Stress, Coping And Personality In Aging Men And Women

Aging

Investigators

Linked publications & trials

Abstract

Summary of work: For the Personal Profiles of Cultures Project (PPOCP), we recruited collaborators to collect data on personality and on perceptions of national character. We obtained data from 51 cultures representing six continents, using translations of the Revised NEO Personality Inventory (NEO-PI-R) into Indo-European, Hamito-Semitic, Sino-Tibetan, Daic, Uralic, Malayo-Polynesian, Dravidian, and Altaic languages. In each culture, 200 college students were randomly assigned to one of four target conditions asking for NEO-PI-R ratings of a college-age woman, college-age man, adult (over 40) man or adult woman whom the rater knew well. An additional 50 volunteers were asked to rate the typical member of their culture, and the typical American, on 30 items designed to parallel the 30 facets of the NEO-PI-R. The first outcome of PPOCP was an article analyzing data at the individual level. These analyses confirmed the universality of trait properties using observer ratings in place of self-reports. Factor analyses within cultures showed that the normative American self-report structure was clearly replicated in most cultures, and was recognizable in all. African cultures have rarely been included in cross-cultural analyses of personality traits, but in this study observer ratings of personality were collected from students in Burkina Faso, Uganda, Nigeria, Ethiopia, and Botswana. The factor structure was weakly replicated in each of these cultures separately, but when combined, the data showed a clear replication of the universal structure. The validity of observer ratings on the NEO-PI-R across cultures made it reasonable to aggregate scores and attempt to replicate intercultural analyses of aggregate self-reports. These analyses presume the scalar equivalence of NEO-PI-R observer ratings, for which we had no direct evidence. In the second article from PPOCP, we argued that aggregate scores could be validated in much the same way that individual test scores are, and we conducted analyses of generalizability, replicablity of structure, and convergent and discriminant validity at the culture level (that is, using culture means as the unit of analysis). The final sample consisted of 202 subsamples of adult men, adult women, college-age men, and college age women (N = 12,256 targets). We assessed reliability as intraclass correlations, ICC(1, k). These values were .88, .91, .92, .91, and .89 for Neuroticism, Extraversion, Openness to Experience, Agreeableness, and Conscientiousness, respectively. These very high values are understandable, given that each of the 51 data points is based on an average of 238 ratings. Next, we standardized the data within age and gender groups and factored the 30 NEO-PI-R facet means for the 202 subsamples in order to test whether the FFM structure was maintained at the culture level. We rotated five factors to a target defined by the American adult self-report normative structure, and found congruence coefficients high enough to claim replicability for four of the five factors. Congruence for the Extraversion factor is somewhat lower than the rule-of-thumb requires for replication, although it is statistically far above chance. The individual level E and the culture level E differ chiefly in that the latter is broader, including several definers not found at the individual level. Descriptively, this appears to be a very brash, self-confident, and open form of Extraversion, and it appears to be a genuine cultural overlay on the basic trait structure. A similar broad factor had been found in culture-level analyses of self-report data, and in a second analysis of observer-rating data restricted to cultures not included in the original self-report study. Two different methods of measurement and two different samples of cultures yielded the same novel structure. We correlated aggregate means for NEO-PI-R factors and facets with aggregate means from an earlier self-report study; three of the factors and 26 of the 30 facet scales showed significant associations. Within each of the 28 cultures included in both studies, we calculated intraclass correlations between the 30 facets scales; 22 of them were significant. Factors also showed a pattern of convergent and discriminant validity with culture-level measures of belief and values. For example, Openness was negatively related to Conservatism, and Extraversion was positively related to subjective well-being. Aggregate NEO-PI-R scores appear to be valid, and it is therefore meaningful to compare cultures on their levels of these traits. A multidimensional scaling analysis of the data replicated earlier findings, again suggesting that Asians and Africans are more introverted than Europeans and Americans. The study of aggregate personality traits is important for an understanding of health and aging because through them the many associations between personality and health may be writ large. For example, with colleagues in Russia, we recently found that at the individual level within cultures, HIV stigmatization was negatively related to Openness, especially O6: Values. This effect appears to be magnified at the aggregate level: Cultures with very low levels of O6 include South Africa and Zimbabwe, where official reluctance to deal with HIV infection has led to devastating epidemics. The full range of aggregate personality traits might be relevant to a host of social, economic, and health outcomes.