Dissonance Resolution By Grade School Consumers

David B. Kyner, DKG Advertising
Jacob Jacoby, Purdue University
Robert W. Chestnut, Purdue University
ABSTRACT - Children evaluated two candy bar brands, made six purchase decisions involving these brands, then reevaluated the same two brands. Significant (p < .01) pre-to-post attitude shifts occurred as predicted by dissonance theory: shifts in predicted directions were greater when the brands were initially considered to be equally attractive than when they were considered unequally attractive.
[ to cite ]:
David B. Kyner, Jacob Jacoby, and Robert W. Chestnut (1976) ,"Dissonance Resolution By Grade School Consumers", in NA - Advances in Consumer Research Volume 03, eds. Beverlee B. Anderson, Cincinnati, OH : Association for Consumer Research, Pages: 134-137.

Advances in Consumer Research Volume 3, 1976      Pages 134-137


David B. Kyner, DKG Advertising

Jacob Jacoby, Purdue University

Robert W. Chestnut, Purdue University


Children evaluated two candy bar brands, made six purchase decisions involving these brands, then reevaluated the same two brands. Significant (p < .01) pre-to-post attitude shifts occurred as predicted by dissonance theory: shifts in predicted directions were greater when the brands were initially considered to be equally attractive than when they were considered unequally attractive.


Children, as a focus for consumer behavior research, have been relatively neglected (cf., McNeal, 1969; Jacoby and Kyner, 1973). While research in related areas such as advertising impact has helped to provide data on consumer-related experiences during childhood (Ward, Robertson and Wackman, 1971; Robertson and Rossiter, 1974), attention has only recently been given to the cognitive processes by which a child arrives at and possibly re-evaluates a consumer decision (Calder, Robertson and Rossiter, 1975).

Information on such processes would be useful from several standpoints. First, children represent an important market segment. They spend nearly 500 million dollars a year in their own right (cf., Jacoby and Kyner, 1973, p. 4) and exert considerable influence over expenditures made by other family members (Coulson, 1966; Deering and Jacoby, 1971; McNeal, 1969; Munn, 1958; Wells and LoSciuto, 1966). Knowing how a child will evaluate and influence the purchase of a given product should significantly improve the marketer's ability to present the product in a favorable light. Second, governmental interest in consumer safety and education often center on the child as consumer. Consider, for example, the FTC's increasing concern over advertising directed at children. Yet basic information regarding how the child perceives products and acts within his role as a consumer is still lacking, and such knowledge is vital for arriving at appropriate public policy decisions. Finally, from a longitudinal perspective, research on consumer behavior in children can provide insight into how adult buying behaviors are acquired. Consumer behaviorists generally assume that certain aspects of consumer behavior are acquired sometime during childhood, yet they typically fail to show how these behaviors are learned or even if they exist as important elements in the child's behavior as a consumer.

The present investigation focuses upon one cognitive process that has been shown to have a significant effect on adult consumer behavior. The process is known as post-decision dissonance reduction and, in the consumer context, involves a drive toward re-evaluating brand preferences immediately after purchase. Of all the empirical issues and research stimulated by Festinger's dissonance theory, reviewers (e.g., Insko, 1967, p. 283; Kiesler, Collins and Miller, 1969, p. 205) consistently point to post-decision dissonance reduction as one of the more stable and interesting research issues. Although an increasing number of studies have appeared on the subject, a literature review reveals no attempt to extend research on this issue to children.

Early cognitive dissonance investigations (Brehm, 1956) demonstrated a tendency to re-evaluate alternatives immediately following a choice situation: the chosen alternative increased in desirability while the rejected alternative became less desirable. In terms of dissonance theory, it was assumed that, after making a choice, a person would attempt to justify his decision by emphasizing the positive aspects of the chosen alternative and the negative aspects of the rejected alternative. This in turn led to a change in judged desirability. Dissonance theory predicts that the more nearly equal the attractiveness of two alternatives, the more dramatic the post behavioral reevaluation of attitudes toward these alternatives --in a favorable direction for the selected alternative and in an unfavorable direction for the rejected alternative. Another prediction is that, as the cognitive overlap (or actual similarity) of the two alternatives decreased, the dissonance and resultant re-evaluation would increase (Brehm and Cohen, 1962, p. 303). The case of maximal dissonance, therefore, would be a choice between two equally desirable alternatives (i.e., both highly attractive) in which the attributes of one alternative were different from those of the second alternative.

