Análisis del método ordinal de decisión multicriterio del Veto/Analysis of the multi-criteria decision-making Veto method

There is no perfect ordinal decision-making method. This means, one ordinal method cannot meet the five Arrow Axioms (independence of irrelevant alternatives, transitivity, universality, Pareto unanimity and total order). For this reason, researchers have been focusing their efforts on finding metho...

Full description

Saved in:
Bibliographic Details
Published in:Ingeniare : Revista Chilena de Ingenieria 2015-10, Vol.23 (4), p.556
Main Authors: Bou, Ángel Santiago Fernández, deMello, João Carlos Soares
Format: Article
Language:Portuguese
Subjects:
Alternatives
Decision analysis
Decision making
Groups
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:There is no perfect ordinal decision-making method. This means, one ordinal method cannot meet the five Arrow Axioms (independence of irrelevant alternatives, transitivity, universality, Pareto unanimity and total order). For this reason, researchers have been focusing their efforts on finding methods that reduce the influence of irrelevant alternatives without incurring intransitivity. In 2012, Morais and Teixeira published at Omega journal the Veto method, which is an ordinal multi-criteria decision method. It consists in removing the intermediate quartiles of the alternatives matrix after it is ranked by the decision makers in order to calculate the chosen alternative; thus, the upper quartile is used to calculate the strength of each alternative and the lower quartile corresponds to the weakness. The alternative chosen is the one with the greatest value of strength minus weakness. This article analyzes the Veto method for the dependence of irrelevant alternatives. To achieve this, the Copeland method was used as reference, reinforced with other methods and examples. The result was that the Veto method is more sensitive to irrelevant alternatives than the Copeland method. The veto of the intermediate quartiles, postulated to reduce the dependence on the irrelevant alternatives, does not reduce it and it can remove essential information to get a fair ranking. Because of its structure, the Veto method allows the manipulation of the alternatives to influence the result, making possible for the one who selects the available alternatives to build a decision. Therefore, the Copeland method must be preferred to Veto method.
ISSN:0718-3291
0718-3305