Profession Comm. 8

Perceptions of Aging Across 26 Cultures and Their Culture-Level Associates

Corinna E. Löckenhoff, Filip De Fruyt, Antonio Terracciano, Robert R. McCrae, Marleen De Bolle, Paul T. Costa Jr., Maria E. Aguilar-Vafaie, Chang-kyu Ahn, Hyun-nie Ahn, Lidia Alcalay, Juri Allik, Tatyana V. Avdeyeva, Claudio Barbaranelli, Veronica Benet-Martinez, Marek Blatný, Denis Bratko,

Thomas R. Cain, Jarret T. Crawford, Margarida P. Lima, Emı́lia Ficková, Mirona Gheorghiu, Jamin Halberstadt, Martina Hřebı́čková, Lee Jussim, Waldemar Klinkosz, Goran Knežević,

Nora Leibovich de Figueroa, Thomas A. Martin, Iris Marušić, Khairul Anwar Mastor, Daniel R. Miramontez, Katsuharu Nakazato, Florence Nansubuga, V. S. Pramila, Anu Realo, Jean-Pierre Rolland, Jerome Rossier,

Vanina Schmidt, Andrzej Sekowski, Jane Shakespeare-Finch, Yoshiko Shimonaka, Franco Simonetti, Jerzy Siuta, Peter B. Smith, Barbara Szmigielska, Lei Wang, Mami Yamaguchi, and Michelle Yik

Corinna E. Löckenhoff, Antonio Terracciano, Robert R. McCrae, and Paul T. Costa Jr., National Institute on Aging, Baltimore, MD; Filip De Fruyt and Marleen De Bolle, Department of Developmental, Personality, and Social Psychology, Ghent University, Ghent, Belgium. The other contributing members of the Adolescent Personality Profiles of Cultures Project are (in alphabetical order): Maria E. Aguilar-Vafaie, Department of Psychology, Tarbiat Modares University, Tehran, Islamic Republic of Iran; Chang-kyu Ahn, Department of Education, Pusan National University, Busan, South Korea; Hyun-nie Ahn, Department of Psychology, Ewha Womans University, Seoul, South Korea; Lidia Alcalay, Escuela de Psi- cologia, Pontificia Universidad Catolica de Chile, Santiago, Chile; Juri Allik, Department of Psychology, University of Tartu, Tartu, Estonia; Tatyana V. Avdeyeva, Graduate School of Professional Psychology, Uni- versity of St. Thomas; Claudio Barbaranelli, Department of Psychology, La Sapienza University, Rome, Italy; Veronica Benet-Martinez, Department of Psychology, University of California, Riverside; Marek Blatný, Institute of Psychology, Academy of Sciences of the Czech Republic, Prague, Czech Republic; Denis Bratko, Department of Psychology, Faculty of Philosophy, University of Zagreb, Zagreb, Croatia; Thomas R. Cain, De- partment of Psychology, Rutgers University; Jarret T. Crawford, Depart- ment of Psychology, Rutgers University; Margarida P. Lima, Faculty of Psychology and Educational Sciences, University of Coimbra, Coimbra, Portugal; Emı́lia Ficková, Institute of Experimental Psychology, Slovak Academy of Sciences, Bratislava, Slovak Republic; Mirona Gheorghiu, School of Psychology, Queens University, Belfast, United Kingdom; Jamin Halberstadt, Department of Psychology. University of Otago, Dun- edin, New Zealand; Martina Hřebı́čková, Institute of Psychology, Acad- emy of Sciences of the Czech Republic; Lee Jussim, Department of Psychology, Rutgers University; Waldemar Klinkosz, Department of Psy- chology, John Paul II Catholic University of Lublin, Lublin, Poland; Goran Knežević, Department of Psychology, Faculty of Philosophy, Belgrade University, Belgrade, Serbia; Nora Leibovich de Figueroa, Research Insti- tute, Department of Psychology, University of Buenos Aires, Buenos Aires, Argentina; Thomas A. Martin, Susquehanna University; Iris Maru- šić, Institute for Social Research in Zagreb, Zagreb, Croatia; Khairul Anwar Mastor, School of General Studies, Universiti Kebangsaan Malay- sia, Bangi, Malaysia; Daniel R. Miramontez, Department of Psychology, University of California, Riverside; Katsuharu Nakazato, Faculty of Social Welfare, Iwate Prefectural University, Iwate, Japan; Florence Nansubuga, Institute of Psychology, Makerere University, Kampala, Uganda; V. S. Pramila, Department of Psychology, Andhra University, Visakhapatnam, India; Anu Realo, Department of Psychology, University of Tartu; Jean- Pierre Rolland, Université de Nanterre, Paris X, Nanterre, France; Jerome

Rossier, Institute of Psychology, University of Lausanne, Lausanne, Swit- zerland; Vanina Schmidt, Research Institute, Department of Psychology, University of Buenos Aires, Buenos Aires, Argentina; Andrzej Sekowski, John Paul II Catholic University of Lublin; Jane Shakespeare-Finch, School of Psychology and Counselling, Queensland University of Tech- nology, Brisbane, Australia; Yoshiko Shimonaka, Department of Human Studies, Bunkyo Gakuin University, Bunkyo, Japan; Franco Simonetti, Escuela de Psicologia, Pontificia Universidad Catolica de Chile; Jerzy Siuta, Institute of Psychology, Jagiellonian University, Krakow, Poland; Peter B. Smith, Department of Psychology, University of Sussex, Falmar, United Kingdom; Barbara Szmigielska, Institute of Psychology, Jagiello- nian University; Lei Wang, Department of Psychology, Peking University, Beijing, China; Mami Yamaguchi, University of Otago; Michelle Yik, Hong Kong University of Science and Technology, Hong Kong.

This research was supported in part by the Intramural Research Program of the National Institutes of Health, National Institute on Aging. The Czech participation was supported by Grant 406/07/1561 from the Grant Agency of the Czech Republic. The Swiss participation was supported by Grant 100013-108290 from the Swiss National Science Foundation.

Additional help was provided by the following individuals: A. Maglio, I. Injoque Ricle, G. Blum, A. Calero, L. Cuenya, V. Pedrón, M. J. Torres Costa, and D. Vion, Instituto de Investigaciones, Facultad de Psicologia, University of Buenos Aires, Buenos Aires, Argentina; A. Butković, De- partment of Psychology, Faculty of Philosophy, University of Zagreb, Zagreb, Croatia; S. Kouřilová, Institute of Psychology, Academy of Sci- ences of the Czech Republic, Prague, Czech Republic; L. Kööts, Depart- ment of Psychology, University of Tartu; S. Chen, I. Lee, and C. Lo, Division of Social Science, Hong Kong University of Science and Tech- nology, Hong Kong; H. Alavi, Theology Department, Al-Zahra University, Tehran, Islamic Republic of Iran; Marina Brunner-Sciarra, J. Paredes, and Norma Reátegui, Facultad de Psicologı́a, Universidad Peruana Cayetano Heredia; V. E. Oryiol, I. G. Senin, and V. V. Onufrieva, Faculty of Psychology, Yaroslavl State University, Yaroslavl, Russia; C. Rigozzi, Institute of Psychology, University of Lausanne; K. Burgetova, K., Grad- uate School of Professional Psychology, University of St. Thomas; V. Laws, and A. Corydon, Department of Psychology, University of Califor- nia, Riverside; T. R. Cain, Rutgers University; Danka Purić, Department of Psychology, Faculty of Philosophy, Belgrade University, Belgrade, Serbia.

Correspondence concerning this article should be addressed to Corinna E. Löckenhoff, who is now at the Department of Human Development, G35 Martha van Rensselaer Hall, Cornell University, Ithaca, NY 14853. E-mail: [email protected]

Psychology and Aging In the public domain 2009, Vol. 24, No. 4, 941–954 DOI: 10.1037/a0016901

941

College students (N � 3,435) in 26 cultures reported their perceptions of age-related changes in physical, cognitive, and socioemotional areas of functioning and rated societal views of aging within their culture. There was widespread cross-cultural consensus regarding the expected direction of aging trajectories with (a) perceived declines in societal views of aging, physical attractiveness, the ability to perform everyday tasks, and new learning; (b) perceived increases in wisdom, knowledge, and received respect; and (c) perceived stability in family authority and life satisfaction. Cross-cultural variations in aging perceptions were associated with culture-level indicators of population aging, education levels, values, and national character stereotypes. These associations were stronger for societal views on aging and perceptions of socioemotional changes than for perceptions of physical and cognitive changes. A consideration of culture-level variables also suggested that previously reported differences in aging perceptions between Asian and Western countries may be related to differences in population structure.

Keywords: aging, stereotypes, cross-cultural, values, national character stereotypes

Perceptions of aging influence societal behaviors and expecta- tions toward older people (e.g., Pasupathi & Löckenhoff, 2002) as well as older adults’ well-being and coping with the aging process (e.g., Levy, 2003; Levy & Myers, 2004). The majority of studies in this field have focused on individual differences in perceptions of aging within (mostly Western) cultures, but there is growing evidence that views of aging may differ across cultures as well (e.g., Arnhoff, Leon, & Lorge, 1964; Giles et al., 2000). The present study extends previous research by comparing multiple aspects of aging perceptions across 26 cultures and examining their culture-level associates. To provide the background for this work, we review previous research on intercultural differences in perceptions of aging and discuss theoretical perspectives on the causes of such differences.

According to social representations theory (Moscovici, 1984, 1988) the views of aging held within a given culture are a form of shared cultural representation. They constitute systems of ideas, values, and customs related to aging that are treated by members of the society as if they were established reality. Perceptions of aging are multidimensional in nature (e.g., Hummert, 1990), en- compass both positive and negative characteristics (e.g., Heck- hausen, Dixon, & Baltes, 1989; Hummert, 1990; McTavish, 1971), and reflect a mix between accurate depictions of age-related changes and distorted views of older people (Kite, Stockdale, Whitley, & Johnson, 2005). Healthy aging is associated with predictable biological changes (DiGiovanna, 2000) that lead to systematic age differences in physical abilities and cognitive per- formance (for reviews, see Christiansen & Grzybowski, 1999; Salthouse & Davis, 2006). To the extent that aging perceptions reflect such biologically based differences in functioning, one might expect to see comparatively little variation across cultures. Age-related changes in socioemotional characteristics and social status, in contrast, appear to depend less on biology and more on motivational priorities (Fung, Rice, & Carstensen, 2005) and so- cietal roles (Eagly, Wood, & Diekman, 2000). Perceptions of age-related changes in these characteristics may therefore show a greater extent of cross-cultural variation.

