Perceptions of Age-Identity;

A Cross-Cultural Inner-Age
Exploration
Benny Barak, Anil Mathur, Keun Lee, and Yong Zhang
Hofstra University

ABSTRACT
This article presents an exploratory and comparative study of the
nature of cognitive and desired age perceptions among 20-59-year-
old men and women in three East Asian societies: India, Korea, and
China. Research questions were posed about the role of gender and
self-identification with being either "yoiing" or "middle-aged," The
data indicate that gender plays no significant role in age perception
in the three nations surveyed. The great majority of respondents
under 40 considered themselves young and rejected a middle-age
role for themselves irrespective of their domicile. Also, young self-
identifiers* chronological and subjective ages were younger than
those of middle-age identifiers. These findings were surprisingly
similar to those found in "ageless" American society. © 2001 John
Wiley & Sons, Inc.

Americans are known to worship youth, a concept that is quite elastic

in consequence of the reality of America's aging population, which is
now, at the start of the 21st century, in better shape both financially
and health wise than any prior generation. Since the early 1960s, sub-
jective age research consistently reports patterns of age denial among
adult Americans; this denial (gauged as discrepancies between inner
ages and age of birth) increases witb the process of chronological aging
(Barak, 1998; Barak & Rahtz, 1999; Barak & Schiffman, 1981; Barak
& Stern, 1986; Barak, Stern, & Gould, 1988; Cutler, 1982; Peters, 1971;

Psychology & Marketing

© 2001 John Wiley & Sons, Inc. Vol. 18(10): 1003-1029 (October 2001)
1003
Stephens, 1991). The present study set out to investigate if this U.S.
pattern of younger age identification and aspiration also holds true in
non-Western societies. The countries selected—India, Korea, and
China—are all in the Far East and have stereotypical expectations that
one venerates and respects ancestors and elders. Such stereotyping
leads to the expectation that, unlike the West, where the constant pur-
suit of youth leads to great efforts to remain ageless as reflected by
heavy use of cosmetics and cosmetic surgery, denial of one's age of hirth
does not seem so self-evident.
Trends, such as keeping physically fit, reliance on healthy diets, seek-
ing preventive medical care, and exploiting cosmetic help, have hecome
global. Glohal trends of this nature reflect the adoption of an ageless
society, which is especially appealing to higher-educated, well-off, and
more Westernized segments of non-Western societies. These semielite
are more likely than others to participate in a global trend that empha-
sizes an aspiration to he, and appear, youthful and ageless. This partic-
ular trend is further supported by global media that typically show at-
tractive adult men and women who provide clear evidence that beautiful
bodies and faces are no longer clearly age identifiable. Combined with
American media's emphasis on the young and youth, this "ageless so-
ciety" trend markedly contributes to globalization of Americans' obses-
sion with, and craving for, the fountain of youth. Global role models in
their 40s, 50s, 60s (e.g., Cher, Tina Turner, Clint Eastwood, and Arnold
Schwartzenegger), and even 70s (e.g., Paul Newman) provide living
proof to global audiences that those no longer so young, at least from a
chronological perspective, can still lead full, glamorous lives. Many of
these glohal age-role models have been in regular worldwide view since
their baby-boomer performances in the 1960s. They still show consistent
vigor, health, and competence, and remain enormously desirable and
appealing in spite of their chronological age. These age-role models sug-
gest it is feasible to reject traditional stereotype age-role models even if
this involves age denial.
Another aspect of global aging is that with the advent of better med-
icine and education there has been a general growth in hfe-spans in
widely diverse societies. These revised chronological realities of longer
life-spans, especially as experienced by members of middle and higher
social strata in developing or newly developed societies, may very well
grow into relevant social forces that impact on the nature of age per-
ception. Exploration of inner-age concepts, such as dimensions of self-
perceived age (hereafter referred to as cognitive ages) and ideal age di-
mensions (hence referred to as desired ages), is a first step toward
finding out about how age perception occurs in the Far East.
Although an American cultural bias favoring youthfulness is often
assumed operative, there has been little published in the United States
about age research done abroad. Therefore, little is known about per-
ceptions of cognitive and ideal ages outside the United States. A notable

1004 BARAK ET AL.

exception is the cross-cultural study done by Uotinen (1998); she com-
pares Finnish and American data sets in order to contrast cognitive-
and desired-age patterns. Uotinen found that, similar to findings in the
United States, Finns' mean cognitive age (Feel/Age) was younger than
chronological age, and their mean ideal age was even younger. Overall,
Finns expressed greater acceptance of their age of birth; that is, there
was less disparity between their chronological, cognitive, and ideal ages.
Another rare cross-cultural inner-age comparison study by Chua, Cote,
and Leong (1990), surveyed Singaporeans and reported that respon-
dents who spoke English identified with a younger cognitive age than
did those who spoke Chinese. All respondents perceived their inner age
as younger than age of hirth, no matter what their primary language
(i.e., irrespective of cultural background).
Does a similar age-denial phenomenon occur in additional countries
and societies? There is a traditional view that the aged are more re-
spected and admired in the Far East as a consequence of cultural
philosophies stressing admiration for the wisdom of those who achieved
old age. If that were true, one might expect cognitive and desired ages
to be frequently at least equal, if not higher, than chronological age.
The alternative view holds that the phenomenon of perceiving one-
self as younger than one's age of birth is a basic and common human
condition irrespective of cultural background and society, in other
words, an underlying human characteristic. This last viewpoint pro-
poses that cognitive and desired ages function in an analogous manner
throughout the world, irrespective of differences in the cultural envi-
ronment.

Functional Self-Model
To assess the way cognitive- and desired-age dimensions represent fac-
ets of actual and ideal age-role self-concepts, measurement of cognitive
and desired ages in the present study relied on a functional age-role
model. This particular paradigm is derived from original research on
age self-concepts by Kastenbaum, Derbin, Sabatini, and Arrt (1972).
The model proposes four dimensions related to functional areas of the
self: psychological (feelings), biological (looks), social (actions), and cog-
nitive (interests). This functional model has also been used in youth-
fulness research (Staats et al., 1993) to measure respondents' percep-
tions of self as younger than, as old as, or older than their chronological
age, as well as in sex-role research (Stern, Barak, & Gould, 1987} as a
basis for a Sexual Identity Scale that elicits sex-role self-concepts in a
nonclinical undisguised format.
The cognitive and desired ages measured in the present exploratory
study depend on, and adhere to, the Kastenbaum et al. functional
model's age dimensions as follows: r•

PERCEPTIONS OF AGE-IDENTITY 1005

 Cognitive Age Desired Age
(Average of (Average of
Functional Dimensions Dimensions I Dimensions)
Personality/Emotional * Feel/Age * Feel Like/Age
Physical/Biological » Look/Age • Look Like/Age
Societal/Occupational » Do/Age * Do Like/Age
Cognitive/Intellectual * Interest/Age » Interest Like/Age

