Belief Systems and the Differential Role of the Self-Concept

George E. Belch, (student), University of California, Los Angeles
ABSTRACT - This study investigates the relationship between belief systems and the differential role of self and ideal self-image in determining purchase intention. The results indicate that the level of conceptual functioning at which an individual operates appears to influence the differential role of the self-concept on purchase intentions. The findings also show that the purchase intention of individuals who are characterized by a high need for cognitive consistency are more influenced by product ownership than those individuals who are more tolerant of cognitive dissonance and inconsistency.
[ to cite ]:
George E. Belch (1978) ,"Belief Systems and the Differential Role of the Self-Concept", in NA - Advances in Consumer Research Volume 05, eds. Kent Hunt, Ann Abor, MI : Association for Consumer Research, Pages: 320-325.

Advances in Consumer Research Volume 5, 1978      Pages 320-325

BELIEF SYSTEMS AND THE DIFFERENTIAL ROLE OF THE SELF-CONCEPT

George E. Belch (student), University of California, Los Angeles

ABSTRACT -

This study investigates the relationship between belief systems and the differential role of self and ideal self-image in determining purchase intention. The results indicate that the level of conceptual functioning at which an individual operates appears to influence the differential role of the self-concept on purchase intentions. The findings also show that the purchase intention of individuals who are characterized by a high need for cognitive consistency are more influenced by product ownership than those individuals who are more tolerant of cognitive dissonance and inconsistency.

INTRODUCTION

Relationships of product image and self-image have been given much attention in consumer research. The idea that consumers prefer products or brands that are congruent with their self-concept has been investigated by many researchers including Birdwell (1968), Grubb and Hupp (1968), Dolich (1969), Ross (1971), and Landon (1974). [Self-concept, as used in this article, refers to both the "self-image" and "ideal self-image" of the consumer. The latter two terms are viewed as the two components of the self-concept and will generally be given specific reference.] The basic assumption of these studies has been that the self-concept is related to the manner in which consumers perceive and purchase products. To the extent that the self-concept is of importance to the individual, he will direct his behavior towards maintenance and enhancement of his self-concept.

Evidence supporting a congruence relationship between self-image and purchase behavior has been reported by Birdwell (1968) and Grubb and Hupp (1968). These studies dealt with the consumers matching of product-image with self-image. However, the consumer might be attempting to express and "ideal" self-image through his purchase of a particular product or brand rather than his "actual" self-image. Hamm and Cundiff (1969) found that individuals with a large discrepancy between actual and ideal self-image describe themselves differently in terms of products and have different product perceptions from those individuals with a small discrepancy between the two.

Several studies have examined the congruence between self and ideal self-concept and products or brands. Dolich (1969) found that there was a greater congruence between self-concept and brand most preferred than brand least preferred. He also found that both self-and ideal self-image were almost equally congruent with most preferred brand.

Ross (1971) also investigated the role of self-image and ideal self-image in consumer purchase behavior. He hypothesized that individuals would prefer products which they perceived to be similar to their self-concept. Ross also predicted that in some cases self-image would best describe brand preference. He hypothesized that ideal self-image would be more related to brand preference when the product was more, rather than less, conspicuous to others, and conversely, self-image would be more closely related to brand preference for less conspicuous products.

The results of his study strongly supported the prediction concerning subject's preference for brands which were similar to their self-concept. However, little support was found for the predictions regarding the role of self- versus ideal self-image and brand preference. Ross concluded that these results were supportive of a simple congruity relationship between self-concept and brand preference.

Landon (1974) examined the differential role of the self- and ideal self-image and purchase intentions. In this study he was interested in consumer tendencies to match product image with self-image and/or ideal self-image. Landon hypothesized that some individuals would show a dominant influence of self-image/purchase intention correlation over all products, or an actualizing tendency, while some individuals would show a dominant influence of ideal self-image, a perfection tendency. The actualization model views the individual as striving to achieve harmony between his environment and his self-concept by coming to know and accept his self-concept, while the perfection model states that the individual is constantly trying to improve on his self-concept, i.e., achieve his ideal self-concept (Landon, 1974).

