Age-Related Stereotypes - Semantic Scholar

Original Paper Gerontology 2006;52:324–333 DOI: 10.1159/000094614

Received: August 1, 2005 Accepted: May 5, 2006

Age-Related Stereotypes: A Comparison of American and Chinese Cultures Aysecan Boduroglu a Carolyn Yoon b Ting Luo c Denise C. Park d a

Department of Psychology, University of Michigan, Ann Arbor, Mich., b Ross School of Business, University of Michigan, Ann Arbor, Mich., USA; c Chinese Academy of Sciences, Beijing, China; d University of Illinois at Urbana-Champaign, Urbana-Champaign, Ill., USA

Key Words Culture Age Stereotype East-Asian Western

Abstract Background: It is commonly assumed that age-related stereotypes are more positive in East Asian cultures compared to Western cultures. However, research conducted in Western cultures has demonstrated that age-related stereotypes are multidimensional and their valence is content-dependent. Objective: In this study we investigated stereotypes about young and old adults, held by both young and old in the US and in China by focusing on the valence of age stereotypes across two content domains: social/emotional and mental/physical. The goal was to identify whether there were any cultural differences in age-related stereotypes in Chinese and American cultures. Methods: Both young and old Chinese and American participants were asked to describe typical young and typical old people. All responses were then coded for valence (positive/negative/neutral) and for content (mental/physical, social/emotional, other). Descriptors about young and old people were initially analyzed separately; then data were integrated to examine group tendencies to be more positive or negative for each target age group. Results: In both cultures, stereotypes reflected a shift from more positive to increasingly negative views of mental and physical traits as a function of aging. In social and emo-

© 2006 S. Karger AG, Basel 0304–324X/06/0525–0324$23.50/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/ger

tional domains, stereotypes regarding old and young adults were relatively neutral, except for a small positive bias found among the young Chinese adults for both target age groups. Conclusion: Our results indicate that age-related beliefs regarding typical older adults are similar across East-Asian and Western cultures and that a global positive bias for old age in East-Asia is absent. Copyright © 2006 S. Karger AG, Basel

Introduction

Old age has been simultaneously associated with positive and negative traits. Nevertheless, in the real world, the exhibition of ageist attitudes have been more widely documented [1]. Negative stereotypes of old age constitute part of these ageist attitudes. Negative stereotypes associated with aging include illness, uselessness, unattractiveness, mental decline, poverty, and depression. Many assume that such negative stereotypes are more prevalent in Western societies, and this stands in contrast to their views of aging in East-Asian cultures, in which they believe positive age-related stereotypes are more common [2]. Recent research on views of aging in EastAsian compared to Western cultures, however, seems to offer some indication that age-related stereotypes in EastAsia do not reflect greater positivity [3–5]. For instance,

Aysecan Boduroglu Department of Psychology 525 East University 4422 East Hall Ann Arbor, MI 48109-1109 (USA) Tel. +1 734 647 6250, Fax +1 734 763 7480, E-Mail [email protected]

Harwood et al. [3, 4] reported that in both Asian and Western cultures while benevolence increased with old age, vitality decreased, and Ryan et al. [5] documented that both positive and negative beliefs exist about communication skills of older adults in Chinese, Korean and Canadian cultures. These studies have served to highlight the need to consider negative as well as positive beliefs about old age in East-Asia. In the present paper, we seek to add to a growing body of evidence emphasizing age-related stereotypes in both East-Asian and Western cultures are multidimensional. What makes the current study novel is that we explore both positive and negative age-related stereotypes in China and the United States by identifying and comparing content-specific beliefs about young and old target groups. We further consider the interaction of participant and target ages in content-specific age beliefs across cultures. Stereotypes about Aging in Western Cultures A growing body of literature suggests that there are multiple stereotypes about aging [6–9]. These stereotypes vary in the extent to which they are positive or negative. However, two meta-analyses summarizing attitudes about old adults have shown that in Western cultures, old adults are generally rated more negatively than young adults [10, 11]. For instance, from a sample of 99 traits generated by American undergraduates about a typical older adult, Schmidt and Boland [9] noted that only a third were positive. When these traits were sorted according to which of them would be found together in a person, twice as many negative clusters (e.g. severely impaired), compared to positive ones (e.g. ‘perfect grandmother/parent’) emerged [8, 9]. Variables such as participant age [7, 12] and generational differences [13] could not account for the greater negativity observed towards old adults, thus further suggesting that the negative view they reported did not merely reflect an out-group bias held by young undergraduate study participants. Attempts have been made to specify precisely the source of negativity in views towards old adults. In a traitrating task, undergraduates indicated only poor physical abilities as more appropriate descriptors of increased age; no systematic pattern emerged of how appropriate they found attributes describing cognitive and personality variables for young, middle-aged, and old adults [14]. Unfortunately, whether the same pattern holds true for stereotypes held by old adults was not examined in the numerous earlier studies that have investigated perceptions of older adults of their own age group [7, 12]. These studies did not address the issue of whether stereotypes in Age-Related Stereotypes

