Self-Concept and Brand Preference

Self-Concept and Brand Preference
Author(s): Ivan Ross

Source: The Journal of Business, Vol. 44, No. 1 (Jan., 1971), pp. 38-50
Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/2351834
Accessed: 09-11-2016 01:32 UTC
REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/2351834?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to
The Journal of Business
This content downloaded from 134.83.1.244 on Wed, 09 Nov 2016 01:32:37 UTC
All use subject to http://about.jstor.org/terms
SELF-CONCEPT AND BRAND PREFERENCE
IVAN ROSS*
INTRODUCTION might lend support to this hypothesis?

PURPOSE
Empirical studies are few.
Birdwell2 predicted that "an automo-
The basic purpose of this research is
bile owner's perception of his car is es-
to empirically test one of the "self-evi-
sentially congruent with his perception of
dent" truths in contemporary theorizing
himself" and that "the average percep-
about consumer behavior-that people
tion of a specific car type and brand is
go about purchasing one thing or another
different for owners of different sorts of
only if these things are consistent with,
cars" (p. 292). He found support for
enhance, or in some other way fit well
both of these hypotheses and concluded
with the conception they have of them-
selves. Secondarily, the research seeks to that "automobiles are often extensions
distinguish the role of two different of the owner's image of self" (p. 303).
modes of self-concept as they relate Grubb and Hupp3 predicted that "con-
to consumer behavior-"actual self-consumers of a specific brand of a product
cept," the way a person actually ("real- would hold self-concepts similar to the
ly") sees himself to be, and "ideal self- self-concepts they attribute to other con-
concept," the way a person would ideally I See F. E. Fiedler, F. J. Blaisdell, and W. G.
("like to") be. Warrington, "Unconscious Attitudes as Correlates
of Sociometric Choice in a Social Group," Journal of
BACKGROUND Abnormal and Social Psychology 47 (1952): 790-96;
R. M. Lundy, W. Katkousky, R. L. Cromwell, and
The theoretical basis for these notions D. J. Shoemaker, "Self Acceptability and Descrip-
derives from the interpersonal attraction tions of Sociometric Choices," Journal of Abnormal
literature in social psychology. A great and Social Psychology 51 (1955): 260-62; A. J.
Smith, "Similarity of Values and Its Relation to
deal of research suggests that people tend Acceptance and the Projection of Similarity,"
to perceive others whom they like as Journal of Psychology 43 (1957): 251-60; Elizabeth
being more similar to themselves than Alfert, "Two Components of Assumed Similarity,"
Journal of Abnormal and Social Psychology 66
those whom they dislike. Similarly, peo- (1958): 135-38; V. H. Vroom, "Projection, Nega-
ple tend to like others whom they per- tion, and the Self Concept," Human Relations 12
ceive as being more similar to themselves(1959): 335-44; Dorothy M. Kipnis, "Changes in
Self Concepts in Relation to Perceptions of Others,"
than those who are less similar.' Journal of Personality 29 (1961): 449-65; T. M.
Transferring these findings to the con-Newcomb, The Acquaintance Process (New York:
sumer behavior context, then, leads to Holt, Rinehart & Winston, 1961); and Jane A.
Broxton, "A Test of Interpersonal Attraction Pre-
the general prediction that people should dictions Derived from Balance Theory," Journal of
prefer brands which they perceive as Abnormal and Social Psychology 66 (1963): 394-97.
being the sort of brands that would be 2 A. E. Birdwell, "Influence of Image Congruence
consumed by a person who is "similar" on Consumer Choice," in Reflections on Progress in
Marketing, ed. L. George Smith (Chicago: American
to themselves.
Marketing Association, 1965), pp. 290-303.
What prior findings are there which
3 E. L. Grubb and G. Hupp, "Perception of Self,
* Currently associate professor, School of Busi- Generalized Stereotypes, and Brand Selection,"
ness Administration, University of Minnesota. Journal of Marketing Research 5 (1968): 58-63.
3$
SELF-CONCEPT AND BRAND PREFERENCE 39
sumers of the same brand. Further, con- example, Grubb and Grathwohl9 simply
sumers of a specific brand would hold propose that "the consuming behavior
self-concepts significantly different from
of an individual will be directed toward
self-concepts they attributed to con- the furthering and enhancing of his self-
sumers of a competing brand" (p. 59). concept through the consumption of
They used automobiles as the test prod- goods as symbols" (p. 26).
uct, and their results supported the hy- It was a secondary purpose of this re-
potheses. search to clarify the mode or perspective
of self that obtains in this relationship.
ACTUAL VERSUS IDEAL SELF-CONCEPT
In this regard, the researcher hypothe-
But one of the crucial questions seem- sized that in some cases ideal self would
ingly ignored by prior empirical research best describe the brand preferences of
has to do with the kind or perspective consumers while in other cases actual
