A comparative analysis between two statistical deviation–based consensus measures in Group Decision Making problems

Abstract

The mean absolute deviation and the standard deviation, two statistical measures commonly used in quantifying variability, may become an interesting tool when defining consensus measures. Two consensus indexes which obtain the level of consensus in some problems of Group Decision Making are introduced in this paper by expanding the aforementioned statistical concepts. A comparative analysis reveals that the levels of consensus derived from these indexes are close to those obtained employing distance functions when a fuzzy preference relations frame is considered, so they turn out to be a useful tool in this context. In addition, these indexes are different from each other and with the distance functions considered. Thus, they are applicable tools in the calculation of consensus in our context and are different from those commonly used.