Concept Formation, Product Conceptualization and Cognitive Development

Elizabeth C. Hirschman, New York University
ABSTRACT - This paper examines the acquisition of concepts by consumers and the nature of conceptual structure incorporated in semantic memory. Alternative perspectives are provided to the assumptions underlying the multiattribute attitude model. Additionally, a model of ontogenetic information acquisition is presented, which has as its theoretical basis the Piagetian paradigm of cognitive development.
[ to cite ]:
Elizabeth C. Hirschman (1980) ,"Concept Formation, Product Conceptualization and Cognitive Development", in NA - Advances in Consumer Research Volume 07, eds. Jerry C. Olson, Ann Abor, MI : Association for Consumer Research, Pages: 405-410.

Advances in Consumer Research Volume 7, 1980     Pages 405-410

CONCEPT FORMATION, PRODUCT CONCEPTUALIZATION AND COGNITIVE DEVELOPMENT

Elizabeth C. Hirschman, New York University

ABSTRACT -

This paper examines the acquisition of concepts by consumers and the nature of conceptual structure incorporated in semantic memory. Alternative perspectives are provided to the assumptions underlying the multiattribute attitude model. Additionally, a model of ontogenetic information acquisition is presented, which has as its theoretical basis the Piagetian paradigm of cognitive development.

INTRODUCTION

It is extensively documented within cognitive psychology that the process by which concepts are formed and the nature of the concepts an individual possesses bear directly upon perceptions of the environment, responses to stimuli and ability to solve problems (e.g., Newell and Simon, 1972). [Stated most simply, a concept is a category or a classification device, used by the individual to organize and sort knowledge. Most concepts which we possess are designated by the nouns present in language; for example, automobile, house, camera and toothpaste are all product-concepts.] Despite the attention given to concept formation within psychology, there have been few attempts in consumer research to investigate this process. Similarly, there is little published evidence of investigation by consumer researchers into how individual conceptual structure (that is, the interrelationships among concepts acquired by the individual) may influence marketing-related activities such as consumer information processing, product evaluation, product purchasing, post purchase satisfaction and the like.

Much consumer research has been directed towards measuring the viability of alternative models of attitude structure and the various strategies by which consumers may combine, weight or sort attributes associated with a particular product-concept. This attitude and attribute-centered research tradition has dominated much of the consumer behavior literature for over a decade (e.g., Bettman, 1970; Bruno and Wildt, 1975; Holbrook, 1978). It has led to a vast array of empirical findings, models, conjectures and, recently, to efforts of middle-range theorization, (e.g., Capon and Lutz, 1979; Jacoby, 1978).

Present Assumptions

A cogent summary of the paradigm underlying the multi-attribute attitude approach to consumer decision-making was provided recently by Capon and Lutz (1979) within the context of developing effective consumer information programs. As derived from their description of this paradigm, three assumptions may be identified.

Assumption One: The alternative products or brands which the consumer considers in a decision are arrayed mentally in a matrix format with the various alternatives forming the columns of the matrix, and the attributes relevant to the alternatives forming the rows.

Assumption Two: An attitude toward a product or brand is assumed to be a function of the consumer's preferences and/or importance of the attributes "composing" or "typifying" that product or brand.

Assumption Three: Individual differences in information processing capacity and decision-making activities are insignificant and/or unpredictable.

PURPOSE

The purpose of this paper is to examine these three assumptions in light of some recent theoretical and empirical developments in cognitive psychology. The first portion of the paper discusses the first and second assumptions of this commonly-held consumer decision making paradigm; that is, that the consumer uses a matrix-like representation of product alternatives and attributes to conceptualize and evaluate products, and that his/her attitude towards a product is a function of importance and/or preference for the bundle of attributes comprising the product (e.g., Bettman, 1970; Bruno and Wildt, 1975; Fishbein and Ajzen, 1975; Hanson, 1972; Holbrook, 1978; Newman and Lockman, 1975; Summers, 1974; Mazis, Ahtola and Klippel, 1975; Rosenberg, 1956).

