Confirming the multidimensionality of psychological well-being among Korean immigrant mothers in the United States

Abstract

Ryff’s Psychological Well-Being (PWB) Scale is a theoretically derived instrument that focuses on six dimensions of eudaimonic well-being. Although the PWB scale has been widely used in both clinical and general samples with different sociodemographic characteristics, its factor structure within first-generation Korean immigrant mothers require examination, as they often face socio-cultural challenges due to immigration (e.g., experiencing acculturation stress, being socially and culturally segregated, following heritage cultural obligations to be a “wise mother and good wife,” taking less prestige jobs, etc.) that could negatively influence their well-being. This study examined the factor structure of an 18-item version of the PWB Scale in a sample of 169 first-generation Korean immigrant mothers of young children in the U.S. A second-order confirmatory factor analysisdid not support the original factor structure of the PWB scale. More importantly, high correlations among some of the PWB subscales suggested that the PWB factors may not be distinct from each other and the number of factors should be reduced. Exploratory factor analysis was conducted on the four highly correlated factors in subsequent analyses and three new factors (Satisfaction with Life, Goal Orientation, and Positive Self-Perceptions) were identified. The newly derived PWB factors were negativelycorrelated with mothers’ depressive symptoms.These findings advanced our understanding of psychological well-being in first-generation Korean immigrant mothersand highlighted the need to consider socio-cultural factors in mental health research on this population.

Psychological well-being is an important construct in the field of psychological science (e.g., Abbott et al. 2006; Burns & Machin 2009), and has been frequently measured by the Ryff’s Psychological Well-Being Scale (PWB; 1989a) in prior research (e.g., Baker, Soto, Perez, & Lee, 2012). As a multidimensional measure, the PWBscale was developed based on an extensive literature review about individuals’ positive functioning and the integration of clinical, developmental, and mental health theories. The theoretically-derived six dimensions of the PWB scale include one’s positive attitude toward the self and one’s past life (self-acceptance), close relations with others (positive relations with others), a sense of independence (autonomy), competence to manage the environment (environmental mastery), belief in the meaning of one’s present and past life (purpose in life), and sense of continued growth as a person (personal growth) (Ryff, 1989b). Ryff’s PWB dimensions are based on a eudaimonic perspective (i.e., how people perceive aspects of their lives) rather than hedonic perspective of well-being (i.e., how people feel) (Abbott et al., 2006). Several versions of the scale have been developed, including those with 18-, 24-, 39-, 42-, 54-, 84-, and 120 items (Sirigatti et al., 2013), with the 18-item version being the most commonly used (Clarke, Marshall, Ryff, & Wheaton, 2001; Sirigatti et al., 2013).Ryff and Keyes (1995) concluded that a second-order factor model with the six subdimensional well-being factors loading onto a general well-being factor is one of the models that best fit the data, which was also confirmed by subsequent studies (e.g., Ryff & Keyes, 1995; Sirigatti et al., 2013; Springer & Hauser, 2006; van Dierendonck, 2005). These scales are widely used in both clinical and general samples with different sociodemographic characteristics such as gender, age, level of education, and marital status (Clarke et al., 2001; Ryff 1989b; Ryff & Keyes, 1995; Sirigatti et al., 2013).

Despite the widespread use and application of the Ryff’s PWB scale, the factor structure of the scale remain contentious across different versions, partially because of high correlations among the six-factors that were theoretically proposed to be independent (Ryff & Keyes, 1995; Ryff, 1989b). In one study examining a sample of American adults with the 120-item version, inter-correlations among factor scores for the constructs ranged from .32 to .76 (Abbott et al., 2006). The correlations were particularly strong between environmental mastery and self-acceptance (.76); self-acceptance and purpose in life (.72); and personal growth and purpose in life (.72) (Ryff, 1989b). Springer and Hauser (2006) also identified high inter-factor correlations for all dimensions (.72-.97), with the highest correlations found between environmental mastery and self-acceptance as well as purpose in life and self-acceptance. The high inter-correlations among the six dimensions may indicate that the dimensions are not distinct from one another, and the scale may not adequately operationalize the constructs in clinical practice (Abbott et al., 2006).