Previous research has explored several potential explanations for intercultural differences in perceptions of aging. Early studies that focused on socioeconomic predictors found that higher levels of economic development and industrialization are associated with less favorable attitudes toward aging and a lower societal status of older adults (e.g., Arnhoff et al., 1964; Bengtson, Dowd, Smith, &

Inkeles, 1975; Maxwell, 1970; Palmore & Manton, 1974; Sim- mons, 1945). Modernization theory (Cowgill, 1972, 1986) ex- plained such findings by arguing that a shift toward industrialized modes of production undermines the societal status of older adults, devalues their experience-based knowledge, breaks up traditional extended families through urbanization, and shifts control over the means of production from family elders to industrial entities (Cowgill, 1972). Although intuitively appealing, modernization theory has been criticized as an oversimplification (e.g., Quad- agno, 1982). In particular, the theory ignores cross-cultural differ- ences in values and belief systems that may shape the way in which a given culture responds to advanced socioeconomic devel- opment (Inglehart & Baker, 2000).

Research on the influence of cultural values and beliefs on aging attitudes has been dominated by comparisons between Eastern or Asian versus Western cultures (see Giles et al., 2003, for a review). This body of work was inspired by the idea that Asian societies are influenced by Confucian values of filial piety and the practice of ancestor worship, which were thought to promote positive views of aging and high esteem for older adults (e.g., Davis, 1983; Ho, 1994; Sher, 1984; see Sung, 2001, for a review). Western societies, in contrast, were thought to be youth-oriented and to hold more negative views about the aging process and the elderly (e.g., Palmore, 1975). Empirical evidence for the proposed East–West differences is scarce. Although some studies have found support for the notion that aging attitudes are more positive in Asian compared to Western cultures (e.g., Levy & Langer, 1994; Tan, Zhang, & Fan, 2004), others have reported effects in the opposite direction (e.g., Giles et al., 2000; Harwood et al., 2001; Sharps, Price-Sharps, & Hanson, 1998; Zhou, 2007) or failed to find any marked cultural differences (e.g., Boduroglu, Yoon, Luo, & Park, 2006; Chappell, 2003; McCann, Cargile, Giles, & Bui, 2004; Ryan, Jin, Anas, & Luh, 2004).

In summary, there is some evidence that both socioeconomic development and cultural values and beliefs may matter for cross- cultural differences in aging attitudes. However, findings are lim- ited in several important aspects. For one, most previous studies included only small groups of countries. In fact, the vast majority of the literature consists of pairwise comparisons. Among the exceptions are one study that compared five countries (Harwood et al., 2001), four studies that each compared six countries (Arnhoff, Leon, & Lorge, 1964; Bengtson et al., 1975; Giles et al., 2003; Harwood et al., 1996), and one study that compared 11 countries

942 LÖCKENHOFF ET AL.

(Giles et al., 2000). Conclusions that can be drawn from the previous literature are also limited because studies used a wide array of measures, ranging from open-ended descriptions of older and younger adults (Boduroglu et al., 2006) to macroeconomic markers of older adults’ societal status (Palmore & Manton, 1974). This makes it difficult to integrate findings into a generalized cross-cultural pattern of aging attitudes. Moreover, cultural differ- ences in values and beliefs have frequently been inferred from the broad classification of countries into Eastern versus Western (see Giles et al., 2000, 2003, for exceptions). This approach glosses over variations among individual Asian cultures, ignores cultures that cannot be classified as Eastern or Western, confounds cross- cultural differences in socioeconomic factors and value systems, and fails to pinpoint the specific aspects of cultural values that are most relevant in shaping perceptions of aging. Decades of cross- cultural research have yielded comprehensive data regarding cul- turally shared values (e.g., Hofstede, 1980; House, Hanges, Javi- dan, Dorfman, & Gupta, 2004; Inglehart & Norris, 2003; Leung & Bond, 2004; Schwartz, 1994; Smith, Dugan, & Trompenaars, 1996) and aggregate psychological characteristics (e.g., Diener, Diener, & Diener, 1995; McCrae, Terracciano, & the Personality Profiles of Cultures Project, 2005) across a wide range of cultures. To date, this rich body of knowledge has not been adequately linked to cross-cultural differences in aging perceptions.

The present study extends previous research by examining per- ceptions of aging among college students from 26 different cul- tures in six continents. Participants rated their perceptions of age-related changes in physical, cognitive, and socioemotional characteristics and evaluated societal views about aging within their cultures. Because contact with a stereotyped group may influence perceptions of that group (Pettigrew & Tropp, 2006), we also assessed participants’ frequency of contact with older adults.

Going beyond dichotomous comparisons between Eastern and Western cultures, we related ratings of aging perceptions obtained in the present study to culture-level scores of possible associates derived from previous work (see Hofstede, 2001; Leung & Bond, 2004; McCrae et al., 2005, for examples of this analytical ap- proach). This allowed us to disentangle the relative influence of socioeconomic factors (i.e., wealth, population structure, and ed- ucation levels) compared to cultural values and beliefs.

To capture cultural differences in values, we adopted Hofstede’s (1980, 2001) dimensions and Schwartz’s (1994) seven value types, which both cover a large range of cultures and have been cross- validated with alternative classification systems (e.g., Hofstede, 2001; Leung & Bond, 2004). We also included Inglehart and Norris’s (2003) dimensions of secular-rational versus traditional values and self-expression versus survival values, which are thought to capture systematic changes in value systems in response to modernizing influences (Inglehart & Baker, 2000). Finally, to examine culture-level associations between perceptions of aging and other forms of stereotypical beliefs, we included national character stereotypes (Terracciano et al., 2005), which capture people’s stereotypical perceptions of the personality traits of a typical member of their culture.

In general, we predicted that perceptions about aspects of aging that are strongly linked to biological changes (i.e., physical aging and changes in fluid cognitive abilities) would show less variation across cultures and fewer associations with culture-level variables than perceptions of socioemotional aspects of aging (e.g., family

relations and life satisfaction) and societal views of the aging process. For culture-level associations with socioeconomic char- acteristics, we expected to replicate previous research indicating that advanced development is associated with less favorable per- ceptions of aging. To the best of our knowledge, this is the first study to examine culture-level associations between values, na- tional character stereotypes, and perceptions of aging. We there- fore adopted an exploratory approach and did not postulate spe- cific hypotheses regarding the direction of the effects. Also, given the equivocal research record on East–West differences (see above), we did not expect to find strong differences in aging perceptions between Asian and Western countries. Instead, we expected to find a more complex pattern, such that East–West differences in aging perceptions are related to both socioeconomic factors and cultural values and beliefs.

Method

Participants and Procedure

Questionnaires were administered to samples of college students from 26 cultures on six continents. Participants were informed about the general nature of the project and completed the ques- tionnaires in a quiet environment, typically in a group setting.1 All data were collected anonymously, and apart from age, gender, and citizenship status (native-born citizen or not), no personal infor- mation was recorded. Thirty participants were excluded because they missed more than two items on the aging perception measures or failed to indicate their gender, leaving a total of 3,435 partici- pants. The vast majority (94.4%) were native-born citizens of their respective countries. The demographic composition of the samples is described in Table 1. On average, participants were in their early 20s, and about two thirds were female.

Measures

Previous research has indicated that perceptions of aging are multifaceted and involve divergent trajectories for different as- pects of functioning (e.g., Heckhausen et al., 1989; Hummert, 1990; McTavish, 1971). To capture this complexity, we assembled a perceptions-of-aging measure (POA), consisting of a short list of characteristics that would be understood across a wide range of cultural contexts. The specific characteristics were adapted from existing measures assessing attitudes about aging and age-related changes in functioning (i.e., Harwood et al., 2001; Heckhausen et al., 1989; Laidlaw, Power, Schmidt, & the WHOQOL Group, 2007; Rosencranz & McNevin, 1969). It is important to note that we did not aim to create an internally consistent scale but selected a set of conceptually related but distinct items that would allow us to capture divergent patterns in specific aspects of aging percep- tions. The final scale consisted of the following eight characteris- tics: physical attractiveness, ability to do everyday tasks, ability to learn new information, general knowledge, wisdom, respect re- ceived from others, authority in the family, and life satisfaction. Participants were given the following instructions: “Below you see

1 In the same session, participants also completed questionnaires de- scribing the personality traits of a typical adolescent, adult, and older adult within their culture. These findings will be reported elsewhere.

943PERCEPTIONS OF AGING ACROSS 26 CULTURES

a list of characteristics that may or may not change as people grow old. For each characteristic, please place a check mark to indicate whether the characteristic decreases, stays the same, or increases in older people.” Participants responded on a 5-point Likert scale from �2 � decreases a lot to 2 � increases a lot, with a score of 0 indicating the absence of age-related change.2

Participants also rated their society’s views of aging. For this purpose they were asked: “In general, how positively or negatively does your culture view old age?” Responses were given on a 5-point Likert scale from �2 � very negative to 2 � very positive. They further rated their frequency of contact with older adults (“How often do you speak or do things with older people?”) on a 5-point Likert scale from 1 � almost never to 5 � almost every day.

For administration in non-English-speaking cultures, the origi- nal English version of the scale was translated by our collabora- tors, most of whom are bilingual psychologists native to the culture. Independent back-translations were created by a person other than the original translator, reviewed by two of the authors (Filip De Fruyt and Robert R. McCrae), and modified as needed to achieve equivalence across languages.

