Abstract
Recent empirical research suggests that having a strong ethnic identity may be associated with reduced perceived stress. However, the relationship between perceived stress and ethnic identity has not been tested in a large and ethnically diverse sample of immigrants. This study utilized a multi-group latent class analysis of ethnic identity on a sample of first and second generation immigrants (N = 1603), to determine ethnic identity classifications, and their relation to perceived stress. A 4-class ethnic identity structure best fit the data for this immigrant sample, and the proportion within each class varied by ethnicity, but not immigrant generation. High ethnic identity was found to be protective against perceived stress, and this finding was invariant across ethnicity. This study extends the findings of previous research on the protective effect of ethnic identity against perceived stress to immigrant populations of diverse ethnic origins.
Keywords: Ethnic Identity, Perceived Stress, Immigrant generation, Ethnic Minority, Immigrant
Introduction
Immigrants are the fastest growing segment of the U.S. population. Currently, one in five people living in the United States is a first or second-generation immigrant [1], and by the year 2060 first generation immigrants are projected to constitute 18.8 % of the total population in the U.S. [2]. Immigration is a profound life transition requiring extensive adaptation [3, 4]. For example, immigrants are often required to adjust to new cultures, and environmental conditions (e.g. language, climate, food, values and customs). Also, a multitude of stressors often co-occur with migration, including pre-migration trauma, loss of social support, loss of family home, and social/racial oppression [5]. Numerous studies have documented the link between these types of adjustment problems and vulnerability to stress in the immigrant population [6–8]. For example, acculturation stressors such as, legal status, discrimination, language barriers and cultural differences, have been associated with adverse physical and mental health outcomes including headaches, high blood pressure and depression [9–12].
Given a multitude of stressors associated with immigration, it is likely that some immigrant individuals may experience a more heightened sensitivity to stress. This interpretation is referred to as perceived stress or the degree to which an individual considers the events in their life as stressful [13]. Perceived stress has emerged as a measure of general stress experienced by an individual stemming from the individual’s appraisal of their life as stressful [13, 14]. As with discrete stress, increased perceived stress has been linked to a number of mental health issues, including depression [13], anxiety [15], and substance abuse [16].
Ethnic identity, a multifaceted construct gauging the extent to which individuals conceive themselves as belonging to or identifying with a particular ethnic group [17–19], has been theorized to serve a protective function against mental health issues and predict positive psychological and sociocultural adjustment [20, 21] in particular, among racial ethnic minority (REM) individuals. For instance, strong ethnic identity, characterized by active exploration and commitment to a particular ethnic group, was found to be associated with increased self-esteem among ethnic-minority high school and college students [22], optimism in Latino adolescence [23], and reduced lifetime prevalence of psychiatric disorders in ethnic-minority adults [24].
However, too few studies have explored whether ethnic identity is related to perceived stress. The studies that have assessed the relationship between ethnic identity and perceived stress have nonetheless demonstrated ethnic identity’s protective function for perceived stress. For instance, ethnic identity was found to moderate the relationship between perceived stress and depression in a sample of Latino adults [14]. Specifically, individuals who had a strong ethnic identity and reported high perceived stress experienced less depression than individuals who had high perceived stress but a weak ethnic identity. More importantly for the current study, Lopez et al. demonstrated that a strong ethnic identity directly lead to reduced perceived stress among HIV + minority women [25].
Despite the encouraging findings of the studies just described, their implications for a diverse immigrant population have yet to be verified. In both studies, the samples used represented homogeneous groups, consisting primarily of non-immigrants. Whether or not ethnic identity impacts perceived stress among immigrants and if any ethnic group variations in this relationship exist have yet to be tested. Given the progressively hostile political and social climate currently surrounding immigration in the United States [26, 27], immigrants may be experiencing more perceived stress than in the previous few years. Therefore, it becomes imperative to ascertain any protective effects ethnic identity may have against perceived stress in this increasingly vulnerable population.
Current Study
The primary aims of this study are to: (1) characterize ethnic identity in an ethnically diverse population of first and second generation immigrants; and (2) determine whether ethnic identity is associated with perceived stress in this population.
Methods
Participants
The sample consisted of first and second generation immigrants (N = 1603) who took part in a larger cross-sectional study (N = 2871) designed to investigate social risk factors for subthreshold psychosis in an urban non-treatment-seeking emerging adult population. Participants were recruited from two urban universities in the Northeastern United States through an online recruitment tool, and they completed a battery of self-report questionnaires on computers in a research lab, including but not limited to, the measures described below. Participation time lasted up to an hour and a half. Recruitment strategy at the two universities varied to meet the demographic needs of the larger study which was to maximize representation of immigrants and Black populations.
Namely, the recruitment strategy for one university required participants to be emerging adults (aged 18–29) and self-identified as Black/African American/African descent or as a first or second-generation immigrant. The second university had no exclusion criteria beyond being 18 years old or older. The Institutional Review Boards of both universities approved the study protocol and researchers obtained written informed consent from all participants prior to enrollment. All participants received course credit for their participation in the study.
