An Information Theoretic Approach to Understanding the Consideration Set/Awareness Set Proportion

ABSTRACT - This paper investigates the consideration set/awareness set proportion and proposes an information theoretic model which identifies a 63:37 proportion as "optimal". This proportion may lead to efficiencies in information processing and has been found in various types of evaluative judgments in prior research. This information theoretic model is applied to two product categories and results support the hypothesized relationships. The depth with which consumers process product category information is identified as a possible boundary condition for this phenomenon.


Ayn E. Crowley and John H. Williams (1991) ,"An Information Theoretic Approach to Understanding the Consideration Set/Awareness Set Proportion", in NA - Advances in Consumer Research Volume 18, eds. Rebecca H. Holman and Michael R. Solomon, Provo, UT : Association for Consumer Research, Pages: 780-787.

Advances in Consumer Research Volume 18, 1991      Pages 780-787


Ayn E. Crowley, Washington State University

John H. Williams, Texas Southern University


This paper investigates the consideration set/awareness set proportion and proposes an information theoretic model which identifies a 63:37 proportion as "optimal". This proportion may lead to efficiencies in information processing and has been found in various types of evaluative judgments in prior research. This information theoretic model is applied to two product categories and results support the hypothesized relationships. The depth with which consumers process product category information is identified as a possible boundary condition for this phenomenon.


Before purchasing a product, consumers must first be aware of the product and believe that the product will meet their needs. Indeed, this act of purchase is often conceptualized as a process (cf. Nicosia 1966, Howard and Sheth 1969). In the marketing literature the concepts of awareness set and consideration set are used to help describe one of the basic processes which leads to purchase of a product (Nedungadi 1987). Consumers may eliminate some brands which they are aware of from further consideration in order to reduce the complexity of the decision process.

Marketers attempt to gain entry into the consumer's awareness set through promotional efforts. Many steps are also taken to help ensure that consumers will perceive the product positively and will include the product among those considered for purchase.

The size of the awareness set is fairly large for some product categories. For example, Campbell (1969) reported a mean awareness set size of 15.2 brands of dishwashing liquid. For cake mixes, Jacoby and Olson (1970) found an average of 7.8 brands in the awareness set. The consideration set is a subset of the awareness set, consisting of brands the consumer considers relevant for the purchase occasion (Alba and Chattopadhyay 1985). This approach to consumer decision making is often described as a simplifying heuristic (Hauser and Wernerfelt 1990), or "funneling" process (Howard and Sheth 1969). Several researchers believe that there may be an upper bound on the size of the consideration set (e.g., All a and Hutchinson 1987).

The relative sizes of the awareness and consideration sets have been discussed in the marketing literature (Brown and Wildt 1987, Campbell 1969, Jarvis and Wilcox 1973, Narayana and Markin 1975, Ostlund 1973). Yet, the consideration set/awareness set ratio has not been examined. In particular, the proportion of brands in the awareness set which are included in the consideration set has received little investigation. In the present paper, an information theoretic model is used to predict the tendency of this particular proportion. The consideration set is viewed in this context as the brands in the awareness set that are evaluated positively. Those brands which the consumer is aware of but-does not include in the consideration set are viewed as negatively evaluated brands.

Information theory suggests a positive:negative proportion which may offer several information processing efficiencies. This proportion has been found repeatedly in empirical tests of positive:negative evaluations of persons (Benjafield and Adams-Webber 1976, Rigdon and Epting 1982), words (Adams-Webber 1978), and objects (Crowley 1990). The major contribution of this paper is the identification of a consistent proportional tendency between positive and negative brand evaluations within a product category.

The inclusion of the concept of a consideration set in quantitative models has been shown to improve the predictive ability of several of these models (e.g., Hauser and Gaskins 1984, Louviere 1988, Silk and Urban 1978). The consideration set, however, is very dynamic and may vary from one point in time to another due to the influence of internally and externally generated retrieval cues f(Nedungadi 1987). Also, the measurements of consideration set size and composition may be susceptible to the influence of methodological factors that may produce results that are an artifact of the methods employed rather than valid measures of the consideration set.

The awareness set, on the other hand, is less susceptible to influence from these factors. Therefore, if it can be established that there is a relatively stable relationship between the sizes of the awareness and consideration sets, more reliable estimates of the size of the consideration set may be available to researchers and marketing practitioners. Similarly, the proportion could be considered as the probability that a new product will achieve inclusion in the consumer's consideration set.