The latter portion of the paper examines a third assumption implicit in the multi-attribute attitude paradigm; that individual differences in information processing capability and problem solving behavior are insignificant and/or unpredictable. A case will be presented for adopting an alternative perspective based upon the paradigm of cognitive development. This theoretical perspective posits that sociocultural forces may substantially influence consumer information processing and problem solving abilities.

Attribute List Theory

Presently, there are two competing cognitive paradigms of concept formation. One cognitive processing theory ("attribute list" theory) postulates that concepts are stored in semantic memory [Tulving (1972, p. B86) defines semantic memory as "the memory necessary for the use of language. It is ... organized knowledge a persons possesses about words and other verbal symbols, their meaning and referents, about relations among them, and about rules ... for the manipulation of concepts."] in terms of their representative features (attributes). As Bourne, et. al. (1979) states, "Objects, processes, events, and states of affairs in the world are categorized and encoded conceptually in memory through the extraction of their common features" (p. 163). Thus, when the consumer encounters a new product, for example a new cooking utensil, he or she attempts to classify it in terms of already held concepts (e.g., other cooking utensils) with which it shares attributes. This classification process is performed by matching up the attributes of the product stimulus with those of various concepts residing in semantic memory.

In the attribute-list paradigm, a concept is defined in terms of some relationship among its relevant attributes, such that:

C = R(x,y ...)

where C is the concept, x,y ... are the relevant or defining attributes of the concept, and R is the rule for integrating the attributes (Bourne, 1974). For example, the concept of an automobile can be defined in terms of a collection of attributes such as "has four wheels," "propelled by motor," 'uses gasoline for fuel," and so on.

Within the attribute-list theory a useful distinction is made between the association of relevant attributes with a concept and the learning of the rules to combine those attributes to represent the concept. To form a concept requires that an individual learn or identify, the critical [Critical attributes are those features which are possessed by (most) members of a given concept category and not possessed by members of other concept categories.] (defining) attributes of examples of that concept and also identify the interrelationships among those attributes. It is this model of concept structure which is in, licitly assumed by most applications of the multi-attribute attitude paradigm. That is, that a concept (e.g., a product or a brand) may be defined simply as a function of its attributes; and that brands and/or products being considered by the consumer in a choice situation are homogeneous in that they all possess the same set of salient attributes (albeit in differing combinations and/or levels).

Prototype Theory?

The attribute list theory of concept formation has been criticized by Rosch (1973) and others for its lack of concern with the structure of concepts. The concept stimuli presented to subjects in many laboratory experiments of concept formation, as well as attitudinal and information processing studies conducted by marketing researchers (see Green and Srinivasan, 1978; Wilkie and Pessemier, 1973 for reviews), typically are based on discrete values of a limited number of assumed independent dimensions. Concepts (products) are experimentally defined by a combination of these values. As Rosch (1973) points out, this state of affairs is not typical of concepts (or products) as they are encountered in "real life." Natural concepts, such as the category automobile or toothpaste, are not easily described as arbitrary combinations of discrete values taken from some number of independent attribute dimensions. Neither are all "real" examples of a particular product concept equally useful examples of their respective categories. For example, most people would probably report that a Ford Fairmont is a "better" (or more "typical") example of the concept category "automobile" than is a Honda Civic or a Chevrolet van. Further, the salient attributes of product concepts are often difficult to specify and may themselves be composed of subtle combinations of underlying dimensions. Attribute dimensions salient to a product category typically are continuous rather than discrete, and may be non-orthogonal as well. Typically in consumer research a dimensional or attribute analysis is performed on product concepts in order to describe their composition and/or consumers' attitudes towards the product. Yet this is not necessarily characteristic of the way the consumer conceptualizes the product in everyday life (Bourne, et. al., 1979). Rather, cognitive research on semantic memory indicates that individuals may view concepts in terms of some global pattern or Gestalt organization, (e.g., Craik, 1979; Simon, 1979).