Cognitive-age dimensions are facets of a self-perceived actual self

in the same way that desired age dimensions represent facets of ideal
self (Rosenberg, 1979; Sirgy, 1982). Cognitive and desired ages accord-
ingly refiect important elements of respondents' age-role self-concept.
There is a long history of actual age-role self-research (see Barak, 1998;
Barak & Stern, 1986a; Peters, 1971), with generally less attention paid
to ideal age roles (see Barak, 1998; Barak et al., 1988). Twenty-six
years ago, Landon (1974) pointed out that there is a connection be-
tween self-concept/purchase intention and matching by consumers of
product image with self-image as well as ideal self-image. It is likely
that age roles relate to age-sensitive products/services and suhse-
quent purchase intentions not only in the United States, but also else-
where.
Since the 1980s, there has heen an increase in assessment of self-
perceived age with measures that rely on the Kastenbaum et al. func-
tional model. Barak's (1979, 1987, 1998) cognitive-age scale, which es-
timates Feel/Age, Look/Age, Do/Age, and Interest/Age dimensions, was
found to be internally reliable and to meet high validation standards
(Barak, 1987, 1998; Barak & Gould, 1985; Barak & Rahtz, 1990, 1999;
Barak & Schiffman, 1981; Staats et al., 1993; Stephens, 1991; Van Au-
ken & Barry, 1995; Van Auken, Barry, & Anderson, 1993). Although
Wilkes (1992) found statistical support for Barak's age scale, his inter-
pretation excludes the Look/Age dimension (a consequence of his data
showing close association between that particular dimension and chron-
ological age).
Research on ideal age-role perceptions (i.e., that image of the age-
self one aspires to) is more limited. However, support for the reliability
and validity of the desired age dimensions, based on measurement of
Feel Like/Age, Look Like/Age, Do Like/Age, and Interest Like/Age di-
mensions, has heen established for diverse U. S. populations (Barak,
1987, 1998; Barak et al., 1988; Staats et al., 1993; Van Auken & Barry,
1995).
This article presents the first findings of an exploratory investigation
of patterns of cognitive and desired ages in India, Korea, and China. It
examines whether inner ages are perceived similarly in non-Western
societies.

1006 BARAK ET AL.

RESEARCH QUESTIONS :

Gender
Because age stereotyping in the United States traditionally was viru-
lent toward American women (Barak & Stern, 1985/86; Nuessel, 1982),
there tended to he an expectation of a relationship hetween gender and
age identity. Nonetheless, research findings to date typically do not sup-
port the idea that women and men differ in their perceptions of cognitive
and/or ideal (desired) ages (Barak, 1998; Barak & Gould, 1985; Barak
& Rahtz, 1990, 1999; Barak & Stern, 1986a, 1986b; Barak et al., 1988;
Baum & Boxley, 1983; Goldsmith & Heiens, 1992; Linn & Hunter, 1979;
Mutran & George, 1982; Zola, 1962). Just a few report on women having
younger inner-age perceptions than male age-peers studies (e.g., Hen-
derson, Goldsmith, & Flynn, 1995; Montpare & Lachman, 1989; Ross,
1982). In other words, U. S. age-perception research found that gender
is not a factor in age perception. On the other hand, it is worth noting
that Uotinen (1998) found that Finnish men and women (aged 25-65)
did not differ in their mean Fell/Age identification. This was not the
case for ideal age, hecause mean female ideal age was older than that
of the men. Uotinen also cities the work of Ruopilla (1992), who had
similar findings for elderly Finns (65-85-years-old). Men wanted to he
20 years younger than their age of hirth, and Finnish women desired to
be 12 years younger. Uotinen's own findings, hased on data from an
ongoing gerontological research project, show that Finnish cultural fac-
tors may have impacted on ideal age perception. Another recent study
that found women to have a significantly younger cognitive age than
men was in the United Kingdom (Szmigin & Carrigan, 2000). This sup-
ports the idea that cultural factors may influence gender identity and
its relation to age perception and identity. In the three countries sur-
veyed, India, Korea, and China, women traditionally tend to he subor-
dinate to the men in their hves. Moreover, the aging process tends to be
more demanding of women in these societies. The socialization of sex
roles in these non-Western cultures is clearly different than in the United
States, and this suggests that a link might occur hetween gender and
inner-age perceptions. This led to the exploration of gender in relation to
inner ages in these societies, and the following question was asked.

Research Question 1: Is there a relationship between gender and

cognitive and desired age perceptions in In-
dia, Korea, and China?

Identity Age
The most commonly used measure of cognitive (actual self age is the
age-identity scale. It asks whether an individual perceives him- or her-

PERCEPTIONS OF AGE-IDENTITY 1007

seif to be either young, middle-aged, elderly, or old (Bloom, 1961; Bre-
ytspraak, 1984; Cutler, 1982). Identification with one of these self-per-
ceived age categories has changed in the latter part of the 20th century
as a consequence of societal forces such as improved health and well
heing (both economic and spiritual) of American society and led to re-
interpretation of the meaning ofyoung and middle-aged (Barak & Stern,
1985/86; Stephens, 1991) with Americans searching for the fountain of
youth and adopting ever-younger age identities (Barak & Rahtz, 1999).
If one follows the same argument used for the first research question,
it seems reasonable to expect that cultural conditioning may affect the
link between gender and inner age. To investigate whether such a link
exists between sex and self-attribution as either young or middle-aged,
the second question was posed.

Research Question 2: Is gender associated with identification of

self as either young or middle-aged in India,
Korea, and China?

It is inevitable that with the march of time people will shift their age
identity from young to middle-age, even if this takes place at an ever-
older chronological age. In the mid-1980s, Barak and Stern (1985/1986)
considered how young and middle-aged identity related to two chrono-
logical age cohorts: those 20-39 and those 40-69-years-old in an U.S.
eastern metropolitan region. Not surprisingly, they found that those
under 40 were more likely to perceive themselves as young. Yet, a sub-
stantial majority (89%) of under-40-year-olds considered themselves
young, as well as 22% of those chronologically 40-59 (Barak & Stern,
1985/1986). To find out if a similar pattern of young identification occurs
in Far Eastern societies, the third research question was posed.

Research Question 3: What proportion of under-40-year-olds (as con-

trasted with 40+ poptilation samples within
India, Korea, and China) identify themselves
as young rather than middle-aged?

Age Interrelationships
More than 30 years of studies have established that chronological age
is positively correlated with both cognitive and desired ages (Barak,
1987 1998; Barak & Gould, 1985; Barak & Rahtz, 1999; Barak & Stern,
1985/86, 1986a; Henderson et al., 1995; Peters, 1971; Zola, 1962). This
finding, particularly the correlation between cognitive and chronological
ages, has also been established in subjective-age research in Finland
(see Uotinen, 1998), the United Kingdom (Smigen & Carrigan, 2000),
and Singapore (Chua et al., 1990). Yet, although all this research sup-
ports the association between chronological age and the subjective ages.