of self-concept that mediates the rela- self would have the best fit. The product
tionship between self and brand prefer- dimension that was hypothesized to be
ence. Martineau4 and Britt5 suggest that critical in determining which perspective
"ideal self-concept," the way a person of self-concept would be most explana-
wants to be or would like to see himself, tory was the conspicuousness with which
more closely corresponds to or explains the product was consumed.
his consumption preference than "actual When a person engages in an act of
self-concept," the way a person says he consumption because he intends to im-
is or "really" sees himself. Dichter6 and press others by that act of consumption,
Grossack and Schlesinger,7 on the other is he more concerned with portraying his
hand, imply that "actual self-concept" actual or his ideal self? Conversely, when
more closely corresponds to or explains he consumes something because it satis-
his preferences. Levy8 maintains that fies him whether or not others are ob-
both actual and ideal self-concepts are serving his consumption, is he more con-
important ("to behave in ways that are cerned with his actual or his ideal self?
consistent with a set of ideas he has I hypothesized that ideal self-concept is
about the kind of person he is or wants more closely related to consumption pref-
to be" [p. 410]). erence than actual self-concept when
In general, most researchers make no the object of consumption is more rather
mention at all of this distinction. For than less conspicuous to others and, con-
versely, that actual self-concept is more
4 P. Martineau, Motivation in Advertising (New
York: McGraw-Hill Book Co., 1957), p. 52. closely related to consumption preference
5 S. H. Britt, The Spenders (New York: McGraw- than ideal self-concept when the object
Hill Book Co., 1960), pp. 105-6. of consumption is less rather than more
6 E. Dichter, Handbook of Consumer Motivations conspicuous to others.
(New York: McGraw-Hill Book Co., 1964), p. 6.
SELF-CONCEPT AND PERSONALITY
7 M. M. Grossack and L. Schlesinger, "The
Social Scientist and New Product Development," in Self theorists essentially define self-
Understanding Consumer Behavior, ed. M. M. Gros-
sack (Boston: Christopher Publishing House, 1964),
concept as an attitude one holds about
pp. 253-63, esp. p. 255. or toward one's person (self), this at-
8 S. J. Levy, "Symbols by Which We Buy," in I E. L. Grubb and H. L. Grathwohl, "Consumer
Advancing Marketing Efficiency, ed. Lynn H. Stock-Self-Concept, Symbolism and Market Behavior:
man (Chicago: American Marketing Association, A Theoretical Approach," Journal of Marketing 31
1959), pp. 409-16. (1967): 22-27.
40 THE JOURNAL OF BUSINESS
titude consisting of cognitive compo- are inferred to exist by others from the
nents (knowledge, belief), affective com- individual's responses to stimuli purport-
ponents (evaluations), and behavioral- ing to measure these constructs). The
motivational components (predisposi- former is what he means by self-concept;
tions or tendencies to respond). One com- the latter more generally describes that
monly quoted definition of self-concept which is called personality. It is intuitive-
is by Combs and Snygg: "By concepts ly obvious that a person may not per-
of self we mean those more or less dis- ceive himself (self-concept) in the same
crete perceptions of self which the indi- way that some personality test describes
vidual regards as part, or characteristic him. A personality inventory may classi-
of, his being. They include all percep- fy one as being extremely neurotic-but
tions the individual has differentiated as that person may not describe himself
descriptive of the self he calls I or me."'0 that way at all.
Some theorists, particularly Sym- In my opinion, much of the confusion
onds," Chein,12 and Murphy,'3 prefer generated in the consumer behavior re-
to discuss the motivational components search literature relating personality and
of self as ego functions rather than as consumer behavior derives from a mud-
self functions, although the distinction isdying of this distinction. As a matter of
somewhat ambiguous. In fact, Sherif fact, most of the empirical studies have
and Sherif do not make any distinction dealt with measures of inferred self.'5
whatsoever: "Ego or self is a develop- Yet, many of these researchers in talking
mental formation in the psychological about their research findings imply that
make-up of the individual consisting ofthey are discussing "self" and consumer
interrelated attitudes which are acquired behavior. This is not correct.
in relation to his own body, to objects,
HYPOTHESES
family, persons, groups, social values,
and institutions and which define and The individual seeks to maintain con-
regulate his relatedness to them in con- gruency between the way in which he
crete situations."''4 views himself (self-concept) and the way
This investigator would make a dis- in which he views objects of consumption
tinction between perceived self (self asinitterms of his preferences for them.
is phenomenally perceived and reported 1. The individual will prefer to con-
