments in Formation of Group Decision (on the Example of Priorities Assessment of University Competitiveness Factors)

Abstract

The relevance of the study is determined by the assessment peculiarities of poorly formalized factors of various aspects of economic activity, including formation of integral competitiveness indicators of goods and services. It is proposed to use T.Saati's hierarchy analysis method to form the integral indicator, which makes it possible to carry out group assessment of competitiveness factors priorities. On the empirical basis of expert survey of leading university professors, low sensitivity of expert participation vector in group assessment of competitiveness factors priorities to the consistency of paired comparisons in the hierarchy analysis method was shown. It is advisable to evaluate the limit values of group assessment vectors and competence coefficients using the matrix of individual expert assessments and matrix of expert competence weights respectively in the software environment of the Mathcad mathematical package. It is shown that comparison of Euclidean distance between experts calculated with hierarchical cluster analysis methods is an effective method for analyzing the proximity of individual expert judgments to group assessment. This makes it possible not only to determine the proximity of individual expert judgments to group assessment, but also to identify their homogeneous subgroups, equidistant from the group assessment in multidimensional space of university competitiveness indicators. Thus, the specifics of expert assessment formation are concretized in the light of integral competitiveness indicator.