An Application of the Analytic Hierarchy Process: A Rank-Ordering of Computer Interfaces

Abstract

This paper presents the analytic hierarchy process (AHP) as a methodology for developing ratio scales from paired comparison data. The AHP offers several advantages over traditional psychophysical approaches for generating measurement scales. One advantage is its ability to readily quantify consistency in human judgments. Another is the ability of the AHP to provide useful empirical results in the event of a small sample of subjects and when the likelihood of obtaining meaningful statistical results may be restricted. Finally, the methodology requires no statistical assumptions regarding the distribution of human judgments. This paper demonstrates that the AHP can be applied to types of subjective data frequently acquired during human factors experimentation. In an empirical scenario, subjects performed paired comparison judgments on a set of five computer interfaces designed for an automated part recognition system. The objective was to rank these interfaces on the basis of users' perceptions regarding two interface characteristics: usability and learnability. Results of the analysis provide ratio scales for evaluating both interface usability and learnability.