A multiple criteria decision model with ordinal preference data

This paper develops a model for aggregating ordinal preference rankings across multiple criteria. Criteria weights are developed as part of the optimization process. The model utilizes three levels of ordinal priorities: (1) preferences among alternatives on the basis of each criterion; (2) preferences among the criteria to reflect the importance of those criteria; and (3) preferences among the criteria which express the degree to which the respondent can clearly distinguish among the alternatives. The model is based upon a Data Envelopment Analysis structure which permits a flexible weighting mechanism for each alternative being ranked.