Culture-Level Associates

Indicators of socioeconomic development assessing wealth, population structure, and education levels were drawn from the 2007–2008 Human Development Report (United Nations Devel- opment Programme, 2007). As an economic indicator, we used per

capita gross domestic product (GDP; in U.S. dollars). As an indicator of demographic composition, we examined the percent- age of the population age 65 and over. We selected this measure over the more commonly used indicator of life expectancy be- cause, unlike life expectancy, it is relatively independent of infant mortality, shows a weaker association with GDP, and presents a direct assessment of the proportion of older adults in the popula- tion. As a measure of education levels, we examined the combined gross enrollment ratio for primary, secondary and tertiary educa- tion.

Hofstede (1980) defined four value dimensions based on world- wide responses to IBM surveys: power distance (i.e., acceptance of differences in status and power), uncertainty avoidance (i.e., low tolerance for ambiguity and endorsement of strict laws and rules as a means of stress reduction), individualism (i.e., emphasis on self and low integration into group or family), and masculinity (i.e.,

2 Initially, two negatively phrased and reverse-coded items assessing age-related changes in health problems and social isolation were included as well. However, preliminary analyses found unusually high standard deviations and bimodal distributions for these items across several cultures. It is possible that the lower anchor of the scale, decreases a lot, carries an implication of worsens a lot in some languages. As a result, participants who believed that health problems and social isolation increased with age may have erroneously marked the lower end of the scale. Because of this concern, these two items were excluded from further analyses.

Table 1 Sample Characteristics

Culture N Age

M (SD) % Male Contact with older adults

M (SD)

Argentina 131 23.2 (5.8) 15.3% 4.0 (1.1) Australia 98 21.6 (7.6) 20.4% 3.8 (0.9) Chile 95 19.4 (1.7) 48.4% 3.7 (1.0) Croatia 102 21.0 (1.3) 15.7% 3.6 (1.0) Czech Republic 222 22.5 (2.1) 22.5% 3.7 (0.9) United Kingdom 95 20.3 (2.5) 20.0% 2.9 (1.1) Estonia 113 22.1 (4.0) 27.4% 3.7 (1.0) France 103 22.3 (6.3) 24.3% 2.8 (1.0) Hong Kong 182 20.7 (1.5) 61.5% 2.6 (1.2) India 50 19.4 (1.5) 0.0% 4.1 (1.0) Islamic Republic of Iran 124 28.0 (5.5) 41.9% 3.2 (1.1) Italy 101 25.3 (4.1) 49.5% 3.6 (1.2) Japan 293 19.7 (1.7) 18.4% 3.0 (1.4) Mainland China 98 21.6 (2.7) 38.8% 2.8 (0.9) Malaysia 111 22.0 (2.5) 27.0% 4.0 (0.8) New Zealand 100 19.5 (3.0) 25.0% 3.2 (1.0) Peru 139 19.0 (2.8) 47.5% 3.9 (1.1) Poland 202 23.6 (4.4) 28.2% 3.6 (1.2) Portugal 103 23.4 (3.7) 23.3% 3.7 (1.0) Russian Federation 100 23.0 (6.8) 28.0% 3.7 (1.0) South Korea 120 25.9 (8.1) 45.0% 2.7 (1.2) Serbia 100 20.5 (2.3) 17.0% 3.8 (1.0) Slovakia 144 20.0 (1.7) 31.3% 3.6 (1.0) Switzerland (French-speaking) 101 20.9 (4.1) 19.8% 3.1 (1.1) Uganda 99 23.3 (3.1) 44.4% 3.0 (1.1) United States 309 21.1 (4.2) 43.0% 3.5 (1.2)

Total 3,435 21.8 (4.5) 31.3% 3.4 (1.2)

Note. Contact frequency was rated on a scale from 1 � almost never to 5 � almost every day.

944 LÖCKENHOFF ET AL.

focus on assertive and egoistic versus socially oriented roles).3 The scores used in the present study were drawn from Hofstede (2001).

Schwartz (1994) characterized cultures according to seven value types: embeddedness (emphasis on tradition and embeddedness in a collective), affective autonomy (focus on individual pursuit of pleasure), intellectual autonomy (creativity, tolerance), hierarchy (deference to power and authority), mastery (focus on success and self-assertion), egalitarian commitment (valuing freedom and so- cial justice), and harmony (focus on environmental protection and unity with nature). Recent value type data that corrected for endorsement frequency differences were obtained from the Israel Social Sciences Data Center (Schwartz, 2005).

Inglehart and Baker (2000) described two broader dimensions derived from the World Values Surveys: Traditional values (em- phasis on religion, national pride, and respect for authority) versus secular-rational values and survival values (emphasis on material well-being) versus self-expression values. The scores used in the present study were drawn from Inglehart and Norris (2003).

National character stereotypes representing people’s stereotyp- ical perceptions of the personality traits of a typical member of their culture were drawn from Terracciano et al. (2005).4

When integrating culture-level variables drawn from different studies we had to account for recent changes in national bound- aries as well as differences in how narrowly different authors defined the boundaries of cultures. When multiple matches were possible, we used the most specific available matches (e.g., French-speaking Swiss compared to French-speaking Swiss). Hof- stede’s (2001) data for Yugoslavia were paired with Croatia and Serbia, and data from the Soviet Union were matched with Russia but not Estonia. Finally, data from Uganda were matched with Hofstede’s (2001) East African region.

With the exception of the demographic variables, the culture- level associates described above were measured at the interval level. Table 2 shows intercorrelations among culture-level markers of demographics, values, and national character stereotypes. Spe- cific scores on the culture-level associates for each individual culture are available from the authors.

Results

Preliminary Analyses

Intercorrelations among the eight POA items in the pooled sample were all positive but generally low, with a mean correlation of r � .21. This suggests that, consistent with our intentions, the different items assess conceptually related but separate aspects of aging perceptions. Given these considerations, we report item- level results for the POA instead of creating summary scores.

As seen in Table 1 (fourth column), the relative proportions of male and female respondents varied across cultures, and in one culture (India) only female raters were available. To examine the extent to which cross-cultural patterns of aging perceptions differ by gender, we computed separate culture-level means on each of the individual POA items for each gender. For each POA item, we then examined the correlations between male and female profiles across the 26 cultures. Correlations were moderate to high (ranging from r � .46 to r � .89, mean r across items � .77). However, to account for any gender differences, further culture-level analyses used means that controlled for unequal gender distributions. In the case of India, the mean response for women was used.

Whereas culture-level means on the POA items were normally distributed (Kolmogorov-Smirnov Z � .8), individual-level POA scores showed considerable deviations from normality. Attractive- ness, everyday tasks, new learning, knowledge, wisdom, and re- spect showed substantial skewness and underwent logarithmic transformation. Authority and life satisfaction showed moderate skewness and underwent square-root transformation. Transformed scores were used for all analyses requiring normal distribution of the scores (i.e., analyses of variance and mixed models). For illustrative purposes, descriptive reports (e.g., Table 3) use un- transformed scores.

Consensus Across Cultures

Table 3 shows gender-weighted means for perceived societal views of aging and POA scores across the different cultures. Respondents from most cultures thought that aging was viewed negatively within their societies (Table 3, last column). The ex- ceptions were Mainland China, India, Malaysia, Russia, and New Zealand, where participants reported neutral or slightly positive views. The most negative perceived societal views were reported in Serbia, Argentina, the Czech Republic, and the United King- dom.

Culture-level scores on individual POA items suggested that participants held divergent expectations for age trajectories in specific aspects of life (see Figure 1). As indicated by the consis- tently negative scores for attractiveness, everyday tasks, and new learning (cf. Table 3), there was cross-cultural consensus that these characteristics decrease with age. In turn, consistently positive scores for general knowledge, wisdom, and respect indicate that across cultures, these characteristics were believed to increase with age. The picture was less clear-cut for the remaining items, but in general, family authority was seen as increasing and life satisfac- tion was viewed as stable.

To examine if some aspects of aging perceptions show greater cross-cultural variability than others, we computed for each of the POA items the squared deviations of the mean scores in each culture from the cross-cultural mean. Using these deviation scores as the dependent variable, we then performed an analysis of variance with the eight POA items as the repeated-measures variable. Because we were interested in relative variability across cultures, we performed this analysis at the culture level (i.e., each of the 26 cultures was treated as a single case). Consistent with our expectations, some aspects of aging per- ceptions showed significantly less variability across cultures

3 A fifth dimension of long-term orientation identified by Hofstede and Bond (1988) was not included in the analyses, because the overlapping number of cultures was small (n � 16).

4 Note that although national character stereotypes are described in terms of Five Factor personality traits, they are not empirically related to aggre- gate personality profiles (i.e., mean personality ratings of actual members of a given culture; Terracciano et al., 2005). For comparison purposes, we conducted preliminary analyses that examined associations between aging perceptions and aggregate personality. Only sporadic associations were found, and the effects of aggregate personality profiles derived from self ratings (McCrae, 2002) and observer ratings (McCrae et al., 2005) did not converge. Aggregate personality profiles were therefore not considered in further analyses.

945PERCEPTIONS OF AGING ACROSS 26 CULTURES

T ab

le 2

C o rr

el a ti

o n s

A m

o n g

C u lt

u re

-L ev

el A

ss o ci

a te

s o f

A g in

g P

er ce

p ti

o n s

C ri

te ri

on 1

2 3

4 5

6 7

8 9

10 11

12 13

14 15

16 17

18 19

20 21

S oc

io de

m og

ra ph

ic ch

ar ac

te ri

st ic

s (n

� 26

)

1. G

D P

— .6

1� �

.6 5�

� �

.4 9�

� .1

1 .6

5� �

.3 3

� .6

1� �

.5 2�

� .5

7� �

� .3

8 �

.3 1

.4 7�

.1 4

.3 2

.6 3�

� .0

3 �

.1 3

� .1

8 �

.2 3

.3 7

2. %

� 65

.6 1�

� —

.5 4�

� �

.2 7

.3 8

.4 6�

.2 2

� .7

1� �

.6 3�

� .7

1� �

� .6

8� �

� .2

9 .3

0 .5

4� �

.5 1�

� .2

6 .1

8 �

.1 8

� .2

9 �

.1 7

.1 0

3. E

du c.