Implicit in the research on multi-attribute attitude models is the idea that products may be defined in terms of a direct listing of their salient attributes. However, Rosch (1975) suggests an alternative paradigm based upon the notion that concepts are represented in memory by their "most typical" examples or "prototypes." This is a Gestalt or holistic interpretation which is centered around the role played in memory by an archetypical or "ideal-type" [The "ideal-type" referred to here is descriptive, rather than normative, in nature. It implies merely that a particular pattern of attributes is an accurate representation of a given concept category; not that this pattern is more desirable or preferable than another.] concept image. Research (e.g., Brown, 1976; Rosch, 1973) indicates that individuals are quickest to identify (and learn most rapidly) stimuli which are closest to the concept prototype. In a marketing setting, these findings would suggest that consumers most rapidly recognize and bring to mind products that are highly representative of a given product class. Products which are somehow "different" or "atypical" of the product class would be less quickly recognized during external search or brought to mind by an internal search process.

The notions of an internalized product prototype concept as discussed by Rosch have their roots in the early work by Bartlett (1932) on memory schemata. Put simply, schemata are abstract analogical representations of concepts composed of a basic set of facts which capture the essence or gist of the stimulus (Attneave, 1957; Evans and Edmonds, 1966; Posner and Keele, 1968). Evidence from studies by these researchers indicates that "during learning, the subject acquires information not only about the prototypical stimulus, but also about the variability among instances of a given class ... subjects learn to identify the best instance of the category, plus something about the allowable variability or distance among admissible stimuli." (Bourne, et. al., 1979, p. 2Ol).

This leads to the conclusion that product concepts may be represented in the consumer's conceptual or semantic memory not in terms of a specific set of attributes, but rather in terms of some schema or pattern of attribute dimensions which captures the "gross characteristics" or gist of the products belonging to that category, (Reed 1972). Further, it has been found that the information contained in these cognitive schemas may permit the individual to recognize novel stimulus configurations (for example, a new design of automobile) and to order these configurations in terms of their distance from the abstracted prototype (Franks and Bransford, 1971).

A further consideration within prototype theory is how the individual extracts the analog structure or pattern from a set of product stimulus examples. Rosch and Mervis (1975) examined the hypothesis that members of a category (e.g., automobiles) come to be viewed as typical to the extent they bear a "family resemblance" (i.e., cue validity) to other members of the same category. Family resemblance was defined as the possession of attributes which are common to other members of the category. The converse of this hypothesis implies that products which are most typical or characteristic of a given category will be those which have the fewest attributes in common with members of other categories.

Therefore, a Ford Fairmont sedan might more closely approximate the prototype of the product concept "automobile," because its features are common to this category of products. A Chevrolet van would possess less family resemblance and be perceived as more distant from the prototype because it possesses attributes not common to this category and which overlap with other product categories. For example, the van has a shape and seating arrangement akin to that of a bus. From their experiments, Rosch and Mervis (1975) were able to confirm that the salient attribute structure of a category tended to reside not in attributes which were common to all members of the category, but rather in attributes which were true of some, but not all, category members. They found, for instance, that the five most typical members of each category held significantly more attributes in common than did any five "less typical" members.

These findings would appear to suggest that the product-by-attribute matrix presented at the outset of this discussion and commonly adopted in marketing research may be an inappropriate paradigm. The notions from prototype cognitive theory suggest that information may be absent from some attribute/alternative cells of the product decision matrix not because the consumer was unable or unwilling to acquire information to fill them, but rather because a particular attribute is absent or irrelevant for a given member of the product class. It follows then, that the usual attribute equation used to describe consumer attitudes towards products may be similarly inappropriate in that it would be filled with "gaps" or "holes" representing missing attributes not characterizing a specific product alternative. Valid mathematical comparisons of consumer attitudes towards alternative products would be difficult (or impossible) given the varying length and composition of the algebraic function.