Empirical support for these hypotheses was adduced by Brehm and Cohen (1959, 1962) and Brock (1963). These first tests of the theory often utilized children. Brock, for example, tested 141 children from three to twelve years of age. In one condition, children were to choose between similar objects (two crackers or two toys), while in the other condition, they were to choose between dissimilar objects (a cracker and a toy). In the usual comparison of pre vs. post-ratings of desirability, children choosing between dissimilar objects reflected a greater change in attitudes and, thus, were assumed to have experienced more dissonance.

Brock's experiment would seem to indicate that post-decision dissonance reduction via a change in perceived desirability plays an important role in a child's decision-making and that it manifests itself at a very early age. Unfortunately, this study has never been replicated in a consumer choice-situation.

Consumer studies which did replicate Brehm's findings were all concerned solely with adult behavior. Anderson, Taylor and Holloway (1966), varying both the number of alternatives and relative attractiveness, showed the predicted changes in desirability on a wide variety of products. LoSciuto and Perloff (1967) provided additional supportive evidence through the use of a ranking task which required people to rank . order the desirability of nine record albums. A high dissonance condition was created by arranging a choice between the third and fourth ranked alternatives, while a low dissonance condition involved a choice between the third and eighth ranked alternatives. As predicted, a greater shift in desirability ratings was noted in the high dissonance condition. Sheth (1970) replicated this finding on a sample containing both housewives and male graduate students.

Despite criticisms that the LoSciuto and Perloff design incorporates a biasing "regression effect," (Oshikawa, 1971), rejoinders (Sheth, 1971; Cummings and Venkatesan, 1974) suggest that the early studies by Brehm and his associates appear to have carried over quite well to the consumer context when adult consumers are considered. Re-examination of data collected earlier for another purpose (cf., Jacoby and Kyner, 1973), permits a test of the following hypothesis: as the desirability of two purchase alternatives becomes more nearly equal, dissonance will increase for children engaged in a consumer decision making task.



The subjects were 80 six-to-nine year old children, 32 males and 48 females. [Four subjects rated brands L and S as equal in the initial taste test and, thus, were forced into making a MPB selection. Since this represents a non-volitional choice (i.e., outside the realm of dissonance theory predictions), data from these subjects were excluded from the analysis.] Previous studies (Vogl, 1964; Marshall and Magruder, 1960) had indicated that children of this age were both selective in the brands they chose and knowledgeable in the use of money. This finding was confirmed in pilot tests with subjects similar to those used in the present study (Kyner, 1971), as well as in discussions with the subjects employed in the present investigation.

Design and Procedure

The experiment was conducted in three segments: a pre-rating of brand preference, a set of six purchase trials, and a post-rating of preference. Each child was tested independently.

Children were first exposed to two different brands of candy bars. These brands had similar appearance and were identified only by the labels L and S -- letters found to elicit relatively neutral affect and low candy associations in the pilot studies (cf., Kyner, 1971). In actuality, brands L and S were the same candy bar. The pilot studies had shown, however, that children still managed to perceive differences and that these differences were randomly distributed. After presenting the supposedly different brands, children were asked to taste-test the candy bars and rate them via two scales: 'a seven-point Smiling Faces Scale (SFS; cf., Wells, 1965) and a ten-point Method of Constant Sum (MCS). Each child's most preferred brand (MPB) and second preferred brand (SPB) was identified via these ratings.

Immediately following this initial rating, the children were engaged in a series of six purchase trials. These trials were administered at four minute intervals during a videotaped cartoon show. For the first five trials the child was confronted with a short, mildly persuasive videotaped advertisement for his MPB after four minutes of cartoons had elapsed. The child was then given a dime and asked to use it in making a purchase decision between the two brands. The context of the first five trials, therefore, tended to reinforce brand loyalty to the MPB. This was successful in that 89% of the 400 choices (80 subjects x 5 trials) were for the MPB. Finally, a sixth trial was included that followed a longer advertisement which contained a strong appeal against the MPB and advocated purchase of the SPB.