certain domains (e.g. cognitive, personality) were more or less positive, but instead focused primarily on identifying major stereotypical archetypes. In one of the most comprehensive studies exemplifying this approach, Hummert et al. [12] compared how young, middle-aged and old adults grouped traits into clusters based on whether they would belong to the same person. They concluded that all participants agreed on 7 archetypes (e.g. ‘golden ager’, ‘severely impaired’, ‘perfect grandmother’). Even though these archetypes were informative about the sorts of stereotypes people have of old adults, the study precluded a clear classification of positive and negative traits across content domains because it allowed for stereotypic traits from different content domains to be included within one archetype. For instance, Hummert et al. [12] reported that in the category of ‘perfect grandmother’, both young and middle-aged adults included the traits ‘loving’ and ‘intelligent’, both of which are positively valenced but potentially descriptive of different domains (interpersonal relationships and cognitive functioning, respectively). In the present study, we focus on specifying the distribution of positive and negative age-related stereotypes within certain content domains across both age and cultural groups. Accordingly, we examine both positive and negative beliefs about aging, consistent with the approach by Ryan et al. [5] in their examination of communication skills of older adults. In addition, we aim to lend further insights by distinguishing between traits descriptive of social and emotional domains, and those descriptive of mental and physical domains. These content domains were found to represent the key dimensions of interest, based largely on initial inspection of the data that were collected in the present study (to be described later on) as well as the literature on age related stereotypes [3, 4], and stereotypic content [15]. Stereotypes about Aging in East-Asian Cultures Age-related stereotypes are believed to reflect cultural norms of social interaction and consequently it is not surprising that the nature of age-related stereotypes show some variation across cultures. Over the years, these cultural variations in age-related stereotypes have been characterized by the existence of more positive views regarding older adults in East-Asian cultures compared to the negative stereotyping of them in Western cultures [2, 16, 17]. Particularly, the highly influential work by Levy and Langer [16] investigating the relationship between age stereotypes and memory performance in Chinese and American cultures has brought widespread recogniGerontology 2006;52:324–333

325

tion of this view. Levy and Langer suggested that old Chinese performed significantly better than old Americans on memory tasks, a difference that the authors attributed to the lack of ageism in East-Asian cultures [16, but see 18]. However, this predominantly positive view about aging in East-Asian cultures is being increasingly challenged by evidence suggesting a co-existence of both positive and negative associations [3, 4, 19], as in Western cultures [12]. In a series of studies, involving a large number of cultures around the Pacific Rim, Harwood et al. [3, 4] attempted to delineate the nature of age-related stereotypes across a number of East-Asian (e.g. Chinese, Thailand, Philippines) and Western (e.g. American, Australian) cultures. In both these studies, older adults were evaluated on 9 traits (active, attractive, liberal, strong, healthy, kind, generous, and wise). The ratings of these traits by both young and old adults revealed that in all cultures, ‘vitality’ was perceived to decrease, while ‘benevolence’ was perceived to increase with old age. Notwithstanding the insights that were generated from these studies about some basic trends in age-related attitudes across cultures, some methodological issues rendered the results difficult to interpret.1 Furthermore, as the authors themselves acknowledged, results from a trait-rating task may not have provided a comprehensive account of age-related stereotypes across these different cultures. Thus, instead of imposing a predetermined set of traits on study participants, we adopt an open-ended approach for studying age-related stereotypes cross-culturally. In order to advance further our understanding of cross-cultural differences in East Asian and Western cultures, we do not constrain subjects’ choice of descriptors or traits associated with aging. Moreover, we systematically investigate the extent of negativity (or positivity) of age stereotypes held by young and old adults across a number of content domains. Specifically, based on descriptors elicited from participants and the review of the existing literature, we identify two content domains – social/emotional, mental/physical – that are then compared across age and cultural groups. For stereotypes about old adults in the US, we predict, that both young and old participants will generate both positive and negative descriptors. However, based on pri1

These difficulties center mainly on translation and interpretation of traits across cultures, and convenience sampling resulting in wide age distributions of the samples across cultures (e.g. age range of 48–96 years for the old adult sample).