by the perceiver) and inferred self (assume a product or brand of product which
pects of self and/or personality which is perceived by him to be more rather
than less similar to his own self-concept.
II A. W. Combs and D. Snygg, Individual Be-
havior: A Perceptual Approach to Behavior, rev. ed. 15 For example, F. B. Evans, "The Brand Image
(New York: Harper & Bros., 1959), p. 10. Myth," Business Horizons 4 (Fall 1961): 19-28;
11 P. M. Symonds, The Ego and the Self (New F. B. Evans, "Correlates of Automobile Shopping
York: Appleton-Century-Crofts, Inc., 1951). Behavior," Journal of Marketing 26, no. 4 (1962): 47-
77; R. Westfall, "Psychological Factors in Predicting
12 I. Chein, "The Awareness of Self and the
Product Choice," Journal of Marketing 26 (1962):
Structure of the Ego," Psychological Review 51
(1944): 304-14.
34-40; A. Koponen, "Personality Characteristics of
Purchasers," Journal of Advertising Research 1
13 G. Murphy, Personality: A Biosocial A approach
(1960): 6-12; P. C. Vitz and D. Johnston, "Mascu-
to Origins and Structure (New York: Harper & linity of Smokers and the Masculinity of Cigarette
Bros., 1947).
Images," Journal of Applied Psychology 49 (1965):
14 M. Sherif and Carolyn W. Sherif, An Outline of
155-59; and W. T. Tucker and J. J. Painter, "Per-
Social Psychology (New York: Harper & Row, 1956), sonality and Product Use," Journal of Applied
p. 581. Psychology 45 (1961): 325-29.
2. Ideal self-concept will be more The assumption was made that auto-
closely related to consumption preferencemobiles were more "conspicuously con-
than actual self-concept when the object sumed" than magazines. One might ar-
of consumption is more rather than less gue that magazine purchasing is very
conspicuous to others, and conversely, much a decision in which some consumers
actual self-concept will be more closely reflect upon what "others" would say
related to consumption preference than about the sort of person who would read
ideal self-concept when the object of different magazines. The problem in lo-
consumption is less rather than more cating a product which few, if any, con-
conspicuous to others. sumers would say had conspicuous char-
acter is quite simply the fact that such
THE STUDY
products turn out to be things such as
SUBJECTS, PRODUCTS, AND BRANDS canned foods and detergents-products
The selections of subjects, products, which few respondents (in the prelimi-
and brands within products for use in nary studies) conceived of as having sym-
this research were not independent con- bolic or image character. In other words,
siderations. Budgetary limitations con- products which are seen to have sym-
strained me to convenience samples of bolic character are also those products
students at Purdue University; hence, which most agree are conspicuously con-
the choice of products was at the outset sumed. At least in a relative sense, how-
limited to those with which young adults ever, automobiles were seen to have much
would have had consumption experience. more conspicuous character than mag-
Both hypotheses 1 and 2 required that azines.
products, and brands within products, Brands of these two products were
be chosen that had symbolic or "image" chosen that satisfied the following cri-
character for the subject sample, and teria: (1) female subjects saw the brands
hypothesis 2 required that two products as having psychological-symbolic char-
be selected which differed in their con- acter-that is, could describe "the kind
sumption conspicuousness. Preliminary of a girl" who would read this brand of
studies were conducted with male and magazine or who would drive this brand
female student samples to determine sub- of automobile; (2) there was wide varia-
ject familiarity with various brands of tion in the nature of the image attrib-
several products. The results of these uted by these subjects to consumers of
studies, as well as a review of the litera- the various brands; and (3) there was
wide variation in brand preferences
ture relevant to the symbolic character
among subjects for the products selected.
of various products,'6 resulted in the
These criteria were applied to data ob-
decision to use female subjects and the
tained in the preliminary study, and six
selection of automobiles and magazines
brands of magazines and six brands of
as relevant products.
automobiles were selected. These brands
16 Especially T. Veblen, The Theory of the Leisure
were, for magazines, Mademoiselle, Sev-
Class (New York: Modern Library, 1934); Francis
enteen, Vogue, Reader's Digest, Saturday
S. Bourne, "Different Kinds of Decisions and Refer-
ence-Group Influence," in Marketing and the Be- Evening Post, and Look, and for auto-
havioral Sciences, ed. P. Bliss (Boston: Allyn & mobiles, Cadillac, Lincoln, Imperial,
Bacon, 1963), pp. 247-55; and W. A. Woods, "Psy-
Ford Falcon, Chevrolet Corvair, and
chological Dimensions of Consumer Decision,"
Journal of Marketing 24 (1960): 15-19, Dodge Dart.
MEASURES OF SELF-CONCEPT AND BRAND IMAGE descriptive polar-adjectival scales on a