1008 BARAK ET AL.

there is not much specific research concerning the relationship between
cognitive and desired (ideal) ages. In 1962, Zola reported a lack of as-
sociation between cognitive and ideal ages, with both being younger
than chronological age. Although in 1985 Barak and Gould found
women in their 30s to possess ideal ages older than their cognitive ages,
this finding reversed itself among 40+ respondents. In contrast, Barak
(1998) found chronological, cognitive, and desired ages to differ in a
consistent pattern with mean chronological age older than that of cog-
nitive age and mean age-of-birth even older than that of desired age. As
mentioned earlier, Uotinen (1998) found differences between male and
female desired-age perception, yet she also reports that Finns exhibit
some of the same age denial as Americans.
Research in a variety of U. S. populations supports the finding that
chronological age and cognitive- and desired-age dimensions interrelate
positively (Barak, 1987, 1998; Barak & Stern, 1995/86, 1986a; Barak et
al., 1988; Van Auken & Barry, 1995). This research implies that the
older one's real age, the older one's cognitive- and desired-age concepts,
which is particularly true when one considers membership in self-as-
cription age groupings such as young and middle-aged. This led to three
additional research questions concerned with interrelationships be-
tween chronological, cognitive, and desired ages. The fourth research
question seeks to investigate the nature and strength of association be-
tween these ages, the fifth explores their in-between distances, and the
sixth estimates potential relationships between the real and the two
subjective ages relative to respondents' self-identification as either
young or middle aged.

Research Question 4: Do cognitive- and desired-age dimensions sig-

nificantly relate to each other and to chrono-
logical age in India, Korea, and China?
Research Question 5: Do subjective cognitive- and desired-age di-
mensions differ from each other and from
real chronological age in India, Korea, and
China?
Research Question 6: Do subjective cognitive- and desired-age di-
mensions differ from each other and from
real chronological age when respondents in
India, Korea, and China identify themselves
as either young or middle-aged?

METHOD , ' • • - • • ( . -

Data Collection
The data for the study were collected from three Asian nations: India,
Korea, and China. Because a main objective of the study was to engage

PERCEPTIONS OF AGE-IDENTITY 1009

in a first exploration of inner-age dimensions in tbese societies, we relied
on convenience samples for ease of data collection. Furthermore, we
limited inclusion of respondents in the present study to those survey
participants who: (a) had an age of birth ranging from 20 to 59, (b)
responded to all items on the cognitive and desired age scales (i.e., the
eight age-dimension questions), (c) identified themselves as either
young, or middle-aged, and (d) identified themselves as either male or
female.
Table 1 provides, after screening the study population sampled on the
basis of the foregoing criteria, the demographic profile of the study pop-
ulations in terms of their age identity as young or middle-aged, as well
as their sex, in India, Korea, and China.

India. For data collection in India, students stud5dng at a university in

eastern India were recruited to assist in data collection. They were in-
structed to ask parents or other adults in their family to complete the
survey questionnaire. In addition, some memhers of the faculty and
staff, as well as their family members, were included in the respondent
pool (only one member per family participated in the survey). A total of
189 completed surveys were obtained; these, 159 were completed and
met the present study's criteria (see also Table 1). •'

Korea. A commercial research firm was contracted to do the field work

for the data collection in Korea. The names and addresses of the poten-
tial respondents were obtained from the firm's panel list. Respondent
selection relied on sex and age quotas to ensure adequate age and sex
distribution of the sample. Of the 300 people initially contacted by the
firm, 259 agreed to participate. Interviewers visited respondents' homes
to drop off questionnaires, and the respondents filled out the surveys
while interviewers waited outside. Of the surveys collected, 237 met
study criteria (Table 1).

China. Data collection in China was carried out with the cooperation
of the Circulation Bureau of the Postal Service of the Province of Shan-
dong. An initial sample of about 400 magazine and newspaper subscrib-
ers was selected and contacted for possible participation in the study.
Although this is obviously a convenience sample, it is fairly represent-
ative of the Chinese population. The sample was not unduly compro-
mised by utilization of magazine and newspaper subscribers, because
such readership typically refiects a broad spectrum of Chinese consum-
ers. In view of such a broad representation of Chinese from all walks of
life, the sample generated is the closest the data collection could come
to a probability sample in China. Respondents were informed of a small
monetary reward (about 50(2) offered for responding to the survey. The
questionnaires were delivered to the participants via regular mail, and
the completed questionnaires were subsequently returned. In this man-

1010 BARAK ET AL.

Table 1. Age Identity and Gender: Demographic Profile.
Age-of-Self Identity Gender Totals
Young Mid-Age Females Males % N
India (N = ) ( 88) ( 71) ( 71) ( 88) 159
Education
HS and less 1.1% 4.2% 5.6% 0.0% 2.5% 4
Some college 2.3 1.4 1.4 2.3 1.9 3
College plus 93.2 93.0 91.5 94.3 93.1 148
No response 3.4 1.4 1.4 3.4 2.5 4
Employment
Full-time 29.6% 59.2% 33.8% 50.0% 42.8% 68
Part-time 6.8 2.8 4.2 5.7 5.0 8
Not employed 60.2 35.2 56.3 43.2 49.1 78
No response 3.4 2.8 5.6 1.1 3.1 5
Marital status
Married 40.9% 88.7% 74.6% 52.3% 62.3% 58
Not married 56.8 11.3 25.4 45.4 36.5 99
No response 2.3 0.0 0.0 2.3 1.2 2
Korea (A'^ = ) (147) ( 90) (120) (117) 238
Education
HS and less 22.4% 56.7% 38.3% 32.5% 35.4% 84
Some college 15.6 1.1 5.0 15.4 10.1 24
College plus 61.9 42.2 56.7 52.1 54.4 129
No response 0.0 0.0 0.0 0.0 0.0 0
Employment
Full-time 64.6% 61.1% 50.0% 76.9% 63.3% 150
Part-time 15.0 12.2 22.5 6.8 13.9 33
Not employed 10.2 20.0 21.7 6.0 13.9 33
No response 10.2 6.6 5.8 12.8 8.9 21
Marital status
Married 34.7% 87.8% 45.8% 44.4% 54.9% 130
Not married 65.3 12.2 54.2 55.6 45.1 107
No response 0.0 0.0 0.0 0.0 0.0 0
China IN - ) (146) ( 35) (120) ( 61) 181
Education
HS and less 12.1% 22.9% 15.8% 9.8% 13.8% 25
Some college 10.3 5.7 9.2 9.8 9.4 17
College plus 73.3 65.7 68.3 78.7 71.8 130
No response 4.8 5.7 6.7 1.6 5.0 9
Employment
Full-time 76.0% 91.4% 82.5% 72.1% 79.0% 143
Part-time 2.7 2.9 1.7 4.9 2.8 5
Not employed 17.8 2.9 12.5 19.7 14.9 27
No response 3.4 2.9 3.3 3.3 3.3 6
Marital status
Married 49.3% 88.6% 58.3% 54.1% 56.9% 103
Not married 46.6 5.7 37.6 42.6 38.7 70
No response 4.1 5.7 5.0 3.3 4.4 8

PERCEPTIONS OF AGE-IDENTITY 1011

ner, a total of 181 usable surveys that met the present study's criteria
were collected (see Table 1).

