Virtually combining the analytical hierarchy process and voting methods in order to make group decisions

Abstract

Multi-criteria decision making (MCDM) is represented by utility and outranking methods. Of the utility models, the analytic hierarchy process (AHP) is probably the most popular in group-decision support. The social choice theory (SCT) with its voting systems can be efficiently combined with MCDM, and AHP in particular, in various group-decision contexts. This paper investigates two possible contexts where modeling group decision-making processes in the field of human---computer interaction (HCI) takes place. Being an important part of universal access in the information society, the first context assumes the use of AHP only, associated with two different aggregating techniques to derive the group decision, with and without a consensus. The second context is based on using only SCT systems, i.e., preferential and non-preferential voting systems. The third, novel, approach that is proposed assumes a full AHP application in the first stage to obtain the weights of alternatives and, in the second stage, an interpretation of the AHP's cardinal information as being the ordinal one and the direct application of the SCT voting systems. Comparative analyses show good agreement for the results when the three methodologies are applied as the decision support to ranking by importance (for a group of 14 PhD students) several widely used sources of information for the internet. The method of virtually combining the AHP and SCT voting systems could be efficiently implemented in real decision-making situations in HCI and related sectors, as well as in cross-sector settings.