Evoked Set Size and the Level of Information Processing in Product Comprehension and Choice Criteria

Frederick E. May, University of Missouri - St. Louis
Richard E. Homans, University of Missouri - St. Louis
ABSTRACT - The relationship between evoked set size and the level of information processing was studied for 111 new car buyers in a cross-section of St. Louis. The results show that the main effects of the level of information processing--choice criteria and product class comprehension are significant and additive.
[ to cite ]:
Frederick E. May and Richard E. Homans (1977) ,"Evoked Set Size and the Level of Information Processing in Product Comprehension and Choice Criteria", in NA - Advances in Consumer Research Volume 04, eds. William D. Perreault, Jr., Atlanta, GA : Association for Consumer Research, Pages: 172-175.

Advances in Consumer Research Volume 4, 1977   Pages 172-175

EVOKED SET SIZE AND THE LEVEL OF INFORMATION PROCESSING IN PRODUCT COMPREHENSION AND CHOICE CRITERIA

Frederick E. May, University of Missouri - St. Louis

Richard E. Homans, University of Missouri - St. Louis

[The collection of the data was supported by the Ford Foundation during Professor May's tenure as a Research Professor. This assistance is gratefully acknowledged.]

ABSTRACT -

The relationship between evoked set size and the level of information processing was studied for 111 new car buyers in a cross-section of St. Louis. The results show that the main effects of the level of information processing--choice criteria and product class comprehension are significant and additive.

INTRODUCTION

Research on the determinants of evoked set size suggests that individual differences in cognitive functioning or information processing are involved in the formation of evoked sets, (Campbell, 1969; Jarvis and Wilcox; 1973, Gr°nhaug, 1975; Ostlund, 1973). Jarvis & Wilcox (1973), show that the relative size of an individual's evoked set for one product is positively correlated with the size for other product classes. Evoked set size is positively correlated with education; that is, an indicator of the level of information processing, (Campbell, 1969, Gr°nhaug, 1975). Also, it has been shown that evoked set size is inversely correlated with brand loyalty, (an indicator of concrete information processing, (Campbell, 1969; Ostlund, 1973)), and that brand loyalty is inversely correlated with education across product categories (Frank, Douglas, and Polli, 1968).

In their Theory, Howard and Sheth (1968) suggest that a cognitive style may influence the size of the evoked set (p. 98). Cognitive style research has only recently become of intense interest to psychologists and consumer behavior researchers. It is now apparent that the cognitive style concept is a multidimensional concept.

COGNITIVE STYLES

May (1973) and Pinson (1975) recently reviewed the literature on cognitive styles that are of interest to consumer behavior and marketing research. Pinson (1975) classified cognitive styles by major categories: (1) information processing complexity, referring to the complexity of an individual's cognitive structures; and (2) accommodative cognitive styles, referring to an individual's controls for information and selectivity of perceptions.

Three aspects of information processing complexity have been identified as separate factors. They are: (1) differentiation (i.e., the number of dimensions used in processing) , (2) discrimination (i.e., the number of separated conceptual categories on a dimension), and (3) integration (i.e., the degree of inter-relatedness of elements within a particular cognitive domain).

Ostlund (1973) tested the suggestion by Howard and Sheth (1968, p. 98) that evoked set size is influenced by discrimination in information processing complexity. He used a frequently employed measure of discrimination; that is, Pettigrew's category width scale with a sample of sub-compact car purchasers. That study found no support for a relationship between evoked set size and discrimination.

However, Stiles (1972) work suggests that integrative complexity may influence evoked set size. He found that the number of offers considered by purchasing agents are correlated with the level of information processing, and that the level of information processing is also correlated with integrative information processing complexity measured by the Paragraph Completion Test (Schroder, Driver and Streufert, 1967).

THE LEVEL OF INFORMATION PROCESSING

The "level" refers to concrete-abstract performance in a task. "In a concrete response the specific object and its unique properties dominate, but in an abstract response it is relationships between objects or dimensions that cut across objects that dominate." (Vinacke, 1974, p. 179).

