- Email: [email protected]
A cross-lagged model of self-esteem and life satisfaction: Gender differences among Chinese university students
Personality and Individual Differences 52 (2012) 546–551
Contents lists available at SciVerse ScienceDirect
Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid
A cross-lagged model of self-esteem and life satisfaction: Gender differences among Chinese university students Shengquan Ye a,⇑, Lu Yu b, Kin-Kit Li a a b
Department of Applied Social Studies, City University of Hong Kong, Hong Kong Department of Applied Social Sciences, Hong Kong Polytechnic University, Hong Kong
a r t i c l e
i n f o
Article history: Received 25 June 2011 Received in revised form 17 November 2011 Accepted 22 November 2011 Available online 24 December 2011 Keywords: Life satisfaction Self-esteem Cross-lagged model Gender difference
a b s t r a c t Past research has shown a close relationship between self-esteem (SE) and life satisfaction (LS). Using cross-lagged structural equation modeling, the current longitudinal study examined the directionality of the relationship among a sample of Chinese university students (214 males and 134 females). In the model, SE and LS at Time 1 were used to predict SE and LS at Time 2, with socioeconomic status (SES) as the independent variable of both constructs. Results showed that SE consistently predicted subsequent LS among both genders, whereas no significant effects of LS on subsequent SE were found. Furthermore, SES had significant effects on LS among both genders. Its influence on SE, however, was significant among females only. Implications and limitations are discussed. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Life satisfaction (LS) is the cognitive component of subjective well-being (SWB), which reflects an individual’s overall evaluation about his or her quality of life (Diener & Diener, 1995). Unlike the emotional components of SWB (pleasant and unpleasant affective experience) that can be influenced by transient mood, life satisfaction is a more enduring assessment of one’s life and therefore is commonly used as the indicator of one’s subjective well-being (Diener, Suh, Lucas, & Smith, 1999). Consequently, the construct of LS and its correlates have attracted much research attention in recent decades (Eid & Larsen, 2008). Existing literature has documented that high LS is associated with less psychological problems as well as positive attributes and outcomes (e.g., Diener, 1994; Diener, Suh, & Oishi, 1997). Among the constructs, self-esteem (SE) has been found to be a robust correlate of LS for people across cultures and gender (e.g., Chen, Cheung, Bond, & Leung, 2006; Diener & Diener, 1995). For instance, in a large cross-national study by Diener and Diener (1995), SE and LS were significantly correlated at r = .47. Moreover, the association was stronger among people in individualistic cultures than those in collectivistic cultures. Diener and Diener explained that individualists were typically socialized to attend to internal over external attributes and thus tend to evaluate lives based on personal attributes (such as SE); while collectivists were often socialized to value social norms and make judgment about lives ⇑ Corresponding author. Tel.: +852 3442 8114; fax: +852 3442 0283. E-mail address: [email protected] (S. Ye). 0191-8869/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.paid.2011.11.018
according to external sources (such as relationships with others). The finding is congruent with self-construal theory (Hofstede, 1980; Markus & Kitayama, 1991; Triandis, 1995), which argues that people in collectivistic cultures tend to develop an interdependent self-construal, whereas people in individualistic cultures usually adopt an independent self-construal. The framework can also be used to explain gender differences in the association between SE and LS. Most cultures encourage males to develop an independent self-construal and females to form an interdependent self-construal (Cross & Madson, 1997; Markus & Oyserman, 1989). Therefore, when people make life evaluations, males (like individualists) would put more weight on internal sources and females would emphasize more on external sources. Consistent with this prediction, Reid’s (2004) study among 206 American college students showed that self-esteem predicted men’s well-being to a larger extent than it predicted women’s well-being, whereas relationship harmony had a significant effect on well-being among women but not men. Similar gender difference can also be found in research on self-efficacy and life satisfaction (e.g., Vecchio, Gerbino, Pastorelli, Del Bove, & Caprara, 2007). Though previous studies have established clear association between SE and LS and provided sound explanation for cultural as well as gender differences in the association, little is known about the directionality of the relationship. In their comprehensive review of self-esteem literature, Baumeister, Campbell, Krueger, and Vohs (2003) acknowledged that previous research had not clearly established causation of SE and happiness, although people often assume or believe that higher SE leads to greater happiness. The authors cautioned that intervention based on incorrect
S. Ye et al. / Personality and Individual Differences 52 (2012) 546–551
assumption of causality could bring about undesirable effects. They explained, with the relationship between SE and academic performance as an example, if self-esteem is not the cause but effect of academic performance, intervention aiming at enhancing self-esteem would not benefit academic performance. On the contrary, it can harm academic performance because students may pay less effort when they feel more satisfied with themselves. Similarly, it is also dangerous to assume that SE causes LS and disregard the other possibility before clear evidence is obtained from empirical studies. The present study endeavors to contribute to the current understanding in the following ways. First, the study is to test the directionality of the relationship between SE and LS. Although researchers often regard happiness as the outcome, recent empirical studies have shown happiness can be the cause as well. For instance, it was found that happiness could lead to various successes (Lyubomirsky, King, & Diener, 2005) and positive changes in people’s behaviors (Salmela-Aro & Tuominen-Soini, 2010). To address the issue, this study applies longitudinal design and analyzes data with cross-lagged structural equation modeling, which is to be described later in details (see also Finkel, 1995; Taris, 2000). Furthermore, the study examines the relationship in a Chinese context, which is regarded as a typical collectivistic culture (Oyserman, Coon, & Kemmelmeier, 2002). As the majority of previous studies were conducted among individualistic cultures, findings from the study will add useful information to the current literature and facilitate future systematic examination (e.g., meta-analysis) of the phenomenon. Second, the study attempts to investigate the differences and similarities of the directional relationship between SE and LS across gender. Researchers have argued that in collectivistic cultures, where social norms and relationship harmony are highly valued, both males and females tend to form an interdependent self-construal (Markus & Kitayama, 1991). Therefore, it is possible that gender differences in the relationship between SE and LS reported in previous Western studies may not be pronounced for people in collectivistic cultures. So far, only one study has examined the moderating effect of gender on the SE–LS relation among Chinese populations (Zhang & Leung, 2002) and reported a nonsignificant gender effect. More evidence on this issue is needed in order to fully understand the moderating effect of gender on the SE–LS relation in different cultures. Third, when testing the directionality of the relationship between variables, researchers need to take into account the effect of the common cause. Previous work has often shown that higher social economic status (SES) is associated with both higher levels of SWB (Kousha & Mohseni, 1997) and self-esteem (Orth, Trzesniewski, & Robins, 2010; Twenge & Campbell, 2002). In order to disentangle the relationship between SE and LS, the effect of SES must be taken into account carefully. Current literature has shown that the effect of SES varies across countries as well as gender. For example, SES was more strongly correlated with LS in poorer countries than in wealthy countries where people have satisfied their basic needs (Diener & Oishi, 2000; Oishi, Diener, Lucas, & Suh, 1999). Several studies have shown that the prediction of SES on SWB is stronger for men than for women (Adelmann, 1987; Aryee, 1999), due to the different gender roles. Because of males’ traditional role as household breadwinners and females’ as caregivers, men are likely to find more life satisfaction in their achievements in occupation-related fields whereas women would derive more happiness from relationships and family. Considering the rapid changes of social and economic systems in modern China, one may wonder to what extent these findings could be applied to Chinese context to help us to understand the relationship between SE and LS among males and females in China. As numerous recent studies have shown, contemporary Chinese people’s conceptuali-
547
