LIMITATIONS

The present findings are also subject to several limitations which should be addressed in future research. First, the prototypicality of the competitor was operationalized, and empirically confirmed in our manipulation checks, as the market leader or as an otherwise prominent brand. As a result, the majority of the brands used in the study were high-share, popular brands (for example, Bic and Lipton in Study 1, and Canon and IBM in Study 2), but they were not necessarily the best-quality brands in the extension category. Therefore, the superiority claims of our advertisements may have been plausible arguments in the context of the ads provided, especially since the brands that were extended were of high quality and had accumulated considerable brand equity which could be leveraged in brand extensions. Our findings may not hold true for comparisons with the best-quality brands in the extension category. In those circumstances, a superiority claim of a brand extension may not be plausible and may not be successful in associating the new brand with the successful brand in the extension category.

Second, the comparison claim in the ads was a superiority claim rather than a parity claim. As mentioned earlier, similar effects have been obtained in prior research on comparative advertising for superiority and parity claims (Dröge and Darmon 1987; Gorn and Weinberg 1984; Walker, Swasy, and Rethans 1985); therefore it seemed unnecessary to include both types of claims in our studies. In some cases, (e.g., in the case of a comparison with the best brand in the extension category) a parity claim may be more believable and therefore more desirable than a superiority claim. Future research should test whether our results also hold for parity claims or under what conditions parity and superiority claims are most successful for brand extensions.

Another limitation concerns the operationalization of perceived fit in our studies. Perceived fit--i.e., the similarity between the brand and the extension--may be determined in various ways, aside from the global, overall similarity judgment that was used in the present research. Tauber (1988), after scrutinizing a sample of 276 brand extensions, distinguished several dimensions of fit. Park, Milberg, and Lawson (1991) drew an important distinction between concept and attribute fit and found that the most favorable brand-extension evaluations occurred in cases of high brand concept consistency and high product-feature similarity. Moreover, Herr, Farquhar, and Fazio (1996) have recently demonstrated the need to examine the fit between the parent and the extension categories with greater sophistication. Specifically, the notion of fit consists of not only strength but direction as well--the latter which was not addressed in this study. As the extent brand's dominance in the parent category has been shown to influence cognition and affect towards the extension category, futureresearch should examine the viability of comparative and noncomparative positioning that accounts for different types and directions of fit between the brand and the extension.

Finally, future research should examine more closely the cognitive processes involved in consumers’ judgments of brand extensions in the context of comparative brands. In addition to the outcome measures as in the present studies, cognitive-response measures as well as experimental manipulations should be employed to provide further evidence for the cognitive mechanisms underlying extension-evaluations such as a reduction in the perceived distance between the brand and the extension in the case of comparative advertising.