| Abstract: | Recent factor-analytic studies demonstrating marked concept-rater-scaleinteraction using evaluative semantic differential scales posea problem for opinion researchers who lack the time, money,subjects, or trained personnel to perform a pretest factor analysisas the basis for selecting reliable scales. In this study, 225raters drawn from 10 sub-populations rated 9 concepts on 39evaluative scales and 2 marker scales. When these data weresubjected to 19 different principal factor analyses based onvarious concept-rater combinations, 6 scales loaded above .60on the dominant first factor of all analyses, 9 additional scalesloaded above .50 on the same factor for all analyses, and 7scales failed only once to load above .50. These scales arerecommended for use as a measure of general evaluation undercertain conditions |
