Effects of Verbal and Visual Information on Brand Attitudes

Yehoshua Tsal, Cornell University
ABSTRACT - Subjects were presented with eight advertisements varying in argument strength and picture attractiveness. Although little argument and picture information could be recalled, both had significant effects on brand attitudes under both high and low involvement conditions. The findings emphasize the role of nonfactual information in mediating attitude formation even under conditions that maximize attention and involvement.
[ to cite ]:
Yehoshua Tsal (1985) ,"Effects of Verbal and Visual Information on Brand Attitudes", in NA - Advances in Consumer Research Volume 12, eds. Elizabeth C. Hirschman and Moris B. Holbrook, Provo, UT : Association for Consumer Research, Pages: 265-267.

Advances in Consumer Research Volume 12, 1985      Pages 265-267


Yehoshua Tsal, Cornell University


Subjects were presented with eight advertisements varying in argument strength and picture attractiveness. Although little argument and picture information could be recalled, both had significant effects on brand attitudes under both high and low involvement conditions. The findings emphasize the role of nonfactual information in mediating attitude formation even under conditions that maximize attention and involvement.


The relative significance of content and non-content information in mediating attitude formation toward messages has been the focus of much interest in persuasion and advertising research. One route of investigation has led to specifying the conditions under which one type of information becomes dominant over the other. These studies (e.g., Petty, Cacioppo and Schumann 1983) have shown that under high level of personal involvement attitudes are influenced by evaluation of the message content, whereas under low involvement conditions attitudes are affected by peripheral cues (e.g., source attractiveness) which are not related to the content of the message. However, in many of these studies motivational involvement has been confounded with the quality of processing information. For example, the minimal effects of source attractiveness on attitude under high involvement conditions could either be caused by minimal processing of the source while subjects are engaged in extensive processing of the message content, or by motivational factors which minimize the influence of processed peripheral information on attitudes. While several studies have shown that peripheral information is equally remembered under high and low involvement conditions, it has been argued (e.g., Hutchinson and Moore 1984) that postexposure recall and recognition measures do not necessarily reflect the depth of information processing during exposure. Thus, in the absence of direct measures of processing quality it is difficult to assess the role of nonfactual information in forming attitudes under high involvement conditions. Indeed, several studies have shown that under conditions which encourage extensive processing, peripheral information can influence attitude toward the message. For example, Mitchell and Olson (1983) instructed subjects to concentrate on remembering advertisements and found that pictures which provided no explicit brand information influenced attitudes toward the advertised Product.

The present study focuses on situational rather than personal involvement. An attempt was made here to devise an experimental manipulation in which both high and low involvement conditions would entail extensive processing of factual and pictorial information from advertisements. Furthermore, the high involvement condition was characterized by processing information for the purpose of forming brand attitudes and inferring the true quality of the advertised brand. This manipulation is similar to the high involvement brand processing strategy used by Mitchell, Russo and Gardner (1980). The purpose was to investigate whether under this type of high involvement, in the process of forming brand attitudes subjects would maximize the processing of content information and minimize the influence of picture information, since the latter was not inherently related to the quality of the brand. This manipulation of involvement is different than those used in the past. In fact, one could argue that according to some definitions the present low involvement is a form of high involvement since it entails extensive processing of the advertisements. Nevertheless, for the purpose of the present study, it is sufficient that the high involvement instructions produced higher involvement than the low involvement instructions by requiring subjects to form brand attitudes during exposure to the advertisements.

The second purpose of the present experiment was to explore the possibility that the advertisements could influence the formation of brand attitudes at some later point in time following exposure, when information from the advertisements could not be adequately retrieved. It was hypothesized that affective reactions generated during exposure to the advertisements could outlast the representation of the specific information that had originally produced them, thereby influencing brand attitudes. This hypothesis is indirectly suggested by findings (e.g., Greenwald 1968) showing that subjects' recall of abstractions and inferences of factual information are better predictors of attitude change than recall of the factual information itself


Subjects and Stimuli

The subjects were 16 undergraduate business students enrolled in marketing research courses. Eight subjects were assigned to the high involvement condition and eight were assigned to the low involvement condition. Both groups were presented with all eight advertisements.

