Some Confounding Effects of Forced-Choice Scales

than by known characteristics of the scaling devices. Under proper experimental conditions the forced-choice scale has the advantages of easier data processing and ability to force an attitude from respondents who, because of laziness, may not express one. However, forcing an attitude from a respondent who has no knowledge of the attitude object or considers the scale to be irrelevant confounds unawareness and indifference. The experiment described here shows the effect of this confounding on the central tendency and variance of attitude measurements. The ability to predict physical behavior from attitude measures is greatly reduced when the measures confound several mental states. The attitude scored by the respondent is determined by more than his evaluation of the object. Additional determinants of his attitude score are his awareness of the attitude object, his intensity of feeling toward the object, his ability to discriminate between scale items, and the social rewards (or risks) in verbalizing his attitudes. Also, the respondent's action tendency toward the object would seem to be determined by his attitude structure, which includes not only his attitudes but also their relative weights in the decision process. This discussion is limited to the confounding of unawareness and indifference.