The Role of Context in Consumers' Category Judgments: a Preliminary Investigation

Craig J. Thompson, University of Tennessee
[ to cite ]:
Craig J. Thompson (1989) ,"The Role of Context in Consumers' Category Judgments: a Preliminary Investigation", in NA - Advances in Consumer Research Volume 16, eds. Thomas K. Srull, Provo, UT : Association for Consumer Research, Pages: 542-547.

Advances in Consumer Research Volume 16, 1989      Pages 542-547


Craig J. Thompson, University of Tennessee

It is proposed that the major structural models of categorization have difficulty in accounting for the contextual variability exhibited in consumers' category judgments. A contextually based approach to categorization is offered and described in terms of a figure/ground metaphor. Excerpts from interviews with consumers are used to illustrate possible figure/ground relationships in category judgments. The implications of this contextually based approach and its relations to contemporary categorization research are discussed.


The ability to categorize has been described as the "keel and backbone of our thinking" (James 1890) and as "the most ubiquitous of cognitive activities" (Bruner, Goodnow, & Austin 1956). As cognitive psychologists have become increasingly interested in the structure of human thought, there has been a resurgence in the study of categorization (Barsalou 1987; Jenkins 1977; Kahneman, Slovic, & Tversky 1982; Reed 1982; Rosch 1975; Rosch & Mervis 1975; Smith & Medin 1981; Wattenmaker, Dewey, Murphy, & Medin 1986). Consumer researchers, too, have been exploring the role that categorizing plays in marketing phenomena such as expertise (Alba & Hutchinson 1987; Sujan, Sujan, & Bettman 1988), consumer judgments (Sujan 1985), and brand recall (Alba & Chattopadhyay 1985).

This paper seeks to explore the view that categories are outcomes of judgment processes which occur in a given context. The discussion will be organized into four sections. First, the three major structural approaches to categorization will be briefly described and some proposed limitations of these models will be discussed. Secondly, a more contextually based approach to categorization will be offered. Third, texts derived from interviews with consumers will be used to illustrate how contextually based category judgements might occur. Finally, the implications of this contextual approach and its relations to contemporary research in categorization will be discussed.


Research in categorization has generally employed one of three basic models of category representation: a classical model where category membership is determined by the presence of necessary and sufficient attributes possessed by class members, a probabilistic or prototype model where category membership is a function of family resemblance, and an exemplar model where category membership is determined by resemblance to a specific instantiation which has come to represent the entire class (Cohen & Basu 1987; Smith & Medin 1981). These three models all presume that "categorical structure" is defined by attributes or features possessed by group members. While classical models presume that category boundaries are rigid (an object either is or is not a member of the category and, therefore, all included members are equally representative of that category) and prototype and exemplar models assume that category boundaries are graded (category membership is a continuum where an object can be more or less representative of a category), the three models ascribe to the "Aristotelian position" which treats categories as being self-contained entities defined by some combination of properties or attributes (Verbrugge 1977). The logic behind the decontextualized nature of the Aristotelian models is straightforward and rather compelling. A robin, for instance, is still seen in the category of "bird" whether-it is flying, nesting, being kept in a cage, or an infinite number of other contextual variations. Thus, it would seem appropriate that our conceptual categories should reflect this seemingly inherent stability.

A growing number of theorists, however, are questioning if the influence of context on category judgments can be adequately accounted for by either the classic, prototype, or exemplar views (Barsalou 1987; Jenkins 1977; Lakoff & Johnson 1980; Roth & Shoben 1983; Wattenmaker et al 1986). That is, in conceptualizing categories as relatively stable, preexisting structures, these three models minimize the influence of contextual setting and judgment purpose on categorization. Higgins & Lurie (1987) found evidence that some category judgments involve a comparison process between the object to be categorized and a contextually derived referent. Rather than comparing the to-be-categorized entity to a stored mental representation, subjects made category judgments on the basis of contextual information and, by altering the contextual setting, systematic differences in subject's category judgments occurred. Roth & Shoben (1983) studied the typicality judgments made in comprehending a written text and found that typicality ratings are highly dependent on the textual context. For instance, robin is rated as a typical bird when no contextual information is provided but can be atypical in certain contexts such as, "the hunter shot at the bird flying overhead." Barsalou (1983) has provided evidence that people often spontaneously construct "ad hoc" categories for use in specialized contexts. Finally, Medin, Wattenmaker, and Hampson (1987), found evidence that family resemblance--a core concept of both prototype and exemplar models--is contextually variable and furthermore, is dependent on background knowledge that is brought to the judgment task. The implication being that category judgments involve more than references to self-contained categorical structures. On this point they conclude:

