An Examination of Ego-Involvement As a Modifier of Attitude Changes Caused From Product Testing

Larry M. Newman, (student), The Pennsylvania State University
Ira J. Dolich, University of Nebraska-Lincoln
ABSTRACT - Ego-involvement as a modifier of attitudes has recently become a topic of great interest to marketers. An experiment was conducted to measure levels of ego-involvement using Sherif's Social Judgement Theory and the effects of an automobile demonstration as a communication to create changes in attitudes. The results demonstrate the usefulness of this theoretical perspective as a schema for examining communication effectiveness.
[ to cite ]:
Larry M. Newman and Ira J. Dolich (1979) ,"An Examination of Ego-Involvement As a Modifier of Attitude Changes Caused From Product Testing", in NA - Advances in Consumer Research Volume 06, eds. William L. Wilkie, Ann Abor, MI : Association for Consumer Research, Pages: 180-183.

Advances in Consumer Research Volume 6, 1979      Pages 180-183


Larry M. Newman (student), The Pennsylvania State University

Ira J. Dolich, University of Nebraska-Lincoln


Ego-involvement as a modifier of attitudes has recently become a topic of great interest to marketers. An experiment was conducted to measure levels of ego-involvement using Sherif's Social Judgement Theory and the effects of an automobile demonstration as a communication to create changes in attitudes. The results demonstrate the usefulness of this theoretical perspective as a schema for examining communication effectiveness.


The concept of an individual's involvement or commitment to a product class or brand has been an expanding area of interest to marketers. Beginning with the early work of Herbert Krugman (1965, 1966, 1971), there has been speculation in this literature about the relationship between level of involvement and an individual's processing of persuasive communications. However, little research exists investigating attitude changes caused by persuasive communications and the moderating effects of involvement. The present study was conceived to examine the relationship of involvement to the amount and direction of attitude change resulting from a demonstration in a Ford Pinto.

Models to evaluate attitude change caused by communications have been proposed by Osgood and Tannebaum, Festinger, and Sherif. The Congruity Model of Osgood and Tannebaum recognizes the position of stimulus (positive or negative) and the position of the receiver (pro or con) which affect evaluation of attitude in the context of the individual's frame of reference (Zajonc, 1960). However, their model does not address itself to how involvement effects the initial position of the attitude. And, Festinger's Dissonance Theory is directed to the nature of dissonance and not attitude adjustment mechanisms. Sherif's model (Sherif and Sherif, 1969) explains the direction and amount of attitude change (assimilation/contrast) based on three factors: level of ego-involvement in an attitude (high or low); direction of attitude held (pro or con); and nature of the stimulus (positive or negative). Their model was selected for this study because of its advantages over Congruity and Dissonance models.

Application of Sherif's Model

Sherif's "Social Judgment Involvement Approach" (Sherif, 1969) has primarily been used in the measurement of attitude change to communications concerning social issues. However, it is used here to evaluate attitude change toward a product. Specifically, it has been adapted to analyze direction and extent of attitude change resulting from the positive communication of an automobile demonstration. The powerful "attention getting, desire motivating" quality of the demonstration makes it an effective tool for changing attitudes (Tilman and Kirkpatrick, 1968).

Ego-involvement is measured by the number of items the respondent accepts (latitude of acceptance) versus those he rejects (latitude of rejection) or those toward which he is neutral (latitude of noncommitment). In this study, mini-compact cars are the involvement items. Where the car being demonstrated falls within the latitude of acceptance, there is a strong likelihood of an assimilation effect. Also, the communication will appear nearer to the subjects own position than it actually is, and thus cause a positive shift in attitude. A contrast effect will occur if the car being demonstrated falls into the latitude of rejection. The communication will be perceived as farther away from the person's own stand than it really is, thus causing a negative shift in attitude (Sherif and Sherif, 1969). Furthermore, the degree of assimilation or contrast effect is dependent upon the level of ego-involvement.

As an example, consider a hypothetical Volkswagen owner (A) who is highly ego-involved with his car and has placed the Pinto in his latitude of rejection (PR)' The Sherif model would predict that this subject would react negatively to the communication (positive car demonstration in a Pinto) causing an attitude shift farther away from his previous position and thus deem the Pinto as more unacceptable (contrast effect).

Conversely, he might be a VW owner (B) who is low ego-involved and has placed the Pinto in his latitude of rejection (PR)' After a demonstration ride in the Pinto (a positive communication), it is likely that he would see the merits of the Pinto and now view it closely to those cars which he accepts (assimilation effect). The theoretical relationship of these expected shifts are presented in Figure 1.





A 2 x 3 (two levels of ego-involvement x three initial attitudes) between subject design was used. Subjects were self selected into the appropriate cell based on the results of an ego-involvement test. The subjects were classified as either low or high ego-involved. Their initial direction of attitude toward the Pinto was either within the region of acceptance, rejection, or noncommitment.


