ABSTRACT
Public opinion surveys often ask respondents to place themselves and political parties, candidates, and public figures on issue scales. Issue scales are often combined into Likert scales and Guttman scales. This chapter covers several widely used methods for the analysis of issue scales: Aldrich-McKelvey scaling, Blackbox scaling, and Blackbox transpose scaling. These methods are included in the basicspace package. The chapter demonstrates the Ordered Optimal Classification method for the analysis of issue scale data and how anchoring vignettes can be used to facilitate the cross-comparability of survey responses. The basicspace package includes several functions for the recovery of latent dimensions of choice and judgment from issue scale data. Basic Space scaling does not explicitly estimate respondent and stimuli locations simultaneously in the manner that Aldrich-McKelvey does, but the locations recovered by scaling one group can be overlaid on the estimated locations of the other group. Ordered Optimal Classification, meanwhile, offers a non-parametric approach to multiple issue scales.
