Ideal Age Concepts: an Exploration

Benny Barak, Hofstra University
Barbara B. Stern, Rutgers, The State University of New Jersey
Stephen J. Gould, Rutgers, The State University of New Jersey
ABSTRACT - This paper presents an exploratory study of the meaning of ideal age to adult men and women in three age cohorts: Baby-Boomers (25-39), Pre-Boomers (4054) and Matures (55-69). Four perceived age measures, a subjective one and three new ideal age scales, elicited the following findings: male and female age perceptions do not differ; age cohort effects influence both subjective and ideal age perception (the older the cohort, the older the perceived ages); personal and cultural definitions of age differ.
[ to cite ]:
Benny Barak, Barbara B. Stern, and Stephen J. Gould (1988) ,"Ideal Age Concepts: an Exploration", in NA - Advances in Consumer Research Volume 15, eds. Micheal J. Houston, Provo, UT : Association for Consumer Research, Pages: 146-152.

Advances in Consumer Research Volume 15, 1988      Pages 146-152

IDEAL AGE CONCEPTS: AN EXPLORATION

Benny Barak, Hofstra University

Barbara B. Stern, Rutgers, The State University of New Jersey

Stephen J. Gould, Rutgers, The State University of New Jersey

[The authors thank three anonymous ACR reviewers for their helpful comments.]

ABSTRACT -

This paper presents an exploratory study of the meaning of ideal age to adult men and women in three age cohorts: Baby-Boomers (25-39), Pre-Boomers (4054) and Matures (55-69). Four perceived age measures, a subjective one and three new ideal age scales, elicited the following findings: male and female age perceptions do not differ; age cohort effects influence both subjective and ideal age perception (the older the cohort, the older the perceived ages); personal and cultural definitions of age differ.

INTRODUCTION

Ideal age is a facet of the ideal self, an important component of one's self-concept (Sirgy 1982; Rosenberg 1979). The ideal self - that image of the self one aspires to be like - has received some research attention in the context of consumer behavior (e.g. Belch 1978; Onkvisit and Shaw 1987; Sirgy 1982, 1980; Sirgy and Danes 1982). Consumer research has paid surprisingly little attention to the ideal age component of the self: only one set of studies (Barak and Gould 1985; Barak and Stern 1985186; 1985) has considered ideal age in such a consumer behavior context.

Advertisers of age-sensitive products such as cosmetics and fashion apparel particularly need information about ideal age to understand the youthful aspirations underlying many products' appeal. The lack of published research indicates a common assumption that most Americans, especially women, consider ideal age to be rather young: probably under 30. This assumption is implied by the under-representation of aspirational figures over 30 in advertising: groups of consumer researchers (Caballero and Solomon 1985; England, Kuhn, and Gardner 1981; Harris and Feinberg 19-77; Rotfield, Reit, and Wilson 1982) have repeatedly found that advertisers infrequently employ over-30 models in either print or electronic media. For example, England, Kuhn, and Gardner (1981), report that over 7?% of magazine ads in the past two decades feature women under 30. This relative lack of mature female promotional figures supports the notion that those who create advertising perceive aspirational referents to be no older than 29.

The chronological aging of the American population, however, indicates that a re-evaluation upward of the concept of ideal age may be in order. By 1980, the American median age was 35 in metropolitan areas such as New York, and was over 30 nation-wide (U.S. Bureau of the Census 1982). Cohort effects of the large and visible population group of the Baby-Boomers, whose oldest members are just now turning 40, appear to be influencing cultural determination of desirable age. Although Americans still worship youth, the concept has become so elastic as a consequence of the new reality of an aging population that reconsideration of the ill-defined youth orientation becomes necessary. Attractive adult men and women provide evidence that beautiful bodies and faces can extend well beyond 40. Trends such as physical fitness, healthy diets, and preventive medical care contribute to the attractiveness of aging Americans. Role models such as Jane Fonda, Joan Collins, Elizabeth Taylor, and the Reagans, are living proof that those no longer "young-young" are still appealing enough to be considered ideal. The concept of ideal age, therefore, deserves examination in the light of revised chronological reality which has become a powerful social force. While an American cultural bias favoring youthfulness is often assumed operative, no investigation of the precise nature of the relationship between the youth bias and individual perceptions of ideal age has taken place.

This paper presents the first findings of an exploratory investigation of ideal age to shed light on the extent to which the American youth bias influences consumers' age perceptions. The paper will examine age in terms of the following questions:

1. How has ideal age been measured in past research?

2. Does biological sex cause differences in perceptions of ideal age?

3. Do members of different age cohorts - Baby-Boomers (25-39), Pre-Boomers (40-54), and Matures (55-69) - differ in their age perceptions?