.6 5�

� .5

4� �

— �

.6 0�

� .3

0 .5

8� �

.0 0

� .4

6� .4

0� .3

2 �

.5 4�

� �

.4 5�

.6 0�

� .3

6 .0

9 .5

3� �

� .0

9 .0

9 .0

4 �

.0 6

.0 6

H of

st ed

e (2

00 1)

va lu

e di

m en

si on

s (n

� 26

)

4. P

ow er

� .4

9� �

.2 7

� .6

0� �

— .0

6 �

.5 9�

� .0

6 .3

7 �

.3 1

� .1

4 .1

9 .0

2 �

.4 3�

� .1

1 .1

6 �

.7 2�

� .1

6 �

.0 5

.0 6

.3 1

� .1

7 5.

U nc

er t.

� .1

1 .3

8 .3

0 .0

6 —

� .2

2 �

.3 0

� .1

5 .0

2 .2

4 �

.4 2�

� .3

0 .2

0 .4

4� �

.0 9

� .1

7 .4

0� .0

6 �

.0 6

� .0

2 �

.5 0�

6. In

di v.

.6 5�

� .4

6� .5

8� �

� .5

9� �

� .2

2 —

.3 7

� .4

1� .5

4� �

.3 4

� .4

4� �

.3 8

.4 3�

.3 4

.0 4

.6 6�

� �

.2 3

.1 0

.1 7

� .1

4 .3

1 7.

M as

c. .3

3 .2

2 .0

0 .0

6 �

.3 0

.3 7

— �

.2 6

.2 2

.3 2

� .0

2 .0

5 �

.1 2

.1 1

.3 1

.2 5

.0 9

� .1

4 �

.1 7

.1 3

.2 8

S ch

w ar

tz (1

99 4)

va lu

e ty

pe s

(n �

26 )

8. E

m b

� .6

1� �

� .7

1� �

� .4

6� .3

7 �

.1 5

� .4

1� �

.2 6

— �

.8 2�

� �

.8 5�

� .6

1� �

.0 0

� .4

6 �

.3 7

� .2

7 �

.5 0�

.1 3

.0 3

.3 0

.0 9

� .1

0 9.

A ff

. au

t. .5

2� �

.6 3�

� .4

0� �

.3 1

.0 2

.5 4�

� .2

2 �

.8 2�

� —

.6 5�

� �

.4 9�

.1 0

.1 9

.2 1

.2 7

.4 4�

� .2

9 .1

0 �

.0 7

.0 1

.2 0

10 .

In t.

au t.

.5 7�

� .7

1� �

.3 2

� .1

4 .2

4 .3

4 .3

2 �

.8 5�

� .6

5� �

— �

.5 9�

� �

.2 3

.4 3�

.3 9�

.3 4

.3 7

.0 9

� .1

7 �

.3 7

� .0

8 .1

0 11

. H

ie ra

rc hy

� .3

8 �

.6 8�

� �

.5 4�

� .1

9 �

.4 2�

� .4

4� �

.0 2

.6 1�

� �

.4 9�

� .5

9� �

— .4

7� �

.5 7�

� �

.6 4�

� .0

0 �

.3 1

� .1

6 �

.1 6

.0 5

.2 0

.2 4

12 .

M as

te ry

� .3

1 �

.2 9

� .4

5� .0

2 �

.3 0

� .3

8 .0

5 .0

0 .1

0 �

.2 3

.4 7�

— �

.5 5�

� �

.6 1�

� �

.0 6

� .1

1 �

.3 1

.2 8

.0 4

.1 6

� .0

2 13

. E

ga l.

.4 7�

.3 0

.6 0�

� �

.4 3�

.2 0

.4 3�

� .1

2 �

.4 6�

.1 9

.4 3�

� .5

7� �

� .5

5� �

— .5

2� �

� .3

5 .5

6� �

� .0

3 .0

9 .0

6 �

.0 9

� .1

2 14

. H

ar m

on y

.1 4

.5 4�

� .3

6 �

.1 1

.4 4�

.3 4

.1 1

� .3

7 .2

1 .3

9� �

.6 4�

� �

.6 1�

� .5

2� �

— .0

7 .1

5 .0

9 �

.1 8

� .0

5 .1

1 �

.0 7

In gl

eh ar

t &

N or

ri s

(2 00

3) va

lu es

(n �

24 )

15 .

S ec

. .3

2 .5

1� .0

9 .1

6 �

.0 9

.0 4

.3 1

� .2

7 .2

7 .3

4 .0

0 �

.0 6

� .3

5 .0

7 —

� .2

5 �

.0 4

� .4

5� �

.2 8

.0 9

.5 4�

�

16 .

S el

f .6

3� �

.2 6

.5 3�

� �

.7 2�

� �

.1 7

.6 6�

� .2

5 �

.5 0�

.4 4�

.3 7

� .3

1 �

.1 1

.5 6�

� .1

5 �

.2 5

— �

.1 3

.1 5

� .0

3 �

.0 6

� .0

1

T er

ra cc

ia no

et al

. (2

00 5)

na ti

on al

ch ar

ac te

r st

er eo

ty pe

s (n

� 25

)

17 .

N N

C S

.0 3

.1 8

� .0

9 .1

6 .4

0� �

.2 3

.0 9

.1 3

� .2

9 .0

9 �

.1 6

� .3

1 �

.0 3

.0 9

� .0

4 �

.1 3

— �

.4 3�

� .4

9� �

.5 5�

� �

.3 6

18 .

E N

C S

� .1

3 �

.1 8

.0 9

� .0

5 .0

6 .1

0 �

.1 4

.0 3

.1 0

� .1

7 �

.1 6

.2 8

.0 9

� .1

8 �

.4 5�

.1 5

� .4

3� —

.7 7�

� .2

6 �

.3 2

19 .

O N

C S

� .1

8 �

.2 9

.0 4

.0 6

� .0

6 .1

7 �

.1 7

.3 0

� .0

7 �

.3 7

.0 5

.0 4

.0 6

� .0

5 �

.2 8

� .0

3 �

.4 9�

.7 7�

� —

.3 8

� .0

6 20

. A

N C

S �

.2 3

� .1

7 �

.0 6

.3 1

� .0

2 �

.1 4

.1 3

.0 9

.0 1

� .0

8 .2

0 .1

6 �

.0 9

.1 1

.0 9

� .0

6 �

.5 5�

� .2

6 .3

8 —

.1 6

21 .

C N

C S

.3 7

.1 0

.0 5

� .1

7 �

.5 0�

.3 1

.2 8

� .1

0 .2

0 .1

0 .2

4 �

.0 2

� .1

2 �

.0 7

.5 4�

� �

.0 1

� .3

6 �

.3 2

� .0

6 .1

6 —

N o te

. G

D P

� g

ro ss

d o

m es

ti c

p ro

d u

ct ;

% �

6 5

� p

er ce

n ta

g e

o f

th e

p o

p u

la ti

o n

ag e

6 5

an d

ab o

v e;

E d

u c.

� ed

u ca

ti o

n al

en ro

ll m

en t;

P o

w er

� p

o w

er d

is ta

n ce

; U

n ce

rt . �

u n

ce rt

ai n

ty av

o id

an ce

; In

di v.

� in

di vi

du al

is m

; M

as c.

� m

as cu

li ni

ty ;

E m

b �

em be

dd ed

ne ss

; A

ff .a

ut .�

af fe

ct iv

e au

to no

m y;

In t.

au t.

� in

te ll

ec tu

al au

to no

m y;

E ga

l. �

eg al

it ar

ia n

co m

m it

tm en

t; S

ec .�

se cu

la r-

ra ti

on al

; S

el f

� se

lf -e

xp re

ss io

n; N

N C

S ,

E N

C S ,

O N

C S ,

A N

C S ,

C N

C S

� na

ti on

al ch

ar ac

te r

st er

eo ty

pe s

of ne

ur ot

ic is

m ,

ex tr

av er

si on

, op

en ne

ss ,

ag re

ea bl

en es

s, an

d co

ns ci

en ti

ou sn

es s.

� p

� .0

5. � �

p �

.0 1.

946 LÖCKENHOFF ET AL.

than others, after Greenhouse-Geisser correction, F(4.35, 108.72) � 3.16, p � .01, �2 � .11. Post hoc tests revealed that variability across cultures was smaller for attractiveness than for all of the other characteristics ( ps � .01), smaller for performance in everyday tasks than for family authority and life satisfaction ( ps � .05), and smaller for new learning than for life satisfaction ( p � .05).

Culture-Level Associates of Aging Perceptions

The associations between aging perceptions and culture-level indicators of sociodemographic variables, values, and national character stereotypes were examined using a mixed-model ap- proach nesting individuals within cultures. Initial unconditional means models entered culture as a random factor and aging per- ception scores as dependent variables. In these models, each indi- vidual score on the aging perception measures (Yij) was expressed as the sum of the overall mean (�00), the effect of culture (u0j), and the effect of variations at the participant level (rij). Estimates of the variability among culture means (�00), and the variability among students in cultures (�2) indicated that there were significant effects of culture on each of the dependent variables (see top rows of Table 4). Also, consistent with the analyses reported above, the proportion of total variance in aging perceptions that was ex- plained by culture [�00/(�00 �

2)] varied for different aspects of aging perceptions and ranged from around 3% for attractiveness to more than 15% for wisdom.

In the next step, we added culture-level associates as fixed factors (�01) at the culture level. As seen in Table 2, a considerable portion of the intercorrelations between culture-level variables reached significance, and preliminary analyses revealed moderate to severe collinearity when multiple culture-level variables were included in the same analysis. Therefore, culture-level predictors were examined one at a time. The bottom part of Table 4 presents the estimated coefficients for the effects of each of the culture- level variables on aging perceptions. We also present the propor- tion of explainable variation in culture-level means explained by a given culture-level predictor.5

When we compared the relative strength of associations with culture-level predictors across different aspects of aging percep- tions, the predicted pattern emerged. Perceptions of physical as- pects of aging (i.e., attractiveness and ability to perform everyday tasks) did not show any significant relations to culture-level vari- ables. For perceptions of cognitive aspects of aging (i.e., new learning, general knowledge, and wisdom), only 16% of the pos-

5 To examine if the findings reported in Table 4 would hold up using alternative analytical approaches, we computed Pearson correlations be- tween gender-weighted culture-level means of aging perceptions and each of the culture-level variables (see Hofstede, 2001; Leung & Bond, 2004; McCrae et al., 2005, for examples of such analyses in the context of cross-cultural research). Ninety percent of the significant findings in Table 4 were replicated with this approach.