Measures
Ethnic Identity
The Multigroup Ethnic Identity Measure—Revised (MEIM—R) [28] is a 6-item self-report questionnaire used to determine participants’ level of ethnic identity. Using a 5-point Likert scale (strongly disagree = 1, neutral = 3, strongly agree = 5) the measure’s items capture the sense of self as a group member that is said to develop over time through an active process of exploration and commitment. To obtain a scale score, the responses to all 6 items are usually averaged, with higher scores indicating a stronger ethnic identity. However, for the purposes of this study, we implemented a Latent Class Analysis (described in more detail below) to obtain mutually exclusive ethnic identity categories. The MEIM-R total scale score demonstrated good internal reliability for the sample (Cronbach’s α = 0.87).
Perceived Stress
The Perceived Stress Scale [13] is a 14 item self-report instrument that gauges the degree to which situations that happened within the last month are perceived as stressful. In particular, items focus on the degree to which participants felt past situations were unpredictable, and whether they lacked control over such situations. Answers range from Never (0) to Very Often (4). The scale is scored by reverse scoring the positively stated items and then adding the responses from all items. PSS scale scores range from 0 to 56, with higher scores indicating more perceived stress. Because the scale is not a diagnostic tool, there are no cutoff scores available. Internal reliability for this measure was also acceptably high (Cronbach’s α = 0.79).
Immigrant Generation
Immigrant status was determined using information from questions regarding participants’ and parents’ place(s) of birth. Participants born outside of the US were labeled as first generation immigrants, while those born in the US, but with at least one parent born outside the US were labeled as second generation. Participants born in the US with both parents born in the US were excluded from the study.
Socio-Demographic Variables
Four covariates were measured: gender, racial/ethnic identification, age (years), and annual family income. Self-identified race/ethnicity was assessed using a question in which participants were instructed to “Choose one category that best captures how you see yourself” and given several options. These answers were grouped into 5 categories for the present analyses: (1) Black (non-US); (2) Hispanic/Latino; (3) Asian/Pacific Islander; (4) White and (5) Other (includes Biracial, and Middle Eastern). Household income was grouped into two categories: below $20,000 and $20,000 and above.
Analyses
Most studies that investigate ethnic identity in REM populations treat the scale as a continuous measure only to then categorize it using 2 or 3 levels (e.g. low, medium and high). While such an approach is valid among REM populations, the process of migration may intensify ethnic identity development such that the available range is more truncated. To our knowledge, no studies have explored the natural structure of ethnic identity in immigrant populations. Therefore, multi-group latent class analysis was used to determine the structure of ethnic identity, and whether levels of ethnic identity vary by immigrant generation and ethnic group membership.
Latent class analysis uses conditional probability to classify respondents into a series of mutually exclusive categories. Individual members of any given latent class have similar response patterns to categorical variables [29, 30]. Multi-group latent class analysis analyzes different subgroups (i.e. immigrant generation), within the same model, which in turn allows for the comparison of the latent class structure across these subgroups [31]. Standard errors were clustered by university or sampling unit [32, 33].
Model selection was based on the percent change in the Log-likelihood, the relative entropy or classification index, Bayesian Information Criterion (BIC) [34] index and model interpretability. The better fitting models are those which have minimum BIC, sizeable percent change in the Log-likelihood and entropy as close to 1 [30, 31].
Multivariate logistic regression was used to examine distributional differences in ethnic identity classification between immigrant generation and ethnicity. Finally, ordinary least squares regression was used to estimate the relation of ethnic identity with perceived stress. All analyses were carried out using STATA v.14, and the latent class analysis plug in for STATA made available through the Methodology Center at Penn State University [33].
Results
Descriptive Statistics
Sample characteristics appear in Table 1. Black immigrants (23.4 %) were mostly from Africa and the Caribbean. Hispanic immigrants (21.2 %) represented mainly South America and the Caribbean. White immigrants (11.5 %) were from European countries, and Asian immigrants (35.6 %) represented most Asian countries. The group labeled as “Other” (8.3 %) mainly represented biracial and multiracial. Forty five percent of the sample was first generation immigrant. Ethnicity and immigrant generation were significantly related with a moderate effect size (χ (4)2 = 62 7.0, p = 0.00, V = 0.20). In particular, Hispanic and Black immigrants were most likely to be second generation immigrants. Asian and to a lesser extent White immigrants were in contrast more likely to be first generation. The average respondent was 20 years old (SD = 2.2). First generation immigrants were slightly older (M = 20.6, SD = 2.3) than second generation immigrants (M = 19.7, SD = 1.8, t(1,594) = 9.1, pright–tail = 0.00, d = 0.44). Overall, the majority of respondents were female (66 %). The majority of the sample (77 %) had income above $20,000, and a small yet significant difference was observed by immigrant status (χ (1)2 = 5.3, p = 0.02, V = −0.06) whereby first generation immigrants are more likely to earn $20,000 a year or less than second generation immigrants. On average the perceived stress score was 26.3 (SD = 7.6), and while there were no significant differences across immigrant generation, a significant difference in perceived stress was found across sampling unit t(1,600) = 5.4, pright–tail = 0.00, d = 0.27). Furthermore, the distribution of race/ethnicity also varied by sampling unit (χ (4)2 = 155.6, p = 0.00, V = 0.31). However, with the exception of income, differences across immigrant generation discovered in the total sample were replicated within sampling unit.
