Methodological Aspects of Measuring Preferences Using the Rank and Thurstone Scale

Abstract

The fundamental problem with the measurement of preferences is that it not only attempts to measure something that is, by its nature, “unmeasurable”, but also hidden from a direct observation. In addition, a person’s current emotional, material and social situation influences the measurement of preferences resulting from the person’s system of values. The paper is a study on the methodology of preference measurement, a comparison and evaluation of two methods of scale construction. Among various techniques we investigate the two methods: Thurstone procedure for finding scale separations developed by Thurstone and the simplest rank method of scaling. This study examines the relative merits of Thurstone and rank techniques of scale construction.

Keywords

• preference ranking
• case v
• thurstone model
• structure similarity test
• input matrix