ABSTRACT
There are two suppositions for experiments on psychophysical scaling at the global level. First, the human being is considered a measuring instrument whose properties remain the same for all manner of sensory inputs but whose sensitivity depends on the physical attribute being assessed. Second, quantitative judgments are the consequence of this measurement and, as such, can be ordered along psychological scales. Tradition holds that these scales come in four varieties: ratio, interval, ordinal, and nominal (S. S. Stevens, 1946, 1951). The mathematical rules governing the legitimate use of these scales have been well documented by S. S. Stevens (1975), Senders (1958), and others (Baird & Noma, 1978; Gescheider, 1985), and a major aim of the field, at least as an empirical enterprise, is to determine the scale type most appropriate for representing judgments. 1