Table 1.
Demographic characteristics of the sample by immigrant generation, and sampling unit
| Total sample |
1st Gen. n = 717 |
2nd Gen. n = 886 |
Statistical Comparisons |
Univer- sity 1 |
1st Gen. n = 473 |
2nd Gen. n = 543 |
Statistical comparisons |
Univer- sity 2 |
1st Gen. n = 244 |
2nd Gen. n = 343 |
Statistical comparisons |
||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| N = 1603 | 44.7 % | 55.3 % | n = 1016 | 46.6 % | 53.4 % | n = 587 | 41.6 % | 58.4 % | |||||||
| Mean (SD) or % | t/χ2 [df] | Effect Size (d/V) |
Mean (SD) or % | t/χ2 [df] | Effect Size (d/V) |
Mean (SD) or % | t/χ2 [df] | Effect Size (d/V) |
|||||||
| Race/ethnicity | |||||||||||||||
| Black (non-US) | 23.3 % | 19.8 % | 26.1 % | 62.7*** | 0.2 | 27.1 % | 22.0 % | 31.5 % | 72.9*** | 0.3 | 16.7 % | 15.6 % | 17.5 % | 8.5 [4] | |
| Hispanic | 21.2 % | 15.6 % | 25.7 % | [4] | 27.6 % | 19.9 % | 34.3 % | [4] | 10.2 % | 7.4 % | 12.2 % | ||||
| Asian | 35.6 % | 44.9 % | 28.1 % | 32.2 % | 44.4 % | 21.6 % | 41.6 % | 45.9 % | 38.5 % | ||||||
| White | 11.5 % | 12.8 % | 10.5 % | 6.1 % | 7.6 % | 4.8 % | 21.0 % | 23.0 % | 19.5 % | ||||||
| Other | 8.4 % | 6.8 % | 9.6 % | 7.1 % | 6.1 % | 7.9 % | 10.6 % | 8.2 % | 12.2 % | ||||||
| Annual family income | |||||||||||||||
| <$20,000 | 22.6 % | 25.2 % | 20.4 % | 5.3* [1] | −0.06 | 26.5 % | 27.7 % | 25.4 % | 0.7 [1] | 15.8 % | 20.5 % | 12.5 % | 6.8** [1] | −0.11 | |
| Age | 20.1 (2.2) | 20.6 (2.3) | 19.7 (1.8) | 9.1*** [1, 594] | 0.44 | 19.9 (2.1) | 20.3 (2.3) | 19.5 (1.8) | 6.4*** [1, 013] | 0.38 | 20.5 (2.1) | 21.2 (2.3) | 20.0 (1.8) | 7.3*** [579] | 0.58 |
| Gender | |||||||||||||||
| Male | 34.0 % | 35.2 % | 33.2 % | 0.7 [1] | 37.5 % | 37.8 % | 37.2 % | 0.1 [1] | 28.1 % | 29.9 % | 26.8 % | 0.7 [1] | |||
| Female | 66.0 % | 64.9 % | 66.8 % | 62.5 % | 62.2 % | 62.8 % | 71.9 % | 70.1 % | 73.2 % | ||||||
| Perceived Stress | 26.3 (7.6) | 26.2 (7.5) | 26.3 (7.6) | −0.2 [1, 600] | 27.1 (7.4) | 27.1 (7.4) | 27.0 (7.4) | 0.2 [1, 013] | 25.0 (7.6) | 24.6 (7.2) | 25.3 (7.9) | −1.1 [585] | |||
1st Gen. first generation immigrant, 2nd Gen. Second Generation immigrant
p < 0.05,
p < 0.01,
p < 0.00; p value for t test reported is one tailed
Multi-Group Latent Class Analysis
All fit statistics for latent class models appear in Table 2. A four to seven class model provide similar fit for the data. Based on model interpretability, the 4 class solution is the most fitting. According to inter class probabilities (Table 3), classes are labeled as follows: [1] “No Exploration or Commitment” (5.74 %, n = 92), [2] “Low Commitment” (31.5 %, n = 505), [3] “High Exploration and Commitment” (39.3 %, n = 630), and [4] “Very High Exploration and Commitment” (23.5 %, n = 376). The “No Exploration or Commitment” group is composed of participants who were at an increased probability (above 0.5) of replying “Strongly Disagree” to all six MEIM-R items. Similarly, the “Low Commitment” group was characterized by participants that were likely to reply with “Neutral” to 2 out of the 3 commitment items in the MEIM-R. Lastly, the “High Exploration and Commitment” and “Very High Exploration and Commitment” contained participants that largely replied with “Agree” and “Strongly Agree”, respectively, to all ethnic identity items. Size of latent classes did not differ statistically across universities (χ (3)2 = 6.6, p = 0.09).