The results of Landon's investigation showed that factors related to products and to the individual are important determinants of the relative impact of self-image and ideal self-image on purchase intentions. Over all subjects, purchase intention for some products tends to be more correlated with ideal self-image while for some products purchase intention tends to be more correlated with self-image. Over all products, some subjects were characterized by a higher self-image/purchase intention correlation (actualizers) while some tended to be characterized by a higher ideal self-image/purchase intention correlation (perfectionists). However, neither the actualization nor the perfection hypothesis explained all of the data and Landon concluded that further research is needed to focus on conditions which might explain when each model is likely to be operative.

These studies give some indication that the self-concept is related to product perception. However, it is evident that there is still a great deal of uncertainty concerning the relative roles of the self- and ideal self-image. The findings thus far indicate that self-and ideal self-image are related and may be equally congruent with brand preference, or in some cases brand preference may be more related to one than the other.

The questions of why either self-image or ideal self-image is more related to purchase behavior and why there is a high correlation between the two self-concepts are still unanswered. One possible explanation for the relationship between self-concept and purchase behavior, as well as the degree of real-ideal self-discrepancy, lies in the level of conceptual functioning at which an individual operates. Harvey, Hunt, and Schroder (1961) have developed four belief systems, each representing a different level of conceptual functioning along a concreteness-abstractness dimension. Given the varying characteristics of these four belief systems, each group might behave differently in respect to using either their self-image or ideal self-image as a guide to their purchase behavior for certain products. The purpose of this paper is to examine the possible relationships between belief systems and the differential role of the self-concept in consumer behavior. First, however, it is necessary to examine the general nature and function of belief systems and how they serve as different systems of motivation.

BELIEF SYSTEMS

A belief system represents a set of predispositions within an individual to perceive, construe, and interpret stimuli or events in a consistent manner. Each individual has certain central or core concepts which guide him in his efforts at making his world consistent. Individuals have conceptual systems which are ways of construing or dimensionalizing relevant aspects of one's world. These systems are characterized by some form of structure or relation among the various parts of the system.

Individuals are believed to develop conceptually through stages, from the undifferentiated and concrete to the more integrated and abstract. The more concrete end of the dimension represents a state of minimal differentiation while abstractness is represented by high differentiation and integration. Individuals will vary according to the degree of differentiation and relativism they attain and will tend to structure their environment according to their own particular conceptual belief system. The variation in differentiation and relativism determines the openness and closedness of the system, its receptivity to deviant elements and its capacity to admit impingements from the outside. Greater abstractness can be viewed as greater openness and greater tolerance of the different and novel (Harvey, 1966).

Harvey, Hunt and Schroder (1961) presented extensive theoretical and empirical bases from which four basic levels of concreteness-abstractness were deduced. The four basic levels were treated as different conceptual systems, each representing a different level of functioning between the hypothetical and points of greater concreteness and abstractness. A brief review of the four belief systems is given before turning to the relationship of conceptual systems to self-concept.

General Characteristics of the Four Systems

System 1. System 1 functioning represents the most concrete mode of dimensionalizing and construing the world. There is a simple cognitive structure with regard to domains of high involvement and this structure is fairly undifferentiated and poorly integrated. Representatives of System 1 show a greater tendency towards extreme and polarized Judgments. This system is assumed to have evolved from situations in which individuals have been restricted in the exploration of their environment.

System 1 representatives tend to manifest such characteristics as high absolutism and closedness of beliefs, high evaluativeness, strong adherence to rules, high ethnocentrism, dogmatism and authoritarianism (Harvey, 1966). They also have a tendency to hold beliefs in accordance with clear cut definitions. Representatives of this system have a low tolerance for ambiguity and are negatively aroused by cognitive inconsistencies. This level of functioning is disposed towards system maintenance by excluding potentially conflicting stimuli from the system.