The many different procedures utilized seven-point scale.
for the measurement of self-concept have Various lists of adjectives and trait
been reviewed by Strong and Feder,"7 names which had been used in previous
Wylie,"8 and Lowe.'9 The more popular self-concept and/or brand image research
procedures include the Q-sort,20 Likert- were studied in an attempt to derive a
type scale methods,2' and various modi- population of such referents from which
fications of the semantic differential tech-selections could be made to compose
nique.22 In their reviews of the self-con- self-concept and brand image dimensions.
cept literature, however, these investi- A trial list of sixty-six referents (thirty-
gators have generally concluded that the three referents and their antonyms) was
various measures of self-concept are re- selected for use in preliminary studies.
liable and, to a great degree, interchange- Subjects were asked to indicate how of-
able. The semantic differential technique ten they saw themselves as the sort of
(applications of this technique in a con- person described by the word. Referents
sumer research context include Crespi,23 (and their antonyms) were rejected
Kjeldergaard,24 Lyle,25 Manis,26 and which were used either so infrequently
Birdwell27) was chosen for use in this or so frequently in the description of self
research. The semantic differential pro- that social desirability (or a general
cedure involves repeated judgments of a evaluative factor) was the only essential
concept or object against a series of variable being measured.
The fifteen adjectival bipoles finally
17 D. J. Strong and D. D. Feder, "Measurement
of the Self Concept: A Critique of the Literature,"
selected are given in Appendix A. These
Journal of Counseling Psychology 8 (1961): 170-78. fifteen dimensions constitute the self-con-
11Ruth C. Wylie, The Self Concept: A Critical cept (both actual and ideal) and the
Survey of Pertinent Research Literature (Lincoln: brand image measure used in the test of
University of Nebraska Press, 1961).
both hypotheses 1 and 2. In the actual
19 C. M. Lowe, "The Self-Concept: Fact or Arti-
fact?" Psychological Bulletin 58 (1961): 325-36.
instrument completed by the subjects,
20 W. Stephenson, The Study of Behavior (Chi-
however, these dimensions were embed-
cago: University of Chicago Press, 1953). ded in a much larger number of dimen-
21 R. E. Bills, E. L. Vance, and 0. S. McLean, sions in order to make the description
"An Index of Adjustment and Values," Journal of task more realistic and complete from
Consulting Psychology 15 (1951): 257-61.
the subject's point of view.
22 C. E. Osgood, G. J. Suci, and P. H. Tennen-
Definitions of ideal and actual self-
baum, The Measurement of Meaning (Urbana: Uni-
versity of Illinois Press, 1957). concept and of the product image task
23 J. Crespi, "Use of a Scaling Technique in Sur-and brand preference orderings, as well
veys," Journal of Marketing 25, no. 5 (1961): 69-72. as the sequence in which these tasks
24 P. Kjeldergaard, "Attitudes toward News- occurred, are contained in Appendix B.
casters as Measured by the Semantic Differential: A
It will be noted that subjects first
Descriptive Case," Journal of Applied Psychology 45
(1961): 35-40. rated their ideal self-concept and then
25 J. Lyle, "Semantic Differential Scales for News- their actual self-concept. I felt that this
paper Research," Journalism Quarterly 37 (1960): order, rather than the reverse, would
559-62.
minimize the social desirability response
26 M. Manis, "Assessing Communications with
the Semantic Differential," American Journal of
set by subjects in describing their actual
Psychology 36 (1959): 189-98. self-concept.
27 Birdwell. After completing their ratings of ideal
and actual self-concept, subjects were ternatives. In both the self and the brand
given instructions for the magazine con- descriptions, subjects would be describ-
cept (brand image) task. Following their ing a "person." Further research is
ratings of the six magazines, subjects needed in order to determine whether the
rank-ordered their preferences for them. product images derived from these two
The brand preference task was inten- different procedures would or would not
tionally placed after the ratings of brand be different.
images rather than before. The contrary
SUBJECTS
procedure might have disposed subjects
to describe their most preferred magazine
The survey instrument was completed
as more "favorable" (more similar to by 247 female undergraduate and gradu-
self) than less preferred magazines. The ate students. All subjects in the final
study participated on a voluntary basis,
same procedure was followed for the
without monetary compensation, and
automobile image rating task. After
their ratings of the six automobiles, sub- completed the instrument anonymously
and at their own convenience. The final
jects rank-ordered their preferences for
usable sample size was 200, after exclu-
them.
sions of subjects who (1) had not resided
Subjects were instructed to disregard
in the United States for the preceding
their actual automobile ownership and
three years and/or (2) omitted one or
magazine subscriptions in their prefer-
more of the self-concept or brand image
ence rankings of brands. This was done
rating tasks. Fifty-seven different sub-
in order to eliminate cost considerations
jects were used in a test-retest reliability
from their preference orderings so that
study and were paid $3.00 each for their
the opportunity for "self" and "brand"
participation.
image matching would be maximized.
It should be noted that, rather than RESULTS
ask the subject to rate the image of the
TEST OF HYPOTHESIS
brand per se,28 she was asked to "imagine
a girl about your own age who prefers Figure 1 shows the average actual and
to read that magazine (own such an auto- average ideal self-concept profiles for
mobile) more than any other. ... Your the 200 subjects across the fifteen seven-
task is to describe the kind of a girl who point semantic differential bipoles. Fig-
would most prefer to read (own) that ure 2 contrasts the profile of the first
particular magazine (that make of auto- preferred magazine with that of the least
mobile)." This procedure was utilized (sixth) preferred magazine for the 200
by Wells et al.29 in an automobile "per- subjects across the same fifteen dimen-
sonality" study and was chosen as a pro- sions. Similarly, figure 3 contrasts the
cedure in this research simply because profile of the first preferred automobile
it appeared to this investigator to be a with that of the least (sixth) preferred
more logical and meaningful task from automobile.
the subject's point of view than any al- A simple graphical comparison of these
figures suggests support for hypothesis
28 For example, ibid.
1 in that, for both magazines and auto-
29 W. D. Wells, F. J. Andriuli, F. J. Goi, and S. mobiles, subjects preferred the brand of
Seader, "An Adjective Check List for the Study of
'Product Personality,' "Journal of Applied Psychol-
the product which was perceived by them
ogy 41 (1957): 317-19. to be more rather than less similar to
their own self-concept (both actual and dividual be compared for their similarity
ideal self-concept). However, these com- or dissimilarity to each of the various
parisons were based upon average data brand profiles of different preference
for the entire group of 200. An appro- value.
priate test of the hypothesis required The various methods for comparing
that the self-concept profiles for each in-
the similarity of profiles have been re-
actual ideal
Ideal self self Actual
x \ \x
4.975 1. excitable calm 3.525