4. How do "subjective" age perceptions relate to "real" chronological age?

5. What is the nature of association between self-perceived and ideal age in different age cohorts?

Research Background

In the rather limited past history of ideal age research, focus has been primarily on relationships between ideal, chronological, and subjective age (Barak 1987; Barak and Gould 1985; Barak and Stern 1985, 1985/86; Zola 1962).

The first measure of ideal age was "Age Desired" (Zola 1962, p.66); which relied on a straightforward question: "What age would you most like to bet' The sample consisted of 100 men and 118 women over 65 in New England (mean ages respectively 69.7 and 67.09); the mean "Age Desired" was 45.8 for the men and 45.2 for the women, and mean "Age Felt," a measure of subjective age, was respectively 53.4 and 52.3. While Zola found ideal age positively correlated (r = .21, pc .01) with chronological age in both sexes, he did not find such correlations between ideal and subjective ages.

The ideal age concept received little empirical attention until over twenty years later (Barak 1987; Barak and Gould 1985; Barak and Stern 1985/86), when it was measured by the question, "What do you consider to be a persons ideal age? ___years." The sample consisted of 567 adult women surveyed during the early 1980's in the metropolitan New York/New Jersey area. Ideal and chronological age were correlated, repeating the earlier finding. However, unlike the previous finding, ideal and subjective ages were also correlated. One interesting finding related to the elasticity of perceived youth: respondents between 30 and 69 had a mean ideal age of 33.6 years and over 66% indicated an age over 30 as ideal (Barak 1987). Another study of this population found that "93.4 percent of consumers over 25 perceived 25+ as an age more desirable than the teens or early twenties" (Barak and Stem 1985186, p.42). In addition, 68% of Baby-Boomers and 69.9% of women who identified themselves as "young" reported an ideal age over 30. Overall, then, these findings indicate that adult Americans' ideal age, while younger than their chronological age, is typically over thirty.

A problem, however, with these findings is caused by the ambiguous nature of the measure utilized to assess ideal age ("What do you consider to be a person's ideal age?"). The possibility of a disparity between, personally truthful and socially desirable response sets seems built into the question's form. The scale is not clear on semantic grounds: the use of "a person" may cause confusion between the respondent's attribution of ideal to him/herself personally and/or any person in general. The question, therefore, does not pinpoint a respondent's feelings about the age s/he might view as subjectively ideal. The intent of the scale had been to be a direct measure of ideal age, instead the question was phrased so awkwardly that it is unclear whether the "person" represents the respondent of simply a generic American "other," or indeed, whether this question can even be interpreted as a projective measure.

New Measures

To overcome interpretation problems and achieve more accurate measurement and understanding of ideal age, three new measures were developed: (1) Desired Age, (2) Personal Ideal Age, and (3) Generic ideal Age (see Appendix 1).

The first new scale, "Desired Age," was developed as a modification of Barak's (1979) Cognitive Age measure. It relies on a functional model, originally derived from earlier research on age self-concepts (Kastenbaum et al 1972), which develops four dimensions related to functional areas of the self: psychological (feelings), biological (looks), social (actions), and cognitive (interests). This functional model has also been used in sex-role research (Stern, Barak, and Gould 1987) as a basis for a Sexual Identity Scale which elicits sex-role self-concepts in a nonclinical undisguised format.

The cognitive Age scale's modification to the new Desired Age measure adhered to the functional mode's dimensions (see Appendix 1) as shown in figure 1. Desired age, like Cognitive Age, is determined by averaging the responses to the four questions that make up the scale: it also yields a continuous variable, expressed by age in years.

The second new scale is a simple one-component measure, Personal Ideal Age, elicited by the numerical answer to the question: "If you could be any age, what age would you most like to be? The answer must be given as "Age in years." The measure was framed to elicit simply and clearly the respondent's personal perceptions of ideal age: the respondent is addressed as "you" throughout, and the concept of "ideal age" is clearly defined as "the age you would most like to be" to prevent any semantic confusion.

FIGURE 1

NEW DESIRED AGE MEASURES' ADHERENCE TO THE FUNCTIONAL MODE'S DIMENSIONS

The third new ideal age scale, "Generic Ideal Age," was developed to elicit a respondent's perception of how others view ideal age. The question is phrased. "What age do you think most other people would like to be?"; the answer again must be expressed as "Age in Years." This question is also neither projective nor disguised: the respondent is asked directly what s/he considers to be an age judged ideal by our society, the age "most other people would like to be." It was thought that the question's response would indicate the age perceived as a kind of "generic" measure of socio-cultural ideal age.