The Attribute Frequency Paradigm

Bourne, et. al. (1979), and other cognitive researchers (e.g., Neumann, 1974, 1975; Franks and Bransford, 1971) have pointed out that a "middle ground" compromise may be possible between the alternative paradigms of attribute list theory and prototype theory of concept formation. They note that the two positions are not incompatible if one modifies the notion of concept representation via attribute lists to incorporate instead a model based upon attribute frequency. Using a model of relative attribute frequency to describe concepts "associates each stimulus with a set of frequency counts, one for each of its discriminable features," (Bourne et. al., 1979, p. 214).

The frequency counts are compiled from the various "real life" examples of the concept the individual encounters. The number of occurrences of a given attribute among the stimulus set encountered by the individual are tallied and stored in long term memory. One role of differential attribute frequencies is in determining the confidence of the individual's recognition and rating of degree of concept category membership for a new stimulus encountered in the environment.

For example, if a consumer encounters a novel automobile (for example, the gull-wing door DeLorean Car) his or her ability to correctly classify and evaluate that innovation is moderated by the frequency and pattern of attribute overlap with the prototype concept of "automobile" the consumer holds in memory. The DeLorean auto being significantly different from "typical" automobiles along several attribute dimensions will result in reducing the consumer's confidence and competence in evaluating its potential performance as a product.

Thus, a compromise theoretical position may be constructed by merging propositions from the attribute list and prototype paradigms of concept formation. Using this perspective, a concept may be fully defined by a set of potential salient attributes, their naturally encountered frequencies (varying according to individual experience) expressed as an objective probability distribution, and the pattern of relationships among the attributes. This notion is expressed in the function below:

C = R(pA1, pA2, pA3 ... pAn)

where:

C is the concept prototype,

R is the combining rule for forming the attributes into a pattern,

pA1 is the objective probability that a given attribute (A1 ... An) has been encountered by the individual in conjunction with naturally occurring examples of the concept.

The concept prototype is an "idealized" stimulus which contains all (or a maximal number of) modal attributes. Naturally occurring examples of the concept rare rank ordered vis-a-vis the prototype according to how many modal attributes they possess, which one, or how similar their composition is to the modal one.

Cognitive Algebra and Syllogistic Reasoning

In much research, the assumption has been made (see Assumptions one and two, page 1 of this paper), that consumers arrive at a decision by operating on a matrix-like array of alternatives and their attributes. This assumption has been challenged on the basis of its incompatibility with the attribute frequency paradigm of concept representation in cognitive structure. The notion of matrix-process consumer decision making may also be questioned on the grounds of empirical evidence concerning cognitive operations.

By the early 1960's cognitive researchers had determined that the construction of concepts resulting from different logical inferences (i.e., learning rules) varied in difficulty and in the kind of process required for their acquisition or discovery (e.g., Bruner, Goodnow and Austin, 1956; Hunt and Hovland, 1960; Conant and Trakasso, 1964; Neisser and Weene, 1962). Neisser and Weene (1962) have theorized that individuals are, at best, limited to the use of three primitive logical operations: negation (not operations ), conjunction (and operations), and disjunction (and/or operations). Therefore, an individual' s comprehension of any relationship among attributes of concepts must be reduced to some combination of these operations. The difficulty of acquiring and evaluating a concept will be a function of how many primitive logical operations it requires.

Haygood and Bourne (1965) have reformulated the three logical operations of Neisser and Weene into four different pairings of rules. These basic rules are labeled conjunctive, disjunctive, conditional and biconditional (i.e., TT, TF, FT, FF combinations) [Where T = True, F = False for any premise.] and are analogous to the logical bidimensional "truth table" construct of formal logic. The truth table logical format assumes that the individual processes information in a matrix context analogous to the decision model described by Capon and Lutz (1979). However, as Bourne, et. al. (1979, P. 226) note, "The truth table strategy is deceptively simple ... To employ it, the subject must attend only to those dimensions of the stimulus that have been confirmed as relevant and ignore all others. Further, he must focus upon the relevant attributes within those dimensions. Beyond attending selectively, the subject must form all possible combinations of relevant attributes ..., must understand what it means to negate or take the complement of an attribute ... and deal simultaneously with (multiple) coded classes or subsets of the stimuli."