zation of masculinity and femininity, gender role adoption, and attitudes toward women’s social roles are different from those of Western people (e.g., Chang, 1999; Tang & Lau, 1995; Wu, 1996; Yu & Xie, 2008). For example, Wu (1996) reported that parents of only children believed that boys and girls should enjoy equal levels of education, dramatically different from the traditional genderdifferentiated expectations about educational achievement. These characteristics of Chinese society might have important implications on the way Chinese young people evaluate their lives as well as themselves. Therefore, it would be interesting to examine whether the relationship of SES with LS and SE would be different for Chinese men and women nowadays. In sum, the study endeavors to investigate three main questions among a Chinese sample. First, the study aims to clarify the directional relationship between LS and SE. Second, the study attempts to explore gender differences in the relationship. Finally, the role of SES in the relationship between LS and SE is examined. 2. Method 2.1. Participants and procedure Three hundred and forty-eight students (214 males and 134 females) from a large university in Mainland China participated in the study. The students were from a variety of science and art disciplines including history, philosophy, law, information technology, and physics. The age range was from 16 to 24 years (M = 20.00, SD = 0.99). The inventories were administered to the participants on two occasions with an interval of 8 months. The scales were translated from English into Chinese, with back-translation procedure to ensure accuracy and equivalency. Students were told that data were collected for research purposes, that participation was completely voluntary, and that confidentiality would be assured. Small gifts were given to the students as an incentive of participation. 2.2. Measures 2.2.1. Life satisfaction The Satisfaction With Life Scale (SWLS; Diener, Emmons, Larsen, & Griffin, 1985) is a five-item scale that assesses individuals’ life satisfaction, the cognitive component of subjective well-being. The SWLS typically uses a 7-point response format (ranging from not at all true of me to exactly true of me). A higher scale score indicates a higher level of life satisfaction. Since it was developed, the SWLS has been widely used in measuring general life satisfaction and has shown favorable psychometric properties in various languages and samples including Chinese (e.g., Abdallah, 1998; Neto, 1993; Sachs, 2004). For the current study, the Cronbach’s alpha coefficients were .77 and .81 for the two measurement occasions, respectively. 2.2.2. Self-esteem The Rosenberg Self-Esteem Scale (RSES; Rosenberg, 1989) is a 10-item scale that has been widely used for assessing general self-esteem. For consistency, this study applied a 7-point Likert scale (ranging from strongly disagree to strongly agree). The RSES consists of five positively and five negatively worded items. Past research has shown satisfactory internal consistency of the scale among Chinese samples (e.g., Chen et al., 2006) and provided evidence for the cross-cultural invariance of the construct (e.g., Russell, Crockett, Shen, & Lee, 2008). However, recent psychometric studies using various scales and languages have generally suggested that including negatively worded items, although helping to reduce response set, could complicate the factor structure by
548
S. Ye et al. / Personality and Individual Differences 52 (2012) 546–551
producing a method factor (e.g., Marsh, 1996; Tomás & Oliver, 1999; Ye, 2009). Following Marsh’s suggestion (1996), negatively worded items were used in the questionnaire but not calculated in forming the sum score. The internal consistency of the five positive items was found to be satisfactory (i.e., alpha coefficients were .80 and .83 for Time 1 and Time 2, respectively). 2.2.3. Socioeconomic status (SES) To assess SES, four related demographic variables were measured: origin, family income, paternal education, and maternal education. Origin refers to the place that the students had been living before being admitted to the university. Three categories provided were rural area, suburb, and urban city (with the latter indicating higher SES). Family income was rated on a 6-point Likert scale, ranging from very poor to very good. Parental education was rated on a 5-point Likert scale, ranging from primary school to postgraduate. The four variables were significantly correlated (r = .38–.71) and used as indicators of a latent SES variable in subsequent structural equation modeling. Since the four indicators were rather stable over the period of investigation, only information collected at Time 1 was used in the data analysis. 2.3. Data analysis With SPSS 15.0, descriptive statistics and mean difference test were first performed. LISREL 8.80 (Jöreskog & Sörbom, 2004) was then employed to examine the structural models and test their goodness of fit. Estimation method for the models was chosen based on the multivariate normality of the items. Maximum Likelihood (ML) estimation would be used when multivariate normality is confirmed, whereas Robust Maximum Likelihood (RML) estimation would be applied when the assumption is rejected. Goodness of model fit was assessed with several indices. As suggested by Hu and Bentler (1999), RMSEA (Steiger, 1990) 6.06 and SRMR (Bentler, 1995) 6.08 indicate a good model fit, while values of .95 or above for TLI (also known as NNFI; Tucker & Lewis, 1973) and CFI (Bentler, 1990) are required for an acceptable fit. Moreover, NFI (Bentler & Bonett, 1980) with values greater than .90 also indicate the model fit is acceptable. 3. Results The mean, standard deviation, skewness, kurtosis, and range of the sum scores of life satisfaction (LS) and self-esteem (SE) are presented in Table 1. As is shown, the sum scores were overall normally distributed. The standard deviations for the scores of LS and SE were slightly but consistently smaller among females than males, suggesting that female students were more homogeneous in their life satisfaction and self-esteem. To test the mean differences in SE and LS, repeated-measures ANCOVA was conducted, with time as the within-subjects factor, gender as the
Table 1 Descriptive statistics of scale sum scores.
Mean SD Skewness Kurtosis Minimum Maximum
LS1m
LS1f
LS2m
LS2f
SE1m
SE1f
SE2m
SE2f
19.85 5.70 .04 .28 8 35
22.26 5.01 .07 .26 9 34
20.32 5.56 .17 .12 5 35
22.07 4.74 .01 .01 10 35
26.90 4.97 .52 .03 11 35
27.64 4.38 .33 .57 17 35
25.45 5.18 .28 .34 10 35
27.17 4.36 .10 .69 17 35
LS1, life satisfaction at Time 1; LS2, life satisfaction at Time 2; SE1, self-esteem at Time 1; SE2, self-esteem at Time 2. m Male. f Female.
between-subjects factor, and SES as the covariate. Results showed that SE decreased across the period (p < .01, Partial g2 = .038), females had higher LS (p < .05, Partial g2 = .019) and SE (p < .05, Partial g2 = .013) than males, and SES had a significant positive effect on LS (p < .01, Partial g2 = .123) and SE (p < .01, Partial g2 = .013). To test the directional relationship between LS and SE, a crosslagged model was specified, in which LS and SE at Time 2 were predicted by LS and SE at Time 1 and all of these four latent variables were predicted by SES, as demonstrated in Fig. 1 and Fig. 2 (only the structural models are presented for clarity). The model was estimated among male and female samples separately to examine possible gender differences in the relationship (i.e., the moderating effect of gender). Prior to model fitting, multivariate normality of the data should be examined in order to choose an appropriate estimation method. With the PRELIS procedure in LISREL program, it was found that the data did not follow multivariate normal distribution and thus failed to meet the assumption required for the widely used estimation method, maximum likelihood (ML). Therefore, asymptotic covariance matrices were calculated and robust maximum likelihood (RML) was used to estimate the cross-lagged structural equation model. The model provided acceptable fit to the data among both genders (for males: v2(df = 239, n = 214) = 471.59, RMSEA = .068 (90% CI = .059–.077), SRMR = .084, NNFI = .95, and CFI = .95; for females: v2(df = 239, n = 134) = 343.51, RMSEA = .057 (90% CI = .043–.070), SRMR = .078, NNFI = .97, and CFI = .97). As shown in Fig. 1 and Fig. 2, most parameters were quite comparable across genders, apart from a couple of exceptions. Specifically, SES had significant effects on SE among female (i.e., the standardized coefficients were .17 and .14 at T1 and T2, respectively, p < .05) but not male students. In addition, though the path coefficients from LS1 to SE2 were non-significant in both models, the effect among males was somewhat stronger and approached the significance level (i.e., b = .12, t = 1.56). 4. Discussion The present study examined the directionality of the relationship between SE and LS and the roles of gender and SES in the relationship. The findings are organized and discussed in the following three sections: (a) gender differences in SE and LS; (b) gender differences in SE–LS relationship, and (c) gender differences in SES–SE relationship. 4.1. Gender differences in SE and LS For both males and females, their LS remained relatively stable across the period, whereas SE dropped significantly. Previous work has shown that transition to college brings a lot of changes in various domains such as financial situation (Goldrick-Rab & Han, 2011), interpersonal relationship (Kenyon & Koerner, 2009), and life style (Abar & Maggs, 2010). The process of coping with these challenges can be stressful and detrimental to one’s self-esteem. However, it should be noted that, due to the relatively short period being investigated in the present study, it remained uncertain whether the decrease could be understood as a continuous trend throughout the whole university life. Future research covering a longer period will help to clarify the picture. Aside from the similar decrease in SE, significant gender differences were also found. Compared with male students, female students were higher in both SE and LS over the investigated period of time. This pattern deserves a closer examination, considering the dynamics of traditional Chinese culture and the rapidly changing socioeconomic environment. Regarding SE, a meta-analytical study with 216 effect sizes and 97,121 respondents showed that men
S. Ye et al. / Personality and Individual Differences 52 (2012) 546–551
549
Fig. 1. Structural cross-lagged model for male students. Note. LS1 = life satisfaction at Time 1; LS2 = life satisfaction at Time 2; SE1 = self-esteem at Time 1; SE2 = self-esteem at Time 2; SES = socioeconomic status; n.s. = non-significant.
