Abstract

Commonly used categorically or nominally measured data often portray each category as separate from other categories with no inherent ordering. Contrary to this portrayal and the perception that categorical data offer limited analytical capabilities, we examine the philosophical, cognitive, and practical basis for re-thinking categorical measurements in terms of a second conceptual space for mapping out categorical semantics. The article presents a perspective that an inherent ordering of categories can be described through this space that provides a transformation into a numerical measurement domain. The article provides several examples of how geographic information analysis can benefit from adding this second conceptual space to the geographic data, together with a discussion of potential applications.

DOI: 10.1111/j.1749-8198.2007.00026.x