In several experiments with individuals lacking formal training in logical reasoning (i.e., the "average" consumer) it was consistently discovered that most were unable to perform logical evaluations of even simple bidimensional stimuli, (e.g., Dodd, Kinsman, Klipp and Bourne, 1971; Bourne and O'Banion, 1971; Ceraso and Prouitera, 1971; Revlis, 1973). Dominowski, for example, found that 12 percent of his subjects displayed logical processing of information, 74 percent were "partly logical'' and 14 percent were "unmodellable" (or random) in their choice behavior (Dominowski 1976). Thus, it would appear likely that many consumers are not capable of processing information in the form assumed by the multi-attribute information matrix.

INDIVIDUAL DIFFERENCES IN INFORMATION PROCESSING CAPACITY AND PROBLEM-SOLVING ABILITY

A closely-related, third assumption which appears to underlie the multi-attribute attitude paradigm is that consumers do not vary significantly in their abilities to process information and to solve consumption problems requiring cognitive activity. [This assumption has been recently challenged in a very insightful theoretical analysis by Capon and Lutz (1979). These two researchers posit that inter-personal differences in information requirements may be a very useful basis for market segmentation. The discussion presented above addresses this conjecture by suggesting the Piagetian Theory as a framework for such segmentation.] This assumption may be challenged from two complementary perspectives - one psychological in origin and the other sociological. The psychological perspective which may be cited is the theory of hierarchical cognitive development put forward by Piaget (1950, 1953, 1969). The Piagetian paradigm posits that human cognitive development is hierarchical in nature. The lower stages of cognitive development are "embedded in and hierarchically subsumed by" the higher stages (Feldman et. al., 1974). Because the higher stages incorporate the lower ones, one can conceptualize everything at a higher stage that was conceptualized at the lower stage, and more; but in a lower stage one cannot conceptualize all that is possible at a higher stage.

Piaget describes human cognitive development as a sequence of four stages. At the first, or sensori-motor stage, the child's abilities are limited to operating directly on tangible objects without any mediating internal symbolic structures. The rather loosely defined abilities of the preoperational stage become active when the child is first capable of statically representing actions internally in the form of mental imagery. He/she becomes a primitive transformer of the environment, but internal symbolic transformations remain unidirectional and therefore lack an internally consistent structure.

With the onset of concrete operations, the child's internal actions become reversible; he/she is capable of mentally undoing what has just been done. Concrete operations represent an advance from the pre-operational period in that the child develops a complete, systematic scheme for organizing reality states. His/her conceptual apparatus is not, however, powerful enough to generate all of the possible combinations of propositions or all of the possible kinds of relations among relations.

The ability to relate propositions into a coherent system containing the total array of propositional possibilities is the hallmark of formal operations. This stage takes the organized structures of concrete operations and generates their further inherent possible combinations when they are applied to such abstract subjects as propositions. At this level the individual is able to reason about propositions independently of the realities they stand for, and thus is able to extend the range of operations from particular responses to hypothetical propositions (Feldman, et. al., 1974). Recalling the earlier discussion of the relatively incomplete or inadequate use of propositional logic found for many adults in problem-solving activities; it is apparent that not all, perhaps not even a majority of consumers fully master Piaget's fourth stage of cognitive development, -that of formal, propositional operations. [In this regard, it may be best to view the formal 'stage' as continuum of logical development, rather than as a dichotomous condition or plateau.]

An important question which arises at this point is: what factors may contribute to individual ability in forming logical propositions for use in making consumption decisions and the capacity to solve consumption-relevant problems. The research cited indicates that individual differences in cognitive development may be present which will affect the consumer's ability to reach logical decisions. It would seem that ascertaining the causes of these developmental differences may be a fruitful avenue for consumer research.