Fig. 2. Structural cross-lagged model for female students. Note. LS1 = life satisfaction at Time 1; LS2 = life satisfaction at Time 2; SE1 = self-esteem at Time 1; SE2 = selfesteem at Time 2; SES = socioeconomic status; n.s. = non-significant.
scored slightly higher on SE than women (Kling, Hyde, Showers, & Buswell, 1999). Most included studies were from the United States, Australia, and Canada. However, a recent study in China showed that SE was higher among female than male students (Chen & Zhang, 2004). Like SE, similar cross-cultural differences in LS were observed. A cross-cultural study with 17,246 students in 21 countries showed men were more satisfied with life than women in two regions including (1) Western Europe and the United States and (2) Central and Eastern Europe, whereas women were more satisfied with life than men in Pacific Asia (Grant, Wardle, & Steptoe, 2009). In addition, Dorahy and colleagues (1996) found that LS was higher among Indian women than Indian men. The current findings were consistent with the results found in those collectivistic cultures. Since SE and LS shared many common factors theoretically, the gender differences in SE and LS might follow similar mechanisms. The differential gender differences across cultures may be due to the rapid changes in educational opportunities for women in China. Because of the one-child policy, urban Chinese men and women enjoy the same educational opportunities (Tsui & Rich, 2002). This sample of young women might be the first generation of women in their families to receive college education. The contrast
between themselves and the older generations of women may be particularly empowering, which boosts their SE and LS. In addition, female students who grew up in the rural areas (more likely to have siblings) might have overcome many challenges and inequalities to be admitted to college. Higher SE and LS might provide them with the competitive advantages to succeed (Joshi & Srivastava, 2009) or could be the results of successes amid the challenges (Ross & Broh, 2000). 4.2. Gender differences in the SE–LS relationship Regarding the directionality of the SE–LS relationship, this study provided clear evidence that higher SE could lead to higher LS over time. Furthermore, it was observed that the effects of SE on subsequent LS were very similar across gender, suggesting that the evaluation about oneself is an important source of life satisfaction for both males and females. The current study, which aimed to examine whether the moderating effect of gender on the SE–LS relation (Reid, 2004) could be demonstrated in a Chinese population, failed to replicate Reid’s finding that SE predicted LS more strongly in men than women. This may be due to more independent self-construal (i.e., others are separated from the
550
S. Ye et al. / Personality and Individual Differences 52 (2012) 546–551
representation of the self) among Chinese women nowadays or the increased egalitarianism of gender roles in modern Chinese societies (e.g., Chang, 1999). As for the effect of LS on subsequent SE, the coefficients were found to be non-significant among males and females. However, it is noteworthy that the non-significant effect among male students might result from the lack of statistical power considering its effect size (b = .12). Being content and satisfied with life may be a source of SE among male college students. For those with a high level of internal locus of control, life dissatisfaction may reflect personal inadequacy, which may dampen SE. Different from the ascribed SES, LS perhaps more directly holds the person responsible. The traditional dominant role for men in Chinese societies may put men in a difficult position. The expectation for men is still high, while privileges and circumstances that lead to success have become more equal across sexes nowadays. 4.3. Gender differences in the SES–SE relationship In a meta-analysis with 446 samples and 312,940 individuals, Twenge and Campbell (2002) showed that the relationship between SES and SE was small but comparable to the effects of gender and race. The effect size of SES for women became larger than for men since 1940 and kept increasing in subsequent decades. The current results were consistent with this pattern, in which the SES effect was small among college students and only significant among female students. SES is ascribed from the parents for most of the college students; hence it may have minimal effects as a social indictor that boosts SE (Rosenberg & Pearlin, 1978). Because of the labor-force movement for women and the relinquished role as the sole breadwinner for men, Twenge and Campbell suggested that the effect of SES on SE may increase for women and decrease for men over generations. There are a few limitations for this study. First, participants were recruited from one university in Mainland China. Generalizability of the findings is limited and future replications with more representative samples are recommended. Second, the sample size of this study was less than ideal. Only moderate effect sizes could be detected statistically. Third, due to some practical constraints, the data were collected at two time points with a relatively short interval. With more waves over a longer period, researchers will be able to capture a more comprehensive picture and test the relationship in a more sophisticated way. Finally, as an exploratory study, only measures of the key constructs were included. Future studies should examine these relationships with additional explanatory variables such as self-construal and relationship harmony (e.g., Kwan, Bond, & Singelis, 1997; Reid, 2004). References Abar, C. C., & Maggs, J. L. (2010). Social influence and selection processes as predictors of normative perceptions and alcohol use across the transition to college. Journal of College Student Development, 51(5), 496–508. Abdallah, T. (1998). The Satisfaction with Life Scale (SWLS): Psychometric properties in an Arabic-speaking sample. International Journal of Adolescence & Youth, 7(2), 113–119. Adelmann, P. K. (1987). Occupational complexity, control, and personal income: Their relation to psychological well-being in men and women. Journal of Applied Psychology, 72(4), 529–537. Aryee, S. (1999). An examination of the moderating influence of breadwinner role salience on the pay-life satisfaction relationship. Human Relations, 52(10), 1279–1290. Baumeister, R. F., Campbell, J. D., Krueger, J. I., & Vohs, K. D. (2003). Does high selfesteem cause better performance, interpersonal success, happiness, or healthier lifestyles. Psychological Science in the Public Interest, 4(1), 1–44. Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107(2), 238–246. Bentler, P. M. (1995). EQS structural equations program manual. Encino, CA: Multivariate Software. Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness of fit in the analysis of covariance structures. Psychological Bulletin, 88(3), 588–606.