Table 2.
Model fit statistics for multi-group latent class analysis of ethnic identity (N = 1603)
| No. classes | Entropy index | Log-likeli-hood (LL) | % ΔLL | BIC | ΔBIC |
|---|---|---|---|---|---|
| 2 | 0.79 | −12594.82 | – | 7193.1 | – |
| 3 | 0.84 | −11926.82 | −5.30 % | 6049.0 | −1144.1 |
| 4 | 0.84 | −11630.59 | −2.48 % | 5648.4 | −400.6 |
| 5 | 0.81 | −11489.99 | −1.21 % | 5559.0 | −89.3 |
| 6 | 0.80 | −11347.23 | −1.24 % | 5465.4 | −93.6 |
| 7 | 0.81 | −11234.80 | −0.99 % | 5432.4 | −33.0 |
| 8 | 0.82 | −11224.49 | −0.09 % | 5603.7 | 171.2 |
Standard errors were clustered by sampling unit
Change in, LL log-likelihood, BIC Bayesian information criterion
Table 3.
Item response probabilities of the four class solution to multi-group latent class analysis of ethnic identity
| 1st generation | 2nd generation | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| n = 717
|
n = 886
|
|||||||||
| MEIM item | Exploration (E)/commitment (C) | Class | None | Low commitment | High | Very high | None | Low commitment | High | Very high |
| Population shares | 7 % | 34 % | 36 % | 23 % | 6 % | 30 % | 39 % | 25 % | ||
| Sample shares | 45 | 246 | 269 | 157 | 47 | 259 | 361 | 219 | ||
| 1 | E | Strongly | 0.7 | 0.1 | 0 | 0.1 | 0.7 | 0.1 | 0 | 0.1 |
| 2 | C | disagree | 0.6 | 0.1 | 0 | 0 | 0.6 | 0.1 | 0 | 0 |
| 3 | C | 0.5 | 0 | 0 | 0 | 0.5 | 0 | 0 | 0 | |
| 4 | E | 0.7 | 0 | 0 | 0 | 0.7 | 0 | 0 | 0 | |
| 5 | E | 0.8 | 0.1 | 0 | 0 | 0.8 | 0.1 | 0 | 0 | |
| 6 | C | 0.7 | 0.1 | 0 | 0 | 0.7 | 0.1 | 0 | 0 | |
| 2 | C | Neutral | 0.1 | 0.5 | 0.2 | 0 | 0.1 | 0.5 | 0.2 | 0 |
| 6 | C | 0.1 | 0.5 | 0.2 | 0 | 0.1 | 0.5 | 0.2 | 0 | |
| 1 | E | Agree | 0.1 | 0.2 | 0.5 | 0.3 | 0.1 | 0.2 | 0.5 | 0.3 |
| 2 | C | 0.1 | 0.2 | 0.7 | 0.2 | 0.1 | 0.2 | 0.7 | 0.2 | |
| 3 | C | 0.1 | 0.4 | 0.8 | 0.2 | 0.1 | 0.4 | 0.8 | 0.2 | |
| 4 | E | 0.1 | 0.2 | 0.6 | 0.3 | 0.1 | 0.2 | 0.6 | 0.3 | |
| 5 | E | 0.1 | 0.3 | 0.6 | 0.3 | 0.1 | 0.3 | 0.6 | 0.3 | |
| 6 | C | 0.1 | 0.2 | 0.6 | 0.2 | 0.1 | 0.2 | 0.6 | 0.2 | |
| 2 | C | Strongly agree | 0 | 0 | 0.1 | 0.8 | 0 | 0 | 0.1 | 0.8 |
| 3 | C | 0.1 | 0 | 0.1 | 0.8 | 0.1 | 0 | 0.1 | 0.8 | |
| 4 | E | 0 | 0 | 0 | 0.5 | 0 | 0 | 0 | 0.5 | |
| 5 | E | 0 | 0 | 0.1 | 0.5 | 0 | 0 | 0.1 | 0.5 | |
| 6 | C | 0.1 | 0 | 0.2 | 0.8 | 0.1 | 0 | 0.2 | 0.8 | |
Probabilities ≥0.5 are in bold face. The table presents only the items that loaded into a class
Immigrant Generation, Ethnicity and Ethnic Identity Latent Class
Table 4 presents the results of two multinomial logistic regression models predicting ethnic identity class membership. Model 1 is a simple model with no interactions between immigrant generation and ethnicity. Model 2 includes all variables from model 1, and interactions between ethnicity and immigrant generation. For both models, the reference group class is Very High Exploration and Commitment, and the base ethnic group is Asian as they constitute a large ethnic proportion in the sample. The reference group for immigrant generation is first generation immigrant. As shown immigrant generation was not significantly related to ethnic identity class membership, but ethnicity was significantly related.