4.97 2. simple complicated 4.680
1.1425 3. graceful awkward 3.360
14.31 4. conservative liberal 14.005
2.93 5. humorous serious 3.1400
5.555 6. follower leader 4.365
3.610 7. dominating submissive 3.620
1.625 8. popular unpopular 2.770
4.875 9. extravagant economical 4.170
1.1490 10. mature immature 2.820
6.6150 11. unsuccessful successful 5.165
1.295 12. informed uninformed 3.075
5.925 13. weak strong 4.890
1.205 14. interesting dull 2.655
4.690 15. conformist nonconformist 4.110
1 2 3 4 5 6 7
FIG. 1.-Average actual and av
First preferred magazine 6th preferred lst preferred Sixth preferred magazine
Mag. mag.
x x

2.590 10. mature immature 4.160

4.835 13. weak strong 3.690
3.945 15. conformist , nonconformist 2.805
1 2 3 4 5 6 7
FIG. 2.-First preferred magazine
viewed by Cronbach and Gleser30 and a point in k-dimensional space, the more
essentially fall into two categories of similar the two profiles, the closer their
measures. One involves the use of some points will lie in this space, and con-
correlational procedure, and the other versely, the more dissimilar, the farther
employs some kind of distance measure apart the points will lie in this space. The
(distance or discrepancy between pro- dissimilarity, then, of two profiles is the
files). Whereas the correlational proce- linear distance between their respective
dures tend to ignore mean (level) dif- points in k-dimensional space.
ferences between profiles, distance mea- In order to test these hypotheses,
sures reflect both level as well as covari- twenty-four difference scores (D2's) were
ance (shape) similarity between pro- computed for each of the 200 subjects.
files.31 There were six difference scores between
First preferred automobile 6th preferred 1st preferred Sixth preferred automobile
auto auto