Natural categories (basic level categories) may be structured not so much in terms of numbers of characteristic properties as in terms of the web of relationships in which these properties participate (p. 277).


The perspective taken here is that categorization is a generative process where the "category" is not treated as a pre-existing structure but as one created by individuals in a given contextual situation with a particular judgment intention. On this view, an analogy can be drawn between categorization and perception. The "category" is seen as existing in the same way the "percept" exist: as an entity which is the outcome of a process but does not exist separate from that process. The percept only exist in relation to the act of perceiving and, analogously, the category would exist only in relation to the act of categorizing. Such an analogy does not discount or ignore conceptual stability, i.e. consistently judging an item as being in a certain category, because stability is also a characteristic of the perceptual field (Gibson 1977). What this analogy does suggest is a means by which context can fundamentally affect category "structure": a means which can be metaphorically described in terms of figure/ground relationships. Figure/ground relationships were a primary concern of Gestalt psychologists (Kohler 1947, Koffka 1935). The present usage of figure/ground is not wedded to Gestalt theory per se but is seeking to expand upon a major insight of Gestalt psychology: the organization of the perceptual field is fundamental to what is perceived and the "same" entity put in a different contextual relationship will be perceived differently (Koffka 1935, Kohler 1969). Figure and ground is a description of this contextual variability but was never intended as a "theory" of perception (Kohler 1947, Kohler 1969).

In figure/ground terms, "figure" is the entity which has a perceivable shape and is the focus of the perceptual field while "ground" is that part of the perceptual field which has no perceivable shape, seems to fill gaps in the figure, and is not the focus of attention. Figure and ground are reciprocally related such that altering "ground" invariably changes "figure." (Koffka 1935).

Applying a figure/ground interpretation to categorization issues suggests that the structure of the judgment field (context) affords a range of potential category judgments. Which judgment is made i.e., how objects in the field are grouped, would depend on the particular goals and intentions of the person making the judgment. Rather than context somehow altering a pre-existing category structure, the category would be generated appropriate to the judgment context. The category field could be "seen" in different ways depending on the nature of the judgment setting. Which attributes or properties of the objects are figural i.e. which are salient or stand out for individual making the judgment, would be dependent on the object's contextual setting and its relation to the judgment goal or intention.

Figure/Ground in Consumer's Category Judgments

Anyone who has opened a introductory psychological text to the chapter on perception knows what figure/ground diagrams look like and has available direct perceptual evidence that such relationships occur. The issue for our purposes is to show what figure/ground relationships would look like in the conceptual domain. If the figure/ground metaphor describes at least some of consumer's category judgments, certain characteristics should be exhibited:

1) Discussing one object or event in relation to another. The nature of the comparison would afford certain features or attributes of the to-be-categorized object or event being figural (salient) while obscuring other attributes or features.

2) A change in the comparative context would lead to a different categorical assignment by making figural a different set of object attributes or contextual relations.

3) Reversibility should be exhibited in the category judgment. In the conceptual domain, reversibility would be indicated by an individual, in the same categorical judgment, shifting his/her focus between the to-be-categorized object or event and to those which were previously in the background of the judgment field.

Rather than using an experimental paradigm, which is employed by much of categorization research, an interpretative approach is used to illustrate figure/ground effects in consumer category judgments. This approach is prefaced on the assumption that individuals in order to use language must make implicit category judgments (Bruner et al 1956; Lakoff & Johnson 1980). For instance, a simple statement such, "My favorite detergent is Tide," implicitly involves the judgment that Tide is in the category of "detergents."