326

Gerontology 2006;52:324–333

or findings, we would expect a negative bias to be more pronounced in mental and physical descriptors people generate rather than in the social and emotional descriptors. In line with Harwood et al.’s [3, 4] findings of increased benevolence, we propose that, in China, both young and old participants are likely to generate relatively favorable views about old adults in the social and emotional domains. Findings on decreased vitality with aging, on the other hand, suggest that more negative descriptors of aging may be generated by both young and old adults in mental and physical domains. The constrained set of traits that were used in these earlier studies to uncover age-related beliefs, however, does not permit strong predictions. We speculate that whereas old Chinese participants are likely to generate negative stereotypes in mental and physical domains based on their subjective experience of declines in these domains, young Chinese participants may not necessarily focus on negative aspects in their descriptions. Stereotypes about Young Adults Stereotypes about young adults have been studied primarily in the context of the younger target ages serving as a control group for the older adults comprising the main target group of interest. There are very few studies that have directly focused on stereotypes of young adulthood, and most of them have focused only on younger American participants [8, 13]. Further, prior research has neither explored stereotypes across content domains in a detailed fashion nor addressed the question of how age related stereotypes about young adults are different in East-Asian versus Western cultures. The few studies conducted in Western cultures that have included older as well as younger participants, found that stereotypes about typical younger adults are more often positive than negative, independent of the participants’ own age [13, 20, but see 21]. This positivity reflected in stereotypes young adulthood has been attributed to the idealization of youth in Western cultures and/or reflections of a relatively favorable past. The negativity associated with adolescents was evident in ratings by middle-aged and older adults and were limited to domains such as autonomy, personal integrity and life satisfaction, and social acceptability [21]. The extent to which similar ratings would characterize middle-aged and older adults’ views of young adults rather than adolescents is not known. With respect to stereotypes about young adults in China, no study to date, to the best of our knowledge, has directly addressed this issue. However, some research on younger Chinese adults, the so-called Boduroglu /Yoon /Luo /Park

‘s-generation’2, suggest that they have adopted more Western-like values with respect to youth, in contrast to older generations who have not [22]. Overall, we expect to find a positive bias towards young adults in both cultures, with the possibility that the bias is more pronounced in a Western culture which tends to idealize youth over old age. Furthermore, it is likely that this positive bias may be pronounced for stereotypes related to the mental and physical characteristics of young adults and less so for the ones related to the social and emotional dimensions. However, given that only few earlier studies investigate content domains in which positive and negative views about young adults are likely to manifest, our approach can be best characterized as exploratory. Approach In this study, we employed an open-ended approach to data collection. We first collected, from young and older study participants, descriptors of youth and old age in the United States and China, classified them into specific content domains, and then determined their valence (positive, negative or neutral). This approach of asking young and old adults about both youth and old age also enabled direct comparisons of the participant groups’ (young American, old American, young Chinese, and old Chinese) views about target age groups across content domains. In order to develop a comprehensive picture of age-related stereotypes across United States and China, we initially analyzed the data for each target age separately. Then, we described overall group tendencies to be positive vs. negative about the stereotypes generated for each target age. We did so by considering the data generated by each participant group separately, and comparing the relative proportions of positive and negative responses for each content domain.

Table 1. Demographic characteristics for study participants by

culture and age Group

Americans Younger Older Chinese Younger Older

Age

Educationa

Healthb

mean SD

mean SD

mean SD

23 20

18.91 0.73 70.10 4.35

14.00 0.00 16.65 2.23

3.65 3.80

1.03 1.01

18 19

21.22 1.31 66.16 1.86

14.33 1.03 16.00 0.00

3.33 3.53

0.84 0.62

n

a

Number of years. Self-report health measure (1–5; 1 = much worse than average; 5 = much better than average). b

Arbor, Mich., USA and the younger Chinese (8 males, 10 females) were students at the University of Science and Technology, Beijing, China. The older Americans (11 males, 9 females) were community-dwelling residents in the Ann Arbor, Michigan area. Older Chinese (11 males, 8 females) participants were recruited from a pool of retired staff and faculty at the National Academy of Sciences in Beijing, China. The young American group received course credit for study participation; all other participants received monetary compensation. American participants were tested at the University of Michigan, and Chinese participants were tested at the Institute of Psychology, National Academy of Sciences, in Beijing. Sample characteristics are presented in table 1.4 Each culture by age group had at least some college education. All groups also indicated having better general health status compared to an average person in their own age group, with no significant differences found across the four comparison groups.

