A Framework to Choose a Discrete Multicriterion Aggregation Procedure

A review of the multicriterion literature reveals the large number of multiple criteria aggregation procedures (MCAPs). These MCAPs are based on different hypotheses, premises and epistemological perspectives, and deals with different decision-making situations. It is obvious that none of these MCAPs is perfect nor applicable for all decision-making situations. The challenge is then to determine which one is more appropriate to a a specific decision-making situation. This paper assesses a framework for choice engineering of an appropriate MCAP to a decision-making situation. This framework is based on the representation of the decision-making process proposed by Guitouni (1998). Then, we define a set of inputs and a set of outputs in order to characterize different decisionmaking situations (more than 1400). The inputs are used to characterize the information admissible by a MCAP. The outputs are used to characterize the information produced by a MCAP. Systematically, it is then possible to associate different MCAPs with each particular couple of (Input, Output). The obtained matrix constitutes the endeavour of the framework proposed. This matrix is still incomplete and imperfect, but gives a good basis for the selection of a MCAP to a specific decision-making situation.