Some insight into sources of these individual differences may be obtained by returning to the Piagetian hierarchical cognitive model. It is posited by this model that the child gains greater reasoning capacity and improved problem-solving skills through experiences with environmental stimuli. Transferring this notion to consumer behavior, it is straight forward to reason that consumers, whether children or adults, may enhance their information processing and problem-solving capacities through exposure to varied, instructive consumption experiences.

In other words, the more consumption-relevant information the individual has received and processed during his or her life experience, the better able he or she may be to conceptualize different product stimuli and consumption situations. Thus in evaluating a new product, the highly experienced consumer may be able to consider it along more dimensions end across a greater range of potential usage contexts than the less experienced consumer.

The central proposition is that as incremental knowledge is obtained by the consumer through actual product use, from various media, or from interpersonal sources, he/ she shapes propositions of product interrelationships and potential uses into theories of alternative consumption strategies. Having created these theories, the consumer may then apply them toward understanding the ways in which products may solve new or existing consumption problems. Additional information inputs from "hypothesis-testing" consumption activity on the part of the consumer (either actual or vicarious) may serve to further refine the theory, permit the consumer to gain a greater sense of understanding as to how the product can or may function, and to increase his/her control over the environment.

Because additional information gained from experiential sources is cumulative, the consumer will tend to enlarge the number of attributes he or she associates with each product-concept, and thus to increase the number of dimensions relating one product to another. Through this process the consumer may be able to establish connections among products which were originally seen as completely dissimilar (i.e., analogization and correlation) and to differentiate among products which were earlier viewed as identical (i.e., discrimination). The highly experienced consumer who has acquired more consumption-relevant information is, therefore, more likely to arrive at alternative problem-solutions in a more logical and effective manner than the consumer having less experience.

A DEVELOPMENTAL MODEL OF INFORMATION ACQUISITION

A central area of inquiry arising from this reasoning is that of the sources from which consumers acquire information applicable to the solution of consumption problems. An examination of behavioral science literatures, particularly those of societal modernization, cross-cultural cognitive development and trait psychology (e.g., Inkeles and Smith 1974; Dasen 1972; Barron 1969) reveals that several environmental correlates of problem-solving capacity may be identified. [The problem-solving capacity possessed by consumers has been termed consumer creativity. For a more detailed discussion of the conceptual rationale for this designation see Hirschman, 1979.] These variables include parental socioeconomic status (PSES), childhood mental stimulation (CMS), formal education (FE), adult occupational stimulation (OS), adult role accumulation (RA), adult role specialization (RS) and urbanization (U). Although the reader is likely familiar with the first four of these variables, the latter three are relatively uncommon in the consumer behavior literature. These may be defined as: (1) role accumulation, a measure of the number of non-overlapping roles the individual is occupying at a given point in time (e.g., Wallendorf, 1979), (2) role specialization, the amount of specialized expertise or training that is required for the performance of each of the individual's roles, summed across the role set; and (3) urbanization, the population density of the area in which the individual resides.

To summarize the posited relationships of these environmental factors to consumer problem-solving capacity, the following functional formulation is put forward:

1. CPSC = f (PSES, CMS, FE, OS, RA, RS, U)

where:

CPSC = consumer problem-solving capacity

PSES = parental socioeconomic status

CMS = childhood mental stimulation

FE = formal education

OS = adult occupational stimulation

RA = adult role accumulation

RS = adult role specialization

U = urbanization

Thus, consumer problem-solving capacity may be defined as a function of parental socioeconomic status, childhood mental stimulation, formal education, occupational stimulation, role accumulation, role specialization and urbanization. With the exception of parental socioeconomic status, the effects of each of these variables upon consumer problem-solving capacity are expected to be cumulative. In other words, the greater the exposure of the individual to these sources of stimulation, the greater his/her capacity as a logical problem-solver will become. Further, these environmental sources of consumer problem-solving capacity may be organized into a longitudinal model to provide an ontogenetic perspective of information acquisition. This is depicted in Figure 1.

A Description of the Model's Functioning

The longitudinal functioning of this model (Figure 1) can be described as follows: The socioeconomic status of the individual's parents affects his/her problem-solving capacity in a variety of ways. First, parents with above average levels of education and occupational status (e.g., socioeconomic status) have been found to instill in their children a higher need for achievement and motivation to excel in competitive activities than parents with low SES (Smelser 1973). This set of values is acquired by the child from the parents and positively influences his/her performance in school as well as the type of job the individual will seek. Further, children born into higher SES homes generally have more financial support for extended formal education, attend better quality primary and secondary schools, and are exposed to more different stimulating non-school experiences as children (e.g., trips, magazines, movies, sports). Thus, parental SES influences the sources and levels of formal education and informal cognitive stimulation the individual receives as a child (Arieti 1976, Bendix 1967, Gerard 1946).

The problem-solving skills the consumer acquires in childhood carry over into adult life. Job choice, role specialization and role accumulation may be influenced by experiences gained during youth. However, once the individual begins functioning as an adult consumer, these three factors will work to further enhance his/her problem-solving competency. Novel Job experiences which require abstract reasoning and complex decision making serve to sharpen the consumer's decision capabilities (e.g., Kohn and Schooler, 1978).

Additionally, the consumer may occupy many roles in addition to his/her occupational role (e.g., Wallendorf, 1979; Hirschman and Wallendorf, 1979). For example, the individual may be an audiophile, a squash player, coin collector, gardener, marketing consultant and parent, in addition to having an occupation as a university professor. Each of the additional roles which the individual accumulates may require the acquisition of incremental specialized knowledge to be performed successfully (e.g., Smelser, 1973). Hence, the number of the roles accumulated factored by the level of additional specialized knowledge they require will increase the consumer's knowledge. The stimulation provided by role accumulation and role specialization is incremental to that provided by his/her occupation.

A final and important source of stimulation for consumer problem-solving skills is the level of urbanization characterizing his/her environment. The more densely populated the area in which the consumer resides and the larger the absolute size of the surrounding population, the more consumption-relevant stimulation he/she may receive (e.g., Bendix 1967, Fischer 1978, Suzman 1977). This is because greater urbanization leads to more differentiated and complex social structure (e.g., Blau 1975). The consumer is confronted with a more diverse set of products, of retail outlets and of consumption opportunities.

CONCLUSIONS

This paper has reviewed selected findings and theories from cognitive psychology with the intent of providing some novel perspectives of consumer perception and decision making.

As Zaltman and Bonoma (1979) have noted, alternative perspectives and "heretical" notions may often have beneficial effects in causing the reexamination of commonly-held assumptions. The desire here is not to criticize the ideas and models that are currently held and which have been valuable in advancing consumer research; but rather to suggest that alternative perspectives may provide useful insights into this very complex area of consumer behavior.

FIGURE 1

SOURCES OF CONSUMPTION-RELEVANT INFORMATION

REFERENCES

Arieti, Silvana (1976), Creativity: The Magic Synthesis, New York, Basic Books, Inc.

Armer, Michael and Isaac, Larry (June 1978), "Determinants and Behavioral Consequences of Psychological Modernity," American Sociological Review, 43, 316-334.

Armer, Michael and Schnaiberg, Allan(1976), "Individual Modernity, Alienation and Socioeconomic Status: A Replication in Costa Rica," Studies in Comparative International Development, 10, 35-47.

Attneave, F. (1957), "Transfer of Experience with a Class-Schema," Journal of Experimental Psychology, 54, 81-88.

Barron, Frank (1969), Creative Person and Creative Process, New York, Holt, Rinehart and Winston.

Bartlett, F. C. (1932), Remembering: A Study in Experimental and Social Psychology, New York: MacMillan Comp.

Bendix, Reinhard (1967), "Tradition end Modernity Reconsidered," Comparative Studies in Society and History, 3, 292-346.

Berry, J. W. (1976), Human Ecology and Cognitive Style, New York, Wiley.