System 2. System 2 functioning is somewhat more differentiated and abstract than System 1. This level of functioning is characterized by negativism and an anti-rule, anti-authority orientation. Members of System 2 tend to be low in self-esteem and high in alienation and cynicism. This system is assumed to result from experiences with an authority who acts omnisciently and is capricious in the control of rewards and punishment. The arbitrary and unpredictable manner in which these sanctions are administered renders the individual unsure of whether courses of action will be reinforced or punished. This results in higher differentiation than System 1, but also produces a feeling of uncertainty, distrust of authority, and rejection of social norms and guidelines. Representatives of System 2, like those of System 1, have a high need for structure and their tolerance for ambiguity is low. They are also high in dogmatism and absolutism.

System 3. System 3 functioning represents the next to highest level of abstractness treated by Harvey et. al. This system is assumed to result from an environment of over-protection. One or both of the parents act as a buffer between the developing child and his environment, thus preventing him from exploration of his physical and social world. System 3 individuals are fairly proficient in affecting desired outcomes in their environment by having others do things for them.

System 3 individuals have a great need for interaction with others and are concerned with establishing friendships, intragroup consensus, and developing dependency relations in order to avoid the feeling of helplessness and social isolation that would result from their being on their own. They need constant feedback from meaningful people in their environment in order to guide their behavior and attain the acceptance they need. System 3 representatives also develop more positive ties to prevailing social norms than representatives of the other three systems.

System 4. System 4 individuals represent the most abstract and open-minded of the four belief systems. This system is viewed as evolving from an environment in which the child is encouraged to explore his physical and social worlds and to establish and rely on values derived from his own experiences and thought. Members of this system are characterized by a high task orientation, risk taking, creativity, and relativism. They are more tolerant of ambiguity and cognitive inconsistency or conflict and are more flexible and contextual in thought and action.

System 4 individuals have a highly differentiated and integrated cognitive structure. Their set of internal standards are independent of external criteria, in some cases coinciding with social definitions, and in others not.

BELIEF SYSTEMS AND SELF-CONCEPT

An individual's belief system will have an effect on his perceptions, cognitions, and actions and will affect his psychological processing and behavior. It is likely that the acts of individuals involved in some type of decision process will be affected by their system of beliefs and concepts. Thus, individuals of each system should manifest variations relative to their perceptions, beliefs, and actions.

Considering the varying characteristics of each system, there should be differences in the manner in which self-concept operates for each level of conceptual functioning. Specifically, this investigation is concerned with the differential role of the self-concept (i.e., self- and ideal self-image) in determining purchase intention for the four systems.

The relative importance of self- and ideal self-image to purchase intention and the discrepancy between the two components of self-concept may in part be due to the level of conceptual functioning of the individuals. For example, members of the first two systems are characterized by concrete functioning. They have a difficult time acting "as if" and role playing. These individuals might fail to see any real discrepancy between how they are (self-image) and how they would like to be (ideal self-image). Moreover, any discrepancy between the two self-concepts might tend to create a state of cognitive inconsistency for System 1 and 2 individuals. Thus, people in these systems would have a small discrepancy between self- and ideal self-image. Members of Systems 3 and 4 should also show congruence between the two self-concepts. However, the discrepancy between self- and ideal self-image for these groups should be larger than for the first two systems.

The purchase intentions of System 3 individuals should be strongly influenced by their ideal self-image. Members of this system have a strong need for affiliation and friendship and will be constantly striving for the attainment of an ideal self-image that will enhance their position among their peers. Members of this system are concerned with social group consensus and peer group influences and may see a certain product and its attributes as a manifestation of the image they hope to project.

System 4 representatives represent the most abstract functioning of the four belief systems. The developing environment of these persons, having been free and open, has offered them an opportunity to explore and possibly discover their self-concept. This group may be more cognizant of their self-image and would no longer feel a need to seek an ideal self-image. Considering the ability of System 4 individuals to be contextual in their thinking, they may be more capable of relating the attributes and image offered by a product to their self-image.

Of primary interest in this study are the relationships between self-image and ideal self-image for the four systems and the role of each in determining purchase intentions for each group. The following research hypotheses were formulated based on the different characteristics of the four conceptual systems.

H1: The discrepancy between self- and ideal self-image will be smaller for Systems 1 and 2 than Systems 3 and 4.