Participants Eighty participants in two age groups from the US (primarily European-Americans and excluding Asian-Americans) and China were sampled. Younger Americans (10 males, 13 females)3 were undergraduate students at the University of Michigan, Ann

Procedures All materials administered in China were translated from English into Chinese and back translated from Chinese into English by two bilingual Mandarin-English speakers fluent in both languages. Two additional fluent bilingual Mandarin-English speakers translated all of the data generated by the Chinese participants into English. These individuals provided detailed explanations for any phrases that either did not have direct correspondences in English, or were related to specific cultural phenomena to ensure that the coders would be able to judge the content appropriately. The experimental sessions were conducted in groups of two to four participants; all participants belonged to the same age group. At the start of the session, participants provided informed con-

2

Refers to the first generation born shortly before and after the official installation of China’s single child policy in 1978. 3 Of the 23 young American participants, 4 were excluded from analyses with the consensus of two coders. The excluded participants generated stereotypes for targets falling outside the traditional age ranges for young and old adults.

4

Age-Related Stereotypes


Method

Even though there were slight age differences in young and old samples across the cultures, samples were drawn from subject pools representing approximately comparable parts of the age distributions in each culture [23].

327

sent and then were immediately presented with the description/ trait generation task. To generate stereotypic content, participants were presented with the same question used by Schmidt and Boland [9]: ‘How would you describe the typical old (young) person? Please write down everything you typically think about, hear about, or read about the elderly (young) people. Include anything that is typically associated with the elderly (young) regardless of whether it is favorable or unfavorable or whether you personally believe it to be true.’ All participants generated responses about both old and young adults, completing the task about their own age group first. Such ordering ensured that both young and old participants generated stereotypic content about their in-group first. During each phase, participants were told to restrict their responses to a maximum of 10 descriptors or short phrases and were given 5 min to write down their answers on separate sheets during each phase. Upon completion of these tasks subjects filled out a demographic questionnaire, followed by a health questionnaire. The participants were then debriefed, thanked for their participation and dismissed. Coding The coding scheme was developed to distinguish between a mental/physical dimension and a social/emotional dimension of aging. Two independent coders in the US coded each response item for valence (positive, negative, neutral) and for content (mental/physical, social/emotional, other). Descriptors about mental and physical abilities were coded in the same category because for many of the items, it was not easy to discriminate whether an item was more about mental or physical abilities (e.g. ‘slow’). The same was true of social and emotional items (e.g. ‘cheerful’). High interrater reliability was established for both sets of domains (Cronbach’s = 0.98 for valence and 0.89 for content). For the valence dimension, items that had a clearly positive (e.g. ‘wise’) or negative (e.g. ‘ill’) component were coded as positive and negative, respectively. The items that were coded as neutral, either had no strong valence component, or reflected likes or dislikes of the rated group (e.g. ‘likes/dislikes loud music’). For the content dimension, items about mental (e.g. ‘wise, slow in thinking’) and physical abilities (e.g., ‘bad drivers’) as well as physical states (e.g. ‘walks with a stagger’) were coded as ‘mental/ physical.’ Items defining personality traits (e.g. ‘kind’) and individuals’ interactions within the society (e.g. ‘grandmotherly’) were coded as ‘social/emotional.’ Items belonging to neither of these two content domains were labeled as ‘other’.

Results

For each target age group, an approximately equal number of responses were generated (mean = 7.96 and mean = 7.79, for stereotypes about young and old, respectively). For both target ages, more than half of the participants produced eight or more items with only 1% of the participants producing fewer than five items. Proportions of responses, rather than raw scores, were used to categorize the types of responses generated by participants. 328


Fig. 1. The breakdown of valence across age and culture groups for stereotypes about the old. The top panel summarizes the data from the young respondent and the bottom panel from the old respondents.

There were nine possible categories of responses: three valences (positive, negative and neutral) crossed with three content areas (mental/physical, social/emotional, other). The proportion of responses falling into each of these categories constituted the main dependent variable. Data analyses were organized by conducting separate analyses for stereotypes about young adults and old adults. For each set of analyses, a repeated measures ANOVA was performed with Age of Participant (2: young vs. old) and Culture (2: American vs. Chinese) as between-group variables, and Valence (3: positive, negative, neutral) and Content (3: mental/physical, social/emotional, other) as within-subjects variables.5 Stereotypes about Old Adults Analysis of responses for stereotypes about old adults yielded a number of significant main effects and 5

Analyses were also conducted after separating the data into earlier and later responses in a manner consistent with work by Trafimov et al. [24]. The results were generally consistent with our reported findings and are available upon request.