Concrete or simple information processing refers to a cognitive process that depends frequently on personal experience with objects, and that results in the identification of unique whole objects within a particular cognitive domain. In contrast abstract functioning refers to a process that results in the specification of one or more major aspects of objects. The aspects abstracted or "taken away from" objects are: (1) qualities or attributes; (2) functions or activities; and (3) relations among objects such as similarities and differences, part-whole; or reflexive relations (Alexander, 1967).

The concept of concrete-abstract cognitive functioning originally emerged from the work of Piaget (1928; Piaget and Inhelder, 1969). He found that the child's intelligence and cognitive functions develop through three main stages. These are preoperational, concrete operational, and abstract (formal) operational functioning. Adults can function at any level, but those whose intelligence or practice with abstract reasoning is less developed are more likely to function at concrete levels of information processing. The amount of education is, therefore, inversely related to the level of information processing.

The level of information processing in buyer behavior is hypothesized to be a function of both cognitive style, in particular integrative information processing complexity, and situational factors. For example, the complexity of the task, or the importance of the product, or social and organizational factors may also influence the level of information processing.

The purpose if this study is to measure the level of information processing in endogenous constructs of the Theory (Howard and Sheth, 1968). Theoretically Product Class Comprehension and Choice Criteria precede the formation of the evoked set in Brand Comprehension. All of these constructs are cognitive states or structures. The problem is to explain variations in the size of the evoked set by observing consistent differences in the cognitive states of product class comprehension and choice criteria.

Product Class Comprehension

"A cognitive state of the buyer that incorporates his ways of describing the common denotative characteristics of the set of brands constituting the product class and of evaluating those brands. Thus it includes choice criteria." (Howard and Sheth, 1968, p. 417). Howard and Sheth later point out (p. 417) that product class comprehension refers to the buyer's description of the denotative and connotative characteristics of the product class. The denotative characteristics are the set of attributes used to identify and describe the product class. Connotative characteristics refer to the set used to judge the brands, i.e., the choice criteria.

It is well known that industrial buyers use different methods of describing product quality. Lee and Dobler (1971, p. 48) found that 28 percent of items purchased are described by brand name, and most others are specified by product dimensions or commercial standards. Descriptions of product quality by brand name are concrete because they identify a particular object and its unique property in the cognitive structure of the product class. Specifications of attributes or commercial standards are abstract.

Choice Criteria

"A cognitive state of the buyer that reflects those attributes of the brands in the product class that are salient in the buyer's evaluation of a brand." (Howard and Sheth, 1969, p. 416). A buyer's choice criteria are concrete if he ascribes saliency to experience with the brand itself. All other criteria such as brand attributes are abstract.

The following hypotheses are tested:

1. Evoked set size will be directly related to the level of information processing in both the denotative and the connotative (i.e., choice criteria) aspects of product class comprehension.

2. The denotative and connotative effects will be additive. Evoked set size will be smallest when the level of information processing in both denotative and connotative aspects is concrete. Evoked set size will be greatest when the level in both determinates is abstract.

METHOD

The data for this study are from personal interviews with 387 household heads residing in the St. Louis metropolitan area. Selected for the analysis are those 111 household heads whose most recent purchase was a new car; who owned at least one car before the most recent purchase; whose brand choice was not restricted by special deals; and whose answers to the questions used were ascertained.

A standard cluster-area probability sampling procedure was used to select a sample of dwelling units in the metropolitan area. Interviews were conducted by the Public Opinion Survey Unit of the University of Missouri in Spring 1966. Up to four interviewer calls were made at each selected dwelling unit. The response rate of 77 percent of attempted interviews appears reasonable for a large metropolitan area and for a high proportion of nighttime interviews. Respondents were asked to recall and describe all cars owned and previously owned by themselves and their family. Next, each respondent was asked to describe the decision leading to his or her most recent purchase.