HYPOTHESES

No empirical studies (see Barak and Stern 1986) report an association between biological sex and forms of subjective age, although some (e.g. Ross 1981) suggest that women's age concepts might be younger than men's. Ageist stereotyping, especially virulent and negative towards "older" women (Barak and Stern 1985; Nuessel 1982), leads to the expectation that the "Denial of Aging" process would be stronger for women. Even in the mid-1980's, advertisers and marketers still define "young" as well under-30 by featuring models in their teen and twenties. Even when positive "old-age" models are used, they tend to be male. Based on the premise that women are socially conditioned to view themselves as "younger" than comparably aged men, the following hypothesis was made:

H1: Women's subjective and ideal ages are "younger" than those of men, in all chronological age-cohorts.

Since chronological age is positively correlated with subjective and ideal age (Zola 1962; Barak 1987; Barak and Gould 1985), the implication is that the older one's real age, the older one's age-concept should be. To test this phenomenon, which has received little systematic study, the following hypothesis was made.

H2: The older one's chronological age cohort, the older one's subjective and ideal ages.

The difference and/or similarity between subjective age and forms of ideal age has seen very little study; the few findings are inconclusive and contradictory. Zola (1962) did not find subjective an; ideal age to be correlated, but Barak et al did find ideal age positively correlated with subjective age as measured by Cognitive and Identity Age (Barak 1987; Barak and Gould 1985; Barak and Stem 1985/86). Other contradictions concern ideal and subjective mean age scores. Zola (1962) reported men's and women's mean ideal age to be eight years "younger" than their subjective age. Barak and Gould (1985) found ideal age scores in a 30-69 population to be younger than those of subjective age. When only thirty year olds were considered, mean ideal age score (31.64) were older than those of subjective age (29.88), but this situation reversed itself among consumers 40+, where the ideal age mean score (35.29) was younger than subjective age (42.21). Since the findings are ambiguous, this null-hypothesis was postulated:

H3: Subjective and ideal ages will not differ from each other in chronological age cohorts.

METHOD

Instrument & Sample

An 11-page self-report questionnaire was completed by 760 men and women in the greater New York/New Jersey Metropolitan area in 1985/86. While the full instrument incorporates many age-related variables, the ones considered here are limited to: Chronological Age, Cognitive Age, and the three new ideal age scales: Desired, Personal Ideal and Generic Ideal Age. The data was collected from a convenience sample of respondents aged 20 to 80 at their homes and places of work by marketing students. An age and sex quota sampling procedure was followed to select a representative sample of men and women. The study reported on here is restricted to respondents between 25 and 69 (mean = 47.93) who answered all subjective and ideal age questions. These limitations reduced the total sample to 578 respondents (299 men & 279 women). Table 1 provides a demographic profile of the study sample. Reliability assessment helped establish internal stability: both the multi-item Cognitive and Desired Age scales possessed an alpha coefficient greater than .92.

TABLE 1

SAMPLE COMPOSITION: DEMOGRAPHICS

Analyses & Results

The first hypothesis was tested with T-Tests conducted between mean male and female age scores (respondents' mean age scores for the measures of Chronological, Cognitive, Desired, Personal Ideal and Generic Ideal Age in both sexes as well as the differing cohorts are provided in Table 2). For the total sample as well as the various age cohorts H1 was not supported: men and women, rather surprisingly (negative female age stereotypes are more frequent; see Nuessel 1981), do not differ in terms of ideal age perceptions.

The second hypothesis was tested (1) with Multivariate Analysis of Variance (employing the repeated measures design of MANOVA in SPSS/PC+) of age means across the three age cohorts, and (2) with T-Tests between mean Cognitive, Desired, Personal Ideal and Generic Ideal Age scores in one cohort vis a vis those found in the older cohort: Baby-Boomers vs. Pre-Boomers, Baby Boomers vs. Matures, and Pre-Boomers vs Matures (see Table 2 for age means utilized in the MANOVA and T-test analyses). The second hypothesis can be accepted: the "older" one's age-cohort, the "older" one's subjective and ideal ages will be. The MANOVA tests of Between-Subjects Effects using Unique sums of squares clearly established the significant difference between each cohort's age variables. The T-tests further helped establish the nature of these differences. There is only one exception: the mean Generic Ideal Age scores alone did not differ between the Baby-Boomers (24.99) and the Pre-Boomers (25.72). Furthermore, response patterns of male and female subjects in terms of age-decades, see Table 3, were similar the older the chronological age cohort, the greater the proportion of respondents indicating a personal age over 30.