Boduroglu /Yoon /Luo /Park

Fig. 2. The valence distribution in each one of the content do-

mains for the two culture groups (collapsed across age) for stereotypes about old adults. In the top panel, the data for the mental/ physical domain is presented. The lower panel summarizes the data for the social/emotional domain.

interactions. In the interest of brevity, we focus on the interpretation of the two highest order interactions: Age ! Culture ! Valence, F (2,152) = 4.35, MSE = 0.039, p ! 0.02, p2 = 0.054; and Culture ! Valence ! Content, F (4,304) = 3.59, MSE = 0.046, p ! 0.007, p2 = 0.045. The Age ! Culture ! Valence interaction was decomposed by conducting specific contrasts, comparing the proportion of positive and negative responses generated by the four culture by age groups (fig. 1). These analyses revealed that the interaction was due to younger Chinese generating an equivalent proportion of positive and negative responses, t(17) = 0.52, p 1 0.61, and all other groups generating significantly more negative than positive responses about the typical old adult (all p ! 0.002). We had predicted that a positive bias would emerge in Chinese responses, but such a bias was evident only among the young Chinese, indicating that the bias may be a function of the respondent’s age. Since we were particularly interested in the distribution of positive and negative traits across different content domains, the significant Culture ! Valence ! ConAge-Related Stereotypes

Fig. 3. This figure summarizes the significant 3-way interaction reported for the stereotypes about younger adults. The figure depicts the breakdown of the content (mental/physical and social/ emotional) items across the 3 valence categories (positive, negative and neutral), for the four age by culture groups. The top panel summarizes the data about the mental/physical domain and the lower panel summarizes the data about the social/emotional domain.

tent interaction was further decomposed by conducting separate analyses for the two content domains – mental/ physical and social/emotional6 (fig. 2). The analyses revealed that the interaction emerged because of the higher proportions of negative responses generated by Americans than Chinese in the mental/physical domain, F (2,156) = 3.70, MSE = 0.036, p ! 0.03, p2 = 0.054; and greater proportions of positive responses generated by Chinese than Americans, in the social/emotional domain, F (2,156) = 6.19, MSE = 0.048, p ! 0.003, p2 = 0.074.

6

Items that were coded as ‘other’ were not further investigated since this category of items did not represent any conceptual overlap across groups.


329

Stereotypes about Young Adults Analyses of the data on the stereotypes about young adults also revealed a number of significant main effects that were moderated by a 3-way interaction of Culture ! Valence ! Content, F (4,304) = 5.81, MSE = 0.028, p ! 0.001, p2 = 0.071. Simple effects analysis by content was thus conducted (fig. 3). Stereotypes about youth were not exclusively positive, with opposing trends emerging across the two content domains. In the mental/physical domain, disproportionately more positive than negative responses were generated in both cultures t(45) = 7.40, p ! 0.001, and t(38) = 6.63, p ! 0.001, for Americans and Chinese, respectively; and in the social/emotional domain, Americans produced more negative than positive items, t(45) = 3.31, p ! 0.002, while the Chinese produced an equivalent proportion of positive and negative items, t ! 1. Comparison of Stereotypes about Young and Old Adults across Content Domains Finally, in order to examine overall group tendencies to be more positive or negative about the stereotypes generated for each target age, the ratio of proportions of positive to negative responses in the mental/physical and social/emotional domains were calculated. A ratio of 1 indicated that participants were equally positive and negative for a particular domain, and a ratio greater than or less than 1 represented a positive or negative bias, respectively. In order to create a distribution that is compatible with more conventional indicators of neutrality [i.e. (origin = 0), (positivity 1 0), and (negativity ! 0)], natural logarithms of all positive-to-negative proportions were calculated. These are presented in figure 4 for both target groups across both content domains.7 As can be seen in figure 4, the major difference between stereotypes about young (left panel) and old adults (right panel) is in the mental/physical domain. While young age is characterized by positive stereotypes in the mental/physical domain, stereotypes about old age reflect an opposite pattern, with greater proportions of negative stereotypes being generated in this domain. Furthermore, in the mental/physical domain for stereotypes about both young and old adults, the young Americans exhibited the greatest bias, generating the highest pro7

No statistical tests were conducted on these proportions since to our knowledge there are no tests to handle such data. Since these proportions were calculated by taking the ratio of the frequencies in the two content domains for each age by culture group, there was no within group variance, and furthermore we could not use any tests typically used for frequency data, because the final data points we considered were ratios.