The third stage of the experiment consisted of administering a number of measures, including both the SFS and MCS. These procedures are more fully described in Jacoby and Kyner (1973).


Subjects were divided into two groups based upon their initial brand ratings. The "equal attractiveness" group consisted of all subjects who perceived little or no difference between brands in the initial taste-test. Operationally, these were subjects who rated the two brands as 0 or 1 point different on the SFS and 0 to 2 points different on the MCS. The "unequal attractiveness" group was composed of the remaining subjects, all of whom had differentiated between the brands. Four subjects had shown equal preference for the two brands on both the SFS and the MCS and, thus, were excluded from the analysis. For ease of comparison, the same type of analyses as employed by Brehm and Cohen (1959) were employed here.

Table 1 shows the results of a t-test between pre- vs. post-ratings on the seven-point SFS for both groups. As can be seen by the index of change, subjects perceiving no initial brand differences manifested the predicted shifts in preference. They increased their liking for the MPB and decreased their liking for the SPB. This change was significant at the .001 level for the equal attractiveness group and nonsignificant for the unequal attractiveness group. A t-test between the two groups showed a difference significant at the .01 level. [A one-tailed t statistic for non-correlated data.]

Table 2 shows the same results for the ten-point MCS measure. Here again, t-tests for the equal attractiveness group were significant at the .001 level, while being nonsignificant for the unequal attractiveness group. A t-test between groups showed a significant difference at the .05 level. [A one-tailed t statistic for non-correlated data.]






The hypothesis that, as the desirability of the available purchase alternatives becomes increasingly similar, post-decision dissonance reduction tendencies will increase was supported by the finding of this investigation. This was manifested by subjects rating their MPB higher and their SPB lower than they did initially. Such attitude changes occurred to a significantly greater degree when the most and second preferred brand were equally attractive than when they were unequally attractive.

Dissonance theory would reason that the positive attributes of the rejected alternative were greater in number for the equal attractiveness group. These positive attributes would generate greater dissonance for the child, as he was confronted with the fact of actually rejecting the SPB in the six purchase trials. To resolve this dissonance, the child apparently engaged in a re-evaluation of brand preferences.

Alternative explanations for this re-evaluation seem unlikely. Statistical regression, for example, would be inappropriate in explaining the systematic polarization of post-task ratings (cf., Campbell and Stanley, 1963, p. 10). A learning interpretation of the data, although of some predictive value, would not adequately account for the increasing negative evaluation of the SPB. Finally, the reliability of results from the two different rating procedures and the fact that post-task ratings did differ substantially from the end-points of the scales argue against potential ceiling effects having affected the results.

Given a dissonance orientation towards the data, one interesting finding was that, despite a great deal of cognitive overlap (i.e., the candy bars were the same in terns of physical appearance and actual composition), post-decision dissonance reduction still occurred. This would seem to conflict with Brehm's statement that "when the two alternatives have the same attributes --that is, they are identical -- nothing is given up or avoided by choosing one over the other (1962, p. 37)." Dissonance should not occur in the case of identical alternatives.

It can be contended, however, that the brands in question were not perceived to be identical by the subjects. They were presented under two different brand names (L & S) which seemed to generate a psychological reality of difference for the children. Not a single child ever questioned the fact that L and S were different, either during the experiment or in post-experimental interview and debriefing sessions. Taken in conjunction with the fact that post-decision dissonance can be assumed to have occurred, this would suggest that overlap must be defined in terms of "psychological" rather than "objective" reality.

In sum, this study has made two contributions to the previous literature on post-decision dissonance reduction. First, it has generalized the findings of Brehm and his associates to the consumer behavior of children by supporting the hypothesis that the closer the two alternatives are in attractiveness, the greater the subsequent attitude change. Second, it has demonstrated (at least within the context of a child's consumer behavior) that actual product similarity may have little effect upon the arousal of dissonance.

Both findings are important for future research involving dissonance reduction in the purchasing behavior of children. Above all, they. suggest that investigators concentrate on the psychological reality of a brand within a product class. What, for instance, is the relationship between brand loyalty and a child's perception of similarity between brands? Can brand loyalty be reduced by actually informing the child of a lack of objective difference in quality? Because of their potential to provide useful insight into the development of various consumer behavior phenomena in adults (e.g., brand loyalty), s research effort designed to examine the implications of cognitive dissonance in the consumer behavior of children seems appropriate.


