Ordinal Priority Approach (OPA) in Multiple Attribute Decision-Making.

The current study aims to present a new method called Ordinal Priority Approach (OPA) in Multiple Attribute Decision-Making (MADM). This method can be used in individual or group decision-making (GDM). In the case of GDM, through this method, we first determine the experts and their priorities. The priority of experts may be determined based on their experience and/or knowledge. After prioritization of the experts, the attributes are prioritized by each expert. Meanwhile, each expert ranks the alternatives based on each attribute, and the sub-attributes if any. Ultimately, by solving the presented linear programming model of this method, the weights of the attributes, alternatives, experts, and sub-attributes would be obtained simultaneously. A significant advantage of the proposed method is that it does not make use of pairwise comparison matrix, decision-making matrix (no need for numerical input), normalization methods, averaging methods for aggregating the opinions of experts (in GDM) and linguistic variables. Another advantage of this method is the possibility for experts to only comment on the attributes and alternatives for which they have sufficient knowledge and experience. The validity of the proposed model has been evaluated using several group and individual instances. Finally, the proposed method has been compared with other methods such as AHP, BWM, TOPSIS, VIKOR, PROMETHEE and QUALIFLEX. Based on comparisons among the weights and ranks using Spearman and Pearson correlation coefficients, the proposed method has an applicable performance compared with other methods. • The study proposes a new Ordinal Priority Approach (OPA) to solve Multiple Attribute Decision-Making (MADM) problems. • The proposed method can calculate the weights and the ranks of the experts, alternatives, and attributes simultaneously using simple steps. • The proposed method supports group decision making without using averaging methods. • The proposed method needs a simple comparison of experts, alternatives, and attributes and supports decision making with incomplete input data. [ABSTRACT FROM AUTHOR]