Each advertisement consisted of a fictitious brand names a product name,a background picture, and two arguments. To maximize the nonfactual nature of the pictures, none provided any explicit information about the advertised brand. Each picture was either attractive or unattractive. For example, an attractive young woman vs. an average looking middle aged woman. To further enhance the affective difference between the two types of pictures only the attractive ones were colored. The two arguments supporting the brands were either both strong or both weak. For example, the (strong) arguments for a pen brand were (l) precision balanced and hand crafted in Germany, and (2) Ink doesn't blott or run, and the (weak) arguments for a coffee brand were (1) available in new colored cans, and (2) good in the morning, noon, or night. Two advertisements were constructed for each combination of argument strength and picture attractiveness (i.e., strong attractive, strong unattractive, weak attractive, and weak unattractive) resulting in a total of eight advertisements.


Subjects participated in two sessions which were separated by 30 hours. The first session inc' 'ed the instructions and the presentation of the advertisements, and the second consisted of attitude and memory questionnaires.

Session I. Subjects were told that the slides they were about to see were of preliminary pictures and drawings which would be used as a basis for constructing professional advertisements. They were further told that the brands depicted in the advertisements have been used in various parts of the world and that the manufacturers were planning to introduce them to the American market in the near future.

All subjects were instructed to extract as much information as possible from the brand and product names, the arguments and the pictures since they would later be asked questions about them. The high involvement group received additional instructions; they were instructed to form an attitude towards the brand during each advertisement presentation and the following rest interval. They were told that each brand had a given known quality, and that their task was to infer the true quality of the brand on the basis of the information provided in the advertisement. The instructions were followed by the presentation of the advertisements. Each slide was presented on the screen for eight seconds and was followed by an eight-second rest interval.

Session II. In the second session all subjects were presented with a booklet that contained attitude and memory questionnaires. The attitude questionnaires consisted of three seven point binary scales ranging from -3 to +3 (a. extremely bad to extremely good, b. extremely unsatisfactory to extremely satisfactory, and c. extremely poor quality to extremely high quality). For each of these questionnaires subjects were instructed to circle a single number that best represented their attitude towards each brand. The attitude questionnaires were followed by a recall task in which the eight product and brand names were presented and subjects were asked to write down anything they could remember about the picture, the arguments, and any impression they had of each advertisement.


Manipulation Checks

A comparable sample of 16 new subjects completed two manipulation checks. In the first, they were presented with the eight product and brand names and their 16 arguments and were asked to rate on a seven point scale how persuasive was each argument in supporting its brand. In the second test these subjects were presented with the eight pictures and were asked to rate how pleasant each was on a seven point scale. These checks indicated that the manipulations produced the intended effects. The strong arguments were significantly more persuasive than the weak arguments (p < .001), and the attractive pictures were significantly more pleasant than the unattractive ones (p < .0001). Furthermore, no significant differences in persuasiveness were found among the eight strong arguments and among the eight weak arguments, and no significant differences in pleasantness were found among the four attractive pictures and among the four unattractive ones.

Recall and Attitudes Measures

The criteria for scoring recall responses were extremely lenient. Any response indicating recollection of a part of a picture or a part of either one or both arguments, as long as it was uniquely associated with the correct brand was scored as a correct response for this brand. Table l shows the proportion of arguments and pictures correctly recalled for the two involvement conditions. An analysis of variance indicated no significant differences among the four cells of Table l.



Since the intercorrelations among the three attitude scales were very high responses were averaged across them in order to assess the general attitude toward each brand. Figure l presents the attitude judgements for the various combinations of arguments and pictures under the two involvement conditions. An analysis of variance revealed that all main effects were significant. Low involvement subjects, strong arguments, and attractive pictures produced more positive attitudes than high involvement subjects F(1,14 = 5.83, p c .05), weak arguments F(1,14 - 5.9, p < .05), and unattractive pictures F(1,14 = 18.12, p < .001) respectively. Also significant was the interaction between condition and picture attractiveness F(1,14 - 4.62, p < .05) indicating that the latter effect was greater under the high involvement condition. While this interaction seems to be caused by the difference in negative attitudes towards the unattractive pictures, the direction of the effect may be due to individual differences between the two groups. More important is the relative attractiveness effect within each group of subjects. While the effect of argument strength tended to be stronger under the high involvement condition, this interaction did not reach statistical significance.




