Subjective Estimate Approach to Scaling Attitude Statements 1

Abstract

Statements expressing attitudes toward labor unions were scaled by two groups of judges. One group of 54 5s used a subjective estimate procedure which presented stimuli in pairs and S expressed his judgment by regulating his rate of tapping a telegraph key. Another group of 53 Si scaled the same set of stimuli using a conventional paired-comparisons procedure. Approximately 13 wk. later, 21 of the original Ss in the subjective estimate group were again given two judgment trials. A strong linear relation obtained between scales obtained from the two methods, and internal consistency indices from the subjective estimate procedure compared favorably with a similar index from the paired comparison scale. High test-retest reliability and high inter-judge agreement was reported for the subjective estimate data. Individual scales appeared to be identical within a linear transformation.