ABX-LEX: An Argument-Driven Approach for the Digital Facilitation of Efficient Group Decisions

Abstract

The quality of a group decision depends on its members sharing and adopting their various perspectives. Multi-criteria models provide an appropriate vehicle for managing this information sharing because they break down the decision into small, single-issue questions that can be treated independently. We present a new lexicographic decision method with three equivalence classes for use by a group. Our algorithm is the first to combine the hidden profile and multi-criteria paradigms and thereby improving the consensus by identifying critical disagreements and inviting the decision makers to share their mental models of them. A computer simulation demonstrates the algorithm’s effectiveness and efficiency. Unless the problem parameters are very unfavorable, the method achieves the desired choice set or a close approximation to it significantly faster than the commonly used compensatory methods. Our algorithm can either provide step-by-step instructions to a human facilitator, or it can be implemented as a software agent acting as a fully automated digital facilitator, allowing decision makers to participate “on the go” using their internet-connected mobile devices.