The results indicate poor recall performance 30 hours following the presentation of the advertisements. That is, on the average, both argument information and picture information was recalled successfully from about two of the eight advertisements. Moreover, due to the lenient criteria of scoring recall responses, a correct response reflects remembering something about the advertisement rather than a detailed recollection of it. Thus, minimal information about the advertisements could be retrieved by subjects while making attitude judgements about the brands. Yet, brand attitudes were significantly influenced by both the strength of the arguments and the attractiveness of the pictures. This finding indicates that affective reactions generated during exposure to the advertisement could outlast the cognitive representations that had originally produced them. The present results, while limited to the short interval between presentation and attitude judgements, suggest that recall performance is not crucial for advertising effectiveness. Affective responses to an advertisement could be dissociated from their specific source and transformed into a global attitude judgement about the quality of the brand at some later point in time. In fact, one could speculate that under certain conditions the absence of recall rather than its presence may be more effective in generating a positive attitude toward the brand because of the inability to critically evaluate the retrieved information during attitude formation.

The second major finding of the present experiment is that the high involvement instructions did not eliminate the influence of picture attractiveness on brand attitudes. In fact, this effect was greater under the high involvement condition than under the low involvement condition. This unexpected finding could be attributed to the fact that the high involvement instructions intensified the overall processing of the advertisements without enabling subjects to selectively focus on the arguments. However, this result is only suggestive and should be interpreted with caution. The combination of a relatively small sample size and the low significance level (.05) of the interaction between condition and picture attractiveness, and the substantial (unexplained) difference between the two conditions in negative attitudes towards the unattractive pictures suggest that this particular finding should be further substantiated by future research.

Irrespective of the above interaction, the present results clearly indicate a strong effect of picture attractiveness for both conditions. Why should pictures which are unrelated to brand quality influence brand attitude of highly attentive and involved subjects? This result is probably due to the automatic generation of affective reactions toward the salient pictures. Pictures are more perceptually salient than verbal information, their processing requires less cognitive effort, and they are more powerful in generating immediate emotional responses. Since the affective system is immediate, automatic, and inescapable (e.g., Zajonc 1980) these responses are likely to be generated regard-less of the subjects' level of involvement. To exclude these affective reactions from the formation of brand attitudes they must be distinguished from the reaction generated by the verbal information. However, as recent data (e.g., Johnson and Tversky 1983) suggest, once generated, emotional reactions are likely to be dissociated from their informational source, making this distinction rather difficult. Consequently, the nature of the unrelated pictures would produce some effect on brand attitudes regardless of the specific goals of the subjects.


Greenwald, A.G. (1968), "On defining attitude and attitude theory," in A.G. Greenwald, T.C. Brock, and T.M. Ostrom (eds.), Psychological Foundations of Attitudes, New York: Academic Press.

Hutchinson J. Wesley and Danny L. Moore (1986), "Issues surrounding the examination of delay effects in advertising," in Thomas C. Kinnear (ed), Advances in Consumer Research, Vol. 11, Ann Arbor Association for Consumer Research. 650-655.

Johnson, Eric J. and Amos Tversky (1983), "Affect generalization, and the perception of risk," Journal of Personality and Social Psychology, 45, 20-31.

Mitchell, Andrew A. and Jerry C. Olson (1981), "Are product attribute beliefs the only mediator of advertising effects on brand attitude?" Journal of Marketing Research, 18 (August), 318-332.

Mitchell, A.A., J.E. Russo and M. Gardner (1980), "Strategy induced low involvement processing of advertising messages," Working Paper, School of Industrial Administration, Carnegie-Mellon University.

Petty, Richard 4., John T. Cacioppo and Davit Schumann (1983), "Central and peripheral routes to advertising effectiveness: The moderating role of involvement," Journal of Consumer Research, 10 (September), 135-146.

Zajonc, Robert B. (1980), "Feeling and thinking: Preferences need no inferences," American Psychologist, 35. 151-175.