The third hypotheses was assessed across the total sample as well as within each of the cohorts: first through the multivariate MANOVA procedure (Within Subject Effect of age means), second with T-Tests of the mean scores of the five age scales in relation to each other, and next with Pearson correlations between the age dimensions (see Table 4).

Based on these three analyses' results, it is clear that this hypotheses has to be rejected: the various age concept measures clearly differ from each other. In only one paired T-Test (T value = 1.1; d.f. = 168; p = .275) did two age measures not differ significantly: Baby-Boomers' mean Personal Ideal Age (25.48) did not differ from the cohort's mean Generic Ideal Age (24.99). All age dimensions were correlated except Chronological and Personal Ideal Age in the Pre-Boomer generation. The high correlations occurred in the total sample: The highest was between Chronological and Cognitive Age =. 85), and the second highest between Cognitive and Desired Age (r = .65). As Table 4 shows, these correlations decrease in strength when Pearson matrices in the three sub-cohorts are considered.

TABLE 2

AGE MEANS & STANDARD DEVIATIONS: TOTAL BY SEX AND AGE-COHORTS

DISCUSSION AND IMPLICATIONS

Although this exploratory study relied on a NorthEastern convenience sample (thus limiting generalizations), it provides several new insights into American age-role perceptions. The lack of difference between male and female subjective age self-concepts has marketing research (e.g. determination of the product mix offered by manufacturers and retailers) as well as advertising implications. First, pooled-sex samples are probably appropriate for age data analysis. But more interestingly, it may be possible to generalize from the predominantly female samples in prior age research, especially if replications find that the lack of sex differences holds true. As indicated, it is not as yet clear how much these findings can be generalized for the population as a whole. Nonetheless, the only national s probability study known to this paper's authors in which both male and female respondents indicated their (Cognitive Age (conducted by Needham and Harper; see Ross 1982) did not clearly establish if such age perceptions differ. On the other hand, all other studies (utilizing different subjective age measures) support the present findings concerning a lack of gender based differences (see Barak and Stern 1986).

TABLE 3

AGE-DECADE PATTERNS BY COHORT

Currently, marketers are more aware of the 40+ population's economic potential and this in turn has led to an increased interest in aging and the specifics of marketing to aging populations (e.g. see Barak and Stern 1985/86; Business Week 1986). The present findings might also be of particular use to advertisers interested in determining the appropriate age for models that appear in advertisements aimed at older consumers (Pre-Boomers and Matures). If both men and women view their actual and ideal age selves in similar terms, they can be matched more appropriately as spokespersons. Further research is necessary to see whether other assumptions concerning male/female valuation as spokespersons still hold true.

The use of spokespersons congruent with the target audience's age image is perhaps the most important advertising and public policy implication of the redefined concepts of subjective and ideal age. While the youth bias still exists - the respondent's mean perceptual age scores were younger than their Chronological Age in all populations considered personal age concepts seem to move upwards in tandem in each succeeding 15 year cohort (see Table 3). Marketers need to show sensitivity to consumer preferences for spokespeople and advertising models reflective of their own personal age ideals. If such spokespersons and/or models are too far removed in age from the target audience, unrealistic aspirations may alienate the consumer. Age congruence with a target audience's expectations seem especially important for age related product offerings (e.g. fashions and cosmetics) or services (e.g. health care, financial management, and leisure-time offerings). In general, a youthful bias is desirable, since energetic and stimulating spokespersons and models are most attractive, but the definition of "attractive" has as much to do with the target segment's age as that of the models featured.

The findings of an increase in age-role self-concepts as the respondent's chronological age increases might be indicative of a general age schema. While causal relations between objective and subjective forms of age would be reflective of age schema, if any, they are not as yet known. It is feasible that there may, in fact, be different age schemas within each age cohort, as well as a generally youth-biased age schema for all. We feel therefore that the concept and nature of age schemas certainly merits further investigation.

The finding that individuals consistently judge ideal age for others as younger than their personal ideal reinforces the American core cultural concept of youthfulness, yet raises questions as to its relevance in one's own life. While personal ideal age appears to be a function of where we are in time, a general context of admiration for youth also holds true. Americans have achieved a comfortable congruence between the personally subjective objective facts of their chronological existence by defining their own ideal age in relation to their birth certificates, and permitting the generic ideal for others to coexist, yet differ from, the personal. Overall, the study therefore established that consumers individually do not desire as much youth as they assume "others" do. The relationship between these two age ideals, however, is worth further study since it is not, as yet, known whether we merely pay lip service to a mythic Fountain of Youth or apply it in some still unknown way to our individual frame of reference.