The method used in this study is an unstructured interview format where subjects discuss various aspects of their experience and views concerning a topic. The extemporaneous style of the interview allows respondents to give long descriptive passages conducive to linguistic analysis. Such a method provides a conversational setting for gathering data and deals with concepts familiar to consumers.

Our present purpose is not theory testing but illustration. The following excerpts are presented in much the same spirit that the Gestalt psychologists had in presenting their figure/ground diagrams. That is, the Gestaltists sought to demonstrate principles of perceptual organization by having readers experience them. Similarly, excerpts will be presented that seem to demonstrate figure/ground effects in consumer's category judgments and the reader should be able to "see" if such an interpretation is descriptive of the presented category judgments.

Illustrations of Figure/Ground Effects

The following section will present excerpts followed by a discussion of possible figure/ground relationships. Five excerpts will be presented.

"It's a big store and it's fun and it has everything. I don't have to go to this and this store. You know, because it is not as convenient as some of the other stores. Sometimes, we have a grocery store at the top of the hill, and I'll stop there but I never think about going there for my big major weekly. I'll run in there to buy bread or, something they have on sale, or in a pinch to get something for dinner, but as far as going in there spending like a weekly thing, I wouldn't do it because I think it is more economical to go to because they have everything and they have more things on sale."

In this passage, the judgment field is organized in such a way that the one store is "seen" and classified as an economical store and the other as a convenience store. At some points the focus (or figure) of the discourse is "economical" stores with convenience stores serving as ground whereas, at other points in the passage, "economical stores" serve as the ground for "convenience stores." The salient properties of the one category are seen against the ground of the other category. Thus, "economical stores" attribute of being "shopped at weekly" is seen against the ground of convenience stores where the respondent shops on an irregular basis (when in a pinch or to pick up a sale). The attribute of "big store" stands out against the ground of "small store." Being in close proximity i.e., "at the top of the hill" is figural against the ground of being further away. The economical store's "having everything" is seen against the ground of the convenience store, which in this respondent's perception, legitimately offers only bread and occasional sale items.

A second passage from this same interview shows a judgment involving an entirely different set of categories:

"I pretty much clothes shop the same way and it drives my husband nuts. If I see something that I really like, generally, what I do is ask the store to hold it for me and then I go and look at every other store and see if I find anything I like better and then I'll go back and usually buy that first thing. He walks right in and says I need a suit and the guy says how about this one and he says fine. And, he calls that shopping."

In this passage, the respondent is understanding her shopping pattern in terms of what she perceives as the shopping behavior of her husband. The result of this organization seems to be the contrasting categories of shopping behavior versus unshopping behavior. Once again, the judgment intention has led to one organization of the contextual setting and from this organization certain properties become salient to her category judgment. As examples, the salient property of "looking to find a better item" is understood in terms of the ground property, "buying the first acceptable item. Similarly, the property of "relying on the salesperson when making a product choice" is salient or figural against the ground of "seeing something that I really like."

In this next excerpt, a respondent is making category judgments about two stores where she has shopped. (Store name has been changed.)

"I hate Store A. You can never find what you are looking for. You- can never find someone to help you look for what you are looking for. I like going to Store B where there are people there to help me find what I want as soon as walk in the door and they help me find it, I pay for it and I am on my way rather than wandering for twenty minutes and not being able to find anything afterward."

The judgment field has been organized into the categories of hated stores and liked stores. The salient attributes for "liked" stores are being assisted and finding what you want whereas for "hated" stores they are not being assisted and not finding what you want. Once again, the dimensions of the "liked" store are understood against the dimensions of the "hated" store and vice versa. A figure/ground description seems to fit the respondent's mode of discourse as she switches the focus from hated stores to liked stores easily and naturally.