Language of Survey Instruments. Although Indian respondents re-

ceived the original English version of the survey instrument, Chinese
respondents received a Mandarin Chinese version, and for Korean re-
spondents, the questionnaire was in Korean. Although English is not a
native language in India, it is the language commonly used by those
with a background in higher education, business, and India's federal
government. Because Indian respondents typically came from families
with higher education (see Table 1), the use of an English language
questionnaire did not pose any problem. Translation into Mandarin Chi-
nese and Korean were carried out by bilingual professional translators.
Backtranslation was used to cross check the accuracy of translation.
Disagreements between translators and backtranslators were resolved
by mutual discussion.

Measurement, Seoring, and Reliability. Cognitive age was mea-

sured with Barak's (1979) cognitive-age scale, and desired age was as-
sessed with the Barak et al. (1988) desired-age scale. Each of the Kas-
tenbaum et al. four functional age dimensions (feel-look-do-interests) is
tapped by these scales. Each of the four questions in the cognitive and
desired-age scales (i.e., a total of eight) asks respondents to indicate
which age decade (pre-teens to 80s) that they most identify with, or
desire to be like, in terms of each of the functional model's age dimen-
sions. Scoring relied on determination of each scale items' midpoint
score (e.g., a Feel/Age decade of 30s is scored as 35). The four age di-
mensions in each scale were then averaged. This in turn yielded two
continuous multidimensional variables, cognitive and desired age, ex-
pressed as an age in years with 2 and one-half-year intervals. The cog-
nitive- and desired-age scales were appraised for their internal reli-
ability in the three country samples. The internal reliability for the
cognitive and desired multidimensional age measures as reflected by
Cronbach's alpha coefficients were, respectively, 0.87 and 0.89 for the
total Indian sample (A^ = 159), 0.92 and 0.91 for the Korean sample
{N = 237), and 0.92 and 0.77 for the Chinese sample {N = 181).
The other study measures employed—chronological age, sex, and
identity age—relied on straightforward questions. Identity age (Cutler,
1982) relies on response to the question

Which of the following age groups would you use to describe yourself?
Young Middle-Aged Elderly Old

Self-Ascription as Elderly or Old. Not all of the chronologically 2 0 -

60-year-old survey respondents identified themselves as young or mid-

1012 BARAK ET AL.

die-aged. Although these respondents were accordingly excluded from
the study sample, their limited number and chronological ages are wor-
thy of special note. There were 18 respondents who identified them-
selves as elderly: ten in India (one in his 20s, one in her 30s, three in
their 40s, and five in their 50s), five in Korea (all in their 50s), and three
in China (all in their 50s). An additional two Indian respondents had
identified themselves as old (one was in his 30s and the other in his
50s).

ANALYSES AND RESULTS

Research Question 1: Is there a relationship between gender and

cognitive- and desired-age perceptions in In-
dia, Korea, and China?

In order to determine the difference between male and female cog-

nitive and desired ages, t tests (equal variances assumed) as well as
ANOVAs (univariate analyses of variance with chronological age as cor-
relate) were conducted between mean male and female scores (0 and 1)
vis-a-vis cognitive and desired ages (scored in 2 and one-half-year in-
tervals) in India, Korea, and China, separately. Table 2 portrays age
means and standard deviations by sex as well as identification as either
young or middle-aged in each of the country samples.

India (see Table 2)

In the Indian sample (N - 159) the t test at first glance seems to indicate
a difference between age perception on the basis of gender with t values
of 2.65 (p < .009) for cognitive-age (female and male mean cognitive
ages, respectively, 32.99 and 29.33), and 1.97 (p < .05) for desired-age
(female and male mean desired ages, respectively, 29.12 and 26.65) dif-
ferentials. However, because a t test of chronological differences be-
tween male and female Indian respondents (respective mean female and
male chronological ages were 36.28 and 31.24) established these partic-
ular differentials to be significant with a t value of 3.10 (p < .002), the
results showing gender to affect cognitive and desired ages might have
been skewed by these male and female chronological age discrepancies.
Such a significant chronological gap between male and female respon-
dents occurred only in the Indian sample. ANOVA analyses provided
specific evidence that among better-educated Indians (such as the sur-
vey participants), gender is not a factor in age perception. This is also
true in the United States. Tbe ANOVA results manifested a nonsignif-
icant F value of 0.440 for cognitive age, and a nonsignificant F value of
0.363 for desired age.

PERCEPTIONS OF AGE-IDENTITY 1013

 T3 C
C
n "* lo in Ol CO CM CO OJ r - O
-XJ"n i n X CO rH O O CO CO C~ Oi
c
CO
4.;! <
v
O X t>
Q

1o Ol 00
IO 00

O5 CO in <XiOl !N CM in >
Ol t- (N Ol CM CO CO CM
rH
CO o
i> CO !N Ol rH O) CO
CO CO (N CO CO CO cq CO

I
CD o in
"2 ^ CM CM o CD
O
00 00 00 00 CO CO
CO g
mQ
(2
xf-

Ol Ol CO Ol CM o CD CM Oi
Ol Ol CO o CO CO rH CO
Ol

41
iH IN CC (M

35
31
r-l
CO cn m

Ol in OJ to
o OJ
o o
o
CD in CD CO CO CO CO
CD CD CO CD
CO Q

Ol
CM

CO X O O in
CD X Ol O rH
CO
CO in Tf t - IO " ^ CO CO " ^
CM CM CM CM W CM CM CM CM oi
CM

9^
bO bxi
CB
ge

ca
ag

•a

p £ .t: n 1 I'S "O - ^ 13

p c .a
c
o
bo m 4) fcn fcil CO b fac w
O 1^ C - C O CD J3 o 0)
sis :J Q o U O Q O DQ o
-a H

1014 BARAK ET AL.

rH r-l CM X IO X
CM CO
to to
X r-i
X
01 CO to CM to o in X
CO o
X
00 !>; 00 00
o 00 00 to to
o
CD
o to to

o o
CM CM

CO E> in ir- CO I—1 in O I> Ol IO 00 X

to CM CM CM CM O X
to in in CO X oCTl
CM CM oi CO CO t> CO to CO 00 00
CO d '^
CO CO CO CO d CM C<1 ca CJ CM CJ ca CM CM
CO CM CO

O Ol O l> X Ol X CO
O iq CM in 01 to CO IO CO in o
X o
X o X
l> in CD to to CO [-^ CO in to IO CO 00

lO m
CO

01 Ol rH Ol CO
i> to X
iq
CO
rH
lO
CD CM in lO Ol to IO
o o Ol to T-I
rH CD
cri Tji r-i CO CO CO d to to CO IO
to to CM ooi X00 to
^ CO CO • ^ Tt< CO CO CO CO CO CO CO CO CO
CO CO CO

