Attitude Measurement and Prediction of Behavior: An Evaluation of Conditions and Measurement Techniques

This paper reanalyzes the relation of attitude measures to other behavior from a methodological point of view and suggests that the degree of correspondence observed is at least a function of the measurement technique used and the extent to which the criterion behavior constitutes action within the range of common experience.

Abstract

The relation of attitude measures to other behavior continues to be problematic. This paper reanalyzes the issue from a methodological point of view. Reconsideration of available evidence suggests that the degree of correspondence observed is at least a function of (1) the measurement technique used, (2) the extent to which the criterion behavior constitutes action within the range of common experience, and (3) the degree to which the criterion behavior represents a repetitive behavioral configuration. Additional data are brought to bear on the issue with the above considerations in mind. Moderate associations are observed between measured attitude and behavior patterns. The magnitudes of these associations are found to vary with the scale construction technique employed.