Chapter 15 - Graphing Is Believing: Interpretable Graphs for Dual Scaling

Abstract

This chapter focuses on interpretable graphs for dual scaling. The asymmetric mapping of rank order data can be extended to paired comparison data and successive categories data without difficulty. The chapter examines two kinds of interpretable graphs: an interpretable graph for rank order data and an interpretable graph for multiple-choice data. The chapter also discusses a graph that allows the examination of the relations between subjects and chosen options without overlaying two spaces—that is, one for subjects and the other for options. The graph usually provides interpretable clusters in terms of common options that characterize them—that is, the interpretation comes directly from those options. The study suggests that the graphic methods that are presented in the chapter can easily be implemented for practical use and should be preferred to the popular method of joint mapping in principal coordinates.