Chang, L. (1999). Gender role egalitarian attitudes in Beijing, Hong Kong, Florida, and Michigan. Journal of Cross-Cultural Psychology, 30(6), 722–741. Chen, S. X., Cheung, F. M., Bond, M. H., & Leung, J.-P. (2006). Going beyond selfesteem to predict life satisfaction: The Chinese case. Asian Journal of Social Psychology, 9(1), 24–35. Chen, L., & Zhang, J. (2004). General life satisfaction and self-esteem of university students. Chinese Mental Health Journal, 18(4), 222–224. Cross, S. E., & Madson, L. (1997). Models of the self: Self-construals and gender. Psychological Bulletin, 122(1), 5–37. Diener, E. (1994). Assessing subjective well-being: Progress and opportunities. Social Indicators Research, 31(2), 103–157. Diener, E., & Diener, M. (1995). Cross-cultural correlates of life satisfaction and selfesteem. Journal of Personality and Social Psychology, 68(4), 653–663. Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction with life scale. Journal of Personality Assessment, 49(1), 71–75. Diener, E., & Oishi, S. (2000). Money and happiness: Income and subjective wellbeing across nations. In E. Diener & E. M. Suh (Eds.), Culture and subjective wellbeing (pp. 185–218). Cambridge, MA: The MIT Press. Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being: Three decades of progress. Psychological Bulletin, 125(2), 276–302. Diener, E., Suh, E., & Oishi, S. (1997). Recent findings on subjective well-being. Indian Journal of Clinical Psychology, 24(1), 25–41. Dorahy, M. J., Schumaker, J. F., Simpson, P. L., & Deshpande, C. G. (1996). Depression and life satisfaction in India and Australia. Journal of Personality and Clinical Studies, 12(1–2), 1–7. Eid, M., & Larsen, R. J. (2008). The science of subjective well-being. New York, NY, US: Guilford Press. Finkel, S. E. (1995). Causal analysis with panel data. Thousand Oaks: Sage. Goldrick-Rab, S., & Han, S. W. (2011). Accounting for socioeconomic differences in delaying the transition to college. Review of Higher Education: Journal of the Association for the Study of Higher Education, 34(3), 423–445. Grant, N., Wardle, J., & Steptoe, A. (2009). The relationship between life satisfaction and health behavior: A cross-cultural analysis of young adults. International Journal of Behavioral Medicine, 16, 259–268. Hofstede, G. (1980). Culture and organizations. International Studies of Management & Organization, 10(4), 15. Hu, L.-t., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55. Jöreskog, K. G., & Sörbom, D. (2004). LISREL 8.70. Chicago: Scientific Software. Joshi, S., & Srivastava, R. (2009). Self-esteem and academic achievement. Social Science International, 25(1), 75–87. Kenyon, D. Y. B., & Koerner, S. S. (2009). College student psychological well-being during the transition to college: Examining individuation from parents. College Student Journal, 43(4, PtA), 1145–1160. Kling, K. C., Hyde, J. S., Showers, C. J., & Buswell, B. N. (1999). Gender differences in self-esteem: A meta-analysis. Psychological Bulletin, 125(4), 470–500. Kousha, M., & Mohseni, N. (1997). Predictors of life satisfaction among urban Iranian women: An exploratory analysis. Social Indicators Research, 40(3), 329–357. Kwan, V. S. Y., Bond, M. H., & Singelis, T. M. (1997). Pancultural explanations for life satisfaction: Adding relationship harmony to self-esteem. Journal of Personality and Social Psychology, 73(5), 1038–1051. Lyubomirsky, S., King, L., & Diener, E. (2005). The benefits of frequent positive affect: Does happiness lead to success? Psychological Bulletin, 131(6), 803–855. Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98(2), 224. Markus, H. R., & Oyserman, D. (1989). Gender and thought: The role of the selfconcept. In M. Crawford & M. Gentry (Eds.), Gender and thought: Psychological perspectives (pp. 100–127). New York, NY: SpringerVerlag New York Inc. Marsh, H. W. (1996). Positive and negative global self-esteem: A substantively meaningful distinction or artifactors? Journal of Personality and Social Psychology, 70(4), 810–819. Neto, F. (1993). The Satisfaction With Life Scale: Psychometrics properties in an adolescent sample. Journal of Youth & Adolescence, 22(2), 125–134. Oishi, S., Diener, E. F., Lucas, R. E., & Suh, E. M. (1999). Cross-cultural variations in predictors of life satisfaction: Perspectives from needs and values. Personality and Social Psychology Bulletin, 25(8), 980–990. Orth, U., Trzesniewski, K. H., & Robins, R. W. (2010). Self-esteem development from young adulthood to old age: A cohort-sequential longitudinal study. Journal of Personality and Social Psychology, 98(4), 645–658. Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and collectivism: Evaluation of theoretical assumptions and meta-analyses. Psychological Bulletin, 128(1), 3–72. Reid, A. (2004). Gender and sources of subjective well-being. Sex Roles, 51(11–12), 617–629. Rosenberg, M. (1989). Society and the adolescent self-image (Rev. ed.). Middletown, Conn.: Wesleyan University Press. Rosenberg, M., & Pearlin, L. I. (1978). Social class and self-esteem among children and adults. American Journal of Sociology, 84(1), 53–77. Ross, C. E., & Broh, B. A. (2000). The role of self-esteem and the sense of personal control in the academic achievement process. Sociology of Education, 73(4), 270–284. Russell, S. T., Crockett, L. J., Shen, Y.-L., & Lee, S.-A. (2008). Cross-ethnic invariance of self-esteem and depression measures for Chinese, Filipino, and European American adolescents. Journal of Youth and Adolescence, 37(1), 50–61.
S. Ye et al. / Personality and Individual Differences 52 (2012) 546–551 Sachs, J. (2004). Validation of the Satisfaction With Life Scale in a sample of Hong Kong university students. Psychologia, 46(4), 225–234. Salmela-Aro, K., & Tuominen-Soini, H. (2010). Adolescents’ life satisfaction during the transition to post-comprehensive education: Antecedents and consequences. Journal of Happiness Studies, 11(6), 683–701. Steiger, J. H. (1990). Structural model evaluation and modification: An interval estimation approach. Multivariate Behavioral Research, 25(2), 173–180. Tang, C. S.-k., & Lau, B. H.-b. (1995). The assessment of gender role stress for Chinese. Sex Roles, 33(7–8), 587. Taris, T. (2000). A primer in longitudinal data analysis. London: SAGE. Tomás, J. M., & Oliver, A. (1999). Rosenberg’s self-esteem scale: Two factors or method effects. Structural Equation Modeling, 6(1), 84–98. Triandis, H. C. (1995). Individualism & collectivism. Boulder: Westview Press. Tsui, M., & Rich, L. (2002). The only child and educational opportunity for girls in urban China. Gender Society, 16(1), 74–92. Tucker, L. R., & Lewis, C. (1973). A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 38(1), 1–10.
551
Twenge, J. M., & Campbell, W. K. (2002). Self-esteem and socioeconomic status: A meta-analytic review. Personality and Social Psychology Review, 6(1), 59–71. Vecchio, G. M., Gerbino, M., Pastorelli, C., Del Bove, G., & Caprara, G. V. (2007). Multifaceted self-efficacy beliefs as predictors of life satisfaction in late adolescence. Personality and Individual Differences, 43(7), 1807–1818. Wu, D. Y. H. (1996). Chinese childhood socialization. In M. H. Bond (Ed.), The handbook of chinese psychology (pp. 143–154). Hong Kong: Oxford University Press. Ye, S. (2009). Factor structure of the general health questionnaire (GHQ-12): The role of wording effect. Personality and Individual Differences, 46(2), 197–201. Yu, L., & Xie, D. (2008). The relationship between desirable and undesirable gender role traits, and their implications for psychological well-being in Chinese culture. Personality and Individual Differences, 44(7), 1517–1527. Zhang, L., & Leung, J.-P. (2002). Moderating effects of gender and age on the relationship between self-esteem and life satisfaction in mainland Chinese. International Journal of Psychology, 37(2), 83–91.