Moreover, the factor structure of the PWB scale may not be the sameacross different cultures. Previous studies investigating the factor structure of the PWB across various Western cultures showed mixed evidence about its dimensionality (Sirigatti et al., 2013; van Dierendonck, 2005). The results from one study on Canadians showed that confirmatory factor analyses supported a six-factor model with an 18-item version of the scale; however, the researchers found low factor loadings (< .30)for two items(“Some people wander aimlessly in life, people would describe me as a giving person” and “the demands of life often get me down”),and two cross loaded items (“People would describe me as a giving person” and “Life has been a continuous process of learning”) (Clarke et al., 2001).

In contrast, another Canadian study performed a principal-axis factor analysis with the 120-item version but did not produce the six-factor solution proposed by Ryff (Kafka & Kozma, 2002). Instead, fifteen factors were extracted, each with an eigenvalue greater than 1.Moreover, another study conducted on a Dutch sample using the Dutch version of the PWBscale found that the model fit of the six-factor model with a single second-order factor was poor for the 14- and 9-items scale versions. The results were only acceptable for the 3-items per subscaleversion, although the internal consistency of these 3items was low(van Dierendonck, 2005).

The mixed evidence for the factor structure of the PWBscale suggests that the instrument does not measure the same construct across different cultural and linguistic groups (Byrne et al., 2009). Culture is an essential concept to consider in this regard due to the varied views of psychological well-being found across various cultural contexts. For instance, Western cultures emphasizeself-oriented aspects of well-being (e.g., self-acceptance or autonomy), whereas Eastern culturesgenerally place greater importance on other-oriented dimensions of well-being (e.g., positive relations with others) (Ryff, 1995). Because some attributes of psychological well-being may not present in certain cultural groups, the factor structure of the PWB scale may vary across contexts. Moreover, the language-diverse versions of the instrument may not sufficiently measure the same construct even if careful translation is conducted (Sirigatti et al., 2013).

Since previous studies have focused on European cultures, there is need to also evaluate the factor structure of the PWB scale in East Asian cultures. Currently, there is only onestudy conducted on an East Asiansample, specifically, Chinese adults. The authors tested the six-factor model with the 18-item scale and found poor model fit (Cheng & Chan, 2005). However, the results of the Chinese study may not be generalizable to other East Asian groups, particularly those residing in different cultural contexts such as Korean immigrants in the U.S., which is the focus of the present study. More specifically, we examined the factor structure of Ryff’s PWB scale in a Korean-American sample of mothers of young children. In addition, Ryff (1989b) found medium-to-large sized negative associations betweenthe PWB scale dimensions and the severity of depressive symptoms with the strongest correlations for environmental mastery (−.60) and purpose in life (−.60). A Japanese study found small-to-medium sized negative correlations between the PWB scale factors and depressive symptoms with the small correlations for autonomy (−.22), positive growth (−.27), and purpose in life (−.29) and the moderate correlations for positive relationship (−.36), self-acceptance (−.41), and environmental mastery (−.43) (Liu, Shono, & Kitamura, 2009). Moreover, in a sample of Asian Americans, Baker et al. (2012) foundweak-to-strong correlations between the PWB scale factors and depressive symptoms with the weakest correlation for autonomy (−.29) and the strongest correlation for environmental mastery (−.65). Thus, we examined the association between the factors of the PWB scale and mothers’ depressive symptoms to provide evidence of the PWB scale’s validity for Korean immigrant mothers in the U.S.

An examination of the factor structure of the six dimensions of Ryff’s PWB scale in Korean-American populations is significant for several reasons. First, Korean immigrants in the U.S. serve as an interesting sample because they often face cultural challenges that negatively influence their mental health as a group. For example, Korean immigrants are known to have difficulties enduring the loss of social ties when they leave their home country and undergo acculturation stress associated with migration in the U.S. due to large cultural and language distance between their heritage and host contexts (Berry, 2005; Garcia, Ramirez, & Jariego, 2002). Some aspects of acculturation stress could be moresignificant for Korean immigrants than other Asian immigrants because Korean immigrantsare generally less willing to learn English and make friends ofa different ethnicity (Min, 1995). Moreover, compared to other Asian immigrants,Korean immigrants in the U.S. have been found to maintain stronger ethnic attachment and adherence to cultural traditions in order to satisfy their primary group needs and preserve their collective identity (Hurh & Kim, 1990; Min, 1995). Korean immigrants tend to build social relationships primarily around their ethnic organizations (e.g., churches) within the boundaries of a Korean enclave (Hurh & Kim, 1990). In fact, Korean immigrants were found to report that their group do not get along very well with other ethnic groups in the U.S. (Pew Research Center, 2013). Moreover, Korean immigrants have shorter immigration history in the U.S., experience greater difficulty in learning new language, and thus take less prestige jobs than other Asians, even when their education level is comparable (Kuo, 1984). Therefore, Korean immigrants may experience more challenges in adjusting to life in the U.S. than other Asian groups with comparable socioeconomic standing (Kim, Seo, & Cain, 2010).

