|Home | About | Journals | Submit | Contact Us | Français|
This study explored a potential pathway by which perceived discrimination may affect levels of depressive symptoms in a sample of 472 Korean American older adults (M age= 69.9, SD = 7.04). Building upon previous studies demonstrating that perceived discrimination has negative impacts on mental health, we hypothesized that sense of control would mediate the associations between perceived discrimination and depressive symptoms. Our analyses showed that the effects of perceived discrimination on depressive symptoms were not only direct but also mediated through a lowered sense of control. Finding that sense of control serves as an intervening step between perceived discrimination and mental health may help explicate the psychological mechanisms involved in responses to discriminatory experience and has implications for intervention strategies.
The experience of being treated unfairly or discriminated against is often encountered in daily life. According to the national survey from the Midlife Development in the United States (MIDUS), more than 60% of the U.S. population reports experiences of discrimination in their daily life (Kessler, Mickelson, & Williams, 1999). Whether based on race, gender, sexual orientation, or other characteristics, discriminatory experiences may be detrimental to both physical health (e.g., Barnes et al., 2008; Hunte & Williams, 2009; Lewis et al., 2009) and mental well-being (e.g., Barnes et al., 2004; Hwang & Goto, 2008; Kessler et al., 1999; Noh, Beiser, Kaspar, Hou, & Rummens, 1999; Williams, Yu, Jackson, & Anderson, 1997). In particular, members of racial/ethnic minority groups are not only more likely to experience discrimination, but such discrimination represents a major stressor in their lives (Williams & Mohammed, 2009; Williams, Neighbors, & Jackson, 2003). For example, in studies with Korean immigrants in the United States (e.g., Bernstein, Park, Shin, Cho, & Park, 2009) and Canada (e.g., Noh, Kaspar, & Wickrama, 2007), a strong relationship between perceived discrimination and depressive symptoms has been reported. However, there remains a paucity of research examining factors that might intervene the adverse effects of perceived discrimination on mental health.
In the present study, we hypothesized that sense of control would act as a mediator in the relations between perceived discrimination and depressive symptoms in Korean American older adults. Sense of control is defined as the extent to which individuals think that they have personal power and control over their life and environment (e.g., Pearlin & Schooler, 1978; Zarit, Pearlin, & Schaie, 2003). The significant role of sense of control in predicting overall health and well-being of Korean immigrants in North America has been previously reported (e.g., Jang, Kim, & Chiriboga, 2006; Noh & Avison, 1996). Although typically treated as a personal disposition or attribute, it is strongly influenced by social and interpersonal relationships (Zarit et al., 2003). The experience of disrespect and discrimination in interactions with others may erode feelings of control, which in turn may contribute to lowering one’s overall mental health status. A few studies have suggested that feelings of control might serve as an intervening agent in responding to discriminatory encounters (e.g., Branscombe & Ellemers, 1998; Jang, Chiriboga, & Small, 2008; Moradi & Hasan, 2004; Ruggiero & Taylor, 1997).
Given the need for more empirical research on perceived discrimination and its psychological responses, we hypothesized that perceived discrimination would not only have direct effects on depressive symptoms, but also its impact would be mediated through a lowering of feelings of control. Exploring the mediating role of sense of control may advance our understanding of psychological mechanisms underlying discriminatory experience and suggest ways to facilitate intervention efforts.
After obtaining approval from the university’s Institutional Review Board, a survey of older Korean Americans (≥60 years old) was conducted between October 2005 and May 2006 in Tampa and Orlando, Florida. Because ethnic immigrant populations are hard to identify by any single approach, we used a multiple sampling strategy. Data were collected from 472 Korean Americans older adults through a combination of mail surveys and visits to churches, senior centers, senior housing facilities, and community groups. The survey questionnaires, which were in Korean, were developed through a translation/back-translation process that also included pilot-testing with 20 Korean American older adults who were representative of the anticipated sample. Most of the instruments used in the study had been used in previous research and had good psychometric properties. Detailed information on the sampling procedures and validation of the multimethod recruitment strategy is available elsewhere (e.g., Jang, Kim, Hansen, & Chiriboga, 2007). None of the participants had missing information on more than 10% of the variables surveyed, and all were included in the analyses.
A 9-item instrument (Williams et al., 1997) assessed the frequency of maltreatment or disrespects by others in daily life. The scale is not designed to measure specific types of discrimination such as ageism, sexism, or racism. Instead the instrument includes statements such as “You are treated with less courtesy than other people,” “You are treated with less respect than other people,” and “People act as if they think you are dishonest.” Participants were asked to indicate their response with a 4-point scale ranging from 1 (never) to 4 (often). Total scores could range from 9 to 36, with higher scores indicating more frequent experiences of discrimination. Bernstein and colleagues (2009) used this scale with Korean immigrants in New York and found high internal consistency. The present sample also showed a good internal consistency (α = 0.97).