Table 3 Mean Perceptions of Aging by Culture

Culture Attractiveness Everyday

tasks New

learning General

knowledge Wisdom Received respect

Family authority

Life satisfaction

Societal views

Argentina �1.4 (0.7) �0.6 (1.0) �1.1 (0.8) 0.7 (1.2) 1.0 (0.8) 0.3 (1.3) �0.2 (1.2) �0.2 (1.0) �0.9 (0.9) Australia �1.3 (0.7) �1.0 (0.9) �0.6 (1.0) 1.4 (0.8) 1.4 (0.6) 0.9 (1.0) 0.6 (1.3) 0.5 (0.9) �0.3 (0.8) Chile �1.5 (0.7) �0.9 (0.8) �1.0 (0.9) 1.0 (1.1) 1.3 (0.9) 0.7 (1.2) 0.2 (1.2) 0.2 (1.1) �0.8 (0.8) Croatia �1.3 (0.8) �1.0 (0.7) �1.3 (0.7) 0.6 (1.0) 1.4 (0.7) 0.8 (1.1) 0.4 (1.2) �0.6 (0.9) �0.7 (0.7) Czech Republic �1.4 (0.6) �0.8 (0.8) �1.3 (0.8) 0.1 (1.0) 0.6 (1.0) 0.4 (1.1) 0.1 (1.0) �0.4 (0.9) �0.8 (0.7) United Kingdom �1.5 (0.7) �1.0 (0.8) �0.9 (0.6) 1.2 (0.9) 1.3 (0.8) 0.5 (1.2) 0.3 (1.2) 0.1 (0.9) �0.8 (0.9) Estonia �1.2 (0.7) �0.4 (0.9) �0.9 (0.9) 1.2 (0.9) 1.6 (0.6) 1.2 (0.7) 0.4 (1.0) �0.1 (0.9) �0.2 (0.8) France �1.4 (0.8) �1.1 (0.6) �1.0 (0.8) 0.7 (1.1) 1.0 (0.8) 0.5 (1.3) �0.1 (1.1) �0.3 (0.9) �0.8 (0.8) Hong Kong �1.2 (0.9) �0.6 (1.0) �1.2 (0.7) 0.5 (1.0) 0.4 (1.0) 0.8 (1.0) 0.6 (1.0) 0.1 (1.0) �0.3 (0.8) India �1.3 (0.8) �0.6 (1.1) �0.1 (1.4) 0.6 (1.2) 0.5 (1.2) 1.1 (1.2) 0.8 (1.2) 0.1 (1.2) 0.3 (1.3) Islamic Republic of Iran �1.6 (0.7) �1.0 (0.8) �1.4 (0.7) 0.3 (1.2) 0.5 (1.2) 1.3 (1.0) 0.7 (1.2) �0.6 (1.1) �0.3 (1.1) Italy �1.5 (0.9) �1.1 (0.8) �0.8 (1.1) 1.1 (1.0) 1.6 (0.8) 1.2 (1.1) 0.4 (1.3) 0.0 (1.1) �0.5 (1.0) Japan �1.4 (0.8) �0.7 (0.8) �0.9 (1.0) 1.1 (0.9) 1.1 (0.8) 1.0 (0.8) 0.2 (1.0) 0.0 (1.0) �0.5 (0.9) Mainland China �1.4 (0.9) �0.4 (1.2) �1.1 (0.8) 1.1 (0.8) 1.0 (1.0) 1.4 (0.8) 1.3 (0.9) 0.5 (1.1) 0.3 (1.0) Malaysia �1.3 (0.9) 0.0 (1.4) �0.5 (1.1) 0.9 (1.2) 0.4 (1.2) 1.3 (1.0) 1.0 (1.1) 0.5 (1.2) 0.2 (1.0) New Zealand �1.4 (0.8) �0.6 (1.1) �0.7 (0.8) 1.3 (0.9) 1.4 (0.8) 1.1 (1.0) 0.7 (1.2) 0.5 (1.1) 0.1 (1.0) Peru �1.3 (0.8) �0.7 (1.2) �0.6 (1.1) 0.8 (1.3) 1.2 (1.1) 0.9 (1.1) 0.3 (1.3) 0.4 (1.2) �0.5 (0.9) Poland �1.5 (0.6) �0.8 (1.0) �1.2 (0.8) 0.4 (1.1) 0.9 (1.0) 0.7 (1.1) 0.9 (1.0) �0.5 (0.9) �0.5 (0.9) Portugal �1.4 (0.8) �0.7 (0.9) �0.9 (0.6) 1.0 (0.9) 1.2 (0.9) 1.0 (0.9) 0.3 (1.1) �0.2 (0.9) �0.7 (0.8) Russian Federation �1.1 (1.0) �0.4 (1.1) �0.9 (1.1) 1.2 (1.1) 1.6 (0.7) 1.0 (0.9) 0.7 (1.1) �0.3 (1.0) 0.2 (0.9) South Korea �1.5 (0.7) �1.2 (0.7) �1.2 (0.8) 0.5 (1.2) 1.0 (1.0) 0.7 (0.9) �0.2 (1.1) �0.2 (1.0) �0.2 (0.9) Serbia �1.3 (0.8) �0.7 (1.1) �1.2 (0.8) 0.6 (1.2) 0.9 (0.9) 0.7 (1.2) 0.1 (1.3) �0.7 (1.0) �1.1 (0.9) Slovakia �1.3 (0.7) �0.9 (0.6) �1.3 (0.6) 0.1 (1.0) 0.4 (0.8) 0.5 (1.0) 0.2 (1.0) �0.8 (0.8) �0.5 (0.7) Switzerland �1.5 (0.7) �0.7 (1.1) �1.1 (0.7) 0.6 (1.0) 0.9 (0.8) 0.5 (1.2) �0.3 (1.1) �0.2 (0.9) �0.4 (0.9) Uganda �1.6 (0.8) �1.4 (1.1) �0.8 (1.2) 0.6 (1.3) 1.1 (1.2) 1.7 (0.7) 0.9 (1.3) 0.5 (1.2) �0.1 (1.1) United States �1.3 (0.9) �0.6 (1.1) �0.6 (0.9) 1.4 (0.8) 1.5 (0.7) 1.1 (1.1) 0.8 (1.1) 0.4 (1.1) �0.6 (0.9)

Note. SDs are shown in parentheses. Perceptions of aging were rated on a scale from �2 � decreases a lot to 2 � increases a lot. Perceived societal views on aging were rated on a scale from �2 � very negative to 2 � very positive. Mean scores are weighted to account for unequal gender distribution across cultures.

947PERCEPTIONS OF AGING ACROSS 26 CULTURES

sible associations with culture-level predictors reached signifi- cance. For perceptions of socioemotional aspects of aging (i.e., received respect, family authority, and life satisfaction), 30% of the associations were significant; and for societal views of aging, 38% of possible associations reached significance. Thus, associa- tions between perceptions of aging and culture-level markers ap- pear to be weak for physical and cognitive aspects of aging, more pronounced for socioemotional aspects of aging, and strongest for participants’ perceptions of their societies’ views on aging. We now discuss the effects for each type of culture-level associate in more detail.

GDP showed no significant associations with any of the aging perception variables, but the proportion of older adults (age � 65 years) in the population was associated with less favorable per- ceptions of societal views on aging. As seen in Figure 2, which plots the proportion of older adults against societal views of aging, European cultures with high rates of population aging are clustered in the lower right quadrant of the graph, indicating that these cultures are thought to view aging as more negative. In contrast, cultures with younger populations, such as Malaysia, India, or Mainland China, are mostly found in the upper left quadrant, indicating more favorable views of aging. It is particularly inter- esting that Japan, the Asian culture with the highest percentage of older adults, is found in the same cluster as European countries. With regard to specific POA items, participants from cultures with older populations and higher education levels were less likely to report an age-related increase in respect and family authority but more likely to report an increase in wisdom. Participants from

cultures with higher education levels also held more positive views of age trajectories in general knowledge.

Regarding Hofstede’s (2001) value dimensions, participants from cultures with greater uncertainty avoidance reported more negative societal views of aging. Uncertainty avoidance was also associated with less favorable expectations about age-related changes in family authority and life satisfaction. Further, partici- pants from cultures higher in power distance reported less favor- able views of age-related changes in knowledge and wisdom.

Regarding Schwartz’s (1994) value types, higher levels of hi- erarchy and embeddedness and lower levels of intellectual auton- omy were associated with more favorable societal views on aging. Only two of the specific POA items showed significant association with Schwartz’s value types: The perception that received respect and family authority increase with age was associated with higher cultural levels of embeddedness and hierarchy but lower levels of intellectual autonomy. Perceived increases in family authority were also linked to lower levels of egalitarian commitment and harmony, and perceived increases in respect were linked to higher levels of mastery.

Inglehart and Norris’s (2003) values showed no significant associations with perceived societal views of aging, but partici- pants from cultures high in self-expression reported more favor- able aging trajectories for new learning and life satisfaction, whereas participants from cultures high in secular–rational values reported more negative views of new learning in advanced age.

National character stereotypes were significantly associated with several aspects of aging perceptions. In cultures stereotyped

-2 -1.5 -1 -0.5 0 0.5 1 1.5 2

1

Attractiveness

Everyday tasks

Knowledge

New learning

Family authority

Received respect

Wisdom

Life satisfaction

decreases a lot

increases a lot

Figure 1. Means and standard deviations of aging perceptions at the culture level (controlling for unequal gender distribution).

948 LÖCKENHOFF ET AL.

as low in neuroticism there were more favorable perceptions of age trajectories in new learning, general knowledge, received respect, family authority, and life satisfaction. Perceptions about age tra- jectories in new learning were also more favorable in cultures that viewed their typical member as high in openness. Finally, partic- ipants’ ratings of societal views of aging were more positive in cultures who view their typical member as open, agreeable, and low in neuroticism.