2.610 10. mature Limmature 3.970

5.085 13. weak strong 3.700
4.325 15. conformist - - -- nonconformist 3.750
1 2 3 4 5 6 7
FIG. 3.-First preferred automobile
If we regard each of the fourteen pro- ideal self-concept and the six automobile
files (actual and ideal self-concept and brands, six difference scores between ac-
twelve brand images) for each subject as tual self-concept and the six automobiles,
six difference scores between ideal self-
30 L. J. Cronbach and Goldie C. Gleser, "As-
concept and the six magazine brands,
sessing Similarity between Profiles," Psychological
Bulletin 50 (1953): 456-73. and six difference scores between actual
31 From the point of view I developed in discuss- self-concept and the six magazine brands.
ing the theory underlying these hypotheses, it was Each profile contained the same fifteen
felt that a definition of similarity which included
seven-point semantic differential dimen-
level as well as shape similarity was necessary.
Cronbach and Gleser have developed such a gen- sions. If two profiles overlapped exactly,
D2 would equal zero. If two profiles were
eralized distance measure, D2, which derives from the
generalized Pythagorean rule:
entirely dissimilar (a value of 1 assigned
to a dimension in the profile and a value
D2 2
D12 (Xjl - X2) 2
j=l of 7 assigned to the corresponding dimen-
j=1
sion in another profile),