TABLE 4

AGE CORRELATION MATRICES

APPENDIX 1

MEASUREMENT OF IDEAL AGE CONCEPTS

REFERENCES

Barak, Benny (1979), "Cognitive Reference Age Among the Elderly: A New Concept for Marketing," unpublished Ph.D dissertation, The City University of New York.

Barak, Benny (1987), "Cognitive Age: A New Multidimensional Approach To Measuring Age Identity," International Journal for Aging and Human Development, 25(1), 109-128.

Barak, Benny and Stephen Gould (1985), "Alternative Age Measures: A Research Agenda," in Advances in Consumer Research, Vol 12, eds. Elizabeth C. Hirschman and Morris B. Holbrook, Ann Arbor: Association for Consumer Research, 53-58.

Barak, Benny and Barbara Stern (1985), "Fantastic at Forty! The New Young Woman Consumer," Journal of Consumer Marketing, 2 (Spring), 41-54.

Barak, Benny and Barbara Stern (1985186), "Women's Age in Advertising: An Examination of Two Consumer Age Profiles," Journal of Advertising Research, 25 (December/January), 38-47.

Barak, Benny and Barbara Stern (1986), "Subjective Age Correlates: A Research Note," The Gerontologist, 26, 5, 571-578.

Belch, George, E. (1978), "Belief systems and the Differential Role of the Self Concept," in Advances in Consumer Research, Vol. 5, ed. H. Keith Hunt, Ann Arbor, MI: Association for Consumer Research, 320-325.

Caballero, Marjorie and Paul J. Solomon (1985), "A Longitudinal View of Women's Role Portrayal in Television Advertising," Journal of the Association of Marketing Science, 12 (Fall), 93-108.

England, Paula, Alice Kuhn, and Theresa Gardner (1981), "The Age of Men and Women in Magazine Advertisements," Journalism Quarterly, 58(August), 468-470.

Gerbner, George, Larry Gross, Nancy Signorelli, and Michael Morgan (1980), "Aging with Television: Images on Television Drama and Conceptions of Social Reality," Journal of Communication, 30(Winter).

Harris, Adella and Jonathan F. Feinberg (1977), "Television and Aging: Is What You See What You Get?" Gerontologist. 17 (December). 464-468.

Kastenbaum, Robert, Valerie Derbin, Paul Sabatini, and Steven Arrt (1972), "The Ages of Me' Toward Personal and Interpersonal Definitions of Functional Aging," Aging and Human Development, 3 (May), 197-211.

Nuessel, Jr. Frank H. 'The Language of Ageism," The Gerontologist, 22 (June), 273-276.

Onkvisit, Sat and John Shaw (1987), "Self-Concept and Image Congruence: Some Research and Managerial Implications," Journal of Consumer Marketing, 4 (Winter), 13-23.

Rosenberg, Morris (1979), Conceiving the Self, N.Y.: Basic Books.

Ross, Ivan (1982), "Information Processing and the Older Consumer: Marketing and Public Policy Implications" in Advances in Consumer Research, Vol 9, ed. Andrew Mitchell, Ann Arbor: Association for Consumer Research, 31-39.

Rotfield, Herbert I., Leonard N. Reid, and Gary B. Wilcox (1981), "Effects of Age of Models in Print Ads on Evaluation of Product and Sponsor," Journalism Quarterly, 59 (August), 374-381.

Sirgy, M. Joseph (1982), "Self-Concept in Relation to Product Preference and Purchase Intention," in Developments in Marketing Science, Vol 3, ed. V.V. Bellar, Marquette, MI: Academy of Marketing Science, 350-354.

Sirgy, M. Joseph (1982), "Self-Concept in Consumer Behavior: A Critical Review," Journal of Consumer Research, 9 (December), 287-299.

Sirgy, M. Joseph and Jeffrey Danes (1982), "Self-Image/Product Image Congruence Models," in Advances in Consumer Research, Vol 9, ed. Andrew Mitchell, Ann Arbor Association for Consumer Research, 556-561.

Stern, Barbara B., Benny Barak, and Stephen J. Gould (1987), "Sexual Identity Scale: A New Self Assessment Measure, Sex Roles. (17), 503-519.

U.S. Bureau of the Census (1982), Statistical Abstract of the United States: 1982-83, Washington, D.C.: U.S. Government Printing Office.

Zola, I.K. (1962), "Feelings about Age among Older People," Journal of Gerontology, 17, 65-68.

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