STUDY 2: ATTRIBUTE TYPICALITY AND BRAND EXTENSIONS

The attribute-selection decision concerns the choice of one or several product attributes that should be highlighted to consumers. One important issue in this respect is whether the selection of certain types of featured attributes may enhance or hinder the effects induced by comparative and noncomparative advertising formats. For example, could the selection of certain attributes alter the cross-over interaction effect that we observed under high and low fit conditions for the noncomparative format and the comparative ad format with a prototypical competitor?

Based on research by Rosch and Mervis (1975) and Loken and Ward (1987), attributes may be viewed along a continuum of typicality, and for experimental purposes, be classified as either typical or atypical. In Study 1, the featured attributes were all typical attributes; therefore, we could not address the role of how different types of attributes may affect brand evaluations in a comparative or noncomparative context.

How could attribute typicality affect brand evaluations in a comparative and noncomparative context? In a study by Pechmann and Ratneshwar (1991), a comparative ad format with atypical product attributes did not result in any stronger association with the comparison brand, whereas, a comparative ad format with typical attributes did. Pechmann and Ratneshwar explained this finding with reference to the construct of incongruency with the existing product category schema. Since an atypical attribute is typically not associated with the comparison brands, the incongruency will induce consumers to rely more on the attribute information rather than on simple heuristics (e.g., "If the advertiser compared these brands, they must be comparable").

Indeed, in line with this interpretation, schema-incongruent objects have been found to evoke greater cognitive effort than congruent ones (Mandler 1982; Meyers-Levy and Tybout 1989; Fiske and Pavelchak 1986). The underlying theory is that stimuli which match the product category schema will be processed heuristically in line with the existing schema. When there is a mismatch with the existing category schema, a more effortful cognitive processing called "piecemeal processing" takes place in an effort to resolve the incongruity (Meyers-Levy and Tybout 1989).

In the context of brand extensions, when shown ads of brand extensions with typical attributes, the presence of typical attributes is consistent with consumers’ expectations. As a result, consumers should engage in heuristic processing. Therefore, for typical attributes, we should be able to replicate the interaction of Study 1, namely that: H3: Under conditions of high (low) perceived fit, brand extensions with typical attributes will be perceived more positively if presented in a noncomparative (comparative) format than a comparative (noncomparative) format.

When presented with atypical attributes, on the other hand, consumers should engage in piecemeal processing. In noncomparative advertising, there is a discrepancy between the atypical attributes that the new product has and consumers’ normal expectations concerning the features of the product category. Moreover, consumers lack a cue that would facilitate a justification (e.g., "the company wants to provide a differential advantage in the category") via piecemeal processing of the atypical attribute and its context. In contrast, in the case of a comparative ad with a prototypical competitor, consumers may infer that the company wants to focus on the differentiating features of the brand in comparison to its competitor and they are provided with an explicit cue (such as a competitor in the ad) to trigger such an inference. As a result, we expect that consumers evaluate a brand with atypical attributes more positively in a comparative than a noncomparative context. The same argument should apply to both high fit and low fit situations. Therefore, we predict: H4: Under conditions of both high and low perceived fit, brand extensions with atypical attributes will be perceived more positively if presented in a comparative format than a noncomparative format.