Perceived control was measured with Pearlin’s 7-item Mastery Scale (Pearlin & Schooler, 1978). Respondents described their feelings about each item (e.g., “I cannot solve my problems” and “My future mostly depends on me”) on a 4-point scale ranging from 1 (strongly disagree) to 4 (strongly agree). Responses to negatively worded items were reverse-coded. Summary scores ranged from 7 (low sense of control) to 28 (high sense of control). The scale has been used with Korean American older adults and has shown good psychometric properties (e.g., Jang et al., 2006). Internal consistency for the scale in the present sample was satisfactory (α = 0.80).
A 10-item short form of the Center for Epidemiologic Studies-Depression Scale (CES-D; Andresen, Malmgren, Carter, & Patrick, 1994; Radloff, 1977) was included to assess depressive symptoms. On a 4-point scale, respondents rated how often symptoms, such as loneliness, feelings of fearfulness, and restless sleep, were experienced during the past week. Total scores could range from 0 to 30, with higher scores indicating greater levels of depressive symptoms. The CES-D has been translated into the Korean language, and its psychometric properties have been validated in previous studies with Koreans in Korea (e.g., Cho, Nam, & Suh, 1998) and Korean Americans (e.g., Jang, Kim, & Chiriboga, 2005). Internal consistency in the present sample was good (α = 0.80).
Other variables included age (in years), gender (0 = male, 1 = female), marital status (0 = not married, 1 = married), education (0 = <high school, 1 = ≥ high school), and length of stay in the United States (in years).
In addition to descriptive and correlational analyses, hierarchical regression models of depressive symptoms were tested by sequentially entering the following independent blocks of predictors: (a) demographic and background information (age, gender, marital status, education, and length of stay in the U.S.), (b) perceived discrimination, and (c) sense of control. Intercorrelations among study variables and variance inflation factors (VIF) were assessed to determine collinearity.
For the main research question, the mediating effect of sense of control was tested with a series of regression analyses following Baron and Kenny’s criteria for mediation (1986): (a) there is a significant association between the independent and dependent variables, (b) there is a significant association between the independent variable and presumed mediator, (c) there is a significant association between presumed mediator and dependent variable, and (d) the previously significant relationship between independent and dependent variables become non-significant or decreased when the mediator is controlled for. The Sobel test was used to assess the significance of any reduction in regression coefficients resulting from mediation (MacKinnon & Dwyer, 1993).
As shown in Table 1, the sample consisted of 472 Korean Americans ranging in age from 60 to 94 with an average age of 69.9 years (SD = 7.04). More than half (57.9%) of the sample was female, and three quarters (75.2%) were married. About 64% of the sample had received more than a high school education. The number of years lived in the U.S. ranged from 2 months to 51 years with a mean year of 24.9 (SD = 10.8). The average score of perceived discrimination and sense of control was 12.1 (SD = 5.98) and 19.8 (SD = 4.00), respectively. Depressive symptom scores averaged 7.59 (SD = 4.85) out of the possible 30.
In bivariate analysis (not shown in tabular format), higher levels of depressive symptoms were associated with being female (r = .10, p < .05), being unmarried (r = −.14, p < .01), lower levels of education (r = −.17, p < .001), and a shorter residence in the U.S. (r = −.10, p < .05). Perceived discrimination was negatively associated with sense of control (r = −.39, p < .001) and positively associated with depressive symptoms (r = .12, p < .05). None of the coefficients exceeded 0.50, providing only minimal evidence of collinearity.
Table 2 summarizes the results of a hierarchical regression model in which depressive symptoms were the dependent variable. Collinearity diagnostics were conducted with the Variance Inflation Factor (VIF). The computed values of VIF ranged from 1.09 to 1.36, substantially less than the 4.0 suggested as an indicator of multicollinearity (Montgomery & Peck, 1992). Demographic and background variables explained 4% of the variance of depressive symptoms. Lower levels of education were significantly associated with higher levels of depressive symptoms. Inclusion of perceived discrimination increased the explanatory power of the model by 2%. Individuals who reported higher levels of perceived discrimination were found to have higher levels of depressive symptoms. In the final step, sense of control was added to the model and accounted for an additional 20% of the variance, resulting in a total of 26% of the variance explained by the estimated model (F (7, 414) = 20.3, p < .001). Lower levels of sense of control were identified as a significant risk of depressive symptoms. It is notable that in the final model, perceived discrimination became non-significant with the presence of sense of control.
