Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Jul 21.
Published in final edited form as: Eur J Psychol Assess. 2020 Jan 24;36(5):889–900. doi: 10.1027/1015-5759/a000567

The German Version of the Multidimensional Acculturative Stress Inventory (MASI) for Turkish-Origin Immigrants –Measurement Invariance of Filter Questions and Validation

Laura Scholaske a,b, Norma Rodriguez c, Nida Emel Sari b, Jacob Spallek d, Matthias Ziegler b,*, Sonja Entringer a,e,*
PMCID: PMC8293836  NIHMSID: NIHMS1608573  PMID: 34295123

Abstract

The Multidimensional Acculturative Stress Inventory (MASI) is an established measure of acculturative stress for people of Mexican origin living in the United States that has been associated with mental health outcomes in this population. We translated the MASI into German and adapted it for use with Turkish-origin immigrants in Germany. The MASI includes filter questions asking if a potentially stressful event had actually occurred before reporting the stress appraisal of these situations. Measurement invariance testing has become a standard practice to evaluate questionnaire translations, however, measurement invariance of filter questions has been scarcely studied. In Study I, we evaluated measurement invariance of the filter questions between a German-based Turkish sample (N = 233) and the Mexican-origin sample from the original study (N = 174) and could show partial strong factorial invariance for three of the four factors. In Study II, a validation study, relations between the German MASI scores and measures of acculturation and stress indicated discriminant validity. This study contributes to research on measurement invariance of filter questions, thereby providing a measure of acculturative stress that can be used in future research to understand the etiology of health disparities in Turkish-origin immigrants in Germany.

The Multidimensional Acculturative Stress Inventory (MASI; Rodriguez, Myers, Mira, Flores, & Garcia-Hernandez, 2002) is an established measure of acculturative stress for use with Mexican-origin adults living in the United States that was developed to address limitations in the few published acculturative stress measures (e.g., confounding of acculturative stress with minority-status and socioeconomic stress occupation, only assessing stress from the host culture and not from the heritage culture, and not going beyond the measurement of language pressures [Bai, 2016; Cervantes, Fisher, Padilla, & Napper, 2016; Hovey & King, 1996; Sandhu & Asrabadi, 1994]). Since its publication, the MASI has been used in several studies including adults living in the U.S. who trace their origins to Mexico and other Latin American countries (e.g., Basáñez, Dennis, Crano, Stacy, & Unger, 2014; Driscoll & Torres, 2013) and has also been adapted for use with ethnically diverse U.S.-college students (Castillo et al., 2015; Schwartz, Zamboanga, Rodriguez, & Wang, 2007). Until now, the MASI has not been used outside of the United States context. At the same time, there is no measure of acculturative stress available in German language. Thus, an aim of our paper is to present a German-language version of the MASI for the use with Turkish-origin immigrants.

Another aim of our paper is to evaluate the measurement invariance of the filter questions used in the MASI between the original English/Spanish version and the German translation. Psychological inventories typically include only one rating scale per item, but for some measures, such as the MASI, a filter question is added before inquiring about the rating of interest. For each item of the MASI, participants first indicate whether they have experienced a potentially stressful event in the context of acculturation in the past three months (filter question); if they respond affirmatively, they are asked to appraise the stressfulness of that situation (stress appraisal scale). In the case of the MASI, including a filter question is an important first query because some of the potentially stressful scenarios might not apply to all participants or the events may have occurred outside the specified period of time, thus, only participants who have experienced a given event are asked to complete the stress appraisal scale. The use of such filter questions constitutes a specific challenge for psychometrics in general and they have been scarcely addressed in measurement invariance testing so far. In case of the MASI, its filter questions measure the occurrence of events and therefore something different than the stress appraisal scale. Showing measurement invariance of these filter questions across groups ensures that the groups have a common understanding and psychological evaluation of the events in question which is particularly important given that the completion of the stress scale depends on the affirmative answer to the filter question.

The aim of this project was thus to test measurement invariance of filter questions of a German version of the MASI while at the same time providing validity evidence for this measure of acculturative stress that can be used in future research.