Contents lists available at SciVerse ScienceDirect
Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid
A cross-lagged model of self-esteem and life satisfaction: Gender differences among Chinese university students Shengquan Ye a,⇑, Lu Yu b, Kin-Kit Li a a b
Department of Applied Social Studies, City University of Hong Kong, Hong Kong Department of Applied Social Sciences, Hong Kong Polytechnic University, Hong Kong
a r t i c l e
i n f o
Article history: Received 25 June 2011 Received in revised form 17 November 2011 Accepted 22 November 2011 Available online 24 December 2011 Keywords: Life satisfaction Self-esteem Cross-lagged model Gender difference
a b s t r a c t Past research has shown a close relationship between self-esteem (SE) and life satisfaction (LS). Using cross-lagged structural equation modeling, the current longitudinal study examined the directionality of the relationship among a sample of Chinese university students (214 males and 134 females). In the model, SE and LS at Time 1 were used to predict SE and LS at Time 2, with socioeconomic status (SES) as the independent variable of both constructs. Results showed that SE consistently predicted subsequent LS among both genders, whereas no significant effects of LS on subsequent SE were found. Furthermore, SES had significant effects on LS among both genders. Its influence on SE, however, was significant among females only. Implications and limitations are discussed. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Life satisfaction (LS) is the cognitive component of subjective well-being (SWB), which reflects an individual’s overall evaluation about his or her quality of life (Diener & Diener, 1995). Unlike the emotional components of SWB (pleasant and unpleasant affective experience) that can be influenced by transient mood, life satisfaction is a more enduring assessment of one’s life and therefore is commonly used as the indicator of one’s subjective well-being (Diener, Suh, Lucas, & Smith, 1999). Consequently, the construct of LS and its correlates have attracted much research attention in recent decades (Eid & Larsen, 2008). Existing literature has documented that high LS is associated with less psychological problems as well as positive attributes and outcomes (e.g., Diener, 1994; Diener, Suh, & Oishi, 1997). Among the constructs, self-esteem (SE) has been found to be a robust correlate of LS for people across cultures and gender (e.g., Chen, Cheung, Bond, & Leung, 2006; Diener & Diener, 1995). For instance, in a large cross-national study by Diener and Diener (1995), SE and LS were significantly correlated at r = .47. Moreover, the association was stronger among people in individualistic cultures than those in collectivistic cultures. Diener and Diener explained that individualists were typically socialized to attend to internal over external attributes and thus tend to evaluate lives based on personal attributes (such as SE); while collectivists were often socialized to value social norms and make judgment about lives ⇑ Corresponding author. Tel.: +852 3442 8114; fax: +852 3442 0283. E-mail address: [email protected] (S. Ye). 0191-8869/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.paid.2011.11.018
according to external sources (such as relationships with others). The finding is congruent with self-construal theory (Hofstede, 1980; Markus & Kitayama, 1991; Triandis, 1995), which argues that people in collectivistic cultures tend to develop an interdependent self-construal, whereas people in individualistic cultures usually adopt an independent self-construal. The framework can also be used to explain gender differences in the association between SE and LS. Most cultures encourage males to develop an independent self-construal and females to form an interdependent self-construal (Cross & Madson, 1997; Markus & Oyserman, 1989). Therefore, when people make life evaluations, males (like individualists) would put more weight on internal sources and females would emphasize more on external sources. Consistent with this prediction, Reid’s (2004) study among 206 American college students showed that self-esteem predicted men’s well-being to a larger extent than it predicted women’s well-being, whereas relationship harmony had a significant effect on well-being among women but not men. Similar gender difference can also be found in research on self-efficacy and life satisfaction (e.g., Vecchio, Gerbino, Pastorelli, Del Bove, & Caprara, 2007). Though previous studies have established clear association between SE and LS and provided sound explanation for cultural as well as gender differences in the association, little is known about the directionality of the relationship. In their comprehensive review of self-esteem literature, Baumeister, Campbell, Krueger, and Vohs (2003) acknowledged that previous research had not clearly established causation of SE and happiness, although people often assume or believe that higher SE leads to greater happiness. The authors cautioned that intervention based on incorrect
S. Ye et al. / Personality and Individual Differences 52 (2012) 546–551
assumption of causality could bring about undesirable effects. They explained, with the relationship between SE and academic performance as an example, if self-esteem is not the cause but effect of academic performance, intervention aiming at enhancing self-esteem would not benefit academic performance. On the contrary, it can harm academic performance because students may pay less effort when they feel more satisfied with themselves. Similarly, it is also dangerous to assume that SE causes LS and disregard the other possibility before clear evidence is obtained from empirical studies. The present study endeavors to contribute to the current understanding in the following ways. First, the study is to test the directionality of the relationship between SE and LS. Although researchers often regard happiness as the outcome, recent empirical studies have shown happiness can be the cause as well. For instance, it was found that happiness could lead to various successes (Lyubomirsky, King, & Diener, 2005) and positive changes in people’s behaviors (Salmela-Aro & Tuominen-Soini, 2010). To address the issue, this study applies longitudinal design and analyzes data with cross-lagged structural equation modeling, which is to be described later in details (see also Finkel, 1995; Taris, 2000). Furthermore, the study examines the relationship in a Chinese context, which is regarded as a typical collectivistic culture (Oyserman, Coon, & Kemmelmeier, 2002). As the majority of previous studies were conducted among individualistic cultures, findings from the study will add useful information to the current literature and facilitate future systematic examination (e.g., meta-analysis) of the phenomenon. Second, the study attempts to investigate the differences and similarities of the directional relationship between SE and LS across gender. Researchers have argued that in collectivistic cultures, where social norms and relationship harmony are highly valued, both males and females tend to form an interdependent self-construal (Markus & Kitayama, 1991). Therefore, it is possible that gender differences in the relationship between SE and LS reported in previous Western studies may not be pronounced for people in collectivistic cultures. So far, only one study has examined the moderating effect of gender on the SE–LS relation among Chinese populations (Zhang & Leung, 2002) and reported a nonsignificant gender effect. More evidence on this issue is needed in order to fully understand the moderating effect of gender on the SE–LS relation in different cultures. Third, when testing the directionality of the relationship between variables, researchers need to take into account the effect of the common cause. Previous work has often shown that higher social economic status (SES) is associated with both higher levels of SWB (Kousha & Mohseni, 1997) and self-esteem (Orth, Trzesniewski, & Robins, 2010; Twenge & Campbell, 2002). In order to disentangle the relationship between SE and LS, the effect of SES must be taken into account carefully. Current literature has shown that the effect of SES varies across countries as well as gender. For example, SES was more strongly correlated with LS in poorer countries than in wealthy countries where people have satisfied their basic needs (Diener & Oishi, 2000; Oishi, Diener, Lucas, & Suh, 1999). Several studies have shown that the prediction of SES on SWB is stronger for men than for women (Adelmann, 1987; Aryee, 1999), due to the different gender roles. Because of males’ traditional role as household breadwinners and females’ as caregivers, men are likely to find more life satisfaction in their achievements in occupation-related fields whereas women would derive more happiness from relationships and family. Considering the rapid changes of social and economic systems in modern China, one may wonder to what extent these findings could be applied to Chinese context to help us to understand the relationship between SE and LS among males and females in China. As numerous recent studies have shown, contemporary Chinese people’s conceptuali-
547
