Determining optimal group weights for consensus creation in AHP for three conflicting stakeholder groups by vector distance minimization

Abstract

Involving three conflicting stakeholder groups in a decision making process and reaching their consensus is a common problem, especially in the case of public service development decisions, in which the public, the operator and the government participate as group decision makers. This paper presents a methodological novelty which is based on vector distance minimization and so is capable of creating an acceptable consensus among three evaluator groups with different interests and motivations, while it is also avoiding the pitfall of overgeneralization. The proof for the convexity of the problem and the solving algorithm are also demonstrated along with the calculation of the error of approximation. Moreover, we show a numerical example to prove that our proposed method is less sensitive to extremity compared to the most common mean-based aggregation techniques. Investigating the development of urban public transport systems, the paper presents three case studies using real preference data gained by an Analytic Hierarchy Process (AHP) survey involving three real conflicting stakeholder groups. Further, we conduct a comparison analysis based on concordance measure among the proposed consensus creation technique and other two well-proven consensus models: the zero-level GAHP and the Interval-AHP. The results indicate that the new method may be applied to several group AHP problems not only in transportation engineering but also in other fields of theoretical and real-world decision making.