Examines the use of celebrity endorsement in advertising. Reviews the results of a recent study looking at the effect of a celebrity′s attractiveness, trustworthiness and expertise on product purchase intentions, and of one examining the relevance of physical attractiveness and other symbolic attributes of the endorser in relation to product meaning. Considers implications for marketing managers and concludes that further research is necessary.

Considers the results of recent studies of celebrity endorsers in advertisements. Analyses the results of a particular study evaluating three product categories: bath towels, blue jeans, and VCRs; two endorsers: Madonna and Christie Brinkley; and the product when advertised by each celebrity. Concludes that the endorser tends to pass their own image onto the product, particularly if the product has an undefined image, so it is important that the endorser′s image is consistent with the desired product image.

The degree of correspondence that actually obtains between intentions and behavior remains controversial. Two opposing views can be distinguished: the cognitive perspective associated with Fishbein and Ajzen (1975) holds that intentions will be the best predictor of consumer spending, while the behaviorist perspective associated with Foxall (1983) rejects the value of intention measures. This study compares buying plans and the prior behavior of consumers as alternative explanations for subsequent action. The results support a middle position (Triandis 1977); neither intentions alone,nor prior behavior alone, afford an optimal prediction of behavior. Market researchers are advised to question consumers about both their cognitions and their past actions.

AbstractDiscontinuous innovations pose special challenges to attitude measurement. The authors propose a solution which uses Rokeach's (1968) value terms to update Rosenberg's (1956) formulation. This values‐based approach solves problems associated with more conventional procedures based on beliefs or adjective checklists. A measure constructed according to this model was included in a mail questionnaire completed by 950 owners of personal computers. Analyses supportive of the validity of the values‐based attitude measure are reported, and an example is given of the practical utility of the technique. The values‐based approach is recommended when a general measure of attitude toward a highly involving product category is desired.