meanD2
data, would equal repre-
which are graphically
15 X 62, or 540. sented in figure 4, are systematically
Hypothesis 1 states that "an individ- consistent with the hypothesis. The mag-
ual will prefer to consume a product or nitude of the discrepancy scores (D2's)
brand of product which is perceived by does increase as a function of a decrease
him to be more rather than less similar in preference value for brands, without
to his own self-concept." That is, the exception. This is true for both actual
magnitude of the discrepancy between and ideal self-concept and for both prod-
self-concept profiles (both actual and ucts.
ideal) and brand profiles should increase The significance of the difference
as a function of a decrease in the prefer- among profile means was computed using
ence value expressed for that brand. The the standard formula for correlated data.
120
110
Ideal self -concept
Actual self-concept
100
- 90/
80
50 I I r. s
1 2 3 4 8 6
Preferred bran
FIG. 4.-Mean discrep
The results of these tests are contained twelve comparisons relevant to hypothe-
in table 1.32 Comparisons 1 through 20 sis 2.) The D2's between ideal self-con-
in table 1 are the twenty comparisons cept and automobile brand preferences
relevant to hypothesis 1. Sixteen of the were greater than, not less than, the
twenty comparisons were statistically D2's between actual self-concept and
significant (one-tailed t-test, p < .10) automobile brand preferences. In fact,
and were consistent with the hypothesis. five of these six comparisons (compari-
Those comparisons which were not sig- sons 21 through 26 in table 1) were
nificant were nevertheless in the ex- statistically significant (one-tailed t-test,
pected direction. p < .01) and were, as well, in the unpre-
Hypothesis 2 states that "ideal self- dicted direction.
concept will be more closely related to The data were consistent, however,
consumption preference than actual self-
TABLE 1
concept when the object of consumption
SIGNIFICANCE TESTS ON DIFFERENCES
is more rather than less conspicuous to
BETWEEN PROFILE MEANS:
others, and, conversely, actual self-con-
CORRELATED DATA
cept will be more closely related to con-
sumption preference than ideal self-con-
Comparison
cept when the object of consumption is
1. ISC-1 mag:ISC-2 mag...... 3.199***
less rather than more conspicuous to
2. ISC-2 mag:ISC-3 mag ...... 2.563***
others." 3. ISC-3mag:ISC-4mag...... 3.165***
4. ISC-4 mag: ISC-5 mag . . 1.429*
That is, the magnitude of the dis-
5. ISC-5 mag: ISC-6 mag . . 3.764***
crepancy between ideal self-concept pro- 6. ISC-1 auto: ISC-2 auto ...... 2.573***
files and profiles for automobiles should 7. ISC-2 auto: ISC-3 auto ...... 2.274**
8. ISC-3 auto: ISC-4 auto ...... 4.040***
be less (at each level of preference) than 9. ISC-4 auto: ISC-5 auto ...... 3.665***
10. ISC-5 auto: ISC-6 auto ...... 4.318***
the magnitude of the discrepancy be-
11. ASC-1 mag:ASC-2 mag..... 3.311***
tween actual self-concept profiles and 12. ASC-2 mag:ASC-3 mag ..... 2.232**
profiles for automobiles (at the same 13. ASC-3 mag:ASC-4 mag..... 2.264**
14. ASC-4 mag:ASC-5 mag ..... 0. 899NS
level of brand preference), and converse-
15. ASC-5 mag: ASC-6 mag .... 2.982***
16. ASC-1 auto:ASC-2 auto..... 0. 263NS
ly, the magnitude of the discrepancy be-
17. ASC-2 auto:ASC-3 auto..... 0. 946NS
tween actual self-concept profiles and 18. ASC-3 auto:ASC-4 auto ..... 0.994NS
19. ASC-4 auto:ASC-5 auto..... 4.925***
profiles for magazines should be less (at
20. ASC-5 auto:ASC-6 auto..... 3. 588***
each level of brand preference) than the21. ISC-1 auto :ASC-1 auto . 0. 294NS
22. ISC-2 auto: ASC-2 auto .. . 2.852***
magnitude of the discrepancy between
23. ISC-3 auto:ASC-3 auto . . 4.354***
ideal self-concept profiles and profiles 24. ISC-4 auto:ASC-4 auto .... 9.255***
25. ISC-5 auto:ASC-5 auto ... . 8.225***
for magazines (at the same level of
26. ISC-6 auto:ASC-6 auto .... 9.685***
brand preference). 27. ASC-1 mag:ISC-1 mag...... 3.458***
The mean data graphically represented 28. ASC-2 mag:ISC-2 mag...... 4.881***
29. ASC-3 mag:ISC-3 mag...... 5.388***
in figure 4 were not consistent with the 30. ASC-4 mag:ISC-4 mag...... 7.233***
first part of this hypothesis. (Compari- 31. ASC-5 mag:ISC-5 mag...... 7.935***
32. ASC-6 mag: ISC-6 mag ...... 9.171***
sons 21 through 32 in table 1 are the
32 For purposes of convenience, the various D2'sASCNOTE.-NS - not significant;
= actualself-concept.
ISC ideal self-concept;
are abbreviated and referred to as follows: ISC-1 * Significant at or beyond p < .10 (t 5 1.283), one-tailed
mag the D2 between ideal self-concept profile and test.
the profile for the first preferred magazine; ASC-3 ** Significant at or beyond p < .05 (t > 1.645), one-tailed
test.
auto = the D2 between actual self-concept profile
*** Significant at or beyond p < .01 (t 5 2.327), one-
and the third preferred automobile; etc. tailed test.
with the second part of this hypothesis. person I would like to be seen as by
The D2's between actual self-concept and others. It is nevertheless apparent that
magazine preferences were less than the the role of consumption conspicuousness
D2's between ideal self-concept and mag- as a product variable mediating the re-
azine preferences. The differences be- lationship between self-concept and con-
tween these profile means were all sta- sumption preferences needs to be further
tistically significant (comparisons 27 investigated. Not only is there a need
through 32 in table 1, one-tailed t-test, to investigate the relative role of ideal
P < .01). versus actual self-concept as they relate
to more than the two products used in
DISCUSSION AND IMPLICATIONS this research, but there is a further need
Hypothesis 1 was strongly supported to develop some metric for quantifying
by the results. Subjects preferred brands the conspicuousness with which products
of products which were more rather than are consumed.
less similar to their own self-concept. These data cannot be interpreted to
The magnitude of the discrepancy be- support the argument that either product
tween self-concept (both actual and differentiation or market segmentation
ideal) and brand image (for both maga- along "psychological-symbolic" as op-
zines and automobiles) increased as a posed to "objective-rational" dimen-
function of a decrease in subjects' pref- sions should always or ever be the basis
erences for those brands. for marketing strategies. What is sug-
Hypothesis 2, regarding the role of gested by these findings is simply that
actual versus ideal self-concept and some sort of congruity of "matching"
brand preferences, was not, however, mechanism seems to be operative with
supported by these data. Actual self- respect to self-concept and brand pref-
concept was in fact more similar to con- erence. To this extent, these findings
sumption preference than ideal self-con- support those reported by Birdwell33
cept for each of the six brands of both and Grubb and Hupp.34
of the products. An implication of this research relates
Another serious question in this re- to future developments in personality
gard has to do with the measures of self- theory. To the extent that an individual
concept used in this research. One might accumulates goods in ways which are
well argue that the most appropriate consistent with some aspect of his view
measure of "ideal" self-concept should of himself (actual or ideal or some other
have been "self as ideally seen by others" self-concept perspective), it may be pos-
rather than "self as ideally seen by self," sible to derive an index based upon that
the rationale being that, if consumption individual's consumption history (or in-
conspicuousness does manifest itself in tended consumption future) which would
determining which aspect of self-concept provide an "objective" indicator of
will mediate brand preference, it would personality free from the traditional
manifest itself most logically in an biases common to many personality in-
"others" view of self. That is, when I struments.
purchase an automobile, I make my 33 Birdwell.
brand preference in terms of the sort34of