zation of masculinity and femininity, gender role adoption, and attitudes toward women’s social roles are different from those of Western people (e.g., Chang, 1999; Tang & Lau, 1995; Wu, 1996; Yu & Xie, 2008). For example, Wu (1996) reported that parents of only children believed that boys and girls should enjoy equal levels of education, dramatically different from the traditional genderdifferentiated expectations about educational achievement. These characteristics of Chinese society might have important implications on the way Chinese young people evaluate their lives as well as themselves. Therefore, it would be interesting to examine whether the relationship of SES with LS and SE would be different for Chinese men and women nowadays. In sum, the study endeavors to investigate three main questions among a Chinese sample. First, the study aims to clarify the directional relationship between LS and SE. Second, the study attempts to explore gender differences in the relationship. Finally, the role of SES in the relationship between LS and SE is examined. 2. Method 2.1. Participants and procedure Three hundred and forty-eight students (214 males and 134 females) from a large university in Mainland China participated in the study. The students were from a variety of science and art disciplines including history, philosophy, law, information technology, and physics. The age range was from 16 to 24 years (M = 20.00, SD = 0.99). The inventories were administered to the participants on two occasions with an interval of 8 months. The scales were translated from English into Chinese, with back-translation procedure to ensure accuracy and equivalency. Students were told that data were collected for research purposes, that participation was completely voluntary, and that confidentiality would be assured. Small gifts were given to the students as an incentive of participation. 2.2. Measures 2.2.1. Life satisfaction The Satisfaction With Life Scale (SWLS; Diener, Emmons, Larsen, & Griffin, 1985) is a five-item scale that assesses individuals’ life satisfaction, the cognitive component of subjective well-being. The SWLS typically uses a 7-point response format (ranging from not at all true of me to exactly true of me). A higher scale score indicates a higher level of life satisfaction. Since it was developed, the SWLS has been widely used in measuring general life satisfaction and has shown favorable psychometric properties in various languages and samples including Chinese (e.g., Abdallah, 1998; Neto, 1993; Sachs, 2004). For the current study, the Cronbach’s alpha coefficients were .77 and .81 for the two measurement occasions, respectively. 2.2.2. Self-esteem The Rosenberg Self-Esteem Scale (RSES; Rosenberg, 1989) is a 10-item scale that has been widely used for assessing general self-esteem. For consistency, this study applied a 7-point Likert scale (ranging from strongly disagree to strongly agree). The RSES consists of five positively and five negatively worded items. Past research has shown satisfactory internal consistency of the scale among Chinese samples (e.g., Chen et al., 2006) and provided evidence for the cross-cultural invariance of the construct (e.g., Russell, Crockett, Shen, & Lee, 2008). However, recent psychometric studies using various scales and languages have generally suggested that including negatively worded items, although helping to reduce response set, could complicate the factor structure by
548
S. Ye et al. / Personality and Individual Differences 52 (2012) 546–551
producing a method factor (e.g., Marsh, 1996; Tomás & Oliver, 1999; Ye, 2009). Following Marsh’s suggestion (1996), negatively worded items were used in the questionnaire but not calculated in forming the sum score. The internal consistency of the five positive items was found to be satisfactory (i.e., alpha coefficients were .80 and .83 for Time 1 and Time 2, respectively). 2.2.3. Socioeconomic status (SES) To assess SES, four related demographic variables were measured: origin, family income, paternal education, and maternal education. Origin refers to the place that the students had been living before being admitted to the university. Three categories provided were rural area, suburb, and urban city (with the latter indicating higher SES). Family income was rated on a 6-point Likert scale, ranging from very poor to very good. Parental education was rated on a 5-point Likert scale, ranging from primary school to postgraduate. The four variables were significantly correlated (r = .38–.71) and used as indicators of a latent SES variable in subsequent structural equation modeling. Since the four indicators were rather stable over the period of investigation, only information collected at Time 1 was used in the data analysis. 2.3. Data analysis With SPSS 15.0, descriptive statistics and mean difference test were first performed. LISREL 8.80 (Jöreskog & Sörbom, 2004) was then employed to examine the structural models and test their goodness of fit. Estimation method for the models was chosen based on the multivariate normality of the items. Maximum Likelihood (ML) estimation would be used when multivariate normality is confirmed, whereas Robust Maximum Likelihood (RML) estimation would be applied when the assumption is rejected. Goodness of model fit was assessed with several indices. As suggested by Hu and Bentler (1999), RMSEA (Steiger, 1990) 6.06 and SRMR (Bentler, 1995) 6.08 indicate a good model fit, while values of .95 or above for TLI (also known as NNFI; Tucker & Lewis, 1973) and CFI (Bentler, 1990) are required for an acceptable fit. Moreover, NFI (Bentler & Bonett, 1980) with values greater than .90 also indicate the model fit is acceptable. 3. Results The mean, standard deviation, skewness, kurtosis, and range of the sum scores of life satisfaction (LS) and self-esteem (SE) are presented in Table 1. As is shown, the sum scores were overall normally distributed. The standard deviations for the scores of LS and SE were slightly but consistently smaller among females than males, suggesting that female students were more homogeneous in their life satisfaction and self-esteem. To test the mean differences in SE and LS, repeated-measures ANCOVA was conducted, with time as the within-subjects factor, gender as the
Table 1 Descriptive statistics of scale sum scores.
Mean SD Skewness Kurtosis Minimum Maximum
LS1m
LS1f
LS2m
LS2f
SE1m
SE1f
SE2m
SE2f
19.85 5.70 .04 .28 8 35
22.26 5.01 .07 .26 9 34
20.32 5.56 .17 .12 5 35
22.07 4.74 .01 .01 10 35
26.90 4.97 .52 .03 11 35
27.64 4.38 .33 .57 17 35
25.45 5.18 .28 .34 10 35
27.17 4.36 .10 .69 17 35
LS1, life satisfaction at Time 1; LS2, life satisfaction at Time 2; SE1, self-esteem at Time 1; SE2, self-esteem at Time 2. m Male. f Female.
between-subjects factor, and SES as the covariate. Results showed that SE decreased across the period (p < .01, Partial g2 = .038), females had higher LS (p < .05, Partial g2 = .019) and SE (p < .05, Partial g2 = .013) than males, and SES had a significant positive effect on LS (p < .01, Partial g2 = .123) and SE (p < .01, Partial g2 = .013). To test the directional relationship between LS and SE, a crosslagged model was specified, in which LS and SE at Time 2 were predicted by LS and SE at Time 1 and all of these four latent variables were predicted by SES, as demonstrated in Fig. 1 and Fig. 2 (only the structural models are presented for clarity). The model was estimated among male and female samples separately to examine possible gender differences in the relationship (i.e., the moderating effect of gender). Prior to model fitting, multivariate normality of the data should be examined in order to choose an appropriate estimation method. With the PRELIS procedure in LISREL program, it was found that the data did not follow multivariate normal distribution and thus failed to meet the assumption required for the widely used estimation method, maximum likelihood (ML). Therefore, asymptotic covariance matrices were calculated and robust maximum likelihood (RML) was used to estimate the cross-lagged structural equation model. The model provided acceptable fit to the data among both genders (for males: v2(df = 239, n = 214) = 471.59, RMSEA = .068 (90% CI = .059–.077), SRMR = .084, NNFI = .95, and CFI = .95; for females: v2(df = 239, n = 134) = 343.51, RMSEA = .057 (90% CI = .043–.070), SRMR = .078, NNFI = .97, and CFI = .97). As shown in Fig. 1 and Fig. 2, most parameters were quite comparable across genders, apart from a couple of exceptions. Specifically, SES had significant effects on SE among female (i.e., the standardized coefficients were .17 and .14 at T1 and T2, respectively, p < .05) but not male students. In addition, though the path coefficients from LS1 to SE2 were non-significant in both models, the effect among males was somewhat stronger and approached the significance level (i.e., b = .12, t = 1.56). 4. Discussion The present study examined the directionality of the relationship between SE and LS and the roles of gender and SES in the relationship. The findings are organized and discussed in the following three sections: (a) gender differences in SE and LS; (b) gender differences in SE–LS relationship, and (c) gender differences in SES–SE relationship. 4.1. Gender differences in SE and LS For both males and females, their LS remained relatively stable across the period, whereas SE dropped significantly. Previous work has shown that transition to college brings a lot of changes in various domains such as financial situation (Goldrick-Rab & Han, 2011), interpersonal relationship (Kenyon & Koerner, 2009), and life style (Abar & Maggs, 2010). The process of coping with these challenges can be stressful and detrimental to one’s self-esteem. However, it should be noted that, due to the relatively short period being investigated in the present study, it remained uncertain whether the decrease could be understood as a continuous trend throughout the whole university life. Future research covering a longer period will help to clarify the picture. Aside from the similar decrease in SE, significant gender differences were also found. Compared with male students, female students were higher in both SE and LS over the investigated period of time. This pattern deserves a closer examination, considering the dynamics of traditional Chinese culture and the rapidly changing socioeconomic environment. Regarding SE, a meta-analytical study with 216 effect sizes and 97,121 respondents showed that men
S. Ye et al. / Personality and Individual Differences 52 (2012) 546–551
549
Fig. 1. Structural cross-lagged model for male students. Note. LS1 = life satisfaction at Time 1; LS2 = life satisfaction at Time 2; SE1 = self-esteem at Time 1; SE2 = self-esteem at Time 2; SES = socioeconomic status; n.s. = non-significant.