330


Fig. 4. This figure depicts the natural logarithm of the propor-

tions of positive to negative stereotypes generated by each age by culture group for each content category. A value of 0 means the group has a neutral view on the specified dimension; values greater than 0 and less than 0 represent, predominantly positive and negative views, respectively. The left side of the figure depicts stereotypes about young adults and the right side of the figure displays stereotypes about older adults. For each target age group, the first four bars display proportions about the mental/physical dimension, and the next set of 4 bars display proportions about the social/emotional dimension.

portion of both positive and negative responses for the young and old, respectively. In marked contrast to the positive and negative biases present in the mental/physical domain for the young and the old, respectively, we found an approximately equal proportion of positive and negative responses in the social/emotional domain, thus, indicating an overall neutral pattern. More interestingly, the pattern of findings for stereotypes about young and old, yielded similar results. While the young Chinese generated more positive stereotypes for both target age groups in this domain, all the Americans, but only the old Chinese, generated a greater proportion of negative responses. What was an even more striking result, however, was that the only positive bias about older adults was reflected in the responses from young Chinese, and limited to social/emotional qualities of old adults.


This study investigated stereotypes about young and old adults held by both young and old in a Western and East-Asian culture, and focused on the valence as well as content domains of stereotypes. This study is unique in its attempt to investigate cross-cultural age-related stereotypes by simultaneously utilizing an open-ended trait generation procedure, and its inclusion of two target age groups (young and old) within a multi-generational approach. Also, to our knowledge, along with the recent work by Ryan et al. [5], this is one of the first studies to investigate the valence of age-related stereotypes across content domains for stereotypes about old age in Western cultures and for both target age groups in an East-Asian culture. The major finding revealed by our comparisons was that stereotypes about young and old age are mainly characterized by a shift from positive to negative mental/ physical traits as a function of aging. So, whereas our predictions of negative stereotypes by Americans were supported, no evidence of a global positive bias towards old age emerged in the Chinese responses. This finding thus complements that of Harwood et al. [3, 4], who noted that both old and young participants in both East-Asian and Western cultures rated old adults negatively on physical descriptors. Future studies need to replicate and further establish the generalizability of these findings and address some procedural shortcomings of the current research. For instance, it would have been more ideal if all responses were coded both by American and Chinese coders; the coding of the responses by non-Chinese coders may have biased our findings. Specifically, what is considered as positive or negative in one culture may not be considered equally negative or positive in the other. However, the converging nature of the evidence from our study and those reported by others using different procedures [3, 4] speak to the value of our findings. The findings reported here have implications for research investigating resistance to cognitive decline that is typically associated with aging. The negative bias about old adults in the mental/physical domain held by all four groups of participants suggests that the older Chinese adults may not be entirely resistant to effects of ageism as suggested by researchers asserting that positive views about aging in East-Asian cultures serve to mitigate the effects of age-related cognitive declines [16, but also see 18]. However, given the debate on the relationship between age-related stereotypes and aging self-stereotypes [25–28], it is essential that the extent to which the nega-

tivity reported for old adults is reflected in self-stereotypes and that this be investigated cross-culturally. To our knowledge, no cross-cultural research has directly investigated how stereotypes about ‘typical’ old adults differ from ‘self’ stereotypes. The presence of similarly valenced age-related stereotypes about ‘typical’ old adults across Chinese and American cultures does not provide any indication of the valence and content distribution of self-stereotypes. It is possible that there are more positive self-stereotypes among Chinese elderly, and this in turn may impact cognitive performance on certain tasks. Another interesting finding that emerged from our analyses of valence distributions across content domains was with respect to stereotypes about old adults held by young Chinese participants. Our analyses revealed that the young Chinese generated only disproportionately more positive responses in the social/emotional domain; their other responses in the mental/physical domain were mostly negative. Realistic evaluations of cognitive declines and the salient physical changes might have contributed to the more negative mental/physical traits generated. We speculate that the favorable descriptors in the social/emotional domain may have emerged as a function of internalized notions of filial piety. In many Eastern cultures, including Japan, Korea, China, and other Chinese communities, Confucian’s teachings of filial piety are a guiding principle in children’s socialization [29]. Filial piety demands that one should be obedient but rational and ensure the emotional and physical well being of one’s aged parents [29, 30]. Furthermore, adherence to filial piety requires the absolute authority of parents over children and moreover, it also dictates the authority of those senior in generational rank over those who are younger [29]. It is conceivable that such beliefs manifest themselves in positive social/emotional stereotypes held by young Chinese. However, we cannot rule out the possibility that the positive social/emotional responses generated by young Chinese reflect normative rather than actual attitudes and beliefs towards old adults. It is thus important that the existence of such positive attitudes towards aging be corroborated through studies using implicit measures, like the implicit associates test used by Hummert et al. [31]. This is especially important given past research indicating that in some East-Asian cultures both positive and negative attitudes towards old adults co-exist; but with the possibility of the latter being masked by cultural norms of interaction, requiring respectful displays towards old adults [32]. In both cultures, responses from old adults about their own age group also reflected a negative bias, confirming



