Decision-making with the AHP: Why is the principal eigenvector necessary

In this paper it is shown that the principal eigenvector is a necessary representation of the priorities derived from a positive reciprocal pairwise comparison judgment matrix A=(aij) when A is a small perturbation of a consistent matrix. When providing numerical judgments, an individual attempts to estimate sequentially an underlying ratio scale and its equivalent consistent matrix of ratios. Near consistent matrices are essential because when dealing with intangibles, human judgment is of necessity inconsistent, and if with new information one is able to improve inconsistency to near consistency, then that could improve the validity of the priorities of a decision. In addition, judgment is much more sensitive and responsive to large rather than to small perturbations, and hence once near consistency is attained, it becomes uncertain which coefficients should be perturbed by small amounts to transform a near consistent matrix to a consistent one. If such perturbations were forced, they could be arbitrary and thus distort the validity of the derived priority vector in representing the underlying decision.

[1]  L. Nachbin.  Topology and order , 1965 .

[2]  Thomas L. Saaty,et al.  Decision making with dependence and feedback : the analytic network process : the organization and prioritization of complexity , 1996 .

[3]  David W. Lewis,et al.  Matrix theory , 1991 .

[4]  J. A. Gómez-Limón,et al.  Identification of Public Objectives Related to Agricultural Sector Support , 2004 .

[5]  J. Doob.  Stochastic processes , 1953 .

[6]  T. L. Saaty,et al.  Decision making with dependence and feedback , 2001 .

[7]  Thomas L. Saaty,et al.  On the relativity of relative measures - accommodating both rank preservation and rank reversals in the AHP , 2000, Eur. J. Oper. Res..

[8]  P. Harker.  Derivatives of the Perron root of a positive reciprocal matrix: With application to the analytic hierarchy process , 1987 .

[9]  Howard Raiffa,et al.  Games And Decisions , 1958 .

[10]  T. Saaty,et al.  Ranking by Eigenvector Versus Other Methods in the Analytic Hierarchy Process , 1998 .

[11]  Charles R. Johnson,et al.  Matrix analysis , 1985, Statistical Inference for Engineers and Data Scientists.

[12]  Charles R. Johnson.  The Theory of Matrices. Second Edition (with Applications) (Peter Lancaster and Miron Tismenetsky) , 1987 .

[13]  F. R. Gantmakher.  The Theory of Matrices , 1984 .

[14]  W. Greub.  Linear Algebra , 1981 .

[15]  Peter Lancaster,et al.  The theory of matrices , 1969 .

[16]  Luis G. Vargas,et al.  Inconsistency and rank preservation , 1984 .

[17]  Luis G. Vargas.  Analysis of sensitivity of reciprocal matrices , 1983 .

[18]  D. DeTurck.  THE APPROACH TO CONSISTENCY IN THE ANALYTIC HIERARCHY PROCESS , 1987 .