CATEGORY SCALES: ASSESSMENT OF THE CONTRIBUTIONS OF POSITIONS AND WORD MEANINGS TO PERCEPTUAL SCALE VALUES

In five interconnected experiments, each of seven subjects assessed the perceptual magnitudes of electrocutaneous shocks and three sets of verbal descriptors for those shocks. The verbal descriptor sets were: (a) the integers 1-7; (b) the words faint, weak, mild, moderate, strong, intense, and severe (WordsI), and (c) the words faint, very weak, weak, mild, moderate, strong, and intense (WordsII). Shocks ranged between detection threshold and tolerance. In experiments 1-3, subjects estimated the disparity between shocks and descriptors. Conjoint measurement analysis of those estimates produced independent perceptual scales for shock intensities and descriptors. Each subject used essentially the same shock scale in all three experiments--perceived shock intensity grew roughly with current. That comparison permitted comparison of scales for descriptors. The integers 1-7 had roughly uniform perceptual spacing for all subjects, whereas the words did not. The six words common to the WordsI and WordsII lists had roughly the same spacing in both lists. In Experiment 4, subjects located the verbal descriptors on a horizontal line connecting the phrases "no sensation" and "sensation as intense as you can imagine." In Experiment 5, subjects magnitude-estimated the shock intensities singly. Scales for descriptors from Experiment 4 did not match those found in Experiments 1-3. Average magnitude estimates of single shock intensities grew more rapidly with current than did the perceptual scale values for the shock found in Experiments 1-3. The major conclusions are: (a) subjects can judge differences between stimuli from distinct modalities; (b) intermodal difference judgments, like intramodal difference judgments, produce shallower psychophysical functions than do magnitude estimates; (c) subjects using categorical judgment scales are sensitive to the categories' identities rather than simply to their numerousness and ordering; (d) change in the categories' identities does not alter the scale of sensory stimuli being categorized; (e) subjects instructed to use the integers 1-7 to equisection their perceptual ranges do obey their instructions; and (f) spacings between descriptors on a category scale can remain stable even with change of other descriptors on the list.