Second, first-generation Korean Americans have been identified as one of the most understudied populations despite being one of the largest and fastest-growing segments of Asian American population (Jang, Kim, Hansen, & Chiriboga, 2007). Although Korean immigrants in the U.S. are perceived to be highly educated and considered a successful immigrant group (Zhou & Lee, 2017), first-generation Korean immigrant mothersin particular have been found to have greater risk for mental health issues compared to Korean immigrant men and women of other ethnic groups (Ding et al., 2011; Kuo, 1984; Shin, 1993). Thus, more attention on Korean American women’s psychological adjustment in the U.S. in warranted.

Korean immigrant mothers of young children may have additional challenges of immigration in relation to childcare due to the loss of social ties and potential support in the host culture, and the cultural obligation to be a “wise mother and good wife (Hyonmo Yangcho)” (Choi, 2009; Lim, 1997). Due to this cultural obligation, Korean immigrant mothersare expected to be the primary caregiver and housekeeper, even though they participate as wage earners by working outside the home(Lee, Martin-Jearld, Robinson, & Price, 2016). Korean immigrant mothershave been found to generally comply with the traditional unequal relationship between husband and wife and follow the patriarchal traditions in terms of work and responsibilities in the home (Choi, Kushner, Mill, & Lai, 2014; Lim 1997; Shin, 2011), despite their exposure to Western cultural values. In addition, the “wise mother and good wife” adage urges Korean women to be self-sacrificing, soft-spoken, and submissive (Kim, 1998). Therefore, their emotional needs may be less supported because expressing such problem is believed to disrupt family harmony (Kim, 1998). Moreover, explicit expressions of one’s inner feeling bring stigma and shame to the cultural valuing of stoicism (Lee et al., 2016). Koreans often suppress their negative emotions, such as anger and frustration, to others in order to avoid the risk of losing face (chemyun) in their interpersonal relationships (Lee, 2013). Therefore, the suppression of emotions and various challenges that Korean immigrant mothers encounter may lead to chronic distress and contribute to the onset of Hwa-Byung (fire illness), an indigenous psychiatric illness reported among both Korean and Korean immigrant mothers(Lee et al., 2016; Lin et al., 1992).

Although a few studies have examined Korean immigrants’ well-being in the U.S. using the PWB scale (Yoon & Lee, 2010), the factor structure of the PWB scale has not been evaluated. As such, the extent to which the PWB scale truly measures positive psychological functioning among Korean immigrant mothersis still unknown.Examining the factor structure of the PWB scalein a Korean immigrant mother population will advance our understanding of psychological well-being as well as the applicability of the PWB scale in immigration settings and across different cultural contexts.

Method

Participants and Procedure

One hundred and sixty-nine first-generation Korean women (M_age = 36.05 years old, SD = 3.82) residing in Maryland, U.S. participated in this study. The participants had either one child (20.33%), two children (53.30%), or three or more children (26.37%). Most participants were married (97.20%), had at least a college degree (71.4%), and have been in the U.S. for an average of 11.83 years (SD = 9.99).

Participants were recruited from various organizations including Asian supermarkets, shopping malls, Korean churches andlanguage schools, preschools, and public libraries in Maryland.The study procedures were approved by the authors’ Institutional Review Board. Phone screening was conducted with potential participants to verify that both parents were ethnically Korean and primary caregivers to a normally developing child. Trained bilingual researchers who were fluent in Korean and English visited the family and collected the data after informed consent was obtained by the mothers.

Measures

All the questionnaires originally written in English were translated into Korean and back-translated into English using forward- and back-translation procedures by bilingual researchers to ensure the semantic equivalence in both versions (Peña, 2007). All discrepancies were discussed and resolved.