In the present study, it is postulated that the information theoretic proportion (i.e., 63:37) will be found in the consideration set/awareness set relationship. The information theoretic rationale for the hypothesized relationship is described below. This is followed by a brief summary of the related psychological literature and a discussion of the possible boundary conditions for the phenomenon. Related findings from the marketing literature are then discussed. Subsequently, a study used to investigate the hypothesized relationships is reported.


"Information" is an important concept in many disciplines including psychology, communication, and marketing. Information theory offers an approach to quantifying the amount of information contained in a message. Information theoretic concepts were developed within the field of communication by Shannon and Weaver (1949) and Weiner (1948). This research, along with the work of Miller (1956), is now viewed as the precursor of the cognitive approach within psychology.

The concept of uncertainty (often termed entropy) is central to understanding "information" in this quantitative context, as will be explained in detail below. Uncertainty is quite analogous to "surprise". In information theory, the reduction in uncertainty is quantified as information.

The amount of information received, on the average, is dependent upon two factors. First, the number of possible alternative symbols (n) which can be "sent" affects average information. Secondly, the probability (Pi; i=1 to n) associated with each alternative symbol being sent is used to calculate average information. With a given number of alternative symbols, the information quantity is maximized when the alternatives are equiprobable (Young 1971).

In communication theory, information is usually measured in bits, or binary digits. If the alternative messages are equiprobable, the amount of information is:

I = log2n

Often, the alternatives are not equiprobable. The general formula used to calculate entropy is (cf. Gamer 1962, Pierce 1961):



H = average information per symbol sent by a source (expressed in bits).

Pi = the proportion of total symbols sent by a source which are symbol i specifically.

n = the number of possible alternative symbols which the source can send.

The behavior of the information theoretic function over various values of Pi for a dichotomy is illustrated in Figure 1.

Several aspects of the application of the entropy formula to psychological processes are discussed by Berlyne (1971). Berlyne conceptualizes the contribution to average information as an index of the strikingness or "salience" of the information. This measure of "strikingness" is postulated to reflect the psychological impact of an element or signal. Therefore, a low probability event could carry a great deal of information, as would be reflected in the events' contribution to the H function.

To understand the information theoretic predictions described by Berlyne (1971), it is necessary to examine the behavior of part of the entropy formula. In the present application, this would represent one of the two possible affective categories (positive and negative). The form of the function for the contribution of one category to average information is depicted in Figure 2. As noted by Berlyne (1971, p. 231), the maximum contribution to average information occurs at about .37 or 1/e. This is the point at which the minor category represents about 37 percent of the total. At this point, the contribution to total information is .53. When 1-p = .63 (i.e., 1-.37), the value of (1-p)log2(1/(1-p)) is .43. Thus, average information per evaluation in this case is .53 + .43 or .96 bits. At the 63:37 proportion, the minor category contributes the.maximum possible amount to total information. In essence, there is maximum "contrast" between the minor category (or foreground) against the background of the major category.

The average information obtained at the 63:37 proportion is about 4 percent less (i.e., 1.00.96) than the maximum possible information obtained in the equiprobable case. Thus, the decrement in information which would be obtained from evaluations is very small compared to the gains, such as increased salience for one category of information.

Based on the rationale described above, it is posited that the human affective system would gradually evolve through natural selection toward the information theoretic processing proportion. This proportion may contribute to an organism's chances of survival due to the efficiency it contributes to affective processing. For example, maximizing the "strikingness" of negative information (Berlyne 1971) may imply that this information can be processed foster. A speed of processing advantage of even a fraction of a second could lead to evolutionary tendencies toward the information theoretic ratio for evaluative judgments. If this rationale is correct, the information theoretic phenomenon should be a robust one. Indeed, empirical evidence suggests that this processing proportion has been found in a variety of psychological contexts, as described subsequently.


Evaluative Judgments

The 63:37 information theoretic proportion has been found in human affective functioning, and in evaluative judgments specifically. The term "golden section proportion" is often used in the psychology literature, reflecting the special geometric properties of a related 62:38 proportion (Crowley 1990). Due to the large amount of prior research on this topic all relevant findings cannot be presented. Thus, only those findings which directly impact hypothesis development are reported.





The first known psychological application of this information theoretic concept occurred in the field of experimental aesthetics. Several investigators (see Huntley 1970) have found that subjects tend to prefer rectangles based on the proportion. For example, McManus (1980) found evidence for preference for rectangles based on proportions near 63:37.