In this particular context, Store A is clearly in the "hated" category. Throughout the interview, Store A often served as the prototypic "hated store." A conventional hypothesis would be that "Store A" is so categorized because it intrinsically possesses the necessary and sufficient attributes of hated stores or because it shares the highest degree of family resemblance with other members of that category. If a figure/ground description of category judgments is apt, then objects would not be categorized on the basis of intrinsically possessed properties or features but, instead, on the objects relation to a particular context. Changing the context, could lead to the object, in this example Store A, being categorized in an entirely different manner. The next excerpt from this same interview seems to illustrate such a change.

"I bought a sweater at Store A and it wasn't that expensive and I wish I hadn't spent that little bit of money on a sweater from Store A. I could have just done without or waited until I had the money on a more quality item....I also tend to want to pinch pennies I guess and so I look to see what clothes they have (Store A). I've bought things in Store A that I've worn a lot and have fit in with my other clothes."

Unlike the previous passage, the category judgment is not focusing on Store A but, instead, on the respondent's clothes shopping behaviors. Store A serves as background to these judgments. The respondent categorizes her clothes shopping into two sets of categories: the first set being buying quality items vs buying items lacking quality and the second being buying clothes when she has money vs buying clothes when she does not have money. These two category systems are similar to the category relationships previously described. As for background role of store A, it is in the class of "stores where inexpensive clothes can be bought" a different category than that of "hated store". In short, in a different contextual relationship, the salient or figural features of an object or event can change, leading to the object or event being categorized differently. The attributes of the object do not seem to be assessed in an isolated, independent fashion but, rather, in terms of their relationship to the broader judgment setting.

This last excerpt from a third interview shows a respondent making category judgments about classes of products.

"I'll talk about generic aspirin. It's great. It is cheap and it works. One of my kids has juvenile arthritis so he takes 10 aspirin a day and the difference in price is amazing and it works as well as Bayer or whatever and it is half the price. Some other generics don't work as well, but take something like aspirin, where aspirin is aspirin and it works, where some things that are generic are not going to be the same quality. For example, the other night, my husband is a pediatrician, and he had to put one of our kids on an antibiotic and he said this is the only one he uses because it doesn't make kids stomachs upset. So different generic type things work in different ways."

This passage reveals a more complicated set of categorical judgments. The judgment field is organized in such a way that the categories seem to be "generic brands that work as well as name brands" and "generics that are inferior to name brands." This organization affords "price" and "quality" as the salient properties of the categories. The salient features of generic products stand out in reference to the properties of Bayer. For the generic aspirin to "work as well," it must have a product to be compared with. For the generic to be "half the price," there must be a branded product that is seen as costing twice as much. Again, the figure is seen in relation to the ground. The same type of "reversibility" found in figure/ground configurations is present in this passage. The respondent begins by focusing on the "generics that work as well as the brand". Then, the focus of the discourse becomes "generics that are inferior to the brand". The passage ends with the respondent commenting on the overall contrast with the statement, "So, different types of generic things work in different ways."


This preliminary investigation is based on the premise that consumer's category judgments are generated in response to contextual demands and judgment intentions. The metaphor of figure/ground is used to describe and illustrate some possible ways that these types of category judgments could occur. In so doing, context is given a more fundamental role in determining how a particular category is structured. Others have suggested that context may influence what categorizing strategy will be used (Cohen Nc Basu 1987). Thus, in one setting a person may categorize by means of comparison to a prototype and in another may categorize an object by summing attributes ala the classical model. Our assertion goes further by proposing that there may be no pre-existing category structure, in the sense of the classic, prototype, or exemplar models, and that what is to serve as a prototype, or which object attributes are to be summed, is generated during the act of categorizing.

A generative approach to categorization is congruent with current work in categorization theory which questions the inherent stability of concepts. For example, experiments by Wattenmaker et al (1986) indicate that basic or "natural" level concepts are not always basic or natural and they go on to conclude:

The general implication of the results is that it is simply not possible to identify an abstract category structure as natural and easy or as unnatural or difficult independent of the knowledge structures that are brought to bear on the category structure (p.187).

Such conclusions amount to a recognition that categories do not exist independently of judgment context. A number of theorist are contending that the Aristotelian models can not be salvaged by modifying their structural properties (Barsalou 1987; Medin & Wattenmaker 1987).