Table 4.
Multinomial logistic regression results of ethnicity and immigrant generation on latent ethnic identity class (N = 1603)
| Model 1: no interactions
|
Model 2: interactions included
|
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No exploration or commitment
|
Low commitment
|
High exploration and commitment
|
No exploration or commitment
|
Low commitment
|
High exploration and commitment
|
|||||||
| Coefficient | OR | Coefficient | OR | Coefficient | OR | Coefficient | OR | Coefficient | OR | Coefficient | OR | |
| White | 0.74* | 2.10 | −0.51* | 0.60 | −0.90*** | 0.40 | 0.91+ | 2.50 | −0.49 | 0.61 | −0.80* | 0.45 |
| Black | −0.56 | 0.57 | −0.75*** | 0.47 | −0.82*** | 0.44 | −0.67 | 0.51 | −0.63* | 0.53 | −0.60* | 0.55 |
| Hispanic | 0.25 | 1.29 | −0.24 | 0.79 | −0.24 | 0.78 | 0.49 | 1.62 | −0.24 | 0.78 | −0.08 | 0.93 |
| Other | 0.17 | 1.19 | −0.93** | 0.39 | −0.52* | 0.59 | 0.92 | 2.51 | −0.73+ | 0.48 | −0.31 | 0.74 |
| 1st generation | 0.26 | 1.29 | 0.20 | 1.22 | −0.03 | 0.97 | 0.61 | 1.84 | 0.32 | 1.38 | 0.23 | 1.26 |
| 1st generation* White | −0.28 | 0.76 | −0.03 | 0.97 | −0.19 | 0.83 | ||||||
| 1st generation* Black | 0.25 | 1.28 | −0.29 | 0.75 | −0.48 | 0.62 | ||||||
| 1st generation* Hispanic | −0.43 | 0.65 | 0.06 | 1.07 | −0.36 | 0.70 | ||||||
| 1st generation* Other | −2.49* | 0.08 | −0.50 | 0.61 | −0.45 | 0.64 | ||||||
| Constant | 0.59*** | 1.80 | −1.61*** | 0.20 | 0.94*** | 2.56 | −1.81*** | 0.16 | 0.53*** | 1.69 | 0.80*** | 2.22 |
The reference ethnic identity group is the “Very High Exploration and Commitment” latent class. The reference group for ethnicity/race is Asian, and the reference group for immigrant generation is second generation immigrant
OR odds ratio
p < 0.05,
p < 0.01,
p < 0.001,
p (one tail) <0.05
Specifically, Asian immigrants were more likely than White, Black and Other ethnic group immigrants to be in the Low Commitment class than the Very High Exploration and Commitment class (odds ratios = 0.60, 0.47 and 0.39, respectively), and were less likely than the same three ethnic groups to be in the High Exploration and Commitment than in the Very High Exploration and Commitment class (odds ratios = 0.40, 0.44 and 0.59, respectively). This suggests Asian immigrants were divided into two contrasting latent ethnic identity classes—Low Commitment and High Exploration and Commitment, independent of immigrant generation. White immigrants were 2 times more likely than Asian immigrants to be in the lowest ethnic identity latent class—No exploration or Commitment. Hispanic immigrants were not statistically different from Asian immigrants in their propensity to be in one latent class versus another. Black and Other ethnic group immigrants were more likely than White, Asian and Hispanic immigrants to be in the highest ethnic identity category—Very High Exploration and Commitment. With the exception of the interaction between 1st generation immigrant status and Other ethnicity, no interactions between ethnicity and immigrant generation were significant.
Perceived Stress and Ethnic Identity
Results from an ordinary least squares regression predicting perceived stress from ethnic identity, ethnicity and demographic covariates are presented in Table 5.
Table 5.
Linear regression of perceived stress on demographic variables, ethnic identity latent classes and ethnicity
| Predictor variables | Perceived stress | |
|---|---|---|
|
| ||
| b | β | |
| Age | −0.21* | −0.06 |
| Male | −1.55*** | −0.10 |
| Annual family income below $20,000 | 1.47** | 0.08 |
| White | −2.02** | −0.09 |
| Black | −0.30 | −0.02 |
| Hispanic | −0.27 | −0.01 |
| Other | −0.40 | −0.01 |
| No Exploration or Commitment | 2.27* | 0.07 |
| Low Exploration and Commitment | 2.04*** | 0.13 |
| High Exploration and Commitment | 0.66 | 0.04 |
| Constant | 29.98*** | |
| N | 1595 | |
| F | 6.35*** | |
| Adj R2 | 0.03 | |
The reference ethnic identity group is the “Very High Exploration and Commitment” latent class, and the comparison ethnic group is Asian
p < 0.05,
p < 0.01,
p < 0.001
Immigrants in the lowest ethnic identity classes have on average higher perceived stress than immigrants in the top two highest ethnic identity latent classes (b = 2.27, p < 0.05, and b = 2.04, p < 0.001). Further, no statistically significant differences in stress between the two highest ethnic identity latent classifications were found, thereby suggesting that once high commitment and exploration of one’s own ethnicity is reached, increasing it to very high, has no additional impact on perceived stress. White immigrants have lower perceived stress on average than Asian immigrants (b = −2.02, β = −0.09, p < 0.01), and no statistical differences in perceived stress were found between Asian and other REM immigrants. A model including interactions between ethnicity and ethnic identity classes did not return significant interactions, implying that there is no differential effect on the role of ethnic identity on perceived stress by ethnicity. Other predictors of perceived stress include, age, gender and income.