Fig. 2. Structural cross-lagged model for female students. Note. LS1 = life satisfaction at Time 1; LS2 = life satisfaction at Time 2; SE1 = self-esteem at Time 1; SE2 = selfesteem at Time 2; SES = socioeconomic status; n.s. = non-significant.
scored slightly higher on SE than women (Kling, Hyde, Showers, & Buswell, 1999). Most included studies were from the United States, Australia, and Canada. However, a recent study in China showed that SE was higher among female than male students (Chen & Zhang, 2004). Like SE, similar cross-cultural differences in LS were observed. A cross-cultural study with 17,246 students in 21 countries showed men were more satisfied with life than women in two regions including (1) Western Europe and the United States and (2) Central and Eastern Europe, whereas women were more satisfied with life than men in Pacific Asia (Grant, Wardle, & Steptoe, 2009). In addition, Dorahy and colleagues (1996) found that LS was higher among Indian women than Indian men. The current findings were consistent with the results found in those collectivistic cultures. Since SE and LS shared many common factors theoretically, the gender differences in SE and LS might follow similar mechanisms. The differential gender differences across cultures may be due to the rapid changes in educational opportunities for women in China. Because of the one-child policy, urban Chinese men and women enjoy the same educational opportunities (Tsui & Rich, 2002). This sample of young women might be the first generation of women in their families to receive college education. The contrast
between themselves and the older generations of women may be particularly empowering, which boosts their SE and LS. In addition, female students who grew up in the rural areas (more likely to have siblings) might have overcome many challenges and inequalities to be admitted to college. Higher SE and LS might provide them with the competitive advantages to succeed (Joshi & Srivastava, 2009) or could be the results of successes amid the challenges (Ross & Broh, 2000). 4.2. Gender differences in the SE–LS relationship Regarding the directionality of the SE–LS relationship, this study provided clear evidence that higher SE could lead to higher LS over time. Furthermore, it was observed that the effects of SE on subsequent LS were very similar across gender, suggesting that the evaluation about oneself is an important source of life satisfaction for both males and females. The current study, which aimed to examine whether the moderating effect of gender on the SE–LS relation (Reid, 2004) could be demonstrated in a Chinese population, failed to replicate Reid’s finding that SE predicted LS more strongly in men than women. This may be due to more independent self-construal (i.e., others are separated from the
550
S. Ye et al. / Personality and Individual Differences 52 (2012) 546–551
representation of the self) among Chinese women nowadays or the increased egalitarianism of gender roles in modern Chinese societies (e.g., Chang, 1999). As for the effect of LS on subsequent SE, the coefficients were found to be non-significant among males and females. However, it is noteworthy that the non-significant effect among male students might result from the lack of statistical power considering its effect size (b = .12). Being content and satisfied with life may be a source of SE among male college students. For those with a high level of internal locus of control, life dissatisfaction may reflect personal inadequacy, which may dampen SE. Different from the ascribed SES, LS perhaps more directly holds the person responsible. The traditional dominant role for men in Chinese societies may put men in a difficult position. The expectation for men is still high, while privileges and circumstances that lead to success have become more equal across sexes nowadays. 4.3. Gender differences in the SES–SE relationship In a meta-analysis with 446 samples and 312,940 individuals, Twenge and Campbell (2002) showed that the relationship between SES and SE was small but comparable to the effects of gender and race. The effect size of SES for women became larger than for men since 1940 and kept increasing in subsequent decades. The current results were consistent with this pattern, in which the SES effect was small among college students and only significant among female students. SES is ascribed from the parents for most of the college students; hence it may have minimal effects as a social indictor that boosts SE (Rosenberg & Pearlin, 1978). Because of the labor-force movement for women and the relinquished role as the sole breadwinner for men, Twenge and Campbell suggested that the effect of SES on SE may increase for women and decrease for men over generations. There are a few limitations for this study. First, participants were recruited from one university in Mainland China. Generalizability of the findings is limited and future replications with more representative samples are recommended. Second, the sample size of this study was less than ideal. Only moderate effect sizes could be detected statistically. Third, due to some practical constraints, the data were collected at two time points with a relatively short interval. With more waves over a longer period, researchers will be able to capture a more comprehensive picture and test the relationship in a more sophisticated way. Finally, as an exploratory study, only measures of the key constructs were included. Future studies should examine these relationships with additional explanatory variables such as self-construal and relationship harmony (e.g., Kwan, Bond, & Singelis, 1997; Reid, 2004). References Abar, C. C., & Maggs, J. L. (2010). Social influence and selection processes as predictors of normative perceptions and alcohol use across the transition to college. Journal of College Student Development, 51(5), 496–508. Abdallah, T. (1998). The Satisfaction with Life Scale (SWLS): Psychometric properties in an Arabic-speaking sample. International Journal of Adolescence & Youth, 7(2), 113–119. Adelmann, P. K. (1987). Occupational complexity, control, and personal income: Their relation to psychological well-being in men and women. Journal of Applied Psychology, 72(4), 529–537. Aryee, S. (1999). An examination of the moderating influence of breadwinner role salience on the pay-life satisfaction relationship. Human Relations, 52(10), 1279–1290. Baumeister, R. F., Campbell, J. D., Krueger, J. I., & Vohs, K. D. (2003). Does high selfesteem cause better performance, interpersonal success, happiness, or healthier lifestyles. Psychological Science in the Public Interest, 4(1), 1–44. Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107(2), 238–246. Bentler, P. M. (1995). EQS structural equations program manual. Encino, CA: Multivariate Software. Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness of fit in the analysis of covariance structures. Psychological Bulletin, 88(3), 588–606.