X CM Ol CM X IO CM CO CD 00 CO CM
Ol in t H •H CO to in to lO in to
"^
X
X
IO
o
to to CD to
o
to IO to 1}' in IO IO IO IO in

X tD
Ol
to

X <O O Ol 00 to rH 00 oIO
to X
>*•
l>
to l>' X
Ol CO o cq CO
I—(
X
IO 1-1 -* [> I> CO
00 OD to oi oi 00 CD in CD CD in IO IO in t-^
CM CM CM cq CM CM CM C\| CM CM CO CM CM CM CM
CM CM

PERCEPTIONS OF AGE-IDENTITY 1015

Korea (see Table 2)
Assessment of Koreans surveyed (N = 237) for gender differences in
their mean cognitive-age (femaie, 32.25 and male, 33.03) and desired-
age (female, 29.77 and male, 30.81) perceptions involved both t test and
ANOVA analyses. The t test revealed that sex is not a predictor of cog-
nitive and desired ages: respective t values were -0.735 ins) for cogni-
tive age and -1.089 ins) for desired age. This was further corroborated
by the ANOVAs (with chronological age as a covariate), which displayed
nonsignificant F values, respectively. 1.779 ins) for cognitive and 2.332
ins) for desired age.

China (see Table 2)

In the Chinese survey population iN = 181), t Tests between female and
male cognitive (respectively, 27.35 and 28.07) and desired (respectively,
26.88 and 28.94) ages indicated a lack of difference: respective t values
were —0.594 ins) for cognitive and —1.958 ins) for desired age. ANOVA
analyses (with chronological age as a covariate) confirmed this no-dif-
ference finding for cognitive age with a nonsignificant F value of 2.058.
On the other hand, the ANOVA results for desired age exhibited a sig-
nificant F value of 6.560 (p ^ .011). This implies that Chinese women
are likely to have lower desired ages than do men, contradictory to the
t test.

Summary
In the main, the findings support the notion that men and women in
the Far East, as in the United States, do not differ too much from each
other in their perception of inner-age dimensions. Nonetheless, there
were some contradictions in the findings. In India, t test analyses hacked
the idea that gender plays a role in age perception, yet ANOVAs (with
chronological age as a covariate) indicated the opposite. In China, the
reverse occurred in the instance of desired age: ANOVAs corroborated
a gender differential for desired age, whereas t tests did not substantiate
such a disparity.

Research Question 2: Is gender associated with identification of

self as either young or middle-aged in India,
Korea, and China?

Cross-tabulations and their significance in terms of chi-square in each

of the three population samples (see Table 3) helped find answers to this
question.

1016 BARAK ET AL.

Table 3. Cross-Tabulations and Chi-Squares by Country: Age Identity
versus Gender and Age Cohort.
(A) India Gender
Female (A^ - 71) Male (N - 88) Total {N - 159)

Young 40.8 67.0 55.3

Middle-aged 59.2 33.0 44.7
Chi-square = 10.915 (p < .001)
Chronological Age
Under 40 (A^ - 106) 40+ (W - 53) Total iN - 159)

Young 81.1 3.8 55.3

Middle-aged 18.4 96.2 44.7
Chi-square - 92.384 (p < . 000)
(B) Korea Gender
Female (A^ - 120) Male {N = 117) Total (N - 237)

Young 56.7 67.5 62.0

Middle-aged 43.3 32.5 38.0
Chi-square = 2.963 ip < .1)
Chronological Age
Under 40 iN = 162) 40+ (A^ = 75) Total (A^ - 237)

Young 82.7 17.3 62.0

Middle-aged 17.3 82.7 38.0
Chi-square = 108.403 ip < .001)
(C) China Gender
Female (AT = 96) Male iN = 50) Total (A^ - 146)

Young 80.0 82.0 80.7

Middle-aged 20.0 18.0 19.3
Chi-square - 0.100 (n.s
Chronological Age
Under 40 (A^ - 168) 40+ (N = 12) Total (A^ - 180)

Young 86.9 0.0 • 81.1

Middle-aged 13.1 100.0 18.9
Chi-equare - 97.534 (p < .000)

PERCEPTIONS OF AGE-IDENTITY 1017

India
As mentioned earlier, the categorization of oneself as either young or
middle-aged is not as precise an assessment of actual age-role self-con-
cept as that derived by Barak's continuous cognitive-age scale. None-
theless, it is noteworthy that in India (see Table 3) the cross tabulation
indicated that a significantly smaller proportion (41%) of Indian women
perceived themselves young than did the Indian men (chi-square =
10.915;p ^ .001). It is not clear if this was a consequence of the signif-
icant chronological age difference between men and women.

Korea
In Korea (see Table 3), cross-tabulation results of gender and the young/
middle-aged self-classification indicate a pattern similar to that found
in India. That is, a slightly greater percentage of men (67.5%) perceived
themselves as young than the proportion (56.7%) of women who iden-
tified themselves that way (chi-square = 2.963; p = .085).

China
In China (see Table 3) there was no association between gender and
identification of being young or middle-aged: 80% of women and 82% of
men considered themselves young (chi-square = 0.100 ns).

Summary
In both India and South Korea women tended to be less inclined to
identify themselves as young than did men, whereas in China male and
female respondents in equal proportions identified themselves as young
and/or middle-aged.
Research Question 3: What proportion of under-40-year-olds (as
contrasted with 40+ population samples
•: within India, Korea, and China) identify
themselves as young rather than middle-
aged?

Cross-tabulations and their significance in terms of chi-square in each

of the three population samples (see Table 3) helped find answers to this
question. Chi-square results were significant (p < .001) and established
that young and middle-aged identification differed for those chronolog-
ically under and over 40.
Among the three peoples sampled, more than 81% of under-40-year-
olds viewed themselves as young, and middle-aged identification for
that cohort ranged from a low of 13.1% among Chinese to a high of 18.4%
among Indian respondents. The reverse occurred with middle-aged

1018 BARAK ET AL.

identification among the 40 + . This is because 100% of 40+ Chinese and
96.2% of Indian members of that cohort identified themselves as middle-
aged. The proportion of 40+ Koreans who considered themselves mid-
dle-aged was also substantial (82.7%).

Research Question 4: Do cognitive- and desired-age dimensions

significantly relate to each other and to
chronological age in India, Korea, and
China?

Determination of age interrelations relied on simple Pearson corre-

lations (two tailed). The correlations found (see Table 4) refiect the
strength of the interconnection of the subjective ages and the intercon-
nection relative to chronological age in all three population samples. As
is typified by a comparison between the correlation matrices' patterns
in the three dissimilar cultures, these patterns were remarkably alike.
The nature and strength of the correlations between cognitive as well
as desired age dimensions is further refiected by the strength of internal
reliability (reported upon earlier) of these age variables. The answer to
Research Question 4 is therefore a resounding yes!

Research Question 5: Do subjective cognitive- and desired-age di-

mensions differ from each other and from
real chronological age in India, Korea, and
China?