Measures

Evoked Set Size. If the respondent said that makes other than the one bought were considered during the purchase decision, a description of each such make was obtained. The respondent was asked to outline specific information gathering activities connected with his consideration of other makes, and to give reasons for his decision not to buy the make.

Evoked set size was then determined by counting only those other makes about which a substantial portion of information was given by the respondents. [Different models within the same family brand were counted as being one make. Therefore, the respondent who said he considered a Chevrolet Impala and a Chevrolet Corvair was counted as having an evoked set size of one.] Thus evoked set sizes in this study of new automobile purchasers may be somewhat smaller on average, than in other studies of new automobile buyer's evoked sets.

Product Class Comprehension. The following questions were used to categorize the information processing level for the denotative characteristic in product class comprehension:

When you first decided to buy, did you have some specific ideas about what type of car you wanted? (IF YES) What were your ideas?

Respondents who mentioned the name of the make bought, or series of the make bought, and only that make, were categorized as concrete. Respondents who mentioned attributes of the type of car wanted, for example; big, small, heavy, economical, were categorized as abstract if they did not also mention the make purchased. If a respondent mentioned more than one make, making a comparison, the response was categorized as abstract.

Functions, such as for city driving, for hauling, for highway driving, for parking, were also mentioned by some respondents. These respondents were originally categorized separately, because they operate on a higher level of information processing. That is, they describe two aspects of abstracting rather than one. They also appear to have larger evoked set sizes than those who only discuss attributes. However, inter-coder reliability for categorizing two distinct abstract levels of information processing was somewhat below the 85 percent agreement level that is generally required. Therefore, respondents who mentioned functions as well as attributes were included in the same category as those who only mentioned attributes.

Choice Criteria. The following questions were used to categorize the information processing level in the use of choice criteria.

Can you give me more details about what impressed you and made you buy this car?

Can you give more details about what was important or meaningful in your decision to buy this make of car?

Summing it up now - - what do you think was most important in causing you to decide on the make bought?

Respondents who mentioned the name of the make bought, and only that make, or who mentioned previous experience with the make or dealer were categorized as abstract.

Coding reliability. Essentially coders were instructed to pick out responses that included mention of the name of a make or series of car or a reference to previous experience. This task is simple; inter-coder agreement was greater than 85 percent for each measure.

Analysis

The hypotheses were tested by a two-way analysis of variance using the SPSS subprogram ANOVA (Nie and others, 1975). Because of unequal cell sizes, the reduction method using the general linear model (Perreault and Darden, 1975) was employed to first test the effect of the denotative characteristics of product class comprehension and then to test the effects of choice criteria.

RESULTS

Table 1 shows that mean evoked set size and the level of information processing vary as predicted by the hypotheses. Information processors who operate at the concrete level report smaller evoked sets when compared to information processors who operate at the abstract level. The smallest evoked sets are reported by respondents whose level of information processing was concrete in their description of the product class and whose choice criteria include the brand itself as a salient criterion. The largest evoked sets are reported by abstract processors of information in both the denotative and connotative (i.e., choice criteria) aspects of product comprehension.

TABLE 1

MEAN EVOKED SET SIZE AND LEVEL OF INFORMATION PROCESSING

Table 2 shows the results of the analysis of variance.

The effects were tested in the following order:

1) Denotative characteristics of product class comprehension.

2) Choice criteria, eliminating the denotative characteristics of product class comprehension.

3) Interaction, eliminating both main effects as stated above.

TABLE 2

ANALYSIS OF VARIANCE

The results of the ANOVA confirm the hypothese stated earlier. Both main effects are significant, and, since the interaction effect is negligible, the model can be assumed to be additive.

DISCUSSION

The reported study supports the hypothesis that individual differences in cognitive states of product class comprehension (denotative and connotative characteristics) correlate with the size of reported evoked sets. It should be noted that the study is based on post-decisional reports of a cross-section of buyers. A longitudinal design is more valid for causal attribution.