Specifically, older immigrants have on average lower perceived stress than younger immigrants (b = −0.21, β= −0.06, p ≤ 0.05). Similarly, male immigrants have lower perceived stress than female immigrants (b = −1.55, β= −0.10, p ≤ 0.001), and immigrants whose income is $20,000 or below have on average higher perceived stress, than their counterparts (b = −1.47, β = −1.47, p < 0.01).
Discussion
This is the first study of a large ethnically diverse immigrant population to uncover the ethnic identity latent class structure of first and second generation immigrants, and to demonstrate the protective aspect of ethnic identity against perceived stress. In particular, the results demonstrate the ethnic identity of immigrants is best described by 4 latent classes, and the classes characterized by high commitment and exploration or above predicted lower perceived stress, whereas low commitment or below predicted higher perceived stress. This study is also the first to find race/ethnic group differences in perceived stress among immigrants. Specifically, White immigrants reported significantly less perceived stress than other immigrant groups. However, despite group differences in overall perceived stress, ethnic identity was found to function consistently across ethnicity as well as immigrant generation in terms of protection against perceived stress.
The results of this study are consistent with the fairly limited research on perceived stress and ethnic identity. Similar to the findings of Lopez et al., increased ethnic identity was found to directly reduce perceived stress. The positive characteristics associated with having strong ethnic identity, such as increased self-esteem [35] and the increased ability to cope with discrimination [36, 37], may help immigrants better manage the stressors they are exposed to resulting in a more optimistic outlook on their lives. Similarly, the cultural and social resources available to those who are strongly identified with their ethnic group [38] can be further leveraged to deal with stress. Clinicians and mental health professionals with immigrant clients should be aware of ethnic identity’s stress reducing potential, and encourage their clients to maintain or re-affirm their ethnic identity when struggling with a large amount of perceived stress.
Another noteworthy finding of this study was the variation in the distribution of immigrants into ethnic identity latent classes based on ethnicity. Namely, Asian immigrants (and to a lesser degree Hispanic immigrants) were mostly fragmented into two disparate classes—low commitment and high commitment and exploration. Prior studies have shown that relative to Blacks and Hispanics, who reported high levels of ethnic identity, Asians reported intermediate levels of ethnic identity [39]. However, the statistical method used here suggests that Asian immigrants are better captured by two different ethnic identity categorizations. The ethnic-specific effect on ethnic identity class membership suggests that ethnic identity formation is based in part on experiences unique to each ethnic group, irrespective of experiences unique to immigrant generations. Consequently, Asian and Hispanic immigrant experiences are likely to be more varied than that of other ethnic groups, and this contributes to heterogeneity across ethnic identity classes.
Nonetheless, the ethnic group variation in ethnic identity did not subsequently impact the relationship between ethnic identity and perceived stress. However, the results revealed an overall difference in perceived stress across ethnic groups, with White immigrants reporting significantly less perceived stress. White immigrants may not be exposed to the same level or the same type of stress as REM immigrants. For instance, research has shown that White immigrants tend to have higher income than other immigrant groups [40] which likely affects their perceptions of stress. Furthermore, White immigrants represent the majority racial group in the US and may therefore experience less discrimination than REM immigrants. However, despite differences in the experience of White immigrants, ethnic identity offers the same benefits in terms of perceived stress as that of other immigrant groups.
This study is not free from limitations. First, the sample only includes college students in two large urban areas, limiting generalizability. Second, the amount of variance explained for perceived stress was small; suggesting ethnic identity may only have a small contribution to the reduction of perceived stress. Finally, the study does not account for differences in the length of time immigrants intend to stay in the U.S. Future studies should explore the role of variant immigration histories, or at least incorporate them in the analyses. Despite these limitations, we are confident that our analyses have presented significant contributions toward understanding differences in ethnic identity formation among immigrant emerging adults, and the resulting effects on stress and consequently mental health.