Paired t tests conducted in each of the nations helped provide the

answers to Question 5. Table 5 shows the Ttest t values of chronological,
cognitive, and desired mean ages in the three societies.

India
The paired t test results attest that in India the various ages indeed
differed significantly (p < .01) from each other. As can be inferred from
Tables 2 and 5 (showing, respectively, mean ages and t values) mean
chronological ages were older (33.49) than those of cognitive (mean
30.69 and a significant t value of 4.29) and desired (27.75 and a t value
of 7.45) ages. Furthermore, the paired t test analyses also showed de-
sired age younger than cognitive age with a significant t value of 6.10.

Korea
In Korea paired t-test results were basically the same as those of India,
with significant (p < .01) differences found between the various ages.
From Tables 2 and 5 it can agadn be inferred that chronological ages
(mean birth age = 34.27) were older than those of cognitive ages (mean

PERCEPTIONS OF AGE-IDENTITY - 1019

 -4-> he
C <

O
q

a a
o t— m
O CD yD

o m o to
q c~; t> in
•-i d d d

o i> ai i> CO
q 00 00 CO 00
r-4 d d d d

o o
o o

o t- ai CO CO CO w
q in in •'t iq iq in

O (N 05
q ; O CO CD
d d d d d

:3 a S
^^ t*^ CO 3LO i^ l O CO '^* ^^
O --H O ^
O 00 X CD

o t - CO t - a i 00 >O tr- O O O (N CO i n
O X CO X CD i n CO i n CD O Ol Ol Ol CO

o C-J l > CO CD CO 00 i n Ol Ol i-H o Tf (M O X !M

o ( > i n i~~ CD " ^ o 00 I> CO l> t>
d d d d d d d d d d d dddd

Ml
CO

•£5) to
< •
—• o

J Q£
a CJ

1020 BARAK ET AL.

 o
o

O O lO
O p -^
•-1 d

O (M CD
O t>- !>•
d d

o
o O OS
CD CO
O rH r-( (M
O iq CO CD
d d d

O C- CM IO O I—1CD as CM
p OD 0> OD Ol o OD r- CD CO
q
i-H
d d d d

IN CQ m CD (N o
l>. U3 CD CD l> cq CO
d d dd d o d d

as w CD CO m •-H as 00 rf CD CO
CD iq 'd; IO CO "*
CD iO CD CD CD
d d d d d d

rf O iM ca (N lO (N CD t > CO
CD IO CD lO CD o CD CD i q CO TP
d ddd d (5 d d d d

•* 0 0 --H iM C D <D cc CO CO ' ^ CO O CO

CD •**" C D IO <:D O 00 I-' CD CD IO TO T

IO Tf O CO TjH l> « ) .M O - ^ CD
CD C- CO CD CD CO • *
o Oi as X

o lO as t- o o o CD CD l O
t- »O CD in o r~ CD - ^ CO CO
O O O O O O O O O O

^ -f" ^

§ o a g
in o
V I- Q I—
J Q
o

PERCEPTIONS OF AGE-IDENTITY 1021

 O CO CD
-^ oi q
d N CO

a
S
CO CO lO CO
CO
q CO i q
^ I' > CO d rH

" * CO CD
CO CO O

0) o
bp OS X CO O
-a II CM CO as

a
S o •^ o
CO r-] 03 Tp
d c6 CO

CD O
lO IO
Q
13

Co <—!
•w T3 II «

I
CD O IO
CQ X rH
•a

in
,£3 — lO

DO
V

I
ED «
-d -a
c c
« o 4 o R
0. o 03
o
"Sc 'So S ti i; S
o o — in rt -S
q qq 3
V V V .3
3 ^ ^ o
H O UO

1022 BARAK ET AL.

cognitive age ^ 32.25) and those in turn were older than desired ages
(mean desired age = 30.28).

China ;
In China the paired t-test results were not so clear. Chronological ages
(mean = 28.45) were significantly (p < .05) older than cognitive ages,
with a mean cognitive age of 27.60 and a paired ^-test value of 2.34
(p < .05). Yet chronological and desired ages did not differ that greatly,
with a hardly significant (p < .1) paired t value of 1.86 (mean desired
age was 27.57). The paired t tests found no difference between Chinese
cognitive and desired ages.

Summary
In both India and South Korea, women's chronological, cognitive, and
desired ages differed significantly (p < .01), with chronological age the
oldest and desired age the youngest. In China, differences between the
three ages were not as clear. Statistically, chronological age was signif-
icantly older than cognitive age (p < .05), and desired age was almost
statistically younger than chronological age (p < .1). Yet cognitive and
desired ages did not differ at all statistically (p was not significant). This
suggests that from this perspective China does not resemble the popu-
lations surveyed in India and Korea.

Research Question 6: Do subjective cognitive- and desired-age di-

mensions differ from each other and from
real chronological age when respondents in
India, Korea, and China identify themselves
as either young or middle-aged?

Paired t tests conducted in India, Korea, and China furnished the

answer to Question 6. Paired ^-test analyses were employed in each of
the three countries, within both young as well as middle-aged self-iden-
tification groupings (see Tables 2 and 5).

India
Among those self-identified as young, chronological and cognitive ages
did not differ significantly, whereas chronological and desired ages dif-
fered at thep < .01 level. The paired differences between cognitive and
desired ages, as reflected by a ^ value of 2.15, denote a two-tailed dif-
ference of .006. The pattern within the Indian sample population iden-
tif3ang themselves as middle-aged was the same as that found within
the total Indian sample: Ages differed significantly, with chronological
ages the oldest, followed by cognitive and then desired ages.

PERCEPTIONS OF AGE-IDENTITY 1023

Korea
Among those who perceived themselves as young, differences between
chronological and cognitive ages were hardly significant, with p < .1;
yet the paired differences between chronological and desired ages, as
well as those between cognitive and desired ages, had significant t val-
ues {p < .000).

China
The only significant difference among Chinese that self-identified as
young is that chronological age was significantly older than cognitive age
[p < .01). The reverse situation occurred among Chinese middle-aged
self-identifiers, showing no establishment of significant differences be-
tween chronological and cognitive ages, whereas desired ages were sig-
nificantly younger than both chronological and cognitive ages (p < .01).

Summary <
In both India and South Korea, women's chronological, cognitive, and
desired ages differed significantly, with chronological age the oldest, and
desired age the youngest. In China, differentials between the three ages
were not as clear. Cognitive and desired ages did not differ at all, sug-
gesting that in this instance China does not resemble the populations
surveyed in India and Korea.

