Combinatorial multi-criteria acceptability analysis: A decision analysis and consensus-building approach for cooperative groups

In group decisions, the decision-makers’ judgements and preferences will inevitably conflict. These discrepancies are caused by differing mental models, and a cooperative, rational group can usually arrive at a correct decision by sharing these mental models and agreeing on the most convincing one. Acceptability analysis was introduced as a decision analysis tool in cases where input data is missing or ambiguous and is therefore represented by random variables. It computes a statistical index that describes each alternative’s performance. We present a new type of acceptability analysis for group decisions called Combinatorial Multicriteria Acceptability Analysis (CMAA). CMAA is not itself a new decision method, but a framework to be used in conjunction with the decision method of choice. It generates combinations of user inputs, resulting in a discrete computational problem. CMAA identifies the issues where unshared mental models prevent consensus and yields detailed insights into the dependence of acceptability scores on individual user inputs. Information entropy is used to identify the discrepancies whose resolution will improve consensus most quickly, which supports an interactive consensus-building function for a facilitator. An example demonstrates the new analytical and consensus-building possibilities offered by the combinatorial approach.