Research by Frank (1959, 1964) and Berlyne (1971) has been especially important in applying the information theoretic proportion in experimental aesthetics. Frank (cited in Berlyne 1971) performed experiments in which subjects were asked to select squares from a collection offering many squares in several colors and to arrange these squares so as to make one color as "striking" as possible. The subjects used the "striking" color, on the average, between 37 and 38 percent of the time, using various other colors to form the "background" for the striking color.

More recently, Benjafield and Adams-Webber (1976) reported data from five experiments in which subjects made dichotomous judgments of acquaintances (e.g., pleasant:unpleasant). Subjects were asked to rate the acquaintances on several semantic differential dimensions. The hypothesis tested in each study was that, when subjects make dichotomous judgments about a series of acquaintances in terms of bipolar dimensions, they will tend to make these judgments in proportions which approximate the information theoretic/golden section ratio,-using positive adjectives 62 percent of the time and negative adjectives 38 percent of the time. The aggregate level results of each of these five studies were within 0.01 of the hypothesized proportion.

An important extension of the information theoretic/golden section hypothesis resulted from the work of Rigdon and Epting (1982). In their experiment, subjects were asked to generate their own construct dimensions, rather than using a set of provided dimensions. Rating acquaintances on these elicited constructs, the proportion of positive judgments was .63, which did not differ significantly from the predicted proportion of .62.

Perhaps most importantly, Rigdon and Epting asked subjects to rate the usefulness of the constructs in terms of "how useful you feel each construct-contrast pair is to you in accurately describing and characterizing people." Their hypothesis that subjects would rate the constructs which they had used in the information theoretic/golden section proportion as more useful than constructs used in other positive/negative proportions was supported: subjects rated constructs which they had used in this proportion as more useful in describing and characterizing people.

In summary, evidence has been accumulated which indicates that some human attitude structures are constructed in approximately the 63:37 proportion. Information theory provides a mathematical rationale indicating that this particular proportion may offer processing efficiencies by maximizing the contribution of the minor category to total information.

A Test of the Information Theoretic Hypothesis in a Marketing Context

One study has examined proportional relationships in consumers' perceptions of a marketing stimulus (Crowley 1990). Subjects in this study rated a retail store environment on twenty semantic differential scales. Responses at the midpoint of these scales were considered affectively neutral and were eliminated from the analysis.

Results of the Crowley (1990) study indicated that subjects' ratings of the store environment followed a 63:37 positive:negative proportion, with the major category being negative evaluations. In addition, responses were factor analyzed and the 63:37 proportion was found within each major factor comprising the overall judgments of the stimulus. The existence of these relationships within these orthogonal factors comprising more "global" judgments may be considered a form of "nesting" of these proportions.


Evidence supporting the information theoretic proportion in evaluative judgments is far from conclusive. Yet, the evidence is strong enough to warrant further research and application of this concept. The broadest implication of the research summarized m this brief review of the literature is that the proportion may be operative in many types of affective judgments. This may include affective judgments made in a marketing context. Affective processes have been the subject of increasing research interest in marketing. Understanding of the processes has added richness and validity to models of consumer decision making. The information theoretic proportion, if it is found to apply to marketing-related affective judgments, can represent a stride forward in gaining deeper insight into the processes underlying these judgments.

Awareness and Consideration Set Findings

The awareness set for a particular product class can be viewed as a cognitive structure held in long-term memory. This structure consists of brand names, information about brands, attributes, and perhaps even decision criteria for evaluating the brands. When a brand choice decision is made, the individual obtains brand information from the external environment and/or retrieves information from long-term memory. Both positively and negatively perceived brands along with neutral brands may be considered by the consumer. The size and composition of the set of brands considered by the individual may be influenced by the accessibility of brand information contained in the cognitive structure for the product class and the particular retrieval cues available at the time the brand choice decision is being made.

Howard and Sheth (1969) hypothesized that the number of brands of a product that a consumer considers for purchase bears some relationship to the number of brands that s(he) is aware of. Several consumer researchers have examined this relationship. The results of these studies are summarized in Table 1. The mean size of the awareness sets for the various products ranged from 3.5 for mouthwash to 19.3 for laundry detergent. The mean size of the consideration sets ranged from 1.3 for mouthwash to 5.6 for dishwashing liquid. The ratio of the mean consideration set size to the mean awareness set size ranged from .27 for laundry detergent to .64 for table napkins with a mean proportion of .39 across all product categories. Variation in consideration set proportions may have resulted from product category and research design differences.