Chang, L. (1999). Gender role egalitarian attitudes in Beijing, Hong Kong, Florida, and Michigan. Journal of Cross-Cultural Psychology, 30(6), 722–741. Chen, S. X., Cheung, F. M., Bond, M. H., & Leung, J.-P. (2006). Going beyond selfesteem to predict life satisfaction: The Chinese case. Asian Journal of Social Psychology, 9(1), 24–35. Chen, L., & Zhang, J. (2004). General life satisfaction and self-esteem of university students. Chinese Mental Health Journal, 18(4), 222–224. Cross, S. E., & Madson, L. (1997). Models of the self: Self-construals and gender. Psychological Bulletin, 122(1), 5–37. Diener, E. (1994). Assessing subjective well-being: Progress and opportunities. Social Indicators Research, 31(2), 103–157. Diener, E., & Diener, M. (1995). Cross-cultural correlates of life satisfaction and selfesteem. Journal of Personality and Social Psychology, 68(4), 653–663. Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction with life scale. Journal of Personality Assessment, 49(1), 71–75. Diener, E., & Oishi, S. (2000). Money and happiness: Income and subjective wellbeing across nations. In E. Diener & E. M. Suh (Eds.), Culture and subjective wellbeing (pp. 185–218). Cambridge, MA: The MIT Press. Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being: Three decades of progress. Psychological Bulletin, 125(2), 276–302. Diener, E., Suh, E., & Oishi, S. (1997). Recent findings on subjective well-being. Indian Journal of Clinical Psychology, 24(1), 25–41. Dorahy, M. J., Schumaker, J. F., Simpson, P. L., & Deshpande, C. G. (1996). Depression and life satisfaction in India and Australia. Journal of Personality and Clinical Studies, 12(1–2), 1–7. Eid, M., & Larsen, R. J. (2008). The science of subjective well-being. New York, NY, US: Guilford Press. Finkel, S. E. (1995). Causal analysis with panel data. Thousand Oaks: Sage. Goldrick-Rab, S., & Han, S. W. (2011). Accounting for socioeconomic differences in delaying the transition to college. Review of Higher Education: Journal of the Association for the Study of Higher Education, 34(3), 423–445. Grant, N., Wardle, J., & Steptoe, A. (2009). The relationship between life satisfaction and health behavior: A cross-cultural analysis of young adults. International Journal of Behavioral Medicine, 16, 259–268. Hofstede, G. (1980). Culture and organizations. International Studies of Management & Organization, 10(4), 15. Hu, L.-t., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55. Jöreskog, K. G., & Sörbom, D. (2004). LISREL 8.70. Chicago: Scientific Software. Joshi, S., & Srivastava, R. (2009). Self-esteem and academic achievement. Social Science International, 25(1), 75–87. Kenyon, D. Y. B., & Koerner, S. S. (2009). College student psychological well-being during the transition to college: Examining individuation from parents. College Student Journal, 43(4, PtA), 1145–1160. Kling, K. C., Hyde, J. S., Showers, C. J., & Buswell, B. N. (1999). Gender differences in self-esteem: A meta-analysis. Psychological Bulletin, 125(4), 470–500. Kousha, M., & Mohseni, N. (1997). Predictors of life satisfaction among urban Iranian women: An exploratory analysis. Social Indicators Research, 40(3), 329–357. Kwan, V. S. Y., Bond, M. H., & Singelis, T. M. (1997). Pancultural explanations for life satisfaction: Adding relationship harmony to self-esteem. Journal of Personality and Social Psychology, 73(5), 1038–1051. Lyubomirsky, S., King, L., & Diener, E. (2005). The benefits of frequent positive affect: Does happiness lead to success? Psychological Bulletin, 131(6), 803–855. Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98(2), 224. Markus, H. R., & Oyserman, D. (1989). Gender and thought: The role of the selfconcept. In M. Crawford & M. Gentry (Eds.), Gender and thought: Psychological perspectives (pp. 100–127). New York, NY: SpringerVerlag New York Inc. Marsh, H. W. (1996). Positive and negative global self-esteem: A substantively meaningful distinction or artifactors? Journal of Personality and Social Psychology, 70(4), 810–819. Neto, F. (1993). The Satisfaction With Life Scale: Psychometrics properties in an adolescent sample. Journal of Youth & Adolescence, 22(2), 125–134. Oishi, S., Diener, E. F., Lucas, R. E., & Suh, E. M. (1999). Cross-cultural variations in predictors of life satisfaction: Perspectives from needs and values. Personality and Social Psychology Bulletin, 25(8), 980–990. Orth, U., Trzesniewski, K. H., & Robins, R. W. (2010). Self-esteem development from young adulthood to old age: A cohort-sequential longitudinal study. Journal of Personality and Social Psychology, 98(4), 645–658. Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and collectivism: Evaluation of theoretical assumptions and meta-analyses. Psychological Bulletin, 128(1), 3–72. Reid, A. (2004). Gender and sources of subjective well-being. Sex Roles, 51(11–12), 617–629. Rosenberg, M. (1989). Society and the adolescent self-image (Rev. ed.). Middletown, Conn.: Wesleyan University Press. Rosenberg, M., & Pearlin, L. I. (1978). Social class and self-esteem among children and adults. American Journal of Sociology, 84(1), 53–77. Ross, C. E., & Broh, B. A. (2000). The role of self-esteem and the sense of personal control in the academic achievement process. Sociology of Education, 73(4), 270–284. Russell, S. T., Crockett, L. J., Shen, Y.-L., & Lee, S.-A. (2008). Cross-ethnic invariance of self-esteem and depression measures for Chinese, Filipino, and European American adolescents. Journal of Youth and Adolescence, 37(1), 50–61.
S. Ye et al. / Personality and Individual Differences 52 (2012) 546–551 Sachs, J. (2004). Validation of the Satisfaction With Life Scale in a sample of Hong Kong university students. Psychologia, 46(4), 225–234. Salmela-Aro, K., & Tuominen-Soini, H. (2010). Adolescents’ life satisfaction during the transition to post-comprehensive education: Antecedents and consequences. Journal of Happiness Studies, 11(6), 683–701. Steiger, J. H. (1990). Structural model evaluation and modification: An interval estimation approach. Multivariate Behavioral Research, 25(2), 173–180. Tang, C. S.-k., & Lau, B. H.-b. (1995). The assessment of gender role stress for Chinese. Sex Roles, 33(7–8), 587. Taris, T. (2000). A primer in longitudinal data analysis. London: SAGE. Tomás, J. M., & Oliver, A. (1999). Rosenberg’s self-esteem scale: Two factors or method effects. Structural Equation Modeling, 6(1), 84–98. Triandis, H. C. (1995). Individualism & collectivism. Boulder: Westview Press. Tsui, M., & Rich, L. (2002). The only child and educational opportunity for girls in urban China. Gender Society, 16(1), 74–92. Tucker, L. R., & Lewis, C. (1973). A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 38(1), 1–10.
551
Twenge, J. M., & Campbell, W. K. (2002). Self-esteem and socioeconomic status: A meta-analytic review. Personality and Social Psychology Review, 6(1), 59–71. Vecchio, G. M., Gerbino, M., Pastorelli, C., Del Bove, G., & Caprara, G. V. (2007). Multifaceted self-efficacy beliefs as predictors of life satisfaction in late adolescence. Personality and Individual Differences, 43(7), 1807–1818. Wu, D. Y. H. (1996). Chinese childhood socialization. In M. H. Bond (Ed.), The handbook of chinese psychology (pp. 143–154). Hong Kong: Oxford University Press. Ye, S. (2009). Factor structure of the general health questionnaire (GHQ-12): The role of wording effect. Personality and Individual Differences, 46(2), 197–201. Yu, L., & Xie, D. (2008). The relationship between desirable and undesirable gender role traits, and their implications for psychological well-being in Chinese culture. Personality and Individual Differences, 44(7), 1517–1527. Zhang, L., & Leung, J.-P. (2002). Moderating effects of gender and age on the relationship between self-esteem and life satisfaction in mainland Chinese. International Journal of Psychology, 37(2), 83–91.