Bias and variability in purchase intention scales

Although purchase intention scales are widely used, relatively little is known about bias and variability in the estimated purchase proportions. Psychometric techniques have been developed to correct for such problems, and analytical approaches have shown that most predictive errors can be explained as probabilistic variability. However, there is a lack of systematic empirical work in the area. We address this problem using two meta-analyses of published work. Our results show that purchase intention scales are empirically unbiased. Furthermore, the variability is much less than previously assumed. This finding improves the confidence researchers can have in the use of such scales. Interestingly, purchase probability scales performed even better than purchase intention scales. The greater precision of probability scales suggests that they may be more useful both as direct measures of likely behavior and as dependent variables in consumer behavior research.