Fifty university students were selected for the study. The sample was divided among those who owned mini-compact cars and potential buyers of this size car. Car owners were randomly selected from the University's files of registered cars. Potential respondents were seniors and graduate students considering the purchase of a mini-compact car within the next year. Subjects were randomly selected from a list of senior and graduate students. These particular students were selected because they contained the highest proportion of potential new car buyers in comparison to other students. Potential mini-compact car buyers were determined by having the subjects admit it themselves. In no case did either the car owners or potential car owners have prior knowledge of the type of car they would be test driving. All subjects understood they would receive no monetary compensation.

Operationalization of the Independent Variable

Use of the Sherif model required a measure of ego-involvement. Photographs of twelve mini-compact cars were presented to the subjects. Each was taken from a then current issue of an automobile magazine. The photos were placed on white cardboard backgrounds with the car's name printed under the picture. Twelve cars were used to provide sufficient differences in establishing levels of ego-involvement.

An adaptation of the "Ordered Alternative Method" was used (Sherif and Sherif, 1969). After the subject studied the pictures he was put into a buyer's position by being asked to sort cars he would seriously consider buying if he had a chance to do so tomorrow. Next he was asked to sort those cars he definitely would not consider buying. The remaining cars were placed in a third pile. The cars the subjects would buy were operationally defined as the acceptance region (P"), the cars he would not buy his rejection region (PR), and the remainder the noncommitment region (PNC).

Level of ego-involvement is defined by the size of the latitude of acceptance, rejection, and noncommitment regions. Sherif et. al. (1969) state that as ego involvement increases, the size of the latitude of rejection increases compared to the latitude of acceptance. It was arbitrarily decided that those subjects with their latitude of rejection smaller than their latitude of acceptance would be considered low involvement subjects. All other subjects were placed in the high involvement category.

Experimental Procedure

The subjects selected for study reported to a central campus location for their individual sessions. Upon arrival, the subjects were seated and asked to complete the pre-questionnaire which consisted of identical semantic differential instruments for an ideal mini-compact car, the Pinto, and five additional mini-compact cars. The pre-questionnaire required 10 to 15 minutes to complete. As soon as the subject finished he was presented the pictures for the ego-involvement test and given instructions for accomplishing the sorts which took about five minutes. The subject was next introduced to another experimenter who presented positive stimuli about the Pinto and accompanied him on the demonstration drive. The test course was prearranged and standardized for all subjects. Total time for the drive ranged from 15 to 20 minutes. Every effort was made to disassociate the car from the experimenters thus limiting experimental demand effects. A series of statements were made to accomplish this task. A third experimenter administered the post questionnaire to the subject immediately after the demonstration drive. It was described to be similar to the pre-questionnaire. Ten minutes was usually sufficient to complete the post-demonstration drive questionnaire. Finally, the subjects were debriefed and dismissed.

Operationalization of the Dependent Variable: Attitude Change

A semantic differential instrument was used to determine the subject's attitude change toward the Pinto. The semantic differential was composed of 24 adjectival bipolar pairs. The scales used were mostly from Osgood's Thesaurus list (Osgood, Suci and Tannebaum, 1957), although some adaptation was required. The preliminary step in determining specific scales was taken from a tested mini-compact study (Ciotti, 1970) with the elimination of scales that showed no correlation with mini-compacts. Added in place were scales consisting of words considered to be the specific characteristics of mini-compact cars.

Individual scales were randomly assigned positions on the instrument. Then, adjectives were randomly positioned to eliminate constant scale polarity and reduce systematic bias. A similar set of scales was used for each car measured.

The questionnaires were randomly constructed by ordering the cars differently. This procedure was used to eliminate any order bias due to concept position.

The positions between each bi-polar pair were numbered from t to 7. The lower end of the scale denoted the position of the positive adjective. Attitude change of the Pinto was calculated by computing the differences of identical scales on the pre-demonstration questionnaire and the post demonstration questionnaire for each individual. The magnitude of the difference indicated the amount of attitude change. If the difference was a positive number, this indicated an assimilation effect. Alternately, a negative difference indicates a contrast effect.



The results are presented graphically in Figure 2 and summarized in Table 1. Comparison of Figures 1 and 2 indicate study results highly similar to the theoretical propositions. However, in this study, there has been a reversal of effects for individuals with the Pinto in the acceptable range with those for the Pinto in the rejection range. In addition, as compared to the theoretical propositions, there is a strong negative effect throughout.





For individuals who initially place the Pinto in their acceptable range (P"), and are low ego-involved in the product class, no significant change occurred (although slightly positive in direction). A relatively large contrast effect did occur for highly involved individuals with the Pinto initially in heir acceptable range.