Lee K. Anderson, James R. Taylor and Robert J. Holloway, "The Consumer and His Alternatives: An Experimental Approach," Journal of Marketing Research, 3(February, 1966),62-67.

Jack W. Brehm, "Post-Decision Changes in the Desirability of Alternatives," Journal of Abnormal and Social Psychology, 52(May, 1956),384-389.

Jack W. Brehm and Arthur R. Cohen, "Reevaluation of Choice Alternatives as a Function of Their Number and Qualitative Similarity," Journal of Abnormal and Social Psychology, 58(May, 1959),373-378.

3ack W. Brehm and Arthur R. Cohen, Exploration in Cognitive Dissonance (New York: Wiley, 1962).

Timothy C. Brock, "Effects of Prior Dishonesty on Postdecision Dissonance," Journal of Abnormal and Social Psychology, 66(April, 1963),325-331.

Bobby J. Calder, Thomas S. Robertson and John R. Rossiter, "Children's Consumer Information Processing," Communication Research, 2(July, 1975),307-316.

Donald T. Campbell and Julian C. Stanley, Experimental and Quasi-Experimental Designs for Research (Chicago: Rand McNally, 1963).

John S. Coulson, "Buying Decisions Within the Family and the Consumer-Brand Relationship," in Joseph Newman (Ed.), On Knowing the Consumer (New York: Wiley, 1966).

William H. Cummings and M. Venkatesan, "Cognitive Dissonance and Consumer Behavior: A Critical Review," The University of Iowa Bureau of Business and Economic Research, Working Paper Series No. 74-1, (January, 1974).

Barbara J. Deering and Jacob Jacoby, "The Effects of Alternative Relationships and Relative Resources on Consumer Decisions Between Mother and Child," Proceedings. Second Annual Convention, The Association for Consumer Research, (1971),135-142.

Chester A. Insko, Theories of Attitude Change (New York: Appleton-Century-Crofts, 1967).

Jacob Jacoby and David B. Kyner, "Brand Loyalty vs. Repeat Purchasing Behavior," Journal of Marketing Research, 10(February, 1973),1-9.

Charles A. Kiesler, Barry E. Collins and Norman Miller, Attitude Change (New York: Wiley, 1969).

David B. Kyner, "Testing Jacoby's Conceptualization of Brand Loyalty," unpublished Ph.D. dissertation, Purdue University, 1971.

Leonard A. LoSciuto and Robert Perloff, "Influence of Product Preference on Dissonance Reduction," Journal of Marketing Research, 4(August, 1967),286-290.

Helen R. Marshall and Lucille Magruder, "Relations Between Parent Money Education Practices and Children's Knowledge and Use of Money," Child Development, 31 (1960),253-284.

James U. McNeal, Dimensions of Consumer Behavior (New York: Appleton-Century-Crofts, 1969).

Mark Munn, "The Effect of Parental Buying Habits of Children Exposed to Children's Television Programs," Journal of Broadcasting, 2(1958),253-258.

Sadaomi Oshikawa, "Dissonance Reduction or Artifact?," Journal of Marketing Research, 8(November, 1971),514-515.

Thomas S. Robertson and John R. Rossiter, "Children and Commercial Persuasion: An Attribution Theory Analysis," Journal of Consumer Research, 1(1974), 13-20.

Jagdish N. Sheth, "Are There Differences in Dissonance Reduction Behavior Between Students and Housewives?," Journal of Marketing Research, 7(May, 1970), 243-245.

Jagdish N. Sheth, "Dissonance Reduction or Artifact? A Reply," Journal of Marketing Research, 8(November, 1971),516-517.

A. J. Vogl, "The Changing Face of the Children's Market," Sales Management, 93(December 18, 1964), 35-38.

Scott Ward, Thomas S. Robertson and Daniel Wackman, "Children's Attention to Television Advertising," Proceedings. Second Annual Conference, The Association for Consumer Research, (1971),143-156.

William D. Wells, "Communicating with Children," Journal of Advertising Research, 5(June, 1965),2-14.

William D. Wells and Leonard A. LoSciuto, "Direct Observation of Purchasing Behavior," Journal of Marketing Research, 3(August, 1966),227-233.