Acculturative Stress and its Link to Adverse Health Outcomes

Acculturation is defined as the process of adapting to the culture of the host country after migration that involves negotiating differing aspects of the heritage and host culture, and this affects immigrants (i.e., people who live in another country than where they were born) as well as their descendants who were born in the host country (Berry, 1997). Acculturation can involve changes in numerous cultural aspects such as values, beliefs, language use and preference, dietary habits, customs, and identity. Whereas immigrants have internalized their heritage culture and have to “learn” the new culture of the target country, their descendants tend to grow up in a hybrid culture and “learn” their heritage culture from the family. Thus, they move between the poles of two (or more) cultures and differ individually in their personal cultural orientation, that is, whether they are closer to just one of the two cultures or identify with both of them. The sources of stress within this context are based in the potentially conflicting demands of both cultures. For example, both members of the heritage and of the host culture confront immigrants and their descendants with the demand to know about and to practice their languages and ways of living. Not being proficient in one culture (e.g., not having competence in one language) can be stress-inducing when interacting with members of the host or heritage culture who may frown upon, exclude, or stereotype people who do not exhibit proficiency in that culture.

As acculturative stress is, by definition, associated with a reduction in mental health -- such as consequent anxiety, depression, feelings of marginality, and identity confusion -- (Berry, Kim, Minde, & Mok, 1987), it might play a key role in understanding how existing health disparities emerge in immigrant populations. Numerous studies show a link between acculturative stress and health outcomes among Latinos in the United States, including poorer self-rated health (Finch, Hummer, Kolody, & Vega, 2001), more anxiety, depressive somatic symptoms and suicidal ideations (Hovey, 2000; Hovey & Magaña, 2000). To conclude, acculturative stress might be a potential pathway between immigration and adverse health outcomes. Thus, it is warranted to explore its antecedents and correlates in more detail. A necessary first step is to have a psychometrically sound measure of acculturative stress available for different target groups.

The Multidimensional Acculturative Stress Inventory (MASI)

The MASI (Rodriguez et al., 2002) was constructed to measure acculturative stress in Mexican-origin adults in the United States. Its factors capture global components of acculturative stress that can be assumed to be universal for most immigrants, including heritage language pressures and pressure against acculturation as potentially arising in the contact with the heritage culture, and host language pressures and pressure to acculturate as presumably arising in the contact with the host culture. The MASI is suitable for translation into different languages as the phrasing of items are easily adaptable to other target groups by replacing the name of the relevant languages and cultures. Furthermore, the wording addresses respondents independent of generation status, whereas other acculturative stress scales target international students (Sandhu & Asrabadi, 1994; Bai, 2016) or native-born immigrant adolescents (Hovey & King, 1996). The MASI directly assesses stress associated with the acculturation process whereas other scales simultaneously target stress in other areas of life (Cervantes et al., 2016), or only assess the occurrence of the stressors but not the stress appraisal such as the Acculturative Stress Index (Haasen et al., 2007). This later scale is available in German but has some further constraints, for example, it has never been validated, includes subscales on stress in other areas of life (e.g., economic), and the items do not focus on a specific time window from the past.

Regarding the psychometric properties of the MASI scores, Rodriguez et al. (2002) provided evidence for high levels of reliability (i.e., internal consistency indicated by Cronbach’s α values ranging from .77 to .93 and Pearson correlation coefficients for test-retest reliability ranging between .53 and .84) and validity (i.e., strong associations with related constructs such as acculturation and stress) using a sample of 174 mostly women, urban, and highly educated Mexican-origin adults. Furthermore, Castillo et al. (2015) tested measurement invariance of the MASI factors as based on the stress appraisal scale across ethnicity, gender, and nativity in a sample of Latino and Asian American college students in the United States and confirmed the four-factor structure, strong factorial invariance across nativity and strict factorial invariance across ethnicity and gender.

A specific challenge for such an instrument is that it needs to be available in a variety of languages in order to include members of a cultural group with different language skills (e.g., Spanish and English for Mexican-origin immigrants in the U.S.) or to use the instrument with other cultural groups (e.g., Turkish-origin immigrants in Germany). A relevant question is whether the same underlying construct is being measured by the scales in different languages and across cultural groups. Therefore, it is a requirement to show measurement invariance between the language versions (Greiff & Scherer, 2018; Milfont & Fischer, 2010; Ziegler & Bensch, 2013). Only then, cross-cultural comparisons are permissible.

Why Focus on Turkish-Origin Immigrants in Germany?

Most studies in the context of acculturative stress and health have been conducted on Latino immigrants such as people of Mexican-origin and their descendants in the U.S. (Finch et al., 2001; Finch & Vega, 2003; Hovey, 2000; Hovey & Magaña, 2000). German-based immigrants from Turkey and their descendants share many similarities with Mexican-origin immigrants in terms of cultural and sociