An Investigation of Cognitive Structure in a Shopping Context

Marsha L. Richins, Louisiana State University
Peter H. Bloch, Louisiana State University
ABSTRACT - While researchers in consumer behavior have begun to study cognitive structure as a consumer variable, no studies have investigated the relationship between cognitive structure and the presence or absence of strong brand preference in a shopping context. This study examines three forms of cognitive structure: differentiation, discrimination, and attribute importance. Data from a mail survey to a random sample of adult consumers were used to test the hypothesis that shoppers with strong brand preferences would score higher on measures of these cognitive structure than would consumers who hold weaker preferences. Study hypotheses were supported.
[ to cite ]:
Marsha L. Richins and Peter H. Bloch (1983) ,"An Investigation of Cognitive Structure in a Shopping Context", in NA - Advances in Consumer Research Volume 10, eds. Richard P. Bagozzi and Alice M. Tybout, Ann Abor, MI : Association for Consumer Research, Pages: 555-558.

Advances in Consumer Research Volume 10, 1983      Pages 555-558

AN INVESTIGATION OF COGNITIVE STRUCTURE IN A SHOPPING CONTEXT

Marsha L. Richins, Louisiana State University

Peter H. Bloch, Louisiana State University

ABSTRACT -

While researchers in consumer behavior have begun to study cognitive structure as a consumer variable, no studies have investigated the relationship between cognitive structure and the presence or absence of strong brand preference in a shopping context. This study examines three forms of cognitive structure: differentiation, discrimination, and attribute importance. Data from a mail survey to a random sample of adult consumers were used to test the hypothesis that shoppers with strong brand preferences would score higher on measures of these cognitive structure than would consumers who hold weaker preferences. Study hypotheses were supported.

INTRODUCTION

In recent years researchers have investigated cognitive structure and its relationship to consumer behavior. Studies have focused on the relationship between cognitive structure and attitude change (Klippel and Anderson 1974; Lutz 1975), communication effects (Kasulis and Zaltman 1977; Lutz and Swasy 1977) and product familiarity (Hirschman 1981; Marks and Olson 1981; Tan and Dolich 1981), among other variables. Other studies have examined the development of cognitive structures and their measurement (Hirschman and Douglas 1981; Kanwar, Olson, and Sims 1981).

Cognitive structure has been conceptualized in diverse manners and there seems to be no common approach used by researchers. Lutz and Swasy (1977), for instance, used the term to refer to the "expectancy X value" component of multiattribute models. A more common use of the term refers to cognitive complexity, a concept initially developed in the personality and information processing areas of psychology (see Bieri et al. 1966; Schroder, Driver, and Streufert 1967). Cognitive complexity itself has been described differently by various researchers. While most authors have subdivided cognitive complexity into three forms, correspondence of forms among authors tends to be low. While an extensive review of cognitive structure is beyond the scope of this paper (the interested reader is referred to Scott, Osgood, and Peterson 1979), two aspects of cognitive complexity are relevant to this study and are discussed below.

One complexity form which is rather widely accepted among researchers, though with varying names, is cognitive differentiation. This construct refers to the number of categories or dimensions a person uses in processing information (Bieri et al. 1966). For instance, an individual who describes the taste of a soft drink along the dimensions of sweetness, fizziness, tartness, and fruitiness would be more complex than one who describes it merely along a good/bad dimension. It is not sufficient to simply examine the number of dimensions used in an evaluation, however. One must instead examine the number of functional dimensions. If two or more dimensions are used in exactly the same manner by an individual (i.e., the correlation between them is 1.0), the number of listed dimensions is greater than the number of dimensions actually in use.

A second complexity form is articulation or discrimination, the precision of distinctions along a particular dimension in use (Scott 1966. 19695. For instance, an individual who processes information about stimuli only as hot or cold is much lower in discrimination complexity than one who evaluates an object as cold, cool, lukewarm, warm, or hot. It is important to note that this construct does not refer to the ability to make discriminations but whether these discriminations are retained in memory and utilized in further information processing.

The study reported here examines cognitive structure in a shopping context. Researchers in brand loyalty and preferences have recognized that shoppers differ in brand commitment, or the extent to which brand names are important in making purchase decisions and to which a brand preference exists- (Chance and French 1972; may 1969). This study proposes that shoppers with strong brand preferences possess cognitive structures different from those whose brand preferences are weak. There is some support for this hypothesis. Robertson (1976) proposed that consumers who are highly brand committed possess a salient belief system which includes perceptions of product differences (articulation), at least among the salient dimensions of an individual's belief system. In empirical work, Scott (1962; see also Gutman 1980) found that those who are more highly involved in a subject area have more highly developed cognitive structures with respect to that subject. In addition, Lastovicka and Gardner (1978) found that perceptions of consumers highly involved with cars were more accurately represented multidimensionally while those of less involved consumers fit a unidimensional model. While product commitment (involvement) and brand commitment are not identical constructs (Traylor 1981), they do share some characteristics in common and, according to some researchers (Lastovicka and Gardner 1979; Tyebjee 1979), are positively related. Thus, it is reasonable to investigate whether shoppers with strong brand preferences are more cognitively complex than their less committed counterparts. In line with Robertson (1976), it is proposed that those with strong brand preference use more dimensions in evaluating products (differentiation) and are more likely to perceive differences between competing brands in a product class on those dimensions (articulation) than those with weak preferences. Further, brand committed shoppers are believed to consider more attributes or dimensions as important in making a purchase decision. These proposals are summarized in the following formal hypotheses tested in this study:

H1: Shoppers with strong brand preferences utilize more dimensions when evaluating competing brands than do shoppers with weak preferences.

H2: Shoppers with strong brand preferences perceive more differences between competing brands than do shoppers with weak preferences.

H3: Shoppers with strong preferences consider more attributes as important when making purchase decisions than do those with weak preferences.

METHODOLOGY

Sample

For the data collection phase of the study, 900 questionnaires were mailed under the auspices of a state university to a random sample of households in a western urbanized area with a population greater than one million. Because the topic of the questionnaire was bakeware. the cover letter requested that the survey be completed by the household member who does the most baking. Forty-seven of the questionnaires were returned as undeliverable by the post office; 291 completed questionnaires were received for a response rate of 32 percent. The demographic characteristics of the sample are reported in Table 1. The sample contained a high percentage of married female respondents, not surprising given the questionnaire topic. Respondents were also of a slightly higher educational and income level than the general population. The majority of respondents spent at least one hour per week baking, indicating that most respondents were relatively familiar with the product class.

Bakeware was chosen as an appropriate topic for the study. Since bakeware is not an expensive or complex product, it was expected that joint decision-making would not be common in the purchase of the product, simplifying data collection. Also, there are only a limited number of bakeware brand names on the market, simplifying measurement of the relevant variables.

Brand Loyalty Measure

In measuring the extent of brand preference for bakeware, four items were used and summed into an index. One combined responses to two questions, asking respondents whether there was a particular brand they preferred when purchasing bakeware and what their response would be if the store were out of stock for that brand when they were shopping for a bakeware product. This item is very similar to that used by Cunningham (1967) to assess brand preference. The remaining three items were Likert scales. Items and their scoring are shown in Table 2. Coefficient alpha for the four items is .83, showing that the index possesses acceptable internal consistency.

TABLE 1

DEMOGRAPHIC CHARACTERISTICS OF THE SAMPLE

TABLE 2

BRAND PREFERENCE MEASURE

Measures of Cognitive Structure

A number of measures have been developed to assess differentiation or dimensionality of cognitive structure. The most widely accepted is based on Kelly's (1955) Role Construct Repertory Test, sometimes referred to as the Reptest. A number of adaptations of Kelly's original measure have been published (e.g., Bieri et al. 1966; Klippel and Anderson 1974). The adaptation used for this study is similar to that reported by Seaman and Koenig (1974), which is strongly based on Bieri et al.'s work. This measure reflects both the number of dimensions used in evaluation and whether some of the dimensions are used similarly.

The study questionnaire listed three brand names of bakeware with which most respondents were expected to be familiar. Respondents who reported they were familiar with a brand were directed to complete six 7-point semantic differential items for each familiar brand. The particular semantic differential pairs used were: durable/not durable, gives good results/gives bad results, good quality/poor quality, economical/uneconomical, easy to use/not easy to use, and easy to clean/ not easy to clean. These pairs were chosen based on exploratory work in which adult consumers who regularly bake were asked to describe the important characteristics of bakeware.

Responses to the semantic differential items can be placed in a grid similar to that described by Bieri et al. (1966) and Seaman and Koenig (1974), with brands placed on,the horizontal axis and attributes on the vertical axis as shown in Figure 1. The dimensionality measure is derived by counting the number of tied ratings for each brand across the six attributes. Thus, ties are counted down and within each of the grid's brand columns, and total dimensionality score is the sum of the total number of ties within each of the three columns. This sum is actually the inverse of dimensionality (Bieri et al., 1966). A respondent who uses the six attributes identically for the three brands would generate a large number of tied ratings and would thus score high but be considered low in dimensionality. Scores were calculated only for respondents who were familiar with all three brands and who completed all semantic differential items. 109 respondents met this criterion.

FIGURE 1

SAMPLE GRID FOR A HYPOTHETICAL RESPONDENT

To test the second hypothesis, the extent of differences perceived among stimulus objects (brands) was measured in two ways, both shown in Figure 1. The first examined all possible attribute discriminations. For each attribute, a respondent's individual rating of each brand was compared with his or her ratings of the two other brands, making a total of three comparisons for each attribute or a total of 18 comparisons for the six attributes. The absolute values of the difference scores resulting from these comparisons were summed as a measure of total number of discriminations.

The second measure assessed discrimination at a more global level. Each respondent's rating of each of the three brands was summed across the six attributes. The total for each brand was then compared with the totals for the two other brands, resulting in three comparisons. The absolute values of the difference scores resulting from these comparisons were again summed and used as the second. global measure of discrimination.

The third hypothesis concerns the number of attributes consumers consider important when purchasing bakeware. One possible way to measure this is to present a number of attributes and have respondents indicate the importance placed on each. To avoid the demand characteristics of such ratings, in which respondents are likely to indicate that all attributes are important, an open-end question format was used instead. Respondents were asked to list the things they consider important when buying bakeware. The total number o. attributes listed was used as the dependent variable. While such a format may be subject to verbosity bias (Peterson 1982), such a bias was not deemed critical since it would probably serve to obscure relationships rather than influence their direction.

RESULTS AND IMPLICATIONS

Hypotheses were tested by correlating the dependent measures described above with the brand index. All hypotheses were supported at p <.01. The correlation between the brand preference measure and dimensionality was -.35 (p <.001). Since the dimensionality measure is an inverse indicator, H1 was supported, showing that shoppers with stronger brand preferences use more dimensions in evaluating products than consumers whose preferences are less emphatic.

With respect to H2, both measures of discrimination correlated in the hypothesized direction. The relationship i for the measure assessing total discriminations was stronger than that for the more global measure (r' =.36, p <.001, and .25, p <.01, respectively). Thus it is clear that brand committed consumers perceive more differences between brands than those who-are not committed.

Finally, the correlation between brand preference and the number of attributes deemed important was positive and significant (r=. 31, p <.001), supporting H3.

Taken together, these findings are strong evidence that consumers committed to a brand have different cognitive structures than those not so committed. Two of the three measures used concern cognitive complexity, with evidence indicating that the extent of defined brand preference is related to cognitive complexity. This raises a number of questions for consumer behavior scholars. For instance, the causal direction of the relationships described in this study is not known. Are individuals who are cognitively complex with respect to a product class more likely to develop brand preferences than less complex individuals? Or, on the other hand, does the complexity arise in response to the development of brand preferences? Perhaps a more tenable explanation, though as yet untested, is that familiarity with the product class leads to both brand preference and a more complex cognitive structure.

The relationship between complexity, preference, and the nature of the product class is also undetermined. This study investigated a single product class, and it is unclear whether the relationships discovered here would hold for different kinds of products. For instance, is the relationship between brand preference and cognitive complexity stronger when the product class in question is itself more complex, or is the relationship weaker?

The indication that the brand committed shopper is high in cognitive complexity has implications for management as well. Studies by a number of researchers (Bieri 1955; Crockett 1965; Klippel and Anderson 1974) have suggested that attitude change is more difficult to accomplish among individuals with higher cognitive complexity, possibly because these individuals are better able to absorb information discrepant from their original position. Also, since complex individuals process more attributes when evaluating brands, they may not be sensitive to changes in a single product attribute such as price or taste unless it is particularly important. These factors may serve to reduce the persuasibility of cognitively complex individuals. Thus, if preferences for competitive products is particularly strong, a marketing program designed to attract these consumers may have difficulty succeeding.

On the other hand, certain cognitive structures may lead to rapid changes in brand preferences. It is possible to envision the case where an individual may consider one or two product attributes so important that preference may easily switch among brands. For instance, a weight-conscious consumer may consider the attribute of caloric content to be so important that this individual's preference for a particular brand of light beer may instantly change when a competitive beer with fewer calories is introduced. The situations under which this and other changes in brand preferences are likely to occur need further investigation and specification.

REFERENCES

Bieri, James, Atkins, Alvin L, Briar, Scott, Leaman, Robert, L., Miller, Henry, and Tripoldi, Tony (1966), Clinical and Social Judgment: The Discrimination of Behavioral Information, New York: Wiley.

Chance, William A., and French, Norman D. (1972), "An Exploratory Investigation of Brand Switching," Journal of Marketing Research, 9, 226-229.

Cunningham, Scott M. (1967), "Perceived Risk and Brand Loyalty," in Risk Taking and Information Handling in Consumer Behavior, ed. Donald v. Cox, Boston: Harvard University, 507-523

Crockett, W. H. (1965), "Cognitive Complexity and Impression Formation," in Progress in Experimental Personality Research, Vol. 2, ed. B. A. Maher, New York: Academic Press, 47-90.