New Contribution to the Literature
This is the first study to examine ethnic identity in relation-ship to perceived stress in a large and diverse sample of first and second generation immigrants. This study is the first to demonstrate that having strong ethnic identity is directly associated with reduced perception of stress regardless of ethnic group. Furthermore, through the use of latent class analysis, this study highlights contrasting classifications of ethnic identity obtained among Asian immigrants. Such finding redefines the “moderate” or “intermediate” ethnic identity observed among these groups in other studies. Next, we highlight the value of a diverse immigrant sample. A major benefit of this study is that, by drawing conclusions from an ethnically diverse sample of immigrant youth, it allows for comparisons between White and REM immigrants.
Acknowledgments
This study was supported by a grant from the New York State (NYS) Center of Excellence for Cultural Competence at the NYS Psychiatric Institute, NYS Office of Mental Health (Dr. Anglin), a CLA Research Award, Temple University, Philadelphia, PA (Dr. Ellman), NIMH R01 grant MH096478 (Dr. Ellman), and a start-up grant (Dr. Ellman).
Footnotes
Compliance with Ethical Standards
Conflict of Interest The authors declare that no conflicts of interests exist.
Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
References
- 1.Mather M. Children in immigrant families chart new path. Washington, DC: Population Reference Bureau; 2009. [Google Scholar]
- 2.U.S. Census Bureau. National population projections: Allegany County, N Y. http://www.census.gov/population/projections/files/summary/NP2014-T5.csv(2014) Accessed 27 Sep 2015.
- 3.Short KH, Johnston C. Stress, maternal distress, and children’s adjustment following immigration: the buffering role of social support. J Consult Clin Psychol. 1997;65:494–503. doi: 10.1037//0022-006x.65.3.494. [DOI] [PubMed] [Google Scholar]
- 4.Berry J. Immigration, acculturation, and adaptation. Appl Psychol. 1997;46:5–34. [Google Scholar]
- 5.Yakushko O, Watson M, Thompson S. Stress and coping in the lives of recent immigrants and refugees: considerations for counselling. Int J Adv Couns. 2008;30:167–78. [Google Scholar]
- 6.Harris-Reid MA. Coming to America: immigration, stress, and mental health. Diss Abstr Int Sect A. 1999;59:3975. [Google Scholar]
- 7.Levitt MJ, Lane JD, Levitt J. Immigration stress, social support, and adjustment in the first postmigration year: an intergenerational analysis. Res Hum Dev. 2005;2:159–77. [Google Scholar]
- 8.Lara M, Gamboa C, Kahramanian MI, et al. Acculturation and Latino health in the United States: a review of the literature and its sociopolitical context. Annu Rev Public Health. 2005;26:367–97. doi: 10.1146/annurev.publhealth.26.021304.144615. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Finch BK, Hummer RA, Kolody B, et al. The role of discrimination and acculturative stress in the physical health of Mexican-origin adults. Hisp J Behav Sci. 2001;23:399–429. [Google Scholar]
- 10.Finch BK, Vega WA. Acculturation stress, social support, and self-rated health among Latinos in California. J Immigr Health. 2003;5:109–17. doi: 10.1023/a:1023987717921. [DOI] [PubMed] [Google Scholar]
- 11.Flores E, Tschann JM, Dimas JM, et al. Perceived discrimination, perceived stress, and mental and physical health among Mexican-origin adults. Hisp J Behav Sci. 2008;30:401–24. [Google Scholar]
- 12.Lee JS, Koeske GF, Sales E. Social support buffering of acculturative stress: a study of mental health symptoms among Korean international students. Int J Intercult Relat. 2004;28:399–414. [Google Scholar]
- 13.Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav. 1983;24:386–96. [PubMed] [Google Scholar]
- 14.Chavez-Korell S, Torres L. Perceived stress and depressive symptoms among latino adults the moderating role of ethnic identity cluster patterns. Couns Psychol. 2014;42:230–54. [Google Scholar]
- 15.Panayiotou G, Karekla M. Perceived social support helps, but does not buffer the negative impact of anxiety disorders on quality of life and perceived stress. Soc Psychiatry Psychiatr Epidemiol. 2013;48:283–94. doi: 10.1007/s00127-012-0533-6. [DOI] [PubMed] [Google Scholar]
- 16.Ng DM, Jeffery RW. Relationships between perceived stress and health behaviors in a sample of working adults. Health Psychol. 2003;22:638–642. doi: 10.1037/0278-6133.22.6.638. [DOI] [PubMed] [Google Scholar]