Additional analyses added gender (�02) and frequency of con- tact with older adults (�03) as fixed factors at the individual level. Age of the rater was not included because the focus on student samples resulted in a restriction of range for that variable. Com- pared with male respondents, women had more favorable opinions of age-related changes in everyday tasks (�02 � .011, p � .0001). Further, participants who had more frequent contact with older adults had more favorable perceptions of age-related changes in attractiveness (�03 � .012, p � .0001), everyday tasks (�03 � .011, p � .0001), new learning (�03 � .007, p � .05), authority in the family, (�03 � .017, p � .01), life satisfaction (�03 � .009, p � .05), and societal views of aging (�03 � .049, p � .0001). How- ever, even after gender and contact frequency were added to the models, the pattern of findings reported in Table 4 remained largely unchanged. The only effects that were no longer significant were the associations of Schwartz’s (1994) hierarchy scale with societal views of aging and national stereotypes of openness with POA learning.

Because previous research has found that cultural values are associated with sociodemographic variables (e.g., Hofstede, 2001; Leung & Bond, 2004), we examined in supplemental analyses if the associations between aging perceptions, cultural values, and national stereotypes remained significant when demographic char- acteristics were added to the models as culture-level variables. Because of concerns regarding multicollinearity (see Table 2), demographic predictors were added one at a time. When we added GDP, only half of the associations of aging perceptions with values but all of the associations with national character stereotypes remained significant. When we added educational enrollment, only 40% of the associations of aging perceptions with values but 89% of associations with national character stereotypes remained sig- nificant. Finally, after we added population age, 25% of associa- tions with values but 44% of associations with national stereotypes remained significant. This suggests that cross-cultural differences in sociodemographic factors, and particularly in population aging, can largely account for the association between aging perceptions and values. However, they cannot fully explain the association between aging perceptions and national character stereotypes.

Comparing Aging Perceptions in Asian Versus Western Cultures

To integrate findings from the present study with the previous literature, we also examined gross-level Eastern versus Western contrasts. To group the cultures in our sample according to broad Asian or Eastern versus Western categories, we drew on the United Nations geographical regions (United Nations Statistics Division, 2008). According to these guidelines, we classified the following countries as Eastern or Asian: Hong Kong, India, Islamic Republic of Iran, Japan, Mainland China, Malaysia, and South Korea. The Western group was composed of European cultures (Croatia,

Czech Republic, Estonia, France, United Kingdom, Poland, Por- tugal, the Russian Federation, Serbia, Slovakia, and Switzerland) and the United States. Following previous studies (e.g., Giles et al., 2000), Australia and New Zealand were also classified as Western. Because they could not be clearly classified in either category, the South American cultures (Argentina, Chile, and Peru) and the sole African culture in our sample (Uganda) were excluded from this comparison.

Again, we used a mixed model approach nesting individuals within cultures where culture was entered as a random factor and the East–West classification was entered as a fixed factor at the culture level. Compared with Western countries, participants in Eastern countries reported more positive societal views of aging (�EastWest � .42, p � .01) but less favorable perceptions of age-related changes in wisdom (�EastWest � �.10, p � .05).

6

In the next step, we examined to what extent East–West differ- ences in aging perceptions are explained by differences in socio- demographic structure, culturally shared values, and national ste- reotypes. For this purpose, we focused on the culture-level variables that were found to have significant associations with POA wisdom and societal perceptions of aging (see Table 4). East–West differences remained significant even when values, national stereotypes, and education levels were added as culture- level predictors. In contrast, when we added the proportion of older adults in the population, East–West differences were ren- dered insignificant (all p � .3).

Discussion

The present study extends the literature by comparing percep- tions of aging across 26 cultures from six continents. Perhaps the most striking finding was the widespread cross-cultural consensus regarding the direction of aging trajectories in different character- istics (e.g., consistent increases in wisdom vs. consistent decreases in the ability to perform everyday tasks). This suggests that basic patterns of aging perceptions are shared across cultures. Many of the perceived changes are also consistent with age trajectories on standardized measures that suggest age-related declines in fluid cognitive abilities (McArdle, Ferrer-Caja, Hamagami, & Wood- cock, 2002; Salthouse & Davis, 2006) and physical functioning (DiGiovanna, 2000) but relative stability in crystallized intelli- gence (McArdle et al., 2002), socioemotional skills, and well- being (Charles & Carstensen, 2007). This implies that, in contrast to other culturally held stereotypes such as perceptions of national character (e.g., Terracciano et al., 2005), culture-level perceptions of the aging process appear to contain more than just a “kernel of truth” (Brigham, 1971, p. 1). Nevertheless, there was considerable variance in aging perceptions across cultures, warranting a closer examination of the possible causes of such differences.

In general, results supported the prediction that the effects of culture would be strongest for those aspects of aging perceptions (i.e., socioemotional functioning and perceived societal views on aging) that are least influenced by age-related biological changes. An examination of the culture-level associates of aging percep- tions also provided some support for the prediction that advanced

6 Patterns of effects were comparable when using a narrower definition of Eastern cultures that excluded India and the Islamic Republic of Iran or when South American countries were grouped with the Western countries.

949PERCEPTIONS OF AGING ACROSS 26 CULTURES

levels of socioeconomic development would be associated with less favorable perceptions of aging. Our findings suggest that economic growth is less relevant in this regard than educational enrollment and population aging. It has been argued that growing numbers of older adults in Western nations may trigger intergen- erational conflicts over limited resources (e.g., Silverstein, Parrott, Angelelli, & Cook, 2000). In support of this idea, young adults from cultures with a higher proportion of older adults reported less favorable societal views on aging and smaller age-related increases in family authority. However, our findings do not indicate a uniformly negative effect of demographic change on aging per- ceptions: Respondents from cultures with older and more educated populations endorsed larger increases in wisdom with age.

In part, these results are consistent with revised forms of mod- ernization theory that predict an association between advanced

societal development and cultural changes toward rational, toler- ant, and participatory values (Inglehart & Baker, 2000). Conceiv- ably, shifts toward progressive values may erode traditional no- tions that older adults deserve respect and authority just because of their age. Consistent with this idea, perceptions that family author- ity increases with age were positively associated with embedded- ness and hierarchy, but negatively associated with intellectual autonomy and egalitarian commitment. A comparable though weaker pattern was found for perceived respect. At the same time, increasing emphasis on self-expression and individual well-being may lead to greater appreciation of older adults’ experience-based knowledge and greater concern for their well-being. In support of this view, participants from cultures with greater emphasis on self-expression (Inglehart & Norris, 2003) expected more positive age trajectories in knowledge and life satisfaction, whereas cul-

Table 4 Summary of Mixed Models Examining the Effects of Culture-Level Associates on Different Aspects of Aging Perceptions

Models and predictors Attractiveness Everyday tasks New learning General knowledge

Unconditional means models �00 �M1/(�M1 � 2) �00 �M1/(�M1 �

2) �00 �M1/(�M1 � 2) �00 �M1/(�M1 �

2) Culture 0.172� .027 0.302� .077 0.268� .087 1.418� .112

Models adding culture-level predictors �01 (�M1 � �M2)/�M1 �01 (�M1 � �M2)/�M1 �01 (�M1 � �M2)/�M1 �01 (�M1 � �M2)/�M1

Sociodemographic characteristics (n � 26)

GDP .000 .020 .000 .000 .000 .000 .000 .009 % � 65 .001 .045 .000 .000 �.001 .000 .003 .000 Educ. �.010 .029 .000 .000 .000 .000 .003� .149

Hofstede (2001) value dimensions (n � 26)

Power .000 .000 .000 .000 .000 .022 �.002� .181 Uncert. .000 .000 .000 .000 .000 .066 .000 .000 Indiv. .000 .000 .000 .000 .001 .056 .001 .068 Masc. .000 .000 .000 .000 .000 .000 .000 .000

Schwartz (1994) value types (n � 26)

Emb. �.010 .000 .009 .000 .004 .000 �.056 .000 Aff. aut. .023 .057 .013 .000 .012 .000 .046 .036 Int. aut. �.009 .000 �.016 .000 �.023 .000 �.008 .000 Hier. �.011 .000 .011 .000 .001 .000 �.019 .000 Mast. .014 .000 .024 .000 .030 .000 .049 .000 Egal. �.014 .000 �.034 .000 .060 .071 .067 .037 Harm. �.005 .000 �.017 .000 �.011 .000 .001 .000

Inglehart & Norris (2003) values (n � 24)

Sec. .009 .103 .019 .181 �.027� .303 �.015 .000 Self �.010 .042 �.010 .105 .026� .335 .028 .068

Terracciano et al. (2005) national character stereotypes (n � 25)

NNCS �.001 .115 �.001 .000 �.005 � .254 �.007� .227

ENCS .001 .113 �.001 .000 .003 .146 .003 .044 ONCS .000 .096 .000 .000 .004

� .211 .003 .032 ANCS .001 .094 .001 .000 .004 .155 .002 .000 CNCS .001 .113 .002 .000 .002 .077 .000 .000

Note. �00 � estimated intercept for unconditional means models; �01 � estimated fixed effects of culture-level associates; �M1 � �00 for unconditional means models; �M2 � �00 for models including culture-level predictors; �M1/(�M1 �

2) � proportion of total variance in aging perceptions explained by culture; (�M1 � �M2)/�M1 � proportion of explainable variation in culture-level means explained by a given culture-level predictor; Educ. � educational enrollment; Power � power distance; Uncert. � uncertainty avoidance; Indiv. � individualism; Masc. � masculinity; Emb � embeddedness; Aff. aut. � affective autonomy; Int. aut. � intellectual autonomy; Hier. � hierarchy; Mast. � mastery; Egal. � egalitarian committment; Harm. � harmony; Sec. � secular-rational; Self � self-expression; NNCS, ENCS, ONCS, ANCS, CNCS � national character stereotypes of neuroticism, extraversion, openness, agreeableness, and conscientiousness. � p � .05.

950 LÖCKENHOFF ET AL.

tures with a traditional emphasis on status differences (power distance; Hofstede, 2001) reported less favorable views of age trajectories in general knowledge and wisdom.

