Constellation graphs and the role of rank position criteria value frontiers in PROMETHEE analysis

PROMETHEE is an outranking method to identify a rank ordering of alternatives based on their values from a number of different criteria. The uncertainty (sensitivity) in the established rank ordering is an ongoing issue. One recent approach identifies the minimum changes required in an alternative's criteria values to equate or reverse the order of its PROMETHEE based net flow value to that of another alternative. In this study, the understanding of a frontier of the minimum changes in the domain of the criteria values is further exposited with emphasis on the graphical presentation of results. Moreover, the first use of the constellation graph method of data representation with PROMETHEE is introduced here. It offers a clear visualisation of the contribution and marginal changes of the criteria values, along with the incumbent net flow values. Throughout the analysis, the hydroelectric power station data set, first used to exposit PROMETHEE is investigated.