The Role of Memory in Understanding Advertising Media Effectiveness: the Effect of Imagery on Consumer Decision Making

Jolita Kisielius, University of Michigan
ABSTRACT - A new approach is suggested for understanding the nature of the differences between the advertising media. The information processing view of the media that is presented in this paper suggests that media differences can be examined in terms of the differential processing capabilities of their presentation formats of pictures and sentences, which vary in imagery level. The availability - valence hypothesis is proposed as a means of identifying and explaining the learning and evaluative effects of advertisements differing in imagery.
[ to cite ]:
Jolita Kisielius (1982) ,"The Role of Memory in Understanding Advertising Media Effectiveness: the Effect of Imagery on Consumer Decision Making", in NA - Advances in Consumer Research Volume 09, eds. Andrew Mitchell, Ann Abor, MI : Association for Consumer Research, Pages: 183-186.

Advances in Consumer Research Volume 9, 1982      Pages 183-186


Jolita Kisielius, University of Michigan


A new approach is suggested for understanding the nature of the differences between the advertising media. The information processing view of the media that is presented in this paper suggests that media differences can be examined in terms of the differential processing capabilities of their presentation formats of pictures and sentences, which vary in imagery level. The availability - valence hypothesis is proposed as a means of identifying and explaining the learning and evaluative effects of advertisements differing in imagery.


The issue of media effectiveness is of major importance to firms and consumers. Firms spend over 40 billion dollars annually for time and space in broadcast and print media. They also allocate significant human resources to media selection and the associated activities of media buying and monitoring. Furthermore, because the viability of the media depends on advertising revenues, advertisers' media choices affect the entertainment and information made available to consumers.

To address the media selection decision, practitioners typically employ a two step approach. Initially, media are identified that yield efficient exposure to some target audience given reach, frequency, and budget constraints. Then selection among these media is based on their impact as determined by expert judgment. Impact is assessed by examining such factors as the suitability of the editorial environment, production quality, and longevity. In some instances, these decision criteria have been formalized, using subjective estimates of media impact to model the selection problem (e.g., Gensch, 1970).

The appropriateness of this approach is questionable because of the problems associated with the measurement of the impact factors. The factors depend on the judgment of experts, yet the definition of an expert is not delineated and may vary across situations. In addition, the judgments may be biased in reflecting idiosyncracies in the experiences of the experts. Moreover, there has been no systematic attempt to empirically confirm the use of impact factors as a means of distinguishing the media.

Despite these problems with the practitioner's approach to media selection, few improvements have been suggested. In large measure, this may be due to the fact that there has been little inquiry regarding the nature of the differences between the media. Rather, media research has been concerned with establishing the superiority of one medium over another. Although a substantial number of studies have been undertaken on the media, the results have not been informative. Some studies reported that television advertising was more effective than print (e.g., Greenberg, 1966) while others found the reverse true (e.g., Williams, Paul, and Ogilvie, 1957). Similarly, in some studies comparing the effectiveness of the print and ratio media, it was observed that print was more effective than ratio (e.g., Haugh, 1952) while others indicated the opposite (e.g., Wilke, 1934). The same lack of consistency characterizes studies comparing the television and radio media (e.g., Beighley, 1952; Nasser and McEwen, 1976). In addition, a number of studies have also found no differences between the media (e.g., Sawyer, 1955).

Central to the problem of determining media effectiveness is the fact that the majority of media research has been motivated solely by the desire to demonstrate main effects. without any attention to the theoretical constructs producing the effects. Yet the inconsistencies found in the media studies may have been due to the existence of unknown factors that were unwittingly varied from study to study. The main effect focus of this research has precluded any investigation of these factors. To resolve this problem, research needs to be motivated by a theoretical perspective that is able to suggest constructs that can be used to distinguish among various media.


A promising approach for understanding the factors underlying the effects of the media is information processing theory (see Bettman (1979) for a general discussion). Within this perspective, differences in media effectiveness can be examined in terms of the level of imagery associated with each of the media. Specifically, the presentation formats of the media are assumed to differ in their degree of imagery. As a starting point, the radio, print, and television media can be viewed as presenting information through sentences, which are low in imagery level; whereas the print and television media can be assumed to also convey information through pictures, which are high in imagery level. According to an information processing view, the degree of imagery present in the information being conveyed by a particular medium is assumed to affect the manner in which the information will be processed. As a result, pictures, which are high imaginal stimuli, are assumed to differ from sentences, which are low imaginal stimuli, in their impact on learning and evaluation.

