Stochastic Judgments in the AHP: Confidence Interval Construction Using Score Statistics

In multicriteria decision-making methods, such as the Analytic Hierarchy Process (AHP), single
values are used to compare criteria and alternatives. Usually this single value given by decision makers
follows the fundamental scale from 1 to 9. However, a decision maker often does not have complete support
information for his or her making decisions. This lack of information causes the decision maker to become
uncertain about his or her decisions. One of the options to overcome this problem is by using intervals
instead of single-valued pairwise.
This paper presents a methodology for analyzing the interval judgment using confidence intervals, constructed from score statistics. Moreover, inconsistency that can appear in the AHP will be restored using a
consistency improving method (CIM). Data uniformly generated are used for implementing the method. The test showed that this interval judgment approach can be a representative method for covering the uncertainty in the decision-making process.