For individuals who initially placed the Pinto in their noncommitment range (PNC), no significant change (although slightly negative) occurred for low involvement individuals. However, a large contrast effect did occur for highly ego-involved individuals.

The data indicates a strong assimilation effect for low involved individuals who placed the Pinto in the rejection region (PR). A moderate assimilation effect was obtained for highly involved individuals.

The pre-demonstration semantic differential instruments show that individuals in the acceptance region provided the most positive attitudes toward the Pinto, followed next by those in the noncommitted and finally those in the rejection region. This same relationship held for the post demonstration instruments.


The results are nearly a mirror image of what would be expected from a communication perceived as a positive stimuli. However, they are generally more negative (less assimilation and more contrast) than would have been predicted. This leads to the belief that the car demonstration may have actually been a negative stimuli. Evaluation of data from Pinto owners who were used as a control group gives some credence to the negative stimuli hypothesis. Assuming that riding in another Pinto would not significantly affect their attitude toward the Pinto, control subjects (Pinto owners) were tested. Results show a slight negative shift in attitude after the demonstration drive for these Pinto owners.

The Sherif Social Judgment Theory explains attitude change based on three factors: level of ego-involvement in an attitude, direction of attitude held, and nature of the stimulus. The nature of the stimulus may be positive or negative. The expectations initially stated were on the assumption of a positive stimuli. If the communication was considered a negative stimuli by these subjects, the results would be consistent with the theory.

The findings can be interpreted in the following manner:

First, individuals have few connections with the object and do not view the demonstration as a negative experience if they are in the acceptance range with low ego-involvement. In other words, the demonstration provided acceptable congruent information. The demonstration did not have either a significant positive nor negative effect. Secondly, for highly involved individuals, the message was discrepant to their prior attitudes. And, the information was discrepant enough to cause attitude change in a negative direction without being extreme enough in its content to cause the individual to reject the message and move the Pinto to an even more favorable position.

Finally, low ego-involved individuals in the rejection region are the most susceptible to change. The demonstration drive, although not a highly positive communication, was still positive to an individual initially rejecting the Pinto. Actually, it was a quite believable communication, thus causing a large positive shift in attitude. In the case of the highly involved, the assimilation effect occurred due to the somewhat acceptable nature of the communication. The change was not as great as in the case of the low involvement individual.

Therefore, although the results were not as originally expected, they nevertheless support the theoretical base of the Social Judgment Theory when the communication used is not strongly positive.


Marketers consider brand preferences to be based on differential attitudes and their communications attempt to alter these attitudes. Ego-involvement with the product is considered to be a major factor influencing attitude change. Although founded in Social Psychology, Sherif's Social Judgment Involvement Approach is applicable and has been adopted for this study to interpret attitude change based on a specialized type of communication.

It has often been assumed by the seller that if he can place his product before the eyes and ears of the consumer, the product will succeed or fail on its own merits. This attitude is often taken by the car salesman who assumes that the car will sell itself when driven by the prospect. This study shows that the car demonstration, like other forms of communication, is subject to assimilation and contrast effects as indicated in Sherif's model. Theory implies, as do our results, that demonstration drives can be harmful in some instances.

Future research in this area should concentrate on comparing the effects of positive and negative communications. Also, a test drive is a very unique form of communication, and while highly involving, it has many atypical characteristics. Therefore, other forms of persuasive communication could be utilized.

The Sherif's Social Judgment Involvement Approach was shown to be relevant to attitude changes created from automobile demonstrations. The study highlights problems in assuming that attempts to create positive communications from demonstrations are always favorable. Ego-involvement, as defined by the Social Judgment Approach, was shown to have important ramifications.


Alfred P. Ciotti, "Pinto Project," unpublished paper, The Pennsylvania State University, 1970.

Herbert E. Krugman, "The Impact of Television Advertising: Learning Without Involvement," Public Opinion Quarterly, 29 (Fall, 1965), 349-356.

Herbert E. Krugman, "The Measurement of Advertising Involvement," Public Opinion Quarterly, 30 (Winter, 1966), 583-596.

Herbert E. Krugman, "Brain Wave Measurements of Advertising Involvement," Journal of Advertising Research, 11 (February, 1971), 7-9.

Charles E. Osgood, George J. Suci, and Percy H. Tannebaum, Measurement of Meaning. (Urbana, Illinois: University of Illinois Press, 1957).

Muzafer Sherif and Carolyn W. Sherif, Social Psychology (New York: Harper and Row Publishers, Inc., 1969).

R. Tilman and C. A. Kirkpatrick, Promotion: Persuasive Communication in Marketing (Homewood, Illinois: Richard D. Irwin, Inc., 1968).

R. B. Zajonc, "The Concepts of Balance, Congruity, and Dissonance," The Public Opinion Quarterly, 163 (Summer, 1960), 280-296.