The sizes of both the awareness and consideration sets were larger when aided rather than unaided recall was used. However, the mean size of the consideration set relative to the mean size of the awareness set was not significantly different for the two approaches. The average consideration set proportion was .40 when aided recall was used and .39 when unaided recall was used.


As described above, it is hypothesized that the 63:37 proportion will be found in the consideration set/awareness set relationship within product categories. This is reflected in Hypothesis 1A:

H1A: Among the brands a consumer is aware of, the consumer will tend to include either 37 percent or 63 percent of these brands in the consideration set.

This hypothesis is based upon the fact that these product judgments are evaluative, and may be governed by the information processing bias found in several psychological studies described previously.

Further, it is hypothesized that the 63 percent versus 37 percent distinction will be a function of the size of the awareness set. This view posits that the consumer is often an "efficient" information processor who will only want to consider a few brands in depth (Bettman 1986). Thus, the consumer who is aware of many brands in a particular product category may consider a relatively small proportion of this plethora of brands. This could reflect a tradeoff between decision making effort (costs) and the desire to make an optimal decision, as described by Hauser and Wernerfelt (1990). Conversely, the consumer is likely to be aware of only a few brands in certain product categories. We might expect a larger proportion of these brands to be considered, as this would still imply a relatively small consideration set. Hypothesis 1B also reflects the idea that there are limits on the amount of information that can be processed at any given time (Miller 1956). This tendency is also reflected by direct empirical evidence regarding typical sizes of consideration sets (see Hauser and Wernerfelt 1990).

H1B: The consideration set/awareness set ratio will be a function of the size of the awareness set. With a larger awareness set, there will be a tendency for 37 percent of the brands to be included in the consideration set. With a smaller awareness set, there will be a tendency for 63 percent of the brands to be included in the consideration set.

Hypothesis 2 represents an attempt to investigate the boundary conditions of the information theoretic proportion. A review of related studies in the psychology literature indicated that the 63:37 proportion was found when subjects gave in-depth consideration to categories such as acquaintances. This concept is quite analogous to the degree of analysis described by Alba and Hutchinson (1987). Degree of analysis is a continuum and "refers to the extent to which consumers access all and only the information that is relevant and/or important for a particular task" (Alba and Hutchinson 1987, p. 417). With more analytic or "in-depth" processing, information search is effortful and extends beyond the most accessible information.

Hypothesis 2 is based upon the rationale that the 63:37 proportion is more likely to be found when subjects access their product category knowledge and affective structure more completely. Thus, cues will be presented to some subjects prior to the experimental task which may facilitate retrieval of this information. As described by Biehal and Chakravarti (1983, p. 4), "The retrieval of information from memory is known to be cue-dependent, i.e., the cues available at retrieval influence whether or not previously stored information can be accessed from memory."

In the present study, some subjects will consider the product category in-depth, while others will not. This will allow an initial glimpse at a possible boundary condition for the information theoretic proportion. Hypothesis 2 addresses the question: Is in-depth processing a necessary condition for finding the 63:37 proportion in attitude structures?

H2: The 63:37 relationship will only be found when subjects engage in in-depth processing of product category information.



A total of 139 juniors and seniors from an upper-level business class participated in the study. These subjects received extra credit in return for their participation.

Independent Variable

The consideration set/awareness set ratio was examined in this study for two products: automobiles and television sets. Automobiles were selected to represent a product category for which consumers' awareness sets would tend to be relatively large, while it was expected that awareness sets for TVs would be relatively small.

The independent variable was designed to manipulate the depth with which subjects processed their knowledge of the product category. Before completing the dependent measures (awareness set and consideration set size), subjects completed detailed questionnaires about a product category. In this initial task, subjects were asked to rate the importance of several attributes of a product class in terms of the attributes' influence on their decision about which brand to buy. Attribute importance was assessed with a five-point scale anchored by "extremely important influence" and "not an important influence". Next, subjects were asked to rate each attribute in terms of perceived differences among brands using a five-point scale anchored by "extreme differences" and "no differences" (Alpert 1971).

Dependent Variables

The attribute questionnaires were followed by a page which asked, "Please list every (product) manufacturer you are aware of" and, "Of those brands you listed above, which would you consider purchasing if you were going to buy a (product)?" The number of manufacturers listed by the subject represented the dependent measures.