H2: System 3 representatives will show the greatest influence of ideal self-image on purchase intentions of the four belief systems.

H3: System 4 representatives will show the greatest influence of self-image on purchase intentions of the four belief systems.

Also of interest in this study is the manner in which product ownership will affect the four systems. Belch and Landon (1977) found that product ownership affects the self-concept ratings for a product, increasing the likelihood of a product being rated as congruent with both self- and ideal self-image. If a true assessment of the role of self-concept in determining purchase intention is to be made, it is necessary to account for the effect of product ownership.

Considering the variations in the systems, particularly with respect to the need to maintain cognitive consistency, they might be affected differently by product ownership. Previous ownership of a product might influence the self-concept scores for that product or affect the purchase intentions of the individual for that product. As noted by Evans (1968) congruity between ownership and purchase intentions might result from a post-purchase process such as dissonance reduction. Individuals who are less tolerant of cognitive inconsistencies, such as members of System 1 and 2, may be influenced more by product ownership than those who can handle the cognitive conflict. The following hypothesis was formulated to test the effect of product ownership across the four systems:

H4: Product ownership will have a greater influence on purchase intentions for Systems 1 and 2 than for the other two systems.

METHOD

Measurement of Belief Systems

The instrument used in this study has been developed by Harvey et. al. specifically as a measure of belief systems. The "This I Believe Test" (TIB) is a semi-projective sentence completion test. The TIB requests the individual to indicate his beliefs about a number of socially and personally significant concept referents by completing in two or three sentences the phrase, "This I Believe About ", the blank being replaced by such referents as religion, friendship, people, or the American way of life among others.

By evaluating subject's responses with regard to several dimensions, including structure and content, subjects can be classified into one of the four belief systems discussed earlier or into an admixture of two or more systems. Test-retest reliabilities for the TIB within one week and after six months have been in the .80's while the interjudge reliability for trained judges in classifying subjects into one of the four systems has been .90 or greater (Harvey, 1966). For this study, subject's responses on the TIB were read and scored by professional readers in the department of Psychology at the University of Colorado. Subjects were classified as being representative of one of the four systems or as admixtures. Only those subjects whose responses classified them as being representative of a pure system were used in this study.

Measurement of Product Perception and Purchase Intentions

This study measured product perception relative to self-image and ideal self-image using a product anchored self-concept measure. This measure yields a congruence relationship between self-image and product-image. Subjects are asked to rate each product on a nine point scale ranging from "very much unlike me" to "very much like me". Ideal self-image for each product was then measured on a scale ranging from "very much unlike I want to be" to very much like I want to be". A similar measure was used by Landon (1971, 1974) for measurement of product perception relative to the self-concept.

Purchase intentions for each product were measured using a five point scale. The first four points on the scale divided a reasonable time range into four intervals. The time dimension for each product varied according to the type of product. The fifth scale point was used to indicate an intention never to buy the product. Product ownership was determined by having the subjects indicate which of the products they owned at the time of the study.

Subjects and Procedure

Subjects for the study were male students enrolled in undergraduate marketing courses at the University of Colorado. The TIB and questionnaire were administered during the regular class periods of each course. After everyone completed the TIB, the questionnaire was administered. Subjects were given twenty minutes to complete the questionnaire. When everyone was finished, the TIBs and questionnaires were collected and the class was given an explanation of the study.

The twelve products used in this study included coffee, after-shave lotion, adult games, imported wine, snow skis, mouthwash, country club membership, sun tan lotion, beer, TV dinners, and deodorant. These products were chosen from a larger list of products used by Landon (1971). Only those products which had a test-retest correlation of .60 or greater on all three measures were used in this study. Two forms of the questionnaire were given, the first presenting the products in a randomized order, and the second a reverse of the original set.

Four questionnaires were eliminated due to homogeneity of responses by the subjects over all items in the questionnaire. After eliminating these subjects and the 36 admixtures, 124 subjects remained in the sample. The number of subjects in each system was as follows:

System 1 = 53, System 2 = 29, System 3 = 18, and System 4 = 24.

