Studies on Consistency Threshold of Kansei in Paired Comparison

Human Kansei is very complicated and it is not easy to evaluate it. Methods of paired comparison might be useful for evaluating human Kansei because it is generally said that human Kansei is nonlinear while methods of paired comparison are considered to be able to deal with nonlinear problems. In paired comparison it is important to confirm consistency of answers of each subject because if a subject’s answers are inconsistent, it could not be said that evaluation of the subject’s Kansei is validity. There are many methods of paired comparison which have been proposed and AHP (Analytic Hierarchy Process), which was proposed by Tomas Saaty as a method to help us determine priority of alternatives to select, uses paired comparison for obtaining weights to criteria and values for evaluation of alternatives with respect to criteria. In AHP consistency of a subjects’ answers is judged using Consistency Index referred to as C.I. in short. with respect to criteria. In AHP consistency of a subjects’ answers is judged using Consistency Index referred to as C.I. in short. Threshold of C.I. on consistency of subjects’ answers is generally 0.1 or 0.15 by empirical judgment but there is no theoretical or experimental ground about it. Other than C.I. coefficient of consistency is used for judgment on consistency of a subject’s answers in paired comparison. Values of coefficient of consistency are calculated from the number of circular triad which implies inconsistency of a subject’s answers. Here it is studied how to determine proper threshold of C.I. for judgment on consistency of a subject’s answers in paired comparison.