Discussion

331

earlier work showing that old adults do not have in-group favoritism, when the group identity is defined around old age [33]. Regarding the stereotypes generated about young adults, the positive bias about youth seems to be less pervasive than was expected. Rather, positive views about young adults were particularly bound to the mental/physical domain in both cultures, complementing earlier findings that younger Chinese have adopted notions similar to Western ones regarding youth [22]. These relatively more negative views associated with young age, and limitations of positivity to the mental/physical domain is also consistent with earlier reports indicating that young, middle-aged and older adults report relatively more negative views about younger adults on social dimensions such as acceptability, integrity and autonomy [21]. Somewhat more exaggerated negative views about social/emotional qualities of young adults in the U.S. versus China are consistent with cultural differences documented in the literature; namely, that values of getting ahead and achievement are much stronger in the US; and that interdependence and harmony among group members are emphasized in China [34]. Although categorizing the descriptors across two valence and two content domains did simplify stereotypic content, it also served to effectively capture meaningful trends within and across cultures. Furthermore, the two content domains, mental/physical and social/emotional, over which the nature of stereotypes were evaluated, map closely on to the main dimensions considered by models of stereotype content – competence and warmth, respectively [15]. We nonetheless believe that a complete understanding of age-related stereotypes in East-Asian cultures also requires the use of trait sorting procedures that has enabled researchers to identify stereotypic archetypes in Western cultures [12]. It remains an open empirical question whether the similar negative mental/physical traits generated by participants in both cultures would be grouped together to yield the same archetypes in both

cultures. Also, detailed methodological approaches investigating how these stereotypes elicited about the typical older or younger adults relate to real-world beliefs about younger and older adults are needed. In conclusion, the present study advances the notion that mental and physical declines associated with aging are highly salient in both East-Asian and Western cultures. Even though it has long been assumed that the Chinese cultural context discourages ageist attitudes and values towards the elderly, our findings suggest that these assumptions are not totally accurate. In the mental/physical domain, young and old participants in both China and in the United States generated positive items only as descriptive of young adults, and stereotypes about old age reflected a major shift towards the negative. The only positive bias about old adults emerged from the young Chinese group, but it was limited solely to descriptors of social and emotional qualities of old adults. Although filial piety may underlie the positive bias expressed by the young Chinese, a clearer picture has yet to emerge on how filial piety influences explicit and implicit attitudes about aging in China. More importantly, this dissociation of young Chinese responses across the mental/physical and social/emotional domains, demonstrates the importance of analyzing positive and negative beliefs across different content domains.

Acknowledgements This research was supported by a grant from the National Institute on Aging (NIA) Center on Aging and Cognition: Health, Education, and Training and by NIA Grant 5-R01-AG15047 awarded to Denise C. Park and Richard Nisbett. We would like to thank Oscar Ybarra for helpful advice on the coding scheme, and Angela Gutchess for comments on the manuscript. We also thank Julie Haumataki, Wei Huang, Yanfen Zhang and Gang Guo for their help.

References 1 Nelson TD: Ageism: Stereotyping and Prejudice against Older Persons. Cambridge, MIT Press, 2002. 2 Palmore E: What can the US learn from Japan about aging? Gerontologist 1975;15:64– 67. 3 Harwood J, Giles H, McCann RM, Cai D, Somera LP, Ng SH, Gallois C, Noels K: Older adults’ trait ratings of three age-groups around the Pacific rim. J Cross Cult Gerontol 2001;16:157–171.

332

4 Harwood J, Giles H, Ota H, Pierson HD, Gallois C, Ng SH, Lim T, Somera L: College students’ trait ratings of three age groups around the Pacific rim. J Cross Cult Gerontol 1996; 11:307–317. 5 Ryan EB, Jin Y, Anas A, Luh JJ: Communication beliefs about young and old age in Asia and Canada. J Cross Cut Gerontol 2004; 19: 343–360.


6 Brewer MB, Dull V, Lui L: Perception of the elderly: stereotypes as prototypes. J Pers Soc Psychol 1981; 41:656–670. 7 Heckhausen J, Dixon RA, Baltes PB: Gains and losses in development throughout adulthood as perceived by different adult age groups. Dev Psychol 1989; 25:109–121. 8 Hummert ML: Multiple stereotypes of elderly and young adults: a comparison of structure and evaluations. Psychol Aging 1990; 5: 182–193.