As a case in point, Higgins & Lurie (1987) conducted a series of experiments which made explanations based on altering the basic prototype model unlikely. One such adaptation is the "refocusing hypothesis" in which context serves to constrain what would serve as an acceptable prototype or exemplar. Contrary to what would be expected if this hypothesis were correct, order of representativeness (typicality rating) is not determined by resemblance to the exemplar most strongly suggested by the context. Higgins & Lurie (1987) note that one plausible way to account for their results is in terms of the "restructuring hypothesis": a hypothesis they note strongly resembles schema theory. In the restructuring case, the entire category is restructured, with a subsequent change in ratings of typicality, in response to contextual information. Such restructuring requires the use of background knowledge that is often not explicitly provided in the judgment context. A schematic representation of knowledge would allow for filling in missing information and generation of contextual variable categories. In similar vein, Barsalou (1983) has proposed a comparative network model to account for the structural properties of "common" and "ad hoc" categories. Barsalou (1987), more recently, has labeled the notion of concepts as pre-existing, invariant structures as an "analytic fiction" and has proposed that concepts are generated in working memory.

Schema theory is not the only theoretical system which could treat categorization as a generative phenomena. Such an approach can also be found in the ecological school of psychology (Gibson 1977; Lombardo 1987; McCabe & Balzano 1986). This school seeks to explain category judgments without recourse to "in the head" variables such as schemas or categorical structures. On this view, organisms by virtue of their ecological niche "directly" pick up information from the environment: a process known as the theory of affordances (Gibson 1977). While a more inclusive discussion of schema theory or the theory of affordances is outside the domain of this paper, the immediate point is that many findings from category research can be accounted for without assuming a pre-existing category structure.

This preliminary investigation does not propose that figure/ground represents a theory of categorization nor is it necessarily contradictory to the robust findings from categorization studies, such as typicality effects. In the examples provided, respondent's category judgments often were attribute based or prototypic in nature so figure/ground descriptions of category judgments do not exclude these phenomena. What is suggested by the figure/ground description is a need to develop theoretical models which can account for the wide range of contextual variability and flexibility exhibited in category judgments.


Alba, Joseph W. and Amitava Chattopadhyay (1985), "Effects of Context and Part-Category Cue on Recall of Competing Brands," Journal of Marketing Research, 22 (August), 340-349.

Alba, Joseph W. and J. Wesley Hutchinson (1987), "Dimensions of Consumer Expertise," Journal of Consumer Research, 13 (March), 411-454.

Barsalou, Lawrence W. (1983), "Ad hoc Categories," Memory and Cognition, 11, 211-227.

Barsalou, Lawrence W. (1987), "The Instability of Graded Structure: implications for the nature of concepts," in Concepts and Conceptual Development: Ecological and Intellectual Factors in Categorization, ed. Ulric Neisser, New York: Cambridge University Press.

Bruner, Jerome S., Jacqueline J. Goodnow, and George A. Austin (1956), A Study of Thinking, New York: John Wiley dc Sons.

Cohen, Joel B. and Kunal Basu (1987), "Alternative Models of Categorization: Toward a Contingent Processing Framework," Journal of Consumer Research, 13 (March), 455-472.

Ericcson, Anders K. and Herbert A. Simon (1984), Protocol Analysis: Verbal Reports as Data, Cambridge, MA.: M.I.T. Press.

Gibson, James J. (1977), 'The Theory of Affordances," in Perceiving, Acting, and Knowing: Toward an Ecological Psychology, eds. Robert Shaw and John Bransford, Hillsdale, NJ.: Lawrence Erlbaum Publishers.

Higgins, E. Tory, and Liora Lurie (1983), "Context, Categorization, and Recall: The "Change-of-Standard" Effect," Cognitive Psychology, 15 (July), 525-547.

James, William (1890/1983), The Principles of Psychology, Cambridge, MA: Harvard University Press.

