Abstract:

Cross-cultural psychologists study personality dimensions and their measurement across cultures using etic (universal, imported), emic (indigenous, culture specific), or a combined etic-emic approach. The etic approach seeks to identify and assess universal dimensions that generalize well across all cultures. The emic approach attempts to identify and measure indigenous personality dimensions that are particular relevant or specific to a given culture. Several researchers have used a combined etic-emic approach, in which indigenous and hypothesized universal dimensions are related to determine the extent to which the indigenous dimensions are actually culture-specific.

This dissertation was comprised of two studies. In the first study, I critically examined imported (imposed-etic) research on personality structure and measurement, and reviewed indigenous or emic theoretical perspectives, research methods, and measures in Mexico. The second study was an empirical follow-up to the critical review in the first study. I addressed three research questions: (a) Does the Five-Factor Model, a hypothesized universal or etic model of personality, generalize well to the Mexican setting?; (b) Do the personality dimensions identified in indigenous or emic Mexican measures replicate well across samples?; and (c) Are replicable indigenous dimensions of Mexican personality well-encompassed by the hypothesized universal dimensions of the Five-Factor Model or relatively culture-specific? 794 Mexican college students completed nine indigenous measures of personality and the Revised NEO Personality Inventory, a measure of the Five-Factor Model.

I found that the Five-Factor Model replicated well across the total sample (n = 794) and two subsamples used for cross-validation (n = 400; n = 394). Reliability indexes were comparable to those reported for the American normative sample. I used the congruence of factor structures across the Mexican subsamples to determine the number of replicable factors. Although internal consistency reliability estimates were acceptable for most instruments as scored by the test authors, our replication criterion suggested alternative structures of fewer, but more replicable dimensions, for most instruments. I used multiple regression analyses and joint factor analyses to relate the replicable indigenous dimensions to the FFM and found that most of the Mexican indigenous dimensions were well encompassed by the FFM and thus not very culture specific.