9 Schmidt DF, Boland SM: Structure of perceptions of older adults: evidence for multiple stereotypes. Psychol Aging 1986; 3: 255– 260. 10 Kite ME, Johnson BT: Attitudes toward older and younger adults: a meta-analysis. Psychol Aging 1988;3:233–244. 11 Kite ME, Stockdale GD, Whitley BE, Johnson BT: Attitudes toward younger and older adults: an updated meta-analytic review. J Soc Issues 2005;61:241–266. 12 Hummert ML, Garstka TA, Shaner JL, Strahm S: Stereotypes of the elderly held by young, middle-aged and elder adults. J Gerontol [B] 1994;49:240–249. 13 Slotterback CS: Projections of aging: Impact of generational differences and the aging process on the perceptions of adults. Psychol Aging 1996;11:552–559. 14 Slotterback CS, Saarnio DA: Attitudes towards older adults reported by young adults: variation based on attitudinal task and attribute categories. Psychol Aging 1996;11:563– 571. 15 Fiske ST, Cuddy AJC, Glick P, Xu J: A model of (often mixed) stereotype content: competence and warmth respectively follow from perceived status and competition. J Pers Soc Psychol 2002; 82:878–902. 16 Levy B, Langer E: Aging free from stereotypes: successful memory in China and among the American deaf. J Pers Soc Psychol 1994;66:989–997. 17 Chang BL, Chang AF, Shen Y: Attitudes towards aging in the United States and Taiwan. J Comp Fam Stud 1984;15:99–130.


18 Yoon C, Hasher L, Feinberg F, Rahhal TA, Winocur G: Cross-cultural differences in memory: the role of culture-based stereotypes about aging. Psychol Aging 2000; 15: 694–704. 19 Liu JH, Ng SH, Loong C, Gee S, Weatherall A: Cultural stereotypes and social representation of elders from Chinese and European perspectives. J Cross Cult Gerontol 2003;18: 149–168. 20 Matheson DH, Collins CL, Kuehne VS: Older adults’ multiple stereotypes of young adults. Int J Aging Human Dev 2000;51:245– 257. 21 Luszcz MA: Characterizing adolescents, middle-aged and elderly adults: putting the elderly into perspective. Int J Aging Human Dev 1985–1986;22:105–121. 22 Li C: China: The Consumer Revolution. Singapore, Wiley, 1998. 23 Hedden T, Park DC., Nisbett RE, Ji LJ, Jing Q, Jiao S: Cultural variation in verbal versus spatial neuropsychological function across the life span. Neuropsychology 2002;16: 65– 73. 24 Trafimov D, Triandis HC, Goto SG: Some tests of the distinction between the private self, and the collective self. J Pers Soc Psychol 1991;60:649–655. 25 Hummert ML: When is an age stereotype an aging self-stereotype: a commentary. J Gerontol [B] 2003;58:212–213.

26 Levy B: Mind matters: cognitive and physical effects of aging self-stereotypes. J Gerontol [B] 2003;58:203–211. 27 Levy B: Conscious versus unconscious levels of aging self-stereotypes: author’s reply. Gerontol [B] 2003b;58:215–216. 28 Zebrowitz LA: Aging stereotypes-internalization or inoculation? A commentary. J Gerontol [B] 2003;58:214–215. 29 Ho Y: Filial Piety and Its Psychological Consequences; in Bond MH (ed): Handbook of Chinese Psychology. Hong Kong, Oxford University Press, 1996, pp 155–165. 30 Sung K: Elder respect: exploration of ideals and forms in East Asia. J Aging Stud 2001;15: 13–26. 31 Hummert ML, Garstka TA, O’Brien LT, Greenwald AG, Mellott DS: Using the implicit association test to measure age differences in implicit social cognitions. Psychol Aging 2002;17:482–498. 32 Koyano W: Japanese attitudes towards the elderly: a review of research findings. J Cross Cult Gerontol 1989;4:233–244. 33 Kite ME, Wagner LS: Attitudes toward older and younger adults; in Nelson T (ed): Ageism: Stereotyping and Prejudice Against Older Persons. Boston, MIT Press, 2002, pp 129–161. 34 Fiske A, Kitayama S, Markus HR, Nisbett RE: The cultural matrix of social psychology; in Gilbert DT, ST Fiske, Lindzey G (eds): The Handbook of Social Psychology. New York, McGraw-Hill, 1998, pp 915–981.


333