Day, George S. (1969), "A Two-Dimensional Concept of Brand Loyalty," Journal of Advertising Research, 9, 29-35.

Gutman, Jonathan (1980), "Equivalence Range in Categorization," in Advances in Consumer Research, Vol. 7, ed. Jerry C. Olson, Ann Arbor: Association for Consumer Research, 411-416.

Hirschman, Elizabeth C. (1981), "Cognitive Complexity and the Perception of Intangible Attributes," in Educator's Proceedings, ed. K. Bernhardt et al., Chicago: American Marketing Association, 158-161.

Hirschmans, Elizabeth C., and Douglas, Susan P. (1981), "Hierarchical Cognitive Content: Towards a Measurement Methodology," in Advances in Consumer Research, Vol. 8, ed. Kent B. Monroe, Ann Arbor: Association for Consumer Research, 100-105.

Kanwar, Rajesh, Olson, Jerry C., and Sims, Laura S. (1981), "Toward Conceptualizating and Measuring Cognitive Structures," in Advances in Consumer Research, Vol. 8, ed. Kent B. Monroe, Ann Arbor: Association for Consumer Research. 122-127.

Kasulis, Jack J., and Zaltman, Gerald (1977), "Message Reception and Cognitive Complexity," in Advances in Consumer Research, Vol. 4, ed. William D. Perrault, Atlanta: Association for Consumer Research, 93-97.

Kelly, G. A. (1955), The Psychology of Personal Constructs, Vol. 1, New York: Norton.

Klippel, R. Eugene, and Anderson, Robert (1974), "Cognitive Differentiation and Cognitive Unity: Two New Variables Used in the Prediction of Consumer Attitude Change," in Combined Proceedings, ed. R. C. Curhan, Chicago: American Marketing Association. 124-128.

Lastovicka, John L., and Gardner, David M. (1978), "Low Involvement Versus High Involvement Cognitive Structures," in Advances in Consumer Research, Vol. 5, ed. H. Keith Hunt, Ann Arbor: Association for Consumer Research, 87-92.

Lastovicka, John L., and Gardner, David M. (1979), "Components of Involvement," in Attitude Research Plays for High Stakes, eds. J. Maloney and B. Silverman, Chicago: American Marketing Association, 53-73.

Lutz, Richard J. (1975), "Changing Brand Attitudes Through Modification of Cognitive Structure," Journal of Consumer Research, 1, 49-59.

Lutz, Richard J., and Swasy, John L. (1977), "Integrating Cognitive Structure and Cognitive Response Approaches to Monitoring Communication Effects," in Advances in Consumer Research, Vol. 4, ed. William D. Perrault, Jr., Atlanta: Association for Consumer Research, 363-371.

Marks, Larry J., and Olson, Jerry C. (1981), "Toward a Cognitive Structure Conceptualization of Product Familiarity," in Advances in Consumer Research, Vol. 8, ed. Kent B. Monroe, Ann Arbor: Association for Consumer Research, 145-150.

Peterson, Robert A. (1982), Marketing Research, Plano, Texas: Business Publications

Robertson, Thomas S. (1976), "Low Commitment Consumer Behavior," Journal of Advertising Research, 16, 19-24.

Schroder, Harold M. Driver, Michael J., and Streufert, Siegfried (1967), Human Information Processing, New York: Holt, Rinehart. and Winston.

Scott, William A. (1962), "Cognitive Complexity and Cognitive Flexibility," Sociometry, 25, 405-414.

Scott, William A., (1966), "Brief Report: Measures of Cognitive Structure," Multivariate Behavioral Research, 1, 391-395.

Scott, William A., (1969), "Structure of Natural Cognitions," Journal of-Personality and Social Psychology, 12, 261.

Scott William A., Osgood, D. Wayne, and Peterson, Christopher, (1979), Cognitive Structure: Theory and Measurement of Individual Differences, New York: Wiley.

Seaman, Jerrol M., and Koenig, Frederick (1974), "A Comparison of Measures of Cognitive Complexity," Sociometry, 37, 375-390.

Tan, Chin Tiong, and Dolich, Ira J. (1981), "The Moderating Effects of Cognitive Complexity and Prior Product Familiarity on the Predictive Ability of Selected Multi-Attribute Choice Models for Three Consumer Products," in Advances in Consumer Research, Vol. 8, ed. Kent B. Monroe, Ann Arbor: Association for Consumer Research, 140-144.

Traylor, Mark B. (1981), "Product Involvement and Brand Commitment," Journal of Advertising Research, 21, 51-56.

Tyebjee, Tyzoon T. (1979), "Refinement of the Involvement Concept: An Advertising Planning Point of View," in Attitude Research Plays for High Stakes, eds. J. Maloney and B. Silverman, Chicago: American Marketing Association. 94-111.

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