Extensive support for the view that imagery affects learning can be found in verbal learning and cognitive psychology studies. These investigations have documented the imaginal superiority effect on learning; highly imaginal stimuli, such as pictorial representations, have been found to yield significantly higher levels of learning than less imaginal stimuli, such as verbal representations (e.g., Dallet and Wilcox, 1968; Paivio and Csapo, 1969, 1971, 1973; Shepard, 1967; Snodgrass and Asiaghi, 1978). Applied to an advertising context, this finding suggests that persuasive communications in the television and print media, which use both pictures and sentences, may be better learned than in the radio medium, which relies on sentences alone.

The relationship between imagery and learning implies that imagery should be related to evaluation as well. If imagery has an effect on the memorability of information, and if memory plays a role in the evaluation process; then imagery should have an effect on the formation of an evaluation. Although this reasoning appears plausible, the results of a number of marketing studies that have empirically examined the effect of imagery on evaluation question this logic. In these studies imagery has been operationalized in two ways. One involved examining the imagery operationalization of presentation format; that is, pictorial and verbal representations of information were compared (Mitchell and Olson, 1977, 1981; Rossiter and Percy, 1978, 1980). The second operationalization entailed varying the type of instructions given to subjects. Subjects in the imagery conditions were given instructions to create a mental picture or image of information they were being presented with, whereas subjects in the no imagery instruction conditions were given no such instructions (Cialdini and Carpenter, 1980; Mowen, 1980; Wright and Rip, 1980).

In contrast to the studies on learning, the marketing studies undertaken on the evaluative effect of imagery have not yielded consistent results. A number of studies found the expected positive relationship between imagery and evaluation (Cialdini and Carpenter, 1980; Mitchell and Olson, 1977, 1981; Rossiter and Percy, 1978, 1980); but one study found that the imagery condition produced a less favorable evaluation than the no imagery control condition (Mowen, 1980), and another found no effect of imagery on evaluation in a preference judgment task (Wright and Rip, 1980). Despite the importance of understanding the learning effects of imagery as a measure of a medium's effectiveness, the learning of a message alone does not seem to be able to account for attitude formation. Although a high level of imagery has been found to lead to a high degree of learning, it has not necessarily let to the formation of positive attitudes.

In summary, the information processing view is useful in that it suggests the imagery construct for explaining differences in media effectiveness. However, extant research based on information processing theory does not provide an adequate account of the effect of imagery on attitude formation. To address this concern, a theoretical view termed the availability-valence hypothesis is proposed.


The availability-valence hypothesis interprets the learning and evaluation of persuasive communications, such as advertisements, in memorial terms. The availability-valence hypothesis is an extension of the concept of the availability heuristic that was advanced by Tversky and Kahneman (1973). The availability heuristic was defined in terms of the "ease with which instances or associations could be brought to mint" (Tversky and Kahneman, 1973, p. 208), but it did not specify the nature of the relationship between the associations and the process that "brings them to mind." The availability-valence hypothesis is more elaborative about the processes by which information becomes available. In addition, the objective of the availability-valence hypothesis differs from that of the availability heuristic. The goal of the availability heuristic is to explain frequency judgments, whereas the goal of the availability-valence hypothesis is to explain evaluative judgments.

According to the availability-valence hypothesis, the learning and evaluation of a communication depends on the availability of information associated with the communication. The hypothesis suggests that the degree to which information is available depends on two factors. One is the cognitive elaboration of the information conveyed about an object or event. Cognitive elaboration refers to the number of information units that are stored in multiple locations and the number of retrieval pathways that are associated with the information (Bower, 1972; Nisbett and Ross, 1980). According to the hypothesis, the greater the cognitive elaboration of the information being processed, the greater is its availability. Availability also depends on the recency of the information processed such that the more recently processed information is more available. In support of the importance of recency in the processing of information, one study (Higgins, Rholes, and Jones, 1977) found that exposure to positive or negative adjective word lists subsequently affected the evaluation of a person. Subjects exposed to the positive words evaluated the person more highly than those exposed to the negative words.

The availability-valence hypothesis can be evoked to explain the learning and evaluation of a persuasive communication, such as an advertisement. According to this view, the degree to which a communication is learned is dependent on the availability of the information in the communication. The greater the availability of the information, the higher is its memorability. The evaluation that is formed about a communication depends not only on the availability of information about the communication but also on the valence of the information that is available. Valence refers to the affective value of information, which can be positive, negative, or neutral. For any particular communication, the more available is favorable information, the more favorable will be the evaluation of the communication.