RESULTS

The primary interest in this study is the variation in the differential role of the self-concept among the four systems. Therefore, the analyses were designed to focus on the individuals comprising each system. The approach taken in some studies (e.g. Ross, 1970) has been to use the averaged scores of the sample over each product. This method is not as useful for revealing group differences since these variations may be confounded or even lost in the aggregation process. The analyses for this investigation were performed by treating an individual's scores on each product as single data points. By focusing on the scores of individual's comprising each system over the twelve products, it is possible to determine whether any differences are, in fact, due to the various systems rather than the products.

In order to test the hypothesis concerning the discrepancy between the two components of self-concept, difference scores between self- and ideal self-image were computed for each subject for each of the twelve products. The discrepancy scores were used as criterion variables in a multivariate Hotelling's T2 test which is appropriate for testing the significance of the differences between the two self-concept ratings. [For a description of Hotelling's T2 test, see Tatsuoka (1971), p.81.] This test uses a sample mean difference vector and tests whether the population mean difference vector is significantly different from zero.

This statistic was computed separately for each of the four systems. Table 1 shows the T2, F statistics and significance level for each system. The F ratios exceed the .05 level of significance for Systems 1, 2 and 4 indicating that for these three systems the difference between self- and ideal self-image is significant. For System 3, however, the discrepancy between the two components of self-concept is not significant. These results are not supportive of the first hypothesis. In fact, the difference between self- and ideal self-image appears to be greater for Systems 1 and 2 rather than Systems 3 and 4.

TABLE 1

RESULTS OF HOTELLING T2 TEST FOR EACH SYSTEM

To test the hypotheses concerning the differential role of self- and ideal self-image and the effect of product ownership, multiple regressions were run for each system with purchase intention as the dependent variable and self-image, ideal self-image and product ownership as independent variables. Subjects' scores for each product on the four measures were used in the regression equations. Thus, for each system the number of observations is equal to the number of subjects in that group times twelve, the number of products.

By examining the regression coefficients of the three variables it is possible to test the hypotheses concerning the differential role of self-concept and the effect of product ownership on purchase intentions across the different systems. The use of regression analysis in this study is not designed to imply causality but only to determine if there are differences between the systems relative to the association of self- and ideal self-image and purchase intentions. This analysis also allows for a better description of this relationship since it takes into account the effect of product ownership.

Table 2 shows the results of the regression run for each system. The regression coefficients, adjusted R2, and the significance level of each variable are presented in this table. [Both b values (non-standardized) and Betas (standardized) are shown in Table 2. However, the discussion uses b values since this coefficient is required for a subsequent analysis reported below.]

TABLE 2

REGRESSION COEFFICIENTS FOR SELF-IMAGE, IDEAL SELF-IMAGE AND PRODUCT OWNERSHIP FOR EACH SYSTEM

Hypotheses 2 and 3 dealt with the role of self- and ideal self-image in determining purchase intention. More specifically, these hypotheses seated that System 3 representatives will show the greatest relationship between ideal self-image and purchase intention among the four systems, while System 4 will have the strongest relationship between self-image and purchase intention of the four groups. As can be seen in Table 2, the regression coefficient for ideal self-image is the highest for System 3 while the coefficient for self-image is the greatest for System 4. These results are consistent with the hypotheses concerning the differential role of the self-concept across the four systems.

The final hypothesis predicted that product ownership will have a greater influence on purchase intentions for Systems 1 and 2 than the other two groups. Examination of the regression coefficients for product ownership across the four systems (Table 2) supports this hypothesis. Product ownership has large regression weights in Systems 1 and 2 while for Systems 3 and 4 these regression coefficients are smaller in magnitude.

The findings discussed thus far are supportive of the hypotheses concerning the differential role of self- and ideal self-image and the effect of product ownership among the four systems. However, to determine whether or not these results are statistically meaningful, all of the data were pooled in an overall regression equation designed to test for the significance of the differences in these specific regression coefficients. The results of this test are shown in Table 3. [The b values are used in this analysis since the differences in the standard deviation of the variable in Table 3 (e.g., SC and SCD4) makes the Betas non-additive.]