In this context, we also explored the role of intergenerational contact. At the individual level, more frequent contact with older adults was associated with more positive aging perceptions. How- ever, at the culture level, a higher proportion of older adults did not translate into more frequent intergenerational contacts (r � �.11, ns). This finding is consistent with the notion that modernizing influences erode multigenerational family structures.

In contrast to values, which appear to be strongly intertwined with sociodemographic variables, national character stereotypes seem to have an association with aging perceptions that is rela- tively independent of national wealth and population structure. In fact, the single most consistent relation between culture-level

associates and aging perceptions is found for national character stereotypes regarding levels of neuroticism (NNCS). As seen in Table 4, NNCS accounts for close to 40% of the explainable variance in societal views of aging across cultures. It appears that the same mechanisms that lead people to stereotype the typical member of their culture as high in negative emotionality would also lead them to perceive their culture’s views of aging and various aspects of the aging process more negatively. As men- tioned above, national character stereotypes are not associated with aggregate levels of personality, and research on their possible sources is still in its infancy (McCrae, Terracciano, Realo, & Allik, 2007; Terracciano & McCrae, 2007). However, the association between national character stereotypes and aging stereotypes ob- served in the present study raises the possibility that other forms of stereotyping (e.g., regarding gender or minority groups) may fol-

Table 4 (continued )

Wisdom Received respect Family authority Life satisfaction Societal views

�00 �M1/(�M1 � 2) �00 �M1/(�M1 �

2) �00 �M1/(�M1 � 2) �00 �M1/(�M1 �

2) �00 �M1/(�M1 � 2)

1.464� .155 1.436� .101 1.677� .120 1.683� .122 2.591� .136

�01 (�M1 � �M2)/�M1 �01 (�M1 � �M2)/�M1 �01 (�M1 � �M2)/�M1 �01 (�M1 � �M2)/�M1 �01 (�M1 � �M2)/�M1

.000 .000 .000 .102 .000 .056 .000 .000 .000 .077

.007� .123 �.006� .145 �.014� .229 �.008 .071 �.038� .242

.004� .249 �.003� .156 �.005� .140 .000 .000 �.010 .049

�.002� .270 .000 .000 .000 .000 �.002 .091 .004 .000 .000 .043 �.001 .106 �.003� .341 �.002� .213 �.080� .233 .001 .056 .000 .000 .000 .000 .000 .000 �.003 .000

�.001 .074 .000 .000 .000 .000 .000 .000 .000 .000

�.087 .039 .117� .166 .256� .247 .034 .000 .567� .152 .042 .011 �.046 .041 �.094 .062 �.032 .000 �.212 .030 .022 .000 �.090� .122 �.229� .266 �.087 .014 �.520� .175

�.054 .052 .073� .204 .169� .326 .077 .058 .467� .352 �.012 .000 .137� .144 .225 .097 .165 .059 .497 .052

.089 .071 �.069 .049 �.179� .135 .053 .000 �.414 .090

.069 .018 �.086 .095 �.224� .228 �.123 .064 �.409 .079

�.017 .121 �.019 .098 �.029 .053 �.044 .200 .045 .050 .018 .134 �.005 .040 �.012 .017 .052� .279 �.12 .144

�.004 .065 �.006� .167 �.011� .122 �.010� .203 �.047 .395 .002 .025 .002 .009 .005 .000 .002 .035 .010 .000 .002 .014 .004 .069 .008 .040 .003 .038 .032 .138

�.001 .000 .004 .051 .004 .000 .002 .029 .043 .289 �.004 .037 .001 .000 .002 .000 .003 .031 .031 .078

951PERCEPTIONS OF AGING ACROSS 26 CULTURES

low similar cross-cultural patterns. This constitutes an important direction for further research.

Broad East–West comparisons found a small but significant effect, indicating that participants from Asian cultures do indeed hold more positive societal views of aging than their Western counterparts. However, this effect was qualified by the finding that participants from Asian cultures show less favorable perceptions of changes in wisdom. This explains some of the inconsistency in the literature: Previous studies of East–West differences may have arrived at divergent conclusions, depending on how much their measures of aging attitudes emphasized wisdom versus general societal views. The trajectories for wisdom also echo results by Harwood et al. (2001), who found that older adults from mainland China and Hong Kong reported less favorable age trajectories for wisdom than their Western counterparts.7 Our analyses suggest that the observed East–West differences are not only related to differences in values or national character but also linked to population structure. This implies that future studies examining East–West differences need to control carefully for the influence of sociodemographic factors.

There are, of course, a number of important limitations that need to be considered when interpreting our findings. For one, our measure of aging perceptions consists of a closed-ended list of only eight trait ratings. Although the individual characteristics were adapted from existing measures of aging perceptions, their combined use with a uniform response format was not validated in previous studies. Further, although we selected characteristics that cover core aspects of physical, mental, and socioemotional aging, the list is far from comprehensive. Moreover, although widely

used to assess stereotypical perceptions (e.g., Biernat & Crandall, 1996; Rosencranz & McNevin, 1969), quantitative trait lists and etic approaches cannot fully capture the complexity of aging perceptions and need to be supplemented by qualitative and culture-centered approaches (e.g., Sharps, Price-Sharps, & Han- son, 1998).

Another important limitation is the nature of our samples, which were composed almost exclusively of university students and were predominantly female. To some extent, using students was advan- tageous because it ensured comparable distributions of age and education across samples. However, using such young samples limits the generalizability of our findings. It is therefore reassuring to note that in a supplemental sample of 63 Italian adults who were not college students (age range: 22–79, M � 37.13, SD � 12.05), age was not significantly related to POA scores or societal views on aging (range of correlations: �.15 � rp � .09, ps � .25). Nevertheless, future studies should aim to recruit samples from the general population that represent the whole adult life span. Our findings are also limited by an underrepresentation of African and Middle Eastern countries and the fact that even a sample of 26 cultures is still relatively small to examine culture-level associa- tions.

7 Harwood et al. (2001) raised the concern that the Chinese version of their questionnaires translated wisdom as chong ming (i.e., clever or smart) instead of using the more appropriate zhi hui. The present study used the latter term, suggesting that the effect is consistent across alternative trans- lations.

20.015.010.05.00.0

% age 65+

0.50

0.00

-0.50

-1.00

S oc

ie ta

l V ie

w s

of A

gi ng

Argentina

Czech Republic UK

Estonia

France

Hong Kong

New Zealand

Peru

Serbia

Swiss-French Poland

RussiaIndia

Croatia

Portugal

Australia

China

Uganda

Chile

South Korea

Slovakia

USA

Malaysia

Italy Japan

I.R. Iran

Figure 2. Culture-level association between population structure and societal views on aging.

952 LÖCKENHOFF ET AL.

Future studies in larger samples should also include a broader range of value dimensions. The taxonomies by House et al. (2004), Smith et al. (1996), or Leung and Bond (2004), as well as Diener et al.’s (1995) culture-level scores on life satisfaction, could not be included because their overlap with the present sample was too small to allow for meaningful analyses. A related issue concerns discrepancies in assessment times among the culture-level associ- ates. Data on some of the value scales (e.g., Hofstede, 2001) were collected more than a decade before the present study. Value systems may change in response to historical events, and associ- ations with aging perceptions might have differed if concurrent data were available.

Finally, although our analyses address individual differences in aging perceptions within a given culture, we do not account for interindividual variability in cultural values and national character stereotypes because these characteristics are aggregated at the culture level. To address these issues, future studies should assess both aging perceptions and cultural values at the individual level and within the same sample of participants.

In spite of these limitations, our findings contribute to the literature on cross-cultural patterns of aging perceptions by illus- trating widespread similarities in perceptions of aging, describing subtle variations within this general consensus, and exploring culture-level associates of such variations. Most important, our findings emphasize the benefits of considering multiple predictors simultaneously and highlight the need for larger scale cross- cultural studies that employ a broader range of measures. As populations in nations across the world are aging (United Nations, 2007), societal perceptions of older adults and people’s expecta- tions about their own aging process will influence the well-being of older adults as well as policy decisions at the societal level. A better understanding of cross-cultural differences in aging percep- tions is a crucial resource in navigating these challenges.

References

Arnhoff, F. N., Leon, H., & Lorge, I. (1964). Cross-cultural acceptance of stereotypes toward aging. Journal of Social Psychology, 63, 41–58.

Bengtson, V. L., Dowd, J. J., Smith, D. H., & Inkeles, A. (1975). Mod- ernization, modernity and attitudes toward aging: A cross-cultural sur- vey. Journal of Gerontology, 30, 689 – 695.

Biernat, M., & Crandall, C. S. (1996). Creating stereotypes and capturing their content. European Journal of Social Psychology, 26, 867– 898.

Boduroglu, A., Yoon, C., Luo, T., & Park, C. D. (2006). Stereotypes about young and old adults: A comparison of Chinese and American Cultures. Gerontology, 52, 324 –333.

Brigham, J. C. (1971). Ethnic stereotypes. Psychological Bulletin, 76, 15–38.

Chappell, N. L. (2003). Correcting cross-cultural stereotypes: Aging in Shanghai and Canada. Journal of Cross-Cultural Gerontology, 18, 127– 147.

Charles, S. T., & Carstensen, L. L. (2007). Emotion regulation and aging. In J. J. Gross (Ed.), Handbook of emotion regulation (pp. 307–327). New York, NY: Guilford Press.

Christiansen, J. L., & Grzybowski, J. M. (1999). Biology of aging: An introduction to the biomedical aspects of aging. New York, NY: McGraw-Hill.

Cowgill, D. O. (1972). A theory of aging in cross-cultural perspective. In D. O. Cowgill & L. D. Holmes (Eds.), Ageing and modernization (pp. 1–14). New York, NY: Appleton-Century-Crofts.

Cowgill, D. O. (1986). Aging around the world. Belmont, CA: Wadsworth.

Davis, D. (1983). Long lives: Chinese elderly and the communist revolu- tion. Cambridge, MA: Harvard University Press.

Diener, E., Diener, M., & Diener, C. (1995). Factors predicting the sub- jective well-being of nations. Journal of Personality and Social Psy- chology, 69, 851– 864.

DiGiovanna, A. G. (2000). Human aging: Biological perspectives (2nd ed.). Boston, MA: McGraw-Hill.