If the availability-valence hypothesis is viable, it should be able to predict an imagery effect on learning and evaluation. The superiority of high vs. low imaginal stimuli in learning can be explained in terms of the differential availability of information for the two types of stimuli. A communication characterized by a high degree of imagery is expected to elicit a higher availability of information about the communication than a low imagery communication. According to the hypothesis,a highly imaginal communication is assumed to stimulate greater cognitive elaboration of the information conveyed in the communication; that is, the development of more storage locations and sensory pathways in response to the information. As a result, the information associated with the highly imaginal communication should be more available or accessible for being retrieved. Therefore, highly imaginal communications, such as those relying on pictorial information, should exhibit higher degrees of learning than communications characterized by a low level of imagery, such as those presenting verbal information.

The availability-valence hypothesis can also explain the effect of imagery on evaluation. According to the hypothesis, the information generated in response to the high and low imaginal communications will be valenced; that is, it will be favorable, unfavorable, or neutral in regard to the advocacy of the communication. However, the extremity or degree of favorableness of the evaluation will differ in the two communications. The greater cognitive elaboration of the highly imaginal communication should lead to a greater availability of that information. Since the information evoked in response to the highly imaginal communication is valenced, the greater availability of this valenced information should lead to the formation of a different and more extreme evaluation than would be formed for the less imaginal communication. Thus whether imagery enhances or reduces the favorability of an evaluation depends on the relative favorability of the information available in the high and low imaginal conditions.


To assess the usefulness of the availability-valence hypothesis as an explanation of imagery effects, empirical investigations of the availability-valence hypothesis are warranted in future research. In order to reconcile the inconsistencies in imagery research that were noted, tests of the availability-valence hypothesis should include the manipulations of imagery that were previously used in imagery research. Specifically, both the independent variables of presentation format and imagery instructions should be examined. In addition, special emphasis should be placed on assessing the effect of imagery on the dependent measure of evaluation. Whereas the majority of imagery research has found a positive effect of imagery on learning, its effect on evaluation has been far from univocal. Preliminary support for the hypothesis has been found by the author within the context of a print advertisement for a new product. Further research is needed to determine the ability of the availability-valence hypothesis to predict the conditions under which the imagery effect occurs. Currently an investigation of the role of cognitive elaboration in the availability-valence hypothesis is being conducted by the author.

If the availability-valence hypothesis is empirically supported, its contribution will be theoretical and practical From a theoretical viewpoint, the hypothesis will extend the information processing perspective of consumer behavior by specifying the learning and evaluative effects of persuasive communications that differ in imagery. The availability-valence hypothesis presents a theoretical framework for reconciling inconsistencies found in media research and suggests future research directions in this area. Most importantly, it provides a bridge between the research concerned with learning and that focused on evaluative processes. The availability-valence hypothesis underscores the importance of going beyond learning effects in understanding evaluative judgments. Because the availability-valence hypothesis is broad-based in its conceptualization of thought processes in memory, it can also be applied to other social cognitive phenomena in future research.

The availability-valence hypothesis also has significant practical implications for various advertising issues. The hypothesis presents a new perspective for analyzing media selection decisions. According to the availability-valence view, a more complete understanding of media effectiveness can be gained by examining the processing of information in these media. Specifically, the impact of the presentation formats of the media should be addressed. Furthermore, because of the hypothesis' ability to explain pictorial processing, it can also provide insight into strategies concerned with developing effective trademarks. The hypothesis also has implications for the measurement of advertising effectiveness in that it proposes that measures of learning are not sufficient indicators of an ad's impact on the consumer. Lastly, the availability-valence hypothesis can contribute to the formulation of public policy decisions concerning advertising by suggesting that their current focus on copy testing (Eighmey, 1978) may be inappropriate. Copy testing, which relies on the processing of verbal information, has been used for the purpose of measuring the amount of information provided by an advertisement. However, the availability-valence hypothesis suggests that the use of copy testing may be ignoring the more important impact of visual or highly imaginal claims.


Beighley, K. (1952), "An Experimental Study of the Effect of Four Speech Variables on Listener Comprehension," Speech Monographs, 19, 249-258.