If one assumes, however, that the measures are reasonably accurate reflections of the cognitive states of the respondents at the time of interview, one can conclude that indications of concrete information processing are associated with recall of smaller evoked set sizes.

The study results have implications for the Theory of Buyer Behavior. The definitions of the central constructs with which this study is concerned anticipates abstract information processing, but not concrete information processing. This study finds that the proportion of concrete information processing in product class comprehension is very similar to that found by Lee and Dobler (1971) in their survey of industrial buyers. About one-third of the buyers in both surveys identify a brand as the product type wanted. A small survey of housewives who were asked about the type of laundry detergent that they wanted also found that about a third of the respondents identified a brand. A substantial proportion, about 26 percent, of new car buyers processed information concretely in product comprehension (denotative and connotative characteristics).

This study suggests further exploration of the relationship between the level of information processing, brand loyalty, and education. Although it must be recognized that much of the variance in brand loyalty across product classes can be explained by market structure and situational factors, it appears that there is a component of personality influencing repeat brand purchasing behavior, that is mediated by the level of information processing.

REFERENCES

Hubert G. Alexander, Language and Thinking, New York; D. Van Nostrand Company, 1967.

R. Darell Bock, and Ernest A. Haggard, "The Use of Multivariate Analysis of Variance in Behavioral Research", in Dean A. Whitla (Ed.), Handbook of Measurement and Assessment in Behavioral Sciences, Reading, Mass.: Addison-Wesley Pub. Co., 1968.

Brian M. Campbell, "The Existence of Evoked Set and Determinants of Its Magnitude in Brand Choice Behavior," unpublished Doctoral Dissertation, Columbia University, 1969.

Ronald E. Frank, Susan P. Douglas, and Rolando E. Polli, "Household Correlates of 'Brand Loyalty' for Grocery Products," Journal of Business, 41, (April, 1968) pp. 237-245.

Kjell Gr°nhaug, "Education and Buyer Behavior," Acta Sociologica, 17 (1974), 179-189.

John H. Howard, and Jagdish N. Sheth, The Theory of Buyer Behavior, New York: John Wiley and Sons, Inc., 1969.

Lance P. Jarvis, and James B. Wilcox, "Evoked Set Size--Some Theoretical Foundations and Empirical Evidence," Combined Proceedings. Conferences of the American Marketing Association, 1973, 236-240.

Lamar Lee, Jr., and Donald W. Dobler, Purchasing and Materials Management, 2nd Ed., New York: McGraw-Hill, 1971.

Frederick E. May., "Marketing Strategies and Cognitive Styles," Combined Proceedings. Conferences of the American Marketing Association, 1973, 194-197.

Norman H. Nie, and others, Statistical Package for the Social Sciences, 2nd Ed., New York: McGraw-Hill Book Co. 1975..

Lyman E. Ostlund, "Evoked Set Size: Some Empirical Results," Combined Proceedings. Conferences of the American Marketing Association, 1973, 226-230.

William D. Perreault, Jr., and William R. Darden, "Unequal Cell Sizes in Marketing Experiments: Use of the General Linear Hypothesis, "Journal of Marketing Research, Vol.12 (August, 1975), pp. 333-342.

Jean Piaget, Judgment Reasoning in the Child, London: Routledge & Kegan Paul, Ltd., 1928 (Reprinted 1962).

Jean Piaget, and Baerbel Inhelder, The Psychology of the Child, New York: Basic Book, Inc., 1969.

Christian Pinson, "Consumer Cognitive Styles, Review and Implications for Marketers," Research Paper No. 160-The European Institute of Business Administration, 1975.

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

Gerald W. Stiles, "An Information Processing Model of Industrial Purchasing Behavior," unpublished doctoral dissertation, University of Minnesota, 1972.

William E. Vinacke, The Psychology of Thinking, New York McGraw-Hill, 1974.

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