Following Baron and Kenny’s (1986) criteria for mediation, an initial series of regression analyses were conducted. After controlling for demographic and background variables (age, gender, marital status, education, and length of stay in the U.S.), the regression coefficients of the three paths among the independent variable (perceived discrimination), presumed mediator (sense of control), and dependent variable (depressive symptoms) were all significant (regression tables are not shown). However, as shown in Figure 1, when sense of control was controlled, the direct coefficient of perceived discrimination on depressive symptoms (B = .10, p < .05) became non-significant (B = −.05, p > .05). The Sobel test confirmed that the degree of reduction was statistically significant (z = 6.75, p < .001).
Recognizing that perceived discrimination is an important stressor among racial/ethnic minorities linked to adverse mental health outcomes (e.g., Williams & Mohammed, 2009; Williams et al., 2003), the present study explored complexities in the relationship of discriminatory experiences to depressive symptoms. Our hypothesis was that perceived discrimination would not only have direct effects on depressive symptoms, but also have its impact mediated by the individual’s sense of control. Using a sample of Korean American older adults (N = 472), we found support for the proposed hypotheses.
The average score of perceived discrimination in the present sample (M = 12.1, SD = 5.98) was similar to the score found in the Midlife Development in the United States (MIDUS), a national representative survey of adults (M = 12.4, SD = 4.59; Jang et al., 2008). However, the average score of depressive symptoms was higher in the present sample than usually reported in studies using the short form of the CES-D. When applying the appropriate cut-off score (summated score equal to or greater than 10; Andresen et al., 1994), we found that more 34% of the present sample fell into the category of having probable depression. This figure is over twice as high as the 15% reported in a state-wide representative sample of community-dwelling older Whites in Florida (e.g., Jang, Chiriboga, Kim, & Phillips, 2008); however, it is consistent with findings from previous studies of elderly Korean Americans (e.g., Lee, Moon, & Knight, 2005; Mui & Kang, 2006).
In the multivariate analysis, perceived discrimination was found to be a significant predictor of depressive symptoms. This finding is consistent with the previous literature showing that discriminatory experience poses a critical risk to mental health (e.g., Barnes et al., 2004; Kessler et al., 1999; Noh et al., 1999; Williams et al., 1997). We also found that perceived discrimination was associated with a lower sense of control.
Adding to the literature, the present study conceptualized sense of control as an intervening step between perceived discrimination and depressive symptoms and found statistical support for the proposed mediation model. While our analyses suggested that the experience of discrimination may erode one’s feelings of control which in turn could lead to diminished mental health, the intervening role of sense of control has been suggested in studies with diverse populations (e.g., Branscombe & Ellemers, 1998; Jang et al., 2008; Moradi & Hasan, 2004; Ruggiero & Taylor, 1997). Our study showed that the model fits data with Korean American older adults. Specifically, the mediation model suggests that sense of control may protect the mental health of the racial/ethnic minority elders in the face of discriminatory experiences. These findings suggest that personal feelings of control should be targeted for intervention and counseling programs. This is because, unlike personality traits, sense of control is changeable (e. g., Skaff, Pearlin, & Mullan, 1996), and can serve as an intervention target. Bolstering personal sense of control may be a useful intervention for preventing and alleviating negative psychological consequences related to discriminatory experiences. Interventions should help older minorities perceive themselves as positive and worthy members of society and develop positive self-identity. As shown in previous studies (e.g., Branscombe & Ellemers, 1998; Noh et al., 1999), positive group identity helps individuals cope better with stigma or prejudice against certain groups. At a societal level, attention should be paid to ways to eliminate discrimination. Our findings highlight the importance of advocacy and education for appreciation of diversity and fair social environments.
Some limitations to the present study should be noted. Because the present study was based on a cross-sectional design, caution must be exercised in drawing causal inferences. Because the meditation model basically assumes a temporal order of study variables, future assessment should utilize longitudinal study designs. Also, due to the non-representative nature of the sample, the findings are only suggestive and warrant further investigation. Although our study was designed to explore a basic psychological pathway by which discrimination may affect mental health, future work should incorporate other variables of importance, such as acculturation and ethnic identity. Another limitation was that while we conceptualized the experience of discrimination as a life stressor, perceived stressfulness of discriminatory events was not assessed. Given the importance of subjective appraisals in stress process (e.g., Lazarus & Folkman, 1984), future studies should incorporate the extent of perceived stressfulness from the experience of discrimination. Such information may advance our understanding of individual variations in the experience of discriminatory stress. Also, reliance on self-reported measures adds limitations to the study. Finally, it should be noted that the assessment of discrimination in the present study focused on the general experiences of discrimination without specifying sources or reasons of discrimination. Future studies addressing the specific types of discrimination may help develop a more targeted intervention strategy.
This project was supported by the National Institute on Aging (Grant No. 1R03 AG 26332-01; principal investigator, Yuri Jang, PhD).