Bettman, James R. (1979), An Information Processing Theory of Consumer Choice, Reading, Massachusetts: Addison-Wesley.

Bower, G. H. (1972),"Mental Imagery and Associative Learning," in Cognition in Learning and Memory, ed. L. Gregg, New York: Wiley.

Cialdini, R. and Carpenter, K. (1980), "A Test of the Availability Heuristic: Does Imagining Make It So?" Working Paper, Arizona State University, Tempe, Arizona.

Dallet, K. and Wilcox, S. G. (1968), "Remembering Pictures vs. Remembering Descriptions," Psychonomic Science, 11, 139-140.

Eighmey, J. (1978), "A Perspective on Advertising Copy Research and Advertising Law Enforcement," in Attitude Research Plays for High Stakes, eds. J. C. Maloney and B. Silverman, Chicago: American Marketing Association.

Gensch, D. (1970), "Media Factors: A Review Article." Journal of Marketing Research, 7, 216-225.

Greenberg, B. S. (1966), "Media Use and Believability: Some Multiple Correlates," Journalism Quarterly, 43, 655-670.

Haugh, O. (1952), "The Relative Effectiveness of Reading Listening to Ratio Drama as Ways of Imparting Information and Shifting Attitudes," Journal of Educational Research, 45, 489-498.

Higgins, E. T., Rholes, W. S., and Jones, C. R. (1977), "Category Accessibility and Impression Formation," Journal of Experimental Social Psychology, 13, 141-154.

Mitchell, A. and Olson, J. C. (1977), "Cognitive Effects of Advertising Repetition," in Advances in Consumer Re- search, Vol. 4, ed. W. D. Perreault, Jr., Atlanta: Association for Consumer Research. 213-220.

Mitchell, A. (1981), "Are Product Attribute Beliefs the Only Mediator of Advertising Effects on Brand Attitude?" Journal of Marketing Research, 18, 318-332.

Mowen, J. (1980), "The Availability Heuristic: The Effect" of Imagining the Use of the Product on Product Perceptions, in AMA 1980 Educator's Conference Proceedings, ed. R. Bagozzi et al., Chicago: American Marketing Association, 140-142.

Nasser, D. and McEwen, W. J. (1976), "The Impact of Alternative Media Channels: Recall and Involvement with the Message," Audio-Visual Communications Review, 24, 263-272.

Nisbett, R. and Ross, L. (1980), Human Inference: Strategies and Shortcomings of Social Judgment, Englewood Cliffs, N.J.: Prentice-Hall.

Paivio, A. and Csapo, R. (1969), "Concrete Image and Verbal Memory Codes," Journal of Experimental Psychology, 80, 249-285.

Paivio, A. (1971), "Short-term Sequential Memory for Pictures and Words," Psychonomic Science, 24, 50-51.

Paivio, A. (1973), "Picture Superiority in Free Recall: Imagery or Dual Coding?" Cognitive Psychology, 5, 176-206.

Rossiter, J. and Percy, L. (1978), "Visual Imaging Ability as a Mediator of Advertising Response," in Advances in Consumer Research, Vol. 5, ed. K. Hunt, Ann Arbor: Association for Consumer Research, 621-629.

Rossiter, J. (1980), "Attitude Change through Visual Imagery in Advertising," Journal of Advertising, 9, 10-16.

Sawyer, T. (1955), "Shift of Attitude Following Persuasion as Related to Estimate of Majority Attitude," Speech Monographs, 22, 68-78.

Shepard, R. (1967), "Recognition Memory for Words, Sentences and Pictures," Journal of Verbal Learning and Verbal Behavior, 6, 156-163.

Snodgrass, J. and Asiaghi, A. (1978), "The Pictorial Superiority Effect in Recognition Memory," Bulletin of the Psychonomic Society, 10, 1-4.

Tversky, A. and Kahneman, D. (1973), "Availability: A Heuristic for Judging Frequency and Probability," Cognitive Psychology, 5, 207-232.

Wilke, W. H. (1934), "An Experimental Comparison of the Speech, the Radio, and the Printed Page as Propaganda Devices," Archives of Psychology, 25, 5-32.

Williams, D. C., Paul J., and Ogilvie, J. C. (1957), "The Mass Media, Learning, and Retention,f' Canadian Journal of Psychology, 11, 157-163.

Wright, P. and Rip, P. (1980), "Product Class Advertising Effects on First - Time Buyers' Decision Strategies," Journal of Consumer Research, 7, 176-188.