Special Session Summary New Approaches to the Role of Similarity in Consumer Research

Shi Zhang, Columbia University
Kathryn A. Fitzgerald, University of California-Los Angeles
[ to cite ]:
Shi Zhang and Kathryn A. Fitzgerald (1997) ,"Special Session Summary New Approaches to the Role of Similarity in Consumer Research", in NA - Advances in Consumer Research Volume 24, eds. Merrie Brucks and Deborah J. MacInnis, Provo, UT : Association for Consumer Research, Pages: 15-16.

Advances in Consumer Research Volume 24, 1997      Pages 15-16



Shi Zhang, Columbia University

Kathryn A. Fitzgerald, University of California-Los Angeles

Similarity is a central component of many cognitive processes and theories of categorization and concept learning, both in psychology and in applied field such as consumer behavior. It is also an elusive concept resistant to comprehensive yet precise definition. Recent advances in theories of similarity have opened new avenues for consumer researchers to examine a range of empirical problems. The objectives of this special session are twofold: (1) to generate an appreciation of the important role that can be played by the new approaches to similarity in consumer behavior research, and (2) to advance recently developed theories of similarity by providing empirical evidence from the field of marketing and consumer research, thus establishing a direction for coordinated future research in the field of consumer behavior.

Consumer research has been influenced by the contrast model (Tversky, 1977), and has utilized it as the basis for constructing theories of brand-extension, consumer learning and product categorization (Block and Johnson 1995, Dhar and Sherman 1996, Loken and Ward 1989, Park et al 1990, among others). This model represents a significant advance over the mental distance model (e.g. Shepard 1974, Rips, Shoben and Smith 1973). Recent theoretical advances have extended the contrast model by identifying and addressing some incompleteness of the model (e.g. Medin, Goldstone and Gentner 1990). In particular, researchers have questioned the assumption that similarity between two items can be represented by lists of features alone and that judgement of similarity simply requires direct feature mapping between the lists (aside from differential weights given to the features). It is proposed that similarity is best characterized as a comparison of structured representations (e.g. Falkenhainer, Forbus and Gentner 1989, Gentner and Markman 1994, Markman and Gentner 1993). The representations are composed of features (i.e. attributes) of objects, objects, and functions, and more importantly, the relations between these representational elements. In other words, similarity is a function of both attributes and the relations of the attributes between a pair of items.

The papers in this session each explored the use of the above views or closely-related views of similarity in different areas of application: product concept learning, consideration set formation, and cross-cultural variation in similarity-directed conceptual organization. Taken together, the papers suggest the inappropriateness of feature-bundle views of products.

Kathryn Fitzgerald presented a paper that uses theory concerning analogical reasoning, the theoretical underpinning of the structured view of similarity, to examine consumers’ learning aboutnovel products and products with complex feature structures. Learning about new offerings stimulates comparison to familiar products, which in turn involves similarity judgment. As the novelty of a new product introduction increases, detecting similarity becomes less routine and must be created. This is accomplished by constructing a mapping that aligns systems of elements. Such a mapping provides a basis for creating inferences about attribute-consequence linkages (likely price changes over time, and other attitudinal and purchase-relevant issues). The paper showed that the use of analogy was distinguishable from a previously studied form of advertising rhetoric, resonance. Both analogy and resonance were found to have a substantial presence in advertising of products in complex and very new categories.

The second paper by Felcher, Malaviya and McGill draws upon recent research that examines the type of features that objects are perceived to share. One notion of similarity is that judgments are influenced by whether the feature structure consists of attributional or relational features. Other research suggests that objects could also share features that are context-independent or context-dependent. This paper builds on these two observations and examines their implications for various choice processes, including the formation of consideration sets. Similarity is construed on both these distinctions, and it yields distinct results in the formation of consideration sets (via categorization). When taxonomic categories are the basis for the consideration set formation, attributional features and context-independent features are invoked, but when goal-derived categories drive consideration of alternatives, relational aspects and context-dependent features become focal. These differences, as shown by the results, influence consumer learning, memory, and generation of additional alternatives when making within and across-category product comparisons. Results also show that subjects who are more familiar with a consumption situation are likely to construct more narrowly defined, within category choice sets, whereas less familiar consumers construct broader, across-category choice sets.

Finally, the paper by Schmitt and Zhang focused on the process of categorization and similarity judgment, and its impact on judgment and choice. It proposes that cognitive systems of different language-based segments of consumers influence their mental representation of objects, recall of the objects, and judgment of products (attitude towards the product). In this approach to studying consumer categorization, similarity is no longer manipulated, but rather is assumed to be part of the cognitive system of groups of consumers (Chinese vs. Japanese vs. English). One of the paper’s foci centers on the automatic aspect of information processing embodied in language classifiers. We conceptualize classifiers as possessing both attributional and conceptual similarities of objects. For example, the classifier "ba" for items that are graspable or have handles (e.g. knife, umbrella, comb, paper fan) can be said to relate items with a perceptual property, a handle, or with a conceptual property, graspability. Classifiers represent a general classification category that takes as members many diverse objects. These objects fall naturally into the traditional characterization of non-comparable. Experiments 1-3 show how the presence of classifiers in Chinese and their absence in English affect the perceived similarity between objects, attribute accessibility and concept organization. Experiment 4 demonstrates the differential effect of classifier structures in Chinese and Japanese on perceived similarity. Experiments 5-7 demonstrate the role that classifiers play in inference-making, judgments and choice. These results not only provide support for the linguistic relativity hypothesis but also suggest new ways of segmenting consumers on the basis of language-influenced cognition.

Joan Meyers-Levy, as synthesizer, provided directions for future studies by raising a series of questions regarding inference-making in a featural context. Some of these questions are: How do people resolve incmpabilities in inferences suggested by difference factors? Do inferences suggested by visual factors dominate those implied by verbal factors (e.g. analogy-driven advertisements)? What factors qualify when these (inference) effects emerge? She also encouraged future research on the role of common features and their associated inferences, the development sequence of sensitivity to such factors in conjunction with difference features.