Family demographic information.

The Family Description Measure (FDM; Bornstein, 1991) was used to obtain demographic information on mothers’ age, years of U.S. residency, number of children, and level of education.

Psychological well-being.

Participants completed an 18-item version of the PWB scale (Ryff, 1995). Mothers responded to each item on a scale of 1 (strongly disagree) to 7 (strongly agree). The full items of the 18-items version is shown in Table 1. Reliability in Korean samples was limited. The internal consistency coefficients ranged from .65 to .87across the six dimensions in a sample of 510 Korean college students (Kim & Moon, 2017).

Table 1.

Theory Guided Dimensions of Psychological Well-Being and Items

PWB Dimensions
Item # and Content
Self-acceptance
6r. In many ways, I feel disappointed about my achievements in life.
12. I like most aspects of my personality.
18. When I look at the story of my life, I am pleased with how things have turned out.
Positive relations with others
4r. I have not experienced many warm and trusting relationships with others.
16r. Maintaining close relationships has been difficult and frustrating for me.
10. People would describe me as a giving person, willing to share my time with others.
Autonomy
1. I judge myself by what I think is important, not by what others think is important.
7. I have confidence in my own opinions, even if they are contrary to the general consensus.
13r. I tend to be influenced by people with strong opinions.
Environmental mastery
2. I am quite good at mastering the many responsibilities of my daily life.
14. In general, I feel I am in charge of the situation in which I live.
8r. The demands of everyday life often get me down.
Purpose in life
5r. I sometimes feel that I’ve done all there is to do in life.
11. Some people wander aimlessly through life, but I am not one of them.
17r. I live life one day at a time and don’t really think about the future.
Personal growth
3r. I have given up trying to make big improvements or changes in my life.
9. For me, life has been a continuous process of learning, changing, and growth.
15. I think it is important to have new experiences that challenge how you think about yourself and your world.

Note. r indicates reverse-scored items.

Depressive symptoms.

The Beck Depression Inventory-II (BDI-II; Beck, Steer, & Brown, 1996) was used to measure the presence and severity of depressive symptoms. The BDI-II consisted of 21 items. Each item was rated on a four-point Likert-type scale ranging from 0 (e.g., “I do not feel sad”) to 3 (e.g.,” I am so sad or unhappy that I can’t stand it”), based on the severity of each symptom in the last two weeks. The total depressive symptomatology score was the sum of the 21 items, which could potentially range from 0 to 63. Higher scores indicated more severe depressive symptoms. This measure has shown to be valid in both Korean and Korean immigrant samples (e.g., Cho & Haslam, 2010; Lee, Lee, Hwang, Hong, & Kim, 2017). The Cronbach’s alpha was .89 in this study.

Data Analysis

A second-order confirmatory factor analysis(CFA) with maximum likelihood estimation was conducted on the 18-item version using the Mplus (Muthén & Muthén, 1998–2012) to assess the six-factor structure that was based on theory and previous studies (e.g., Clarke et al., 2001; Ryff, 1989a,b).Overall model fit was evaluated using the following goodness-of-fit indices: Root mean square error of approximation (RMSEA), comparative fit index (CFI), and Tucker-Lewis index (TLI), RMSEA value below .08, and CFI and TLI values above .90 indicate acceptable model fit. CFI and TLI values greater than .95, and RMSEA value less than .05, are all considered good fit (Brown, 2006; Hu & Bentler, 1999). In the event that CFA failed to confirm the six-factor structure, exploratory factor analysis (EFA) would have been conducted to explore new factor solutions.The R function swain (Boomsma & Herzog, 2013) was used to calculate Swain-corrected model fit indices (Swain 1975) for structural equation models with small sample sizes. Finally, correlations between factors of the PWB scale and depressive symptomscores were examined.

Results

Bivariate correlations among the 18 items are summarized in Table 2.The Swain-corrected fit indices indicated that the second-order CFA model (Figure 1) did not fit the data well(χ²(129) = 270.71, p< .001, CFI =.79, TLI =.76, RMSEA =.08). Covariance matrix of the six factors was not positive definite due to high correlations among factors. Specifically, EM was too strongly correlated to SA, PR, PL, and PWB (Table 3). This indicates that some factors are not distinct from each other and the number of factors should be reduced.

Table 2.

