ABSTRACT
In Stevens’ (1951) familiar classification of measurement scales, above nominal scales the order goes: ordinal, interval, and ratio. We treat these scale types in this chapter under the collective rubric “ quantitative,” and note that they share the minimum property that the scale numbers assigned to the objects order them with regard to the measured attribute. Ordinal (rank-order) scales have only this property, interval scales add the property of equal units (intervals), and ratio scales have both equal intervals and equal ratios, hence a true scale zero (Section 1.1.2). Despite these differences in the amount of information they yield, we find it convenient to treat them together. This chapter presents several alternative methods of representing (coding) quantitative scales, and although the choice among these methods is not completely unrelated to the level of scaling, the relationship is neither strong nor simple: Most methods may be used for all three types of quantitative scales, and some are more appropriate or useful for one level of scaling than another. We will not focus on these relationships in the course of presenting the alternative methods (Sections 6.2-6.5), but view them from the perspective of scale type afterwards (Section 6.6) and conclude with a summary (Section 6.7).
