عزت نفس و ارتباط آن با افسردگی در نوجوانان چینی، ایتالیایی و کاستاریکا: مطالعه میان فرهنگی

This study investigated the factor structure of the Rosenberg Self-Esteem Scale (RSES), the differences in mean level of self-esteem and its association with depression in adolescents in three different cultures. The RSES and the Children’s Depression Inventory were administered to Chinese (N = 350), Italian (N = 352), and Costa Rican (N = 343) adolescents. Confirmatory factor analyses supported the two-factor model of the RSES and it was demonstrated to be invariant across cultures using multi-group confirmatory factor analyses. MANOVA results indicated that Costa Rican adolescents scored higher on positive and negative self-esteem than their Chinese and Italian counterparts. Furthermore, both positive and negative self-esteem was related to depression across cultures. In conclusion, there are both cultural differences and similarities in self-esteem.

Self-esteem refers to one’s general sense of worthiness (Rosenberg, 1965). People in different cultures have different perceptions about themselves (Markus & Kitayama, 1991), and therefore cross-cultural perspective is an important approach to investigate self-esteem. To date, some debates have not been completely solved. First, the Rosenberg Self-Esteem Scale (RSES), a popular instrument to assess self-esteem, is conceptualized as unidimensional (Schmitt & Allik, 2005), but some studies argue that it consists of two distinct, yet related, components (Greenberger, Chen, Dmitrieva, & Farruggia, 2003). Second, some studies have found that people in individualistic cultures have higher self-esteem than in collectivistic cultures (Heine, Lehman, Markus, & Kitayama, 1999), while others have not supported this view (Schmitt & Allik, 2005). Last, some scholars posited that self-esteem plays a role in numerous outcomes (e.g., depression) only in individualistic cultures, because self-esteem is more emphasized in individualistic than in collectivistic cultures (Heine et al., 1999). However, others contended that the function of self-esteem is equally important in individualistic and collectivistic cultures (Cai, Wu, & Brown, 2009). To our knowledge, there is scant research comparing the factor structure of the RSES, the differences in mean level of self-esteem, and its relationship with depression among East Asian, European, and Latin American adolescents. For example, Farruggia, Chen, Greenberger, Dmitrieva, and Macek (2004) investigated the factor structure of the RSES, the differences in mean level of self-esteem, and the relationship between self-esteem and depression in Chinese, Korean, U.S., and Czech adolescents, but it did not include a Latin American sample. Schmitt and Allik (2005) investigated the factor structure of the RSES and the differences in mean level of self-esteem in 53 countries, but they did not focus on adolescents and did not examine its relationship with depression. To fill these gaps in the literature, the present study investigated the factor structure of the RSES, the differences in mean level of self-esteem, and the relationship between self-esteem and depression among Chinese, Italian, and Costa Rican adolescents. 1.1. Individualism and collectivism Individualism–collectivism is an important framework to investigate cultural similarities and differences. At first, individualism–collectivism is viewed as one bipolar dimension (Hofstede, 1980). However, this classification is criticized for its oversimplification, and some studies proposed more complex models (Freeman and Bordia, 2001 and Triandis and Gelfand, 1998). Triandis and Gelfand (1998) considered individualism and collectivism as two general orthogonal dimensions and they further divided these two dimensions into four categories (i.e., vertical individualism, vertical collectivism, horizontal individualism, and horizontal collectivism) by including whether people emphasize equality or hierarchy, a contention similar to Hofstede’s power distance (defined as the extent to which people within a country accept that power is distributed equally, geert-hofstede.com) although there are some conceptual differences (Shavitt, Torelli, & Riemer, 2010). Triandis and Gelfand (1998) supported that people could be both collectivist and individualist at the same time. Freeman and Bordia (2001) stated that this does not mean people are collectivistic in one context and individualistic in another, but that “people may endorse both individualist and collectivist attitude statement within the same context” (p. 107). On this basis, they found that individualism and collectivism was a general bipolar higher-order construct that included individualism and collectivism within different contexts (i.e., individualism–collectivism within family, individualism–collectivism within peers, individualism–collectivism within nation, and individualism–collectivism within school). Although there is not clear definition, based on the scores of individualism and power distance (http://geert-hofstede.com), China can be roughly viewed as a vertical collectivistic culture (low individualism and high power distance), and Costa Rica can be roughly seen as a horizontal collectivistic culture (low individualism and low power distance). Italy is a representative of individualistic country but it is in the middle of horizontal and vertical dimension (high individualism and medium power distance). According to Triandis and Gelfand (1998), in vertical individualistic cultural contexts, people like to become distinguished and acquire status through competition with others; in horizontal individualistic cultural contexts, people focus on expressing their uniqueness and building one’s ability to be successfully self-reliant and see themselves as equal to others in status; in vertical collectivistic cultural contexts, people highlight the integrity of the in-group and are willing to sacrifice their own personal goals to comply with authorities; in horizontal collectivistic cultural contexts, people tend to view themselves as being similar to others and emphasize sociability and interdependence, but do not easily submit to authority.

We tested one-factor model with 10 items, two-factor model with 10 items, and two-factor model with 9 items (with item 8 deleted because it was found problematic, Schmitt & Allik, 2005). Results showed that the two-factor model with 10 items showed adequate fit in Chinese and Italian samples, and close fit in Costa Rican sample. In addition, because the one-factor model with 10 items and the two-factor model with 10 items were nested, we tested chi-square difference between these two models. Results showed that the two-factor model was better than the one-factor model across Chinese (Δχ2 (1) = 167.323, p < .01), Italian (Δχ2 (1) = 57.968, p < .01), and Costa Rican (Δχ2 (1) = 208.854, p < .01) samples. Therefore, the two-factor model with 10 items was selected as our final model for subsequent analyses ( Table 1).