Correlation Coefficients for 18 Items

Items#	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18
1	1
2	.37	1
3	−.32	−.20	1
4	−.04	−.25	.25	1
5	.09	.15	.17	.26	1
6	−.22	−.41	.29	.34	.03	1
7	.40	.31	−.10	−.05	.02	−.33	1
8	.05	−.15	.07	.28	.03	.34	−.26	1
9	.14	.24	−.32	−.24	−.10	−.31	.37	−.20	1
10	.02	.29	−.02	−.12	−.02	−.14	.26	−.21	.22	1
11	.17	.30	−.23	−.18	−.05	−.40	.33	−.16	.45	.28	1
12	.29	.41	−.12	−.22	.09	−.31	.41	−.41	.31	.45	.34	1
13	−.22	−.15	.24	.04	.02	.22	−.29	.21	−.16	−.09	−.09	−.15	1
14	.09	.38	−.18	−.16	.09	−.20	.19	−.02	.32	.27	.43	.28	−.00	1
15	.23	.27	−.28	−.19	.04	−.26	.40	−.19	.49	.20	.34	.43	−.23	.40	1
16	−.10	−.30	.21	.29	.04	.36	−.26	.31	−.18	−.21	−.22	−.32	.29	−.16	−.15	1
17	−.12	−.30	.21	.27	.14	.22	−.12	.11	−.22	−.18	−.24	−.17	.09	−.25	−.24	.21	1
18	.23	.35	−.15	−.20	.06	−.47	.37	−.18	.33	.18	.38	.43	−.03	.20	.29	−.12	−.14	1

Figure 1.

Standardized parameter estimates from the second-order CFA. Note. SA = Self-acceptance, PR = Positive relations with others, AU = Autonomy, EM = Environmental mastery, PL = Purpose in life, and PG = Personal growth. Standardized coefficients were reported.

Table 3.

Latent Variable Correlation Coefficients for Second-Order Model

	SA	PR	AU	EM	PL	PG	PWB
SA	1
PR	.91	1
AU	−.67	−.64	1
EM	−1.10	−1.05	.78	1
PL	.93	.89	−.66	−1.08	1
PG	.77	.73	−.54	−.89	.75	1
PWB	.98	.93	−.69	−1.13	.96	.79	1

Note. SA = Self-acceptance, PR = Positive relations with others, AU = Autonomy, EM = Environmental mastery, PL = Purpose in life, PG = Personal growth, and PWB = Psychological well-being.

Given the high factor correlations from the CFA model, effort was made to better understand the factor structure in subsequently analyses.First, Item 5 (“I sometimes feel that I’ve done all there is to do in life”) had R-square values.00 in the second-order CFA model, suggesting that it does not indicate the latent factor at all. It is possible that Item 5 may not have been appropriate for Korean immigrant mothers, whose primary reason for coming to the U.S. is to have better educational and economic opportunities (Hong & Hong, 1996; Park, 1998). Therefore, this item was removed.

Second, EFA with principal axis factoring (PAF) method was conducted on the remaining 11 items that were highly correlated within factors (EM, SA, PR, and PL) other than the items from AU and PG and Item 5. Direct oblimin rotation method was used because the current and past research suggested that the factors are correlated (Roberson III, Elliott, Chang, & Hill, 2014). A significant level for Bartlett’s test of sphericity was obtained (379.95, df = 55, p< .000), indicating the adequate correlation among variables in the matrix. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was .79, indicating that a factor analysis may be useful with this data.The screen plot (Figure 2) suggests a three-factor solution, with eigenvalues ranging from 3.56 to 1.10. The descending order of variance showed that 32.39% of the variance in the 11 items was explained by component one, 11.54% by component two, and 10.03% by component three, while three components accounted for 53.96% of the total variance in the items.These components were named satisfaction with life, goal orientation, and positive self-perceptions. The factor correlation between satisfaction with life and goal orientation was .19, satisfaction with life and positive self-perceptions was .54, and goal orientation and positive self-perceptions was .19.

Figure 2.

Scree plot of the results of the principal axis factoring, showing 11 items.

