On the Reliability and Predictive Validity of Purchase Intention Measures #252547

The purpose of the present paper is to report some further analyses and empirical applications of Morrison's model. We begin with a summary of the model, emphasizing its psychometric interpretation and correcting a minor error in one of the original equations. In the second section of the paper, we review the evidence available relating to two threats to the predictive validity of intention measures represented in Morrison's model: random measure ment error or unreliability and systematic error or bias. Some empirical esti mates of the reliability of a seven point intention scale are reported from a small scale test-retest study of new product concepts. An examination of the evidence on bias draws attention to possible differences in the properties of intention scales according to whether they are applied to generic durable goods, a brand of frequently purchased consumer products, or concept test stimuli.