New results on inconsistency indices and their relationship with the quality of priority vector estimation

摘要: New results on inconsistency indices (ICIs) of pairwise comparison matrices (PCM).A first study the relationship between ICIs and priorities estimation (PE) errors.Results allowing for a more profound interpretation well-known ICIs.An introduction new ICI that manifests high correlation with PE errors.A proposal PCM acceptance approach based classical statistical methodology. The article is devoted to problem in comparisons prioritization issue "inconsistency" this context has gained much attention recent years. literature provides us number different suggested measuring matrix (PCM). latter understood as deviation from consistent case - notion formally defined theory. However usage justified only by some heuristics. It still unclear what they really "measure". What even important not known their values "consistency" decision maker's judgments one hand, upon other.In paper we argue it necessary distinguish three following tasks: "measuring" "PCM inconsistency", PCM-based consistency and, finally, usefulness source information priority vector (PV). We present examples showing improving may lead poorer PV results, such situation occur quite naturally. Next focus third above tasks, which very multi-criteria making. For time literature, help Monte Carlo simulations, analyze performance most common indicators final estimates quality. consider two types errors examine distributions well values. presented here allow characteristics. Moreover, analysis, also introduce index. In other ones, index significantly higher errors. This fact enables propose novel supported