Table 4 provides the item-component loadings for the 11 items. Factor 1, satisfaction with life, was comprised of items 2, 4, 6, 8, 16, 17, and 18. Factor 2, goal orientation, was comprised of items 11 and 14. Factor 3, positive self-perceptions, was comprised of items 10 and 12. Item 8 had a cross loading for factor 1 and 2. However, we included Item 8 in Factor 1 because it was negatively correlated to Items 14 and 11 in Factor 2, although all three items had positive loadings on Factor 2 (Table 4). Moreover, it makes more sense to conceptually include Item 8 in Factor 1 than Factor 2 because Item 8 captures an aspect of life functioning rather than directedness and goal orientation in one’s life.

Table 4.

Item Loadings for the PWB Scales on 11 Items. Pattern Matrix and Structure Matrix. Principle Axis Factoring with Direct Oblimin

Item # and content	Pattern Matrix			Structure Matrix
	Factor 1	Factor 2	Factor 3	Factor 1	Factor 2	Factor 3
6. In many ways, I feel disappointed about my achievements in life	−.86	−.05	.13	−.81	−.19	−.34
4. I have not experienced many warm and trusting relationships with others	−.50	.05	.02	−.48	−.05	−.24
16. Maintaining close relationships has been difficult and frustrating for me	−.43	.12	−.11	−.47	.02	−.31
2. I am quite good at mastering the many responsibilities on my daily life	.37	.28	.18	.52	.38	.43
18. When I look at the story of my life, I am pleased with how things have turned out	.35	.17	.19	.48	.27	.41
17. I live life one day at a time and don’t really think about the future	−.32	−.18	.01	−.34	−.24	−.19
14. In general, I feel I am in charge of the situation in which I live	.09	.54	.22	.31	.60	.38
11. Some people wander aimlessly through life, but I am not one of them	.29	.40	.21	.48	.50	.45
8. The demands of everyday life often get me down	−.37	.38	−.29	−.45	.26	−.41
12. I like most aspects of my personality	.02	−.08	.87	.47	.09	.86
10. People would describe me as a giving person, willing to share my time with others	−.06	.12	.54	.25	.21	.53

Finally, correlations were conducted to examine the associations among the original factors of the PWB scale, three new factors derived from EFA, and mothers’ depressive symptoms (M = 9.11, SD = 7.16). As shown in Table 5, there were medium- tolarge-sized correlations betweenthe newly derived PWB factors and depressive symptoms.

Table 5.

Correlations between PWB Factors and Depressive Symptoms

	Six original factors						Three new factors
	SA	PR	AU	EM	PL	PG	SL	GO	PS
BDI-II	−.57***	−.35***	−.30***	−.42***	−.24**	−.19*	−.56***	−.31***	−.33***

Note. SA = Self-acceptance, PR = Positive relations with others, AU = Autonomy, EM = Environmental mastery, PL = Purpose in life, PG = Personal growth, PWB = Psychological well-being, SL = Satisfaction with life, GO = Goal orientation, and PS = Positive self-perceptions;

^***p < .001,

^**p< .01,

^*p< .05.

Discussion

The purpose of the present study was to confirm the theory-guided six dimensions of psychological well-being in a sample of first-generation Korean immigrant mothers in the U.S. Our study could not confirm the second-order factor model proposed by Ryff (e.g., Ryff & Keyes, 1995; Sirigatti et al., 2013) for the 18-item version. This result is consistent with a previous study on a Chinese sample, which also failed to confirm the 6-factor model with a single second-order factor for the 18-item version (Cheng & Chan, 2005).

Among the original factors theoretically operationalized by Ryff, only the autonomy and personal growth factors were distinct from each other. An individual’s psychological well-being is dependent on his or her adjustment to the norms and values defined by culture (Karasawa et al., 2011). For example, psychological well-being in independent contexts such as the U.S. has been shown to be associated with high levels of perceptions of autonomy (Oishi, 2000) and personal achievement and growth (Karasawa et al., 2011; Uchida & Kitayama, 2009). The constructs of autonomy and personal growth may be valued and accepted forms of psychological state that have greater significance for the psychological well-being of Korean immigrant mothers who are exposed to the U.S. culture. According to Self-determination Theory, a sense of autonomy is essential to and beneficial for well-being in all cultures, but tends to be more valued in Western cultures than Eastern cultures (Deci & Ryan, 2000). Ryff’s PWB model has also been found to fit well when assessing the psychological well-being of individuals with affective disorders such as depression, and those who report having low levels of a sense of meaning or goals in life (Moeenizadeh & Salagame, 2010; Rafanelli et al., 2000). Immigrants are at risk of experiencing psychological distress, including depression and loneliness, due to acculturative stress and adaptation challenges (Lai, 2004). However, past studies have found that acculturative challenges can also have positive functions for immigrants, for example, in providing opportunities for Korean immigrants to experience personal growth (e.g., Kim & Kim, 2013; Kim, Malonebeach, Heo, Kim, & Kim, 2015; Kim, Suh, & Heo, 2012). In particular, Korean women who leave their home country to immigrate to a new country despite large cultural distances (Gao, 2016), where gender roles are more likely to favor them compared to their country of origin (Min, 2001) may be more likely to see opportunities for personal growth and development.