Berry, J. W. (1974), and Dasen, P. R. (eds.) Culture and Cognition, (London Metheien).

Bettman, James R. (August 1970), "Information Processing Models of Consumer Behavior," Journal of Marketin6 Research, 7, 370-376.

Blau, Peter M. (1975), "Parameters of Social Structure," in Peter M. Blau (ed.) Approaches to the Study of Social Structure, New York, The Free Press, 135-153.

Bourne, L. E. and O'Banion, K. (1971), "Conceptual Rule Learning and Chronological Age," Developmental Psychology, 5, 525-534.

Bourne, L. E., Dominowski, Roger L. and Loftus, Elizabeth F. (1979), Cognitive Processes, Prentice Hall: Eagle-wood Cliffs, New Jersey.

Bourne, L. E. (1974), "An Inference Model of Conceptual Rule Learning," in Theories in Cognitive Psychology, R. Solso (ed.) Washington D.C., L. Erlbaum and Associates.

Brown, R. (1976), "Reference in Memorial Tribute to Eric Lennebert," Cognition, 4, 125-153.

Bruner, J. S. (1966), "On Cognitive Growth," in J. S. Bruner, R.R. Olver and P.M. Greenfield, Studies in Cognitive Growth, New York: Wiley, 30-67.

Bruner, J. S. and Postman, L. (1949), "On the Perception of Incongruity: A Paradigm," Journal of Personality, 18, 206-223.

Goodnow, J. J. and Austin, G. A. (1956), A Study of Thinking, New York: John Wiley.

Bruno, Albert V. and Wildt, John (September 1975), "Toward Understanding Attitude Structure: A Study of the Complementarity of Multi-Attribute Attitude Models," Journal of Consumer Research, 2, 137-145.

Capon, Noel and Lutz, Richard J. (January 1979), "A Model and Methodology for the Development of Consumer Information Programs," Journal of Marketing, 43, 58-67.

Ceraso, J. and Provitera, A. (1971), "Sources of Error in Syllogistic Reasoning," Cognitive Psychology, 2, 400-410.

Chinoy, Eli (1968), Sociological Perspective, New York, Random House.

Chodak, Szymon (1973), Societal Development: Five Approaches with Conclusions from Comparative Analysis, New York, Oxford University Press.

Conant, M. B. and Trabasso, T. (1964) "Conjunctive and Disjunctive Concept Formation Under Equal Information Conditions," Journal of Experimental Psychology, 67, 250-255.

Craik, Fergus I. M. (1979), "Human Memory," Annual Review of Psychology, Mark R. Rosenzweig and Lyman W. Porter (eds.), Annual Reviews, Inc., Palo Alto, CA, 63-102.

Dasen, P. (1972), "Cross-Cultural Piagetian Research: A Summary," Journal of Cross-Cultural Psychology, 3, 23-41.

Dodd, D. H., Kinsman, R., Klipp, R. and Bourne, L. E. (1971), "Effects of Logic Pretraining a Conceptual Rule Learning," Journal of Experimental Psychology, 88, 119-122.

Dominowski, R. L. (1976), "Reasoning," Inter-American Journal of Psychology, 311-321.

Evans, S. H. and Edmonds, E. M. (1966), "Schema Discrimination as a Function of Training," Psychonomic Science, 5, 303-304.

Feldman, Carol, Lee, Benjamin, McLean, James, Pillemer, David and Murray, James (1974), The Development of Adaptive Intelligence, San Francisco: Jossey-Bass, 9.

Fischer, Claude S. (July 1978), "Urban to Rural Diffusion of Opinions in Contemporary America," American Journal of Sociology, 151-159.

Fishbein, Martin and Icek, Ajzen (1975), Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research, Reading, Mass: Addison Wesley.

Franks, J. J. and Bransford, J. D. (1971), "Abstraction of Visual Patterns," Journal of Experimental Psychology, 90, 65-74.

NOTE: Balance of references omitted due to lack of space. They are available upon request from the author.

----------------------------------------