DISCUSSION AND DIRECTIONS FOR FUTURE RESEARCH

Prior to any discussion, it is relevant to mention some of the limitations

of the research presented. This study relied on convenience samples,
and the equivalency of these samples could not be verified. The samples
also differed from each other in some of their demographic character-
istics. Despite these limitations, the samples did represent the better-
educated component of these countries' middle classes. For example,
sample populations were disproportionally educated, with more than
50% of both sexes having graduated college. A sizable component of
India's 1 billion residents is middle class; assessments published in dif-
ferent media (e.g., CIA World Factbook, 1999) claim the number to be
approximately 300 million. A sizable proportion of this middle class has
finished college, very likely substantially more than the total number
of U.S. college graduates, and most of these Indian college grads are
fiuent in English. Yet the Indian, as well as the South Korean and Chi-
nese populations (and that is true for each country's college graduate
populations as well) are culturally extremely different and unique. This
exploratory research seeks to assess aspects of inner-age perception in

1024 BARAK ET AL.

non-Western societies, a goal not affected by the utilization of conve-
nience samples. Each of these countries' populations differ in their his-
tory, ethnic makeup, religion, first language, food, and economic status.
How people in these countries relate to aging is therefore more than
likely to be different as well. Furthermore, as mentioned earlier, these
varied Asian cultures traditionally relate to age in a more respectful
manner than is common to the West. For instance, in all three countries
ancestor veneration is common and elders are characteristically viewed
as having more wisdom. In Korea, with its Confucian culture, seniority
is highly respected.
The authors therefore stress that although the samples are not truly
representative of the respective countries studied, they nonetheless pro-
vide important insights to marketers and advertisers seeking to reach
potential audiences in these societies for age-sensitive products and
services.
This research attempted to examine the age concepts of cognitive and
desired ages in three culturally dissimilar East Asian countries. The
concept of cognitive age has been well developed and examined in the
West, especially in the United States, yet rarely among other cultures.
The scales utilized for the measurement of cognitive and desired ages
had a high level of reliability: For cognitive age Cronbach alpha coeffi-
cients ranged from a low of 0.87 in India to a high of 0.92 in both Korea
and China, and for desired age they ranged from a low of 0.77 in China
to a high of 0.91 in Korea. The internal reliability of these measures
was comparable to results reported for the same measures in studies
that use a wide variety of samples collected in the United States. This
confirms the four-dimensional model of both the cognitive- and the de-
sired-age scale form. These scales showed themselves to be reliable mea-
sures that can be used in diverse cultures. The findings also validate
the Kastenbaum et al. four-dimensional age model in a global setting.
It also suggests that subjective ages represent innate characteristics
independent of culture. Researchers doing cross-cultural research are
always concerned with the applicability of their measures across cul-
tures, a concern that is reduced by the present findings.
Although this was an exploratory study that did not have any hy-
potheses to test, the research set out to examine six questions. Gender-
role identification and the role of women in society has very strong cul-
tural roots. The present research examined the extent that gender
influences age-related self-concepts in the three Asian countries. One
interesting finding is that the three culturally diverse countries show
stark similarity to the United States regarding the role of gender in age
perceptions. Although the role of women might be very different in dis-
parate societies, their perceptions of their own ages do not differ from
the way men in the same societies see themselves. This is interesting
and relevant to global marketers. For some time, American marketers
have used cognitive age as a key variable in developing advertising ap-

PERCEPTIONS OF AGE-IDENTITY 1025

peals, especially in the marketing of age-sensitive product categories
such as cosmetics and fashion. Global marketers attempting to market
similar age-sensitive goods in Asian countries might feel reassured by
the present findings, which show that they may use similar age- and
gender-based strategies in India, Korea, and China.
An intriguing result occurred relative to age identification as either
young or middle-aged and gender. In India, only 41% of women per-
ceived themselves as young, whereas a significantly larger proportion
of male respondents did so (67%). To some extent, this finding may have
been affected by the 5-year age differential in male and female samples'
chronological ages. Nonetheless, in India, gender seemed to affect this
particular form of age identification.
This study found that, as in the United States, a chronological age of
40 distinguishes age identification, as either young or middle-aged in
all three countries surveyed.
This similarity for so many diverse cultures is information that could
be useful to both cross-cultural researchers and global marketers inter-
ested in understanding age identification in these newly developing con-
sumer markets.
The finding that the dimensions of desired age relate in the same
positive fashion to the dimensions of cognitive age, and that both of
these types of subjective age are also linked in a corresponding manner
to chronological age in the three Asian countries, is again of value to
marketers. This finding demonstrates that the manner in which differ-
ing psychographic dimensions of actual and ideal age-role concepts are
connected with the demographic chronological age is the same in ra-
cially, religiously, and culturally dissimilar societies. These correlations
explain the high-reliability alpha scores and also strengthen the validity
of the cognitive- and desired-age measures in a global context. Global
marketers and advertisers may find these strong associations between
the differing age measures helpful in the preparation of campaigns that
focus on worldwide graying target populations to market age-sensitive
products (e.g., cosmetics and fashion) and services (e.g., tourism) to
newly affluent culturally disparate markets. The most relevant dimen-
sion can be tapped, depending on what is to be marketed. For instance,
if the product is a cosmetic, the cognitive and/or desired Feel/Age di-
mensions may be the most relevant; yet, if the product is a form of
exercise, then the Do/Age or Interests/Age dimensions may be more rel-
evant to advertisers. The similarities in interactive patterns of the age
dimensions within such varied cultures suggests that age can be seen
as a useful global segmentation variable.
The desire to be young and youthful appears to have a universal ap-
peal. Data suggest that both cognitive and desired age are younger than
chronological age across the three cultures. Furthermore, with the ex-
ception of China, desired age tends to be younger than cognitive age.
Marketers might be able to exploit this in developing communication

1026 BARAK ET AL.

strategies by appealing to youthful tendencies and ensuring the use of
models that fit the perceived ages of intended target markets. Of course,
they ought to avoid using models who are younger than the ages that
target audience members see themselves, or would desire to be. Also,
because three culturally diverse countries show similar patterns of age
identification demonstrates that standardized promotional strategies
for certain age-sensitive products (e.g., cosmetics) might be applicable
in these countries.
Future researchers might attempt to use more representative sam-
ples in these countries and attempt to achieve sample equivalency,
which will enable greater external validity. Researchers might also at-
tempt to develop research/data-collection methodologies that account
for cultural, economic, political, and social factors, yet produce samples
and data that could be used in comparative studies. Pragmatic research-
ers claim that cultural difTerences make it impossible to get comparable
samples, so they rely on samples and data that are not at all comparable.
On the other hand, purists might get comparable samples but find that
the results cannot be compared. In view of the importance of global
research, it is desirable to consider both theoretical considerations as
well as practical considerations and develop methodologies that can pro-
duce comparable data and results. The present research findings
strongly suggest that subjective ages such as cognitive and desired ages
tap a psychographic dimension that seems to be quite independent of
cultural background. These age concepts represent a relevant global
segmentation characteristic that is commonly found among the greater
majority of humanity. Future studies will also have to replicate much
of the previous research done in the United States. When that happens,
further evidence will be provided to show that cognitive and desired ages
capture aspects of age that are not adequately refiected by chronological
age in any society on Earth.