The other four factors from the original structure (environmental mastery, purpose in life, positive relations with others, and self-acceptance) were highly correlated with each other, and did not fit the data well. Previous studies have also found similar results. For instance, self-acceptance and environmental mastery were found to be highly correlated (Ryff & Keyes, 1995; Kafka & Kozma, 2002; van Dierendonck, 2005). These scales were known to load on the negative affectivity factor, and the negative affectivity content may have caused them to cluster together (van Dierendonck, 2005). Moreover, it was probably more difficult to obtain coherent measures of these factors in Korean immigrant mothers due to the role of Korean culture and the context of immigration. Specifically, the self-acceptance construct measures the evaluations of one’s past and current self, and the purpose in life construct assesses one’s sense of goals in life. Under the cultural influence of fatalism in Korean society (Morris, Menon, & Ames, 2001), Korean individuals might be compelled to accept their past life while disliking their current self (Cheng & Chan, 2005) and believe that their future is predetermined regardless of their life personal goals. Moreover, Korean immigrant mother’s life may be constrained by cultural obligations to men and children due to the Korean traditional patriarchal norm (Lee, 2007; Min& Kim, 2009; Park, 2007) and gender ideology of “wise mother and good wife” (Cho, 2009). These cultural gender norms may make it difficult for Korean women to shape their own identities (Park & Bernstein, 2008), manage their environments, or set life goals without their husbands’ or elders’ permission.

Interpretations of the items from the positive relations with others and environmental mastery subscales can also be dependent on one’s acculturation level to the U.S. A separation style (Berry, 1997) characterized by high levels of participation in their heritage Korean culture along with low level of acculturation towards the mainstream U.S. culture can result in close relationships formed with other Koreans living in the U.S. and greater competence in managing the environment when tasks can be accomplished within the Korean immigrant subculture. However, not having intimate relationships with Americans in the host culture may lead to lower competence in managing activities that are require cultural and language (e.g., English) skills. In addition, as Korean immigrant mothersare expected to be managers of their children and households (Kim, 2013), they may become more socially isolated from the mainstream U.S. culture. Moreover, due to the cultural influence of filial piety in Confucian societies like Korea (Cheng & Chan, 2005), the first-generation Korean immigrant mothers in this study may have considered their intergenerational relationships with their parents and parent-in-laws when responding to the items from the positive relations with others subscale. In addition, these individuals tend to provide instrumental support (e.g., childcare and financial) for Korean immigrant mothers (Seo et al., 2017), which could have contributed to the high correlation between the positive relations factor with the environmental mastery factor.

Three new factor solutions were achieved on the items from the four highly correlated factors: satisfaction with life, goal orientation, and positive self-perceptions. The different structures of psychological well-being can be explained by the characteristics of Korean immigrants in the U.S. First-generation Asian immigrants including Korean who arrived in the past four decades are positively selected, meaning that they are more likely to have graduated from college than non-migrants from their country of origin, due to the change in the United States immigration law in 1965, which gave preference to highly-educated or skilled applicants from Asia (Zhou & Lee, 2017). In fact, most of the Korean immigrant mothers in this data had at least a college degree. For these mothers, items from the satisfaction with life factor may better describe one aspect of their psychological well-being because material prosperity and improving one’s family lives are fundamentally rooted in many immigrants’ pursuit of the American dream. (Hong & Hong, 1996; Park, 1998). Thus, this group of individuals may emphasize the ability to control their own lives, learn about various resources, gain knowledge, and increase interpersonal skills in evaluating their personal well-being, as it will increase their opportunities for better social integration (Lee, 2013) and chances of success in the new setting.

