Absolute and relative measurement with the AHP. The most livable cities in the United States

Thomas Saaty
Joseph M. Katz Graduate School of Business
University of Pittsburgh
United States

Publication date: Jan, 1986

Journal: Socio-Economic Planning Sciences
Vol.: 20- Issue: 6- Pages: 327-331

Abstract: In this paper it is shown that there are two types of measurement involved in the AHP, absolute and relative. The first requires a standard with which to compare elements, but mostly alternatives at the bottom of the hierarchy. The process leads to absolute preservation in the rank of the alternatives no matter how many are introduced. The second is based on paired comparisons among the elements of a set with respect to a common attribute. This process is essential for comparing intangible attributes for which there are no agreed upon measures. At the level of alternatives new elements (i.e. alternatives) do introduce new information generated by the changing number in the set and by their measurement which essentially rescales the criteria and hence can lead to reversals of previous rank orders. Absolute measurement is used on standardized problems whereas relative measurement is used in new learning situations. Absolute measurement is applied to the Rand McNally study that ranks cities according to livability in the United States.

Keywords: Analytic Hierarchy Process, AHP, Absolute measurement, Measurement