TABLE 3

REGRESSION COEFFICIENTS FOR SELF-IMAGE, IDEAL SELF-IMAGE AND PRODUCT OWNERSHIP ALL SYSTEMS COMBINED

In this analysis dummy variables representing specific systems are multiplied with the original variables and included in the regression analysis. The variables SCD4, ISCD3, and OWND5 are cross products between variables SC (self-image), ISC (ideal self-image), and OWN (product ownership) and dummy variables D4, D3, and D5, respectively. Dummy variable D3 represents System 3, D4 System 4, and D5 systems 1 and 2. If the coefficients for the cross product variables are statistically significant, this means that the difference in the value of the regression coefficient for a specific system from the other systems is significant.

For example, self-image (SC) in general has a regression coefficient of .170. However, the coefficient of self-image for System 4 is somewhat greater than the other systems, specifically by .064 which is the value for the coefficient of SCD4. Moreover, this difference is significant beyond the .01 level. The difference in the value of the regression coefficient for ideal self-image between System 3 and the other systems (ISCD3) is .068. This difference is also significant beyond the .01 level.

These results offer statistical support for hypotheses 2 and 3. The influence of self-image on purchase intentions across the four systems is the greatest for System 4 while ideal self-image has the strongest influence for members of System 3.

The differences in the influence of product ownership across the four systems are also tested in this final equation. Overall, the effect of product ownership (shown by OWN in Table 3) is not significant. However, the difference in the influence of ownership for Systems 1 and 2, represented by OWND5 in Table 3, is quite large (.403). This difference is also significant beyond the .01 level. Clearly, product ownership influences purchase intention for Systems 1 and 2 much more than Systems 3 and 4.

DISCUSSION

The results of the present study are interesting in several respects. The findings suggest that predispositional factors related to the level of conceptual functioning of an individual are related to the differential influence of self- and ideal self-image on purchase intentions. It also appears that the impact of product ownership on purchase intentions is mediated by an individual's level of conceptual functioning.

The evidence shows that for Systems 1, 2 and 4 the difference between self- and ideal self-image is significant. However, for System 3 this discrepancy is smaller and not statistically significant. Concreteness of conceptual functioning does not appear to be an explanation for any congruence between the two components of the self-concept.

One possible explanation for the unpredicted large discrepancy between the components of self-concept for Systems 1 and 2 may be that these two groups are void of a well-developed, ideal self-image. Their concrete level of functioning might dispose them towards viewing the "self-image" as the relevant and meaningful component of their self-concept. However, given the nature of the task, which requires subjects to rate each product first with respect to self-image then ideal self-image, representatives of these two systems are confronted with the problem of defining their ideal self-image. If it is not initially well defined, the ideal self-image for these two systems might be more susceptible to a "contrasting effect", wherein the more well-defined self-image serves as an anchor against which the less well-defined ideal self-image is contrasted. This would result in a large discrepancy between self- and ideal self-image for System 1 and 2.

There are, however, differences among the four systems with respect to the influence of either self- or ideal self-image on purchase intentions. Across the four groups, ideal self-image has the greatest impact on purchase intentions for System 3. The conceptual functioning for System 3 representatives predisposes them to be more concerned with such factors as social group consensus and peer group influence. They also have a strong need for affiliation and friendship and tend to develop more positive ties to prevailing social norms. These characteristics might motivate the System 3 individual to attempt to enhance his self-concept through reliance on an image that he feels is consistent with his idealized self. The ideal self-image will be the source of direction for such persons as they search for approval and guidance from others. This suggests that factors such as other-directedness, social desirability of product or brand, and reference group influence might be important determinants of the role of the ideal self-image in guiding purchase behavior.

The results also indicated that System 4 individuals show the strongest influence of self-image on purchase intentions of the four systems. This domination of self-image for System 4 suggests that persons who have had the opportunity to explore and become aware of their self-concept might tend to rely on a relevant self-image as a guide to their purchase intentions. The conceptual functioning of the System 4 individual may be more in line with the actualization model of personality discussed by Landon (1974). As mentioned previously, this model views the individual as striving to achieve harmony between his environment and his self-concept by coming to know and accept his self-concept.