Eagly, A. H., Wood, W., & Diekman, A. B. (2000). Social role theory of sex differences and similarities: A current appraisal. In T. Eckes (Ed.), The developmental social psychology of gender (pp. 123–174). Mahwah, NJ: Erlbaum.

Fung, H. H., Rice, C., & Carstensen, L. L. (2005). Reactive and proactive motivational changes across adulthood. In W. Greve, K. Rothermund, & D. Wentura (Eds.), The adaptive self: Personal continuity and inten- tional self-development (pp. 171–183). New York, NY: Hogrefe/Huber.

Giles, H., Noels, K., Ota, H., Ng, S. H., Gallois, C., Ryan, E. B., . . . & Sachdev, I. (2000). Age vitality across 11 nations. Journal of Multilin- gual and Multicultural Development, 21, 308 –323.

Giles, H., Noels, K. A., Williams, A., Ota, H., Lim, T.-S., Ng, S. H., . . . & Somera, L. (2003). Intergenerational communication across cultures: Young people’s perceptions of conversations with family elders, non- family elders and same-age peers. Journal of Cross-Cultural Gerontol- ogy, 18, 1–32.

Harwood, J., Giles, H., McCann, R. M., Cai, D., Somera, L. P., Ng, S. H., . . . & Noels, K. (2001). Older adults’ trait ratings of three age- groups around the Pacific Rim. Journal of Cross-Cultural Gerontology, 16, 157–171.

Harwood, J., Giles, H., Ota, H., Pierson, H.D., Gallois, C., Ng, S.H., . . . & Somera, L. (1996). College students’ trait ratings of three age groups around the Pacific Rim. Journal of Cross Cultural Gerontology, 11, 307–317.

Heckhausen, J., Dixon, R. A., & Baltes, P. B. (1989). Gains and losses in development throughout adulthood as perceived by different adult age groups. Developmental Psychology, 25, 109 –121.

Ho, D. Y. (1994). Filial piety, authoritarian moralism, and cognitive conservatism in Chinese societies. Genetic, Social, and General Psy- chology Monographs, 120, 347–365.

Hofstede, G. (1980). Culture’s consequences: International differences in work-related values. Newbury Park, CA: Sage.

Hofstede, G. (2001). Culture’s consequences (2nd ed.). Beverly Hills, CA: Sage.

Hofstede, G., & Bond, M. H. (1988). The Confucius connection: From cultural roots to economic growth. Organizational Dynamics, 16, 4 –21.

House, R., Hanges, P., Javidan, M., Dorfman, P. W., & Gupta, V. (Eds.). (2004). Culture, leadership, and organizations: GLOBE study of 62 societies. Thousand Oaks, CA: Sage.

Hummert, M. L. (1990). Multiple stereotypes of elderly and young adults: A comparison of structure and evaluations. Psychology and Aging, 5, 182–93.

Inglehart, R., & Baker, W. E. (2000). Modernization, cultural change, and the persistence of traditional values. American Sociological Review, 65, 19 –51.

Inglehart, R., & Norris, P. (2003). Rising tide: Gender equality and cultural change around the world. New York, NY: Cambridge Univer- sity Press.

Kite, M. E., Stockdale, G. D., Whitley, B. E., Jr., & Johnson, B. T. (2005). Attitudes toward older and younger adults: An updated meta-analysis. Journal of Social Issues, 61, 241–266.

Laidlaw, K., Power, M. J., Schmidt, S., & the WHOQOL Group. (2007). The Attitudes to Ageing Questionnaire (AAQ): Development and psy- chometric properties. International Journal of Geriatric Psychiatry, 22, 367–379.

Leung, K., & Bond, M. H. (2004). Social axioms: A model for social

953PERCEPTIONS OF AGING ACROSS 26 CULTURES

beliefs in multi-cultural perspective. Advances in Experimental Social Psychology, 36, 119 –197.

Levy, B. R. (2003). Mind matters: Cognitive and physical effects of aging self-stereotypes. Journals of Gerontology, Series B: Psychological Sci- ences and Social Sciences, 58, P203–P211.

Levy, B., & Langer, E. (1994). Aging free from negative stereotypes: Successful memory in China and among the American deaf. Journal of Personality and Social Psychology, 66, 989 –997.

Levy, B. R., & Myers, L. M. (2004). Preventive health behaviors influ- enced by self-perceptions of aging. Preventive Medicine, 39, 625– 629.

Maxwell, R. J. (1970). The changing status of elderly Polynesian society. Aging and Human Development, 1, 127–146.

McArdle, J. J., Ferrer-Caja, E., Hamagami, F., & Woodcock, R. W. (2002). Comparative longitudinal structural analyses of the growth and decline of multiple intellectual abilities over the life span. Developmental Psy- chology, 38, 115–142.

McCann, R. M., Cargile, A. C., Giles, H., & Bui, C. T. (2004). Commu- nicative ambivalence towards elders: Data from North Vietnam, South Vietnam, and the USA. Journal of Cross-Cultural Gerontology, 19, 275–297.

McCrae, R. R. (2002). NEO-PI-R data from 36 cultures: Further intercul- tural comparisons. In R. R. McCrae & J. Allik (Eds.), The five-factor model of personality across cultures. New York, NY: Kluwer Academic/ Plenum.

McCrae, R. R., Terracciano, A., & the Personality Profiles of Cultures Project. (2005). Personality profiles of cultures: Aggregate personality traits. Journal of Personality and Social Psychology, 89, 407– 425.

McCrae, R. R., Terracciano, A., Realo, A., & Allik, J. (2007). Climatic warmth and national wealth: Some culture-level determinants of national character stereotypes. European Journal of Personality, 21, 953–976.

McTavish, D. G. (1971). Perceptions of old people: A review of research methodologies and findings. Gerontologist, 11, 90 –101.

Moscovici, S. (1984). The phenomenon of social representations. In R. M. Farr & S. Moscovici (Eds.), Social representations (pp. 125–147). Cambridge, England: Cambridge University Press.

Moscovici, S. (1988). Notes towards a description of social representa- tions. Journal of European Social Psychology, 18, 211–250.

Palmore, E. (1975). What can the USA learn from Japan about aging? Gerontologist, 15, 64 – 67.

Palmore, E. B., & Manton, K. (1974). Modernization and status of the aged: International correlations. Journal of Gerontology, 29, 205–210.

Pasupathi, M., & Löckenhoff, C. E. (2002). Ageist behavior. In T. D. Nelson (Ed.), Ageism: Stereotyping and prejudice against older persons (pp. 201–246). Boston, MA: MIT Press.

Pettigrew, T. F., & Tropp, L. R. (2006). A meta-analytic test of intergroup contact theory. Journal of Personality and Social Psychology, 90, 751– 783.

Quadagno, J. (1982). Aging in early industrial society: Work, family and social policy in nineteenth century England. New York, NY: Academic Press.

Rosencranz, H. A., & McNevin, T. E. (1969). A factor analysis of attitudes toward the aged. Gerontologist, 9, 55–59.

Ryan, E. B., Jin, Y. S., Anas, A. P., & Luh, J. (2004). Communication

beliefs about youth and old age in Asia and Canada. Journal of Cross- Cultural Gerontology, 19, 343–360.

Salthouse, T. A., & Davis, H. P. (2006). Organization of cognitive abilities and neuropsychological variables across the lifespan. Developmental Review, 26, 31–54.

Schwartz, S. H. (1994). Beyond individualism/collectivism: New dimen- sions of values. In U. Kim, H. C. Triandis, C. Kagitcibasi, S. C. Choi, & G. Yoon (Eds.), Individualism and collectivism: Theory, method, and applications (pp. 85–119). Thousand Oaks, CA: Sage.

Schwartz, S. H. (2005). Schwartz Value Survey. Jerusalem, Israel: Israel Social Sciences Data Center.

Sharps, M. J., Price-Sharps, J. L., & Hanson, J. (1998). Attitudes of young adults toward older adults: Evidence from the United States and Thai- land. Educational Gerontology, 24, 655– 660.

Sher, A. (1984). Aging in post-Mao China: The politics of veneration. Boulder, CO: Westview Press.

Silverstein, M., Parrott, T. M., Angelelli, J. J., & Cook, F. L. (2000). Solidarity and tension between age groups in the United States: Chal- lenge for an aging America in the 21st century. International Journal of Social Welfare, 9, 270 –284.

Simmons, L. W. (1945). The role of the aged in primitive society. New Haven, CT: Yale University Press.

Smith, P. B., Dugan, S., & Trompenaars, F. (1996). National culture and the values of organizational employees: A dimensional analysis across 43 nations. Journal of Cross-Cultural Psychology, 27, 231–264.

Sung, K. (2001). Elder respect: Exploration of ideals and forms in East Asia. Journal of Aging Studies, 15, 13–26.

Tan, P. P., Zhang, N., & Fan, L. (2004). Students’ attitude toward the elderly in the people’s Republic of China. Educational Gerontology, 30, 305–314.

Terracciano, A., Abdel-Khalek, A. M., Adam, N., Adamovova, L., Ahn, C. K., Ahn, H. N., . . . & McCrae, R. R. (2005, October 7). National character does not reflect mean personality trait levels in 49 cultures. Science, 310, 96 –100.

Terracciano, A., & McCrae, R. R. (2007). Perceptions of Americans and the Iraq invasion: Implications for understanding national character stereotypes. Journal of Cross-Cultural Psychology, 38, 695–710.

United Nations. (2007). World population ageing 2007. Retrieved August 11, 2008, from http://www.un.org/esa/population/publications/ WPA2007/wpp2007.htm

United Nations Development Programme. (2007). 2007–2008 Human De- velopment Report. New York, NY: Palgrave Macmillan.

United Nations Statistics Division. (2008). Composition of macro geo- graphical (continental) regions, geographical sub-regions, and selected economic and other groupings. Retrieved August 7, 2008, from http:// unstats.un.org/unsd/methods/m49/m49regin.htm

Zhou, L. (2007). What college students know about older adults: A cross-cultural qualitative study. Educational Gerontology, 33, 811– 831.

Received October 14, 2008 Revision received April 28, 2009

Accepted May 4, 2009 �

954 LÖCKENHOFF ET AL.