For the factor of goal orientation, although there has been a radical increase in Korean immigrant mothers’labor force participation rate and a change towards more egalitarian gender roles in Korean immigrant society (Min, 2001), Korean immigrant mothers continue to take primary responsibility for caring for their young children (Kim, 2008). Korean immigrant mothers feel fully responsible for their children’s behaviors and outcomes because they tend to consider children to be extensions of themselves (Kim & Choi, 1994). Thus, mothers are extensively involved in making decisions for their children, from daily routine needs to their college, profession, and spouse (Lehrer, 1996). Because Korean culture traditionally evaluates women’s successes by their children’s educational successes, many Korean immigrants migrate to the U.S. to pursue educational and life opportunities for their children (Lee et al., 2016). Therefore, this group of individuals who leave their country of origin for one that is culturally very different tend to be more directed to achieve specific goals for their children’s success. Such drive and purpose may lead them to perceive their lives to more meaningful, which contributes to another aspect of their psychological well-being.

For the positive self-perceptions factor, positive attitudes toward the self (e.g., strong sense of self-esteem) and positive social relationships with others are known to be effective coping resources for immigrant women (Delara, 2016; Lee, 2013). In particular, Korean immigrants with higher self-esteem were found to be more likely to experience lower levels of psychological distress (Noh & Avison, 1996) and Hwa-Byung, a culturally bounded mental illness in Korean culture (Lee, 2013). Moreover, having positive social relationships with others through leisure activities also helped Korean immigrant motherscope with acculturative stress (Kim, 2012; Kim, Kim, Henderson, Han, & Park, 2016). Thus, the items from the positive self-perception factor appear to capture various aspects of life that facilitate these women’s ability to manage the challenges of adaption and reduce acculturative stress and loneliness (Kim, Kim, Han, & Chin, 2015).

Finally, previous studies have found that psychological well-being was significantly correlated with depressive symptoms (e.g., Baker et al., 2012; Liu et al., 2009). Consistent with the literature, the newly derived factors of PWB were negatively correlated with mothers’ depressive symptoms, which provided some evidence for the construct validity of these factors.Thus, although the newly derived factors did not fit the original proposed factor structure of the Ryff’s PWB scale, their significant correlations with depressive symptoms in this study indicate the validity of these factors in the measurement of psychological well-being and its associationwith depressive symptomatology.

Limitations and Future Directions

The current study has several limitations. First, the results should be interpreted with caution due to the relatively small sample size. A larger sample size is recommended for future research, with EFA conducted on one-half of the sample and CFA conducted on the other half to confirm the new factor structures (Brown, 2006). Second, because this study only included Korean immigrant mothers who tended to have certain demographic characteristics (e.g., mostly middle-class, middle-aged, married) within the Washington D.C. metropolitan area, researchers should be cautious about generalizing these findings to Korean immigrant motherswith other demographic features residing in other regions of the U.S. In addition, Korean immigrant fathers were not included in this study. As previous studies have suggested that the structure of the PWB scale may vary across gender (e.g., Ryff, 1995; Ryff & Keyes, 1995) and Korean immigrants are known to adhere to traditional gender roles (Kim & Kim, 1998), future research should further evaluate factorial invariance across Korean immigrant mothers and fathers.

Conclusions

The failure to confirm a second-order factor model of Ryff’s PWBscale suggests that the theory-driven scale maynot applicable and generalizable to first-generation Korean immigrant women with children in the U.S.Moreover, results of the EFA support the view that how people define and perceivepsychological well-being may differ across cultures (Ford et al., 2015). For instance, first-generation Korean immigrant mothersmay hold different conceptualizations of psychological well-being because of the culture in their country of origin and their immigration-related experiences in the U.S. Based on these results, we recommend that this instrument be used with caution in this population and further research following our recommendations noted above need to be conducted.

Despite these limitations, the present study is the first to examine the factor structure of the Ryff’s PWB scaleamong Korean immigrant mothers in the U.S. Overall, the present study provided some preliminary validation evidence for the construct of psychological well-being in this immigrant setting, although future research with larger and more representative samples are needed. Moreover, further empirical research in different cultural contexts is needed to examine the factor structure of the PWB scale to advance our understanding of psychological well-being and possible contextual variations in this construct.