The finding that product ownership had the greatest influence on Systems 1 and 2 is particularly interesting. The possible confounding effects of post-purchase measures is an important issue in self-concept research. The congruity principle of Birdwell (1968) was challenged by Evans (1968) on the grounds that it did not test causality since congruence was measured after the purchase decision was made. Evans argued that post-purchase data does not necessarily reflect purchase motivations due to self-image/product-image congruity. After purchase congruity or indication of future purchase intentions might result from a process such as dissonance reduction.

The concreteness of conceptual functioning of System 1 and 2, which disposes them toward a lower tolerance for dissonance, may motivate them to give more consideration to the fact that they presently own the product under consideration. The greater influence of product ownership for these two systems might reflect a way of maintaining cognitive consistency between past and future purchase behavior. Other empirical evidence (Harvey, 1965; Ware and Harvey, 1967; Harvey and Ware, 1967) has shown that dissonance and consistency theories have greater validity for concretely functioning than abstractly functioning individuals. These results appear to be consistent with this evidence.

The purpose of this paper was to determine if belief systems can offer insight into the differential role of the self-concept in determining purchase intentions. The present study has shown that factors related to an individual's level of conceptual functioning are correlates of the differential influence of self- and ideal self-image on purchase intentions. The conceptual system of the individual appears to mediate the relationship between self-concept and purchase behavior.

This study also indicates that product ownership may affect the self-concept/purchase behavior relationship. The significant influence of ownership on the first two systems suggests that further research might examine the effect of dissonance or other consistency theories on the self-product-image congruity of consumers.

Future research on the role of the self-concept may find it fruitful to concentrate on individual differences such as those suggested in this study and their relationship to the differential role of the self-concept. Finally, this study has shown that belief systems can provide some interesting insights into the behavior of the consumer. The conceptual systems approach of Harvey et. al. may have potential for other areas of research in consumer behavior.

REFERENCES

George E. Belch and E. Laird Landon, Jr., "discriminant Validity of a Product-Anchored Self-Concept Measure," Journal of Marketing Research, 14 (May 1977) 252-6.

Al. E. Birdwell, "A Study of the Influence of Image and Congruence on Consumer Choice," Journal of Business, 41 (January 1968), 76-88.

Ira J. Dolich, "Congruence Relationships Between Self Images and Product Brands," Journal of Marketing Research, 6 (February 1969), 80-4.

Franklin B. Evans, "Automobiles and Self-Imagery: Comment," Journal of Business, 41 (October 1968), 445-59.

Edward L. Grubb and Gregg Hupp, "Perception of Self, Generalized Stereotypes and Brand Selection," Journal of Marketing Research, 5 (February 1968), 125-7.

O. J. Harvey, "Some Situational and Cognitive Determinants of Dissonance Resolution," Journal of Personality and Social Psychology, 1965, 1, 349-55.

O. J. Harvey, "System structure, flexibility and creativity,'' in O. J. Harvey (Ed.), Experience, Structure and Adaptability, New York: Springer, 1966.

O. J. Harvey, D. E. Hunt, and H. M. Schroeder, Conceptual Systems and Personality Organization, New York: Wiley, 1961.

O. J. Harvey and R. Ware, "Personality Differences in Dissonance Resolution," Journal of Personality and Social Psychology, 1967, 5, 38-44.

E. Laird Landon, Jr., "Need for Achievement, Self-Concept, and Product Perception," unpublished doctoral dissertation, University of California, Los Angeles, 1971.

E. Laird Landon, Jr., "The Differential Role of Self-Concept and Ideal Self-Concept in Consumer Purchase Behavior," Journal of Consumer Research, 1 (Sept. 1974).

Ivan Ross, "Self-Concept and Brand Preference," Journal of Business, 44, (Jan. 1971).

Maurice M. Tatsuoka, Multivariate Analysis: Techniques for Educational and Psychological Research, John Wiley and Sons, Inc. 1971.

R. Ware and O. J. Harvey, "A Cognitive Determinant of Impression Formation," Journal of Personality and Social Psychology, 1967, 5, 38-44.

----------------------------------------