Jenkins, James J. (1977), Remember That Old Theory of Memory? Well, Forget it!," in Perceiving, Acting, and Knowing: Toward an Ecological Psychology, eds. Robert Shaw and John Bransford, Hillsdale, NJ.: Lawrence Erlbaum Publishers.

Kahneman, Daniel, Paul Slovic, Amos Tversky (1982), Judgment Under Uncertainty: Heuristics and Biases, Cambridge: Cambridge University Press.

Kohler, Wolfgang (1947), Gestalt Psychology, New York: American Library Press.

Kohler, Wolfgang (1969), The Task of Gestalt Psychology, Princeton, NJ.: Princeton University Press.

Koffka, Kurt (1935), Principles of Gestalt Psychology, New York: Harcourt, Brace, and World Inc.

Lakoff, George and Mark Johnson (1980), Metaphors We Live By, Chicago: University of Chicago Press.

Lombardo, Thomas J. (1987), The Reciprocity of Perceiver and Environment: The Evolution of James J. Gibson's Ecological Psychology, Hillsdale, NJ.: Lawrence Erlbaum Associates.

McCabe, Viki and Gerald J. Balzano (1986), Event Cognition: An Ecological Perspective, Hillsdale, NJ.: Lawrence Erlbaum Associates.

Medin, Douglas L. and William D. Wattenmaker (1987), "Family Resemblance, Conceptual Cohesiveness, and Category Construction," Cognitive Psychology, 19, 242-279.

Medin, Douglas L. (1987), "Cognitive Models and Prototype Theory," in Concepts and Conceptual Development: Ecological and Intellectual Factors in Categorization, ed. Ulric Neisser, New York: Cambridge University Press.

Neisser, Ulric (1976), Cognition and Reality, New York: W.H. Freeman And Company.

Pollio, Howard R., Jack M. Barlow, Harold J. Fine, and Marilyn R. Pollio (1977), Psychology and the Poetics of Growth, Hillsdale, NJ.: Lawrence Erlbaum Press.

Reed, Stephen K. (1982), Cognition, Monterey, CA.: Brooks/Cole Publishing.

Rosch, Eleanor (1978), "Principles of Categorization," in Cognition and Categorization, eds. E. Rosch and B.B. Loyd, Hillsdale, NJ.: Lawrence Erlbaum Associates.

Rosch, Eleanor (1985), "Cognitive Representations of Semantic Categories," Journal of Experimental Psychology: General, 104, 192-233.

Rosch, Eleanor and Carolyn B. Mervis (1975), "Family Resemblances: Studies in the Internal Structure of Categories," Cognitive Psychology, 7, 573-605.

Roth, Emile M. and Edward J. Shoben (1983), 'The Effect of Context on the Structure of Categories," Cognitive Psychology, 15 (July), 346-378.

Shaw, Robert and John Bransford (1977), "Psychological Approaches to the Problem of Knowledge," in Perceiving, Acting, and Knowing: Toward an Ecological Psychology, eds. Robert Shaw and John Bransford, Hillsdale, NJ.: Lawrence Erlbaum Publishers.

Smith, Edward E. and Douglas Medin (1981), Categories and Concepts, Cambridge, MA.: Harvard University Press.

Sujan, Mita (1985), "Consumer Knowledge: Effects on Evaluation Strategies Mediating Consumer Judgments," Journal of Consumer Research, 12 (June), 31-46.

Sujan, Harish, Mita Sujan, and James Bettman (1988), "Knowledge Structure Differences Between More Effective and Less Effective Salespeople," Journal of Marketing Research, 25 (February), 81-86.

Verbrugge, Robert R. (1977), "Resemblances in Language and Perception, " in Perceiving, Acting, and Knowing: Toward an Ecological Psychology, -eds. Robert Shaw and John Bransford, Hillsdale, NJ.: Lawrence Erlbaum Publishers.

Wattenmaker, William D., Gerald I. Dewey, Timothy D. Murphy, and Douglas Medin (1986), "Linear Separability and Concept Learning: Context, Relational Properties, and Concept Naturalness," Cognitive Psychology, 18, 158-194.