REFERENCES

Barak, B. (1979). Cognitive reference age among the elderly: A new concept for
marketing. Unpublished Ph.D. dissertation. The City University of New
York, New York.
Barak, B. (1987). Cognitive age: A new multidimensional approach to mea-
suring age identity. International Journal of Aging and Human Develop-
ment, 25(2), 109-127.
Barak, B. (1998). Inner-ages of middle-aged prime lifers. International Joumal
of Aging and Human Development, 46, 189-228.
Barak, B., & Gould, S. (1985). Alternative age measures: A research agenda.
In E. C. Hirshman & M. B. Holbrook (Eds.), Advances in consumer re-
search (Vol. 12, pp. 53-58). Ann Arbor, MI: Association for Consumer Re-
search.

PERCEPTIONS OF AGE-IDENTITY 1027

Barak, B., & Rahtz, D. R. (1990). Cognitive age: Demographic and psycho-
graphic dimensions. Journal of Ambulatory Care Marketing, 3, 51-65.
Barak, B., & Rahtz, D. R. (1999). Perceived youth: Appraisal and characteriza-
tion. Intemational Journal of Aging and Human Development, 49(3), 231 —
257.
Barak, B., & Schiffman, L. G. (1981). Cognitive age: A nonchronological vari-
able. In K. B. Monroe (Ed.), Advances in consumer research (Vol. 9, pp. 602-
606). Ann Arhor, MI: Association for Consumer Research.
Barak, B., & Stern, B. (1986a). Subjective age correlates: A research note. The
Gerontologist, 26(5), 571-578.
Barak, B., &. Stem, B. (1986h). Sex-linked trait indexes among baby-boomers
and pre-boomers: A research note. In R. J. Lutz (Ed.), Advances in consumer
research (Vol. 13, pp. 204-209). Ann Arbor, MI: Association for Consumer
Research.
Barak, B., & Stern, B. (1985/86). Women's age in advertising: An examination
of two consumer age profiles. Joumal of Advertising Research, 25, 38-47.
Barak, B., Stern, B., & Gould, S. (1988). Ideal age concept: An exploration. In
M. Houston (Ed.), Advances in consumer research (Vol. 15, pp. 146-152).
Ann Arbor, MI: Association for Consumer Research.
Baum, S. K., & Boxley, R. L. (1983). Age identification in the elderly. The Ger-
ontologist, 23(5), 532-537.
Bloom, K. L. (1961). Age and the self concept. American Joumal of Psychiatry,
118, 534-538.
Breytspraak, L. M. (1984). The development of self in later life. Boston: Little,
Brown and Company.
Chua, C , Cote, J. A., & Siew, M. L. (1990). The antecedents of cognitive age.
In M. E. Goldberg et al. (Eds.), Advances in consumer research (Vol. 17, pp.
880-885). Ann Arbor, MI: Association for Consumer Research.
CIA World Facthook. (1999). India. Available: http://www.odci.gov/cia/
publications/factbook/in.html.
Cutler, N. E. (1982). Subjective age identification. In D. J. Mangen & W. A.
Peterson (Eds.), Research instruments in social gerontology: Clinical and so-
cial psychology (Vol. 1, pp. 437-461). Minneapolis, MN: University of Min-
nesota Press.
Goldsmith, R. E., & Heiens, R. E. (1992) Subjective age: A test of five hypoth-
eses. The Gerontologist, 32(3), 312-317.
Henderson, K., Goldsmith, R. E., & Flynn, L. E. (1995). Demographic charac-
teristics of subjective age. The Journal of Social Psychology, 135(4), 447-
457.
Kastenhaum, R., Derhin, V., Sahatini, P., & Arrt, S. (1972). "The Ages of Me"
toward personal and interpersonal definitions of functional aging. Aging and
Human Development, 3, 197-211.
Landon, E. L. (1974). Self concept, ideal self concept, and consumer purchase
intentions. Journal of Consumer Research. 1, 44-51.
Linn, M. W., & Hunter, K. (1979). Perception of age in the elderly. Joumal of
Gerontology, 34, 46-52.
Montpare, J. M., & Lachman, M. E. (1989). You're only as old as you feel: Self-
perceptions of age, fears of aging, and life satisfaction from adolescence to
old age. Psychology and Aging, 4(1), 73-84.

1028 BARAK ET AL.

Mutran, E., & George, L. K. (1982). Alternative methods of measuring role/
identity: A research note. Social Forces, 60, 866-875.
Nuessel, F. H., Jr. (1982). The language of ageism. The Gerontologist, 22, 2 7 3 -
276.
Peters, G. R. (1971). Self-conceptions of the aged, age identification and aging.
The Gerontologist, 11, 69-73.
Rosenberg, M. (1979). Conceiving the self. New York: Basic Books.
Ross, I. (1982). Information processing and the older consumer: Marketing and
public policy implications. In A. A. Mitchell (Ed.), Advances in consumer re-
search (Vol. 9, pp. 31-39). Ann Arbor, MI: Association for Consumer Research.
Ruoppila (1992). Attitudes of 65-84 year olds towards age and aging. Geron-
tologia, 6(1), 3-21 (translated title).
Sirgy, M. J. (1982). Self-concept in consumer behavior: A critical review. Jour-
nal of Consumer Research, 9, 287-299.
Staats, S., Heaphey, K., Miller, D., Partlo, C , Romine, N., & Stubbs, K. (1993).
Subjective age and health perceptions of older persons: Maintaining the
youthful bias in sickness and in health. International Journal of Aging and
Human Development, 37(3), 191-203.
Stephens, N. (1991). Cognitive age: A useful concept for advertising? Journal
of Advertising, 20, 37-48.
Stern, B., Barak, B., & Gould, S. J. (1987). Sexual identity scale: A new self-
assessment measure. Sex Roles, 17(9/10), 503-519.
Szmigin, I., & Carrigan, M. (2000). The older consumer as innovator: Does
cognitive age hold the key? The Journal of Marketing Management, 16(5),
505-527.
Uotinen (1998). Age identification: A comparison between Finnish and North-
American cultures. International Journal of Aging and Human Development,
46(2), 109-124.
Van Auken, S., & Barry, T. E. (1995). An assessment of the trait validity of
cognitive age measures. Journal of Consumer Psychology, 4(2), 107-132.
Van Auken, S., Barry, T. E., & Anderson, R. L. (1993). Observations toward
the internal validation of cognitive age measures in advertising research.
Journal of Advertising Research, 82-84.
Wilkes, R. E. (1992). A structural modelling approach to the measurement and
meaning of cognitive age. Journal of Consumer Research, 19, 292-301.
Zola, I. K. (1962). Feelings about age among older people. Journal of Gerontol-
ogy, 17, 65-68.
This research was partly supported by a summer research grant from the Frank
G. Zarb School of Business to the first author. The very helpful editorial input
by Barbara Barak is acknowledged with gratitude.

Correspondence regarding this article should be sent to: Benny Barak, De-
partment of Marketing and International Business, Frank G. Zarb School of
Business, 222 Weller Hall, 134 Hofstra University, Hempstead, NY 11549
([email protected]).

PERCEPTIONS OF AGE-IDENTITY 1029