- 17.Phinney JS. Ethnic identity in adolescents and adults: review of research. Psychol Bull. 1990;108:499–514. doi: 10.1037/0033-2909.108.3.499. [DOI] [PubMed] [Google Scholar]
- 18.Phinney JS, Horenczyk G, Liebkind K, Vedder P. Ethnic identity, immigration, and well-being: an interactional perspective. J Soc Issues. 2001;57:493–510. [Google Scholar]
- 19.Berry J, Sabatier C. Acculturation, discrimination, and adaptation among second generation immigrant youth in Montreal and Paris. Int J Intercult Relat. 2010;34:191–207. [Google Scholar]
- 20.Smith TB, Silva L. Ethnic identity and personal well-being of people of color: a meta-analysis. J Couns Psychol. 2011;58:42–60. doi: 10.1037/a0021528. [DOI] [PubMed] [Google Scholar]
- 21.Umaña-Taylor AJ. Ethnic identity In: Handbook of identity theory and research. New York: Springer; 2011. pp. 791–809. [Google Scholar]
- 22.Umaña-Taylor AJ. Ethnic identity and selfesteem: examining the role of social context. J Adolesc. 2004;27:139–46. doi: 10.1016/j.adolescence.2003.11.006. [DOI] [PubMed] [Google Scholar]
- 23.Roberts RE, Phinney JS, Masse LC, et al. The structure of ethnic identity of young adolescents from diverse ethnocultural groups. J Early Adolesc. 1999;19:301–22. [Google Scholar]
- 24.Burnett-Zeigler I, Bohnert KM, Ilgen MA. Ethnic identity, acculturation and the prevalence of lifetime psychiatric disorders among Black, Hispanic, and Asian adults in the U.S. J Psychiatr Res. 2013;47:56–63. doi: 10.1016/j.jpsychires.2012.08.029. [DOI] [PubMed] [Google Scholar]
- 25.Lopez CR, Antoni MH, Fekete EM, et al. Ethnic identity and perceived stress in HIV+ minority women: the role of coping self-efficacy and social support. Int J Behav Med. 2012;19:23–8. doi: 10.1007/s12529-010-9121-x. [DOI] [PubMed] [Google Scholar]
- 26.Chiacu D. Cruz says would deport illegal immigrants, sharpens immigration stance. Reuters [newspaper on the Internet] http://www.reuters.com/article/usa-election-cruz-immigration-idUSKCN0VW24M(2016) Accessed 2 Feb 2016.
- 27.Diamond J. Donald Trump: Ban all Muslim travel to U.S. CNN [newspaper on the Internet] http://www.cnn.com/2015/12/07/politics/donald-trump-muslim-ban-immigration/(2015) Accessed 2 Feb 2016.
- 28.Phinney J. The multigroup ethnic identity measure: a new scale for use with diverse groups. J Adolesc Res. 1992;7:156–76. [Google Scholar]
- 29.Hagenaars JA, McCutcheon A. Applied latent class analysis. Cambridge: Cambridge University Press; 2002. [Google Scholar]
- 30.Muthén B. Latent variable mixture modeling. In: Marcoulides GA, Schumacher RE, editors. New developments and techniques in structural equation modeling. Mahwah: Lawrence Erlbaum Associates; 2001. pp. 1–33. [Google Scholar]
- 31.Collins LM, Lanza ST. Latent class and latent transition analysis with applications in the social, behavioral, and health sciences. New Jersey: Wiley; 2010. [Google Scholar]
- 32.Skinner CJ, Holt D, Smith TMF. Analysis of complex surveys. New York: Wiley; 1989. [Google Scholar]
- 33.Lanza ST, Dziak JJ, Huang L, et al. LCA Stata plugin users’ guide (Version 1.2) University Park: The Methodology Center, Penn State; 2015. Available from methodology.psu.edu; [Google Scholar]
- 34.Schwarz G. Estimating the dimension of a model. Ann Stat. 1978;6:461–4. [Google Scholar]
- 35.Bracey JR, Bámaca MY, Umaña-Taylor AJ. Examining ethnic identity and self-esteem among biracial and monoracial adolescents. J Youth Adolesc. 2004;33:123–32. [Google Scholar]
- 36.Brittian AS, Kim SY, Armenta BE, et al. Do dimensions of ethnic identity mediate the association between perceived ethnic group discrimination and depressive symptoms? Cult Divers Ethn Minor Psychol. 2015;21:41–53. doi: 10.1037/a0037531. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Tynes BM, Umaña-Taylor AJ, Rose CA, et al. Online racial discrimination and the protective function of ethnic identity and self-esteem for African American adolescents. Dev Psychol. 2012;48:343–55. doi: 10.1037/a0027032. [DOI] [PubMed] [Google Scholar]
- 38.Quintana SM, Scull NC. Latino ethnic identity. In: Villarruel FA, Carlo G, Grau JM, Azmitia M, Cabrera NJ, Chahin TJ, editors. Handbook of U S Latino psychology: developmental and community-based perspectives. Thousand Oaks: SAGE; 2009. pp. 81–98. [Google Scholar]
- 39.Martinez RO, Dukes RL. The effects of ethnic identity, ethnicity, and gender on adolescent well-being. J Youth Adolesc. 1997;26:503–16. [Google Scholar]
- 40.Painter MA, Holmes MD, Bateman J. Skin tone, race/ethnicity, and wealth inequality among new immigrants. Soc Forces. 2015;94:1153–85. [Google Scholar]