Grubb and Hupp.
APPENDIX A
THE FIFTEEN SEMANTIC DIFFERENTIAL BIPOLES

USED FOR TESTING HYPOTHESES 1 AND 2
excitable - . .. . .. - . calm
simple - . . . . . complicated
graceful- ....... . . . awkward
conservative -- . . . ... . liberal
humorous ... . -serious
follower- ....... leader
dominating - . submissive
popular - - .... .. unpopular
extravagant -. - ... . . . . economical
mature ..-.. ... . immature
unsuccessful- ........ successful
informed ..... .. . . ... uninformed
weak- ....... ........ -. strong
interesting- ..... - ........ - dull
conformist - .. ......-.. .. -nonconformist
APPENDIX B
SUMMARY OF RESEARCH INSTRUMENT
1. Ideal self (thirty seven-point semantic 4. Magazine preferences: "Assume that

differential bipoles): "Place an X at the you are offered a free magazine subscription
point on each scale that best describes your to any of the following six magazines of your
'ideal self'-that is, the sort of person you choice: Look, Mademoiselle, Reader's Digest,
would most like to be or the way in which Saturday Evening Post, Seventeen, and Vogue.
you would like to see yourself." Place the number 1 next to the magazine
2. Actual self (thirty seven-point seman- title that would be your first choice, the
tic differential bipoles): "Place an X at the number 2 next to the magazine that would
point on each scale that best describes your be your second choice, and so on through
'actual self'-that is, the sort of person you your sixth choice. (Disregard the fact that
actually are or the way in which you actually you may already subscribe to some of these
see yourself." magazines.)"
3. Magazine concepts (fifteen seven-point 5. Automobile concepts (fifteen seven-
semantic differential bipoles): "The title of a
point semantic differential bipoles): "The
magazine appears at the top of each of the
make of an automobile appears at the top of
following six pages. For each of the maga-
each of the following six pages. For each of
zines, imagine a girl about your own age who
the makes of automobiles, imagine a girl
prefers to read that magazine more than
any other magazine. Your task is to describe about your own age who owns such an
the kind of girl who would most prefer to read automobile and prefers it to any other. Your
that particular magazine. Place an X at that task is to describe the kind of a girl who would
point on each scale that best describes the most prefer to own that particular make of
girl who reads that particular magazine." automobile. Place an X at that point on each
scale that best describes the girl who owns of these automobile makes. Place the num-
that particular make of automobile." ber 1 next to the car that would be your first
6. Automobile preferences: "Assume that choice, the number 2 next to your second
you are in the market for a new automobile. choice, and so on through your sixth choice.
Also assume that you are purchasing this (Disregard the fact that you may already
automobile for your own use and pleasure own one of these automobiles.)" Automo-
and that your decision is independent of the biles: Cadillac, Chevrolet Corvair, Dodge
wishes and/or needs of your family. Finally, Dart, Ford Falcon, Imperial, Lincoln.
assume that you can afford to purchase any

Self-Concept and Brand Preference

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Self-Concept and Brand Preference

Caricato da

Copyright:

Formati disponibili

Self-Concept and Brand Preference

Author(s): Ivan Ross

INTRODUCTION might lend support to this hypothesis?

MEASURES OF SELF-CONCEPT AND BRAND IMAGE descriptive polar-adjectival scales on a

4.975 1. excitable calm 3.525

FIG. 1.-Average actual and av

3.925 1. excitable calm 3.45

5.500 11. unsuccessful successful 4.500

FIG. 2.-First preferred magazine

3.875 1. excitable calm 3.940

5.025 6. follower leader 3.485

5.665 11. unsuccessful successful 14.390

FIG. 3.-First preferred automobile

sion in another profile),

Ideal self -concept

FIG. 4.-Mean discrep

purchase an automobile, I make my 33 Birdwell.

brand preference in terms of the sort34of

THE FIFTEEN SEMANTIC DIFFERENTIAL BIPOLES

SUMMARY OF RESEARCH INSTRUMENT

1. Ideal self (thirty seven-point semantic 4. Magazine preferences: "Assume that

Potrebbero piacerti anche