The product category asked about in the initial questionnaires was either the same as that used for the consideration set/awareness set questioning ("match") or represented a different product category ("mismatch"). For example, a subject in the "match" condition of the automobile experiment would be asked questions about automobile attributes, and subsequently asked to list every automobile manufacturer s(he) could think of. In the "mismatch" condition, subjects were asked attribute questions about an unrelated product category (radio/alarm clocks).

Subjects completed the questionnaires for the study in a classroom environment at their own pace. They were told that their input was needed for developing advertisements for a research study. Subjects were debriefed during a later class period.




The mean consideration set/awareness set proportions for the four groups of subjects are shown in Table 2. The matched conditions did not differ significantly from the hypothesized proportion of .37 for automobiles (t = .32) and .63 for TVs (t = -.81). There was a significant difference between the information theoretic proportion and the actual proportion in the unmatched conditions for both the automobile (t = -2.52) and the television (t = -2.93). These results provide support for Hypotheses 1A and 2.

Hypothesis 1B posited that the 63 percent versus 37 tendency would be a function of awareness set size. The mean awareness set size for automobile manufacturers was 15.23 (matched) and 17.86 (unmatched); for television sets the mean was 6.54 (matched) and 6.88 (unmatched). Thus, Hypothesis 1B was also supported.

These results may indicate that consumers are "efficient" information processors in terms of the information theoretic model described in this paper. When information is processed in sufficient depth, the consideration set/awareness set relationship was found to approximate the 63:37 information theoretic or "golden section" proportion. The division of the awareness set into brands which are considered vs. not considered for purchase is viewed in this context as a division of brands into "positive affect" and "negative affect" subsets. According to the information theoretic model, this proportional division may offer efficiency in terms of speed of processing or maximum "contrast" between these two subsets of the awareness set.

The present study represents a first step in applying this information theoretic model to purchase decisions. This study is limited in the sense that only two product categories were examined among a relatively small number of subjects. Because of this, the depth of processing, or "degree of analysis" is confounded with product class in the research design. A more extensive study utilizing a variety of product classes could provide more conclusive results.

If the 63:37 tendency is indeed operative in consumers' consideration set/awareness set proportions, several additional avenues for research are suggested. For example, the consumer's degree of expertise regarding a product category may be reflected in a relatively large awareness set. Does the consideration set expand correspondingly? Future research could also investigate these proportional relationships for the brand loyal versus variety seeking consumer within various product categories. Studies such as these would help to clarify the boundary conditions for the 63:37 information theoretic proportion.

In a broader context, the proportion may have implications for many other aspects of consumer behavior studied under the current information-processing paradigm. Examples of key areas of marketing inquiry in which related research may prove fruitful include information search, product pricing and positioning, consumer satisfaction, package design, inference formation, categorization phenomena and category-based processing, and attitude change.


Adams-Webber, J. (1978), "A Further Test of the Golden Section Hypothesis," British Journal of Psychology, 69, 439-442.

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

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

Alpert, Mark I. (1971), "Identification of Determinant Attributes: A Comparison of Methods," Journal of Marketing Research, 8 (May), 184-191.

Benjafield, John and Adams-Webber, J. (1976), 'The Golden Section Hypothesis," British Journal of Psychology, 1, 11-15.

Berlyne, D. E. (1971), Aesthetics and Psychobiology, New York: Meredith Corporation.

Bettman, James R. (1986), "Consumer Psychology," Annual Review of Psychology, 37, 257-289.

Biehal, Gabriel and Chakravarti, Dipankar (1983), ''Information Accessibility as a Moderator of Consumer Choice," Journal of Consumer Research, 10 (June), 1-14.

Brown, Juanita J. and Wildt, Albert R. (1987), "Factors Influencing Evoked Set Size", working paper, University of Missouri-Columbia.

Campbell, Brian M. (1969), The Existence of Evoked Set and Determinants of Its Magnitude in Brand Choice Behavior, unpublished Ph.D. Dissertation, Columbia University.

Church! Nancy J., Laroche, Michel, and Rosenblatt, Jerry A. (1985), "Consumer Brand Categorization for Durables with Limited Problem Solving: An Empirical Test and Proposed Extension of the Brisoux-Laroche Model, Journal of Economic Psychology, 6, 231-253.

Crowley, Ayn E. (1990), "The Golden Section: The First Universal Law in Marketing?" Psychology and Marketing, in press.

Frank, H. (1959), Grundlagunprobleme der Informationsasthetik und erste Anwendung auf die mime pure, Quickborn: Schnelle.

Frank, H. (1964), Kybernetische Analysen sebjektiver Scackverhalte, Quickborne: Schnelle.

Garner, Wendell R. (1962), Uncertainty and Structure as Psychological Concepts, New York: John Wiley and Sons.

Hauser, John R. and Gaskin, Steven P. (1984), "Applications of the 'Defender' Consumer Model," Marketing Science, 3 (Fall), 327-351.

Hauser, John R. and Wernerfelt, Birger (1990), "An Evaluation Cost Model of Consideration Sets," Journal of Consumer Research, 16 (March), 393408.

Howard, John A. and Sheth, Jagdish N. (1969), The Theory of Buyer Behavior, New York: Wiley.

Huntley, E. H. (1970), The Divine Proportion, New York: Dover.

Jacoby, J. and Olson, J. C. (1970), "An Attitudinal Model of Brand Loyalty: Conceptual Underpinnings and Instrumentation Research," paper presented at the Conference on Attitude Research and Consumer Behavior (Urbana: University of Illinois, December 5-6, 1970).

Jarvis, Lance P. and Wilcox, James B. (1973), "Evoked Set Size - Some Theoretical Foundations and Empirical Evidence," Combined Proceedings, Thomas V. Greer, ed., Chicago: American Marketing Association, No. 35, 236-240.

Louviere, Jordan J. (1988), "On the Consequences of Mis-specifying Consumers' Choice Sets in Multinomial Logit Choice Models," working paper, University of Alberta, Edmonton, Alberta, Canada T6G 1S2.

McManus, I. C. (1980), 'The Aesthetics of Simple Figures," British Journal of Psychology, 71, 505-524.

Miller, George A. (1956), "The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information," The Psychological Review, 63 (March), 81-97.

Narayana, Chem and Markin, Rom J. (1975), "Consumer Behavior and Product Performance: An Alternative Conceptualization," Journal of Marketing, 39, 1-6.

Nedungadi Prakash (1987), Formation and Use of A Consideration Set: Implications for Marketing and Research on Consumer Choice, unpublished Ph.D. Dissertation, University of Florida.

Nicosia, Franceso M. (1966), Consumer Decision Processes: Marketing and Advertising Implications, Englewood Cliffs, NJ: Prentice Hall.

Ostlund, Lyman E. (1973), "Evoked Set Size: Some Empirical Results," in Thomas V. Greer, ed., Combined Proceedings, Chicago: American Marketing Association, No. 35, 226-230.

Pierce, John R. (1961), Symbols, Signals and Noise: The-Nature and Process of Communication, New York: Harper.

Prasad, V. Kanti (1975), "Evoked Set Size Personality Correlates and Mediating Variables," Journal of Academy of Marketing Science, 3, 272-279.

Rigdon, Michael A. and Epting, Franz R. (1982), "A Test of the Golden Section Hypothesis With Elicited Constructs," Journal of Personality and Social Psychology, 43, 1080-1087.

Shannon, C. E. and Weaver W. (1949), The Mathematical Theory of Communication, Urbana: University of Illinois Press.

Silk, Alvin J. and Urban, Glen L. (1978), "Pre-Test Market Evaluation of New Packaged Goods: A Model and Measurement Methodology," Journal of Marketing Research, 15 (May), 171-191.

Weiner, N. (1948), Cybernetics, New York: Wiley.

Young, John F. (1971), Information Theory, New York: Wiley.



Ayn E. Crowley, Washington State University
John H. Williams, Texas Southern University


NA - Advances in Consumer Research Volume 18 | 1991

Share Proceeding

Featured papers

See More


L3. Categorizing Engagement Behaviors from the Perspective of Customer Resources

Xianfang Zeng, University of Calgary, Canada
James Agarwal, University of Calgary, Canada
Mehdi Mourali, University of Calgary, Canada

Read More


The Victory Effect: Is First-Place Seeking Stronger than Last-Place Aversion?

David Hardisty, University of British Columbia, Canada
Steven Shechter, University of British Columbia, Canada

Read More


L11. Consumer Search Mode Produces Unintended Marketing Consequences

Dan King, University of Texas Rio Grande Valley, USA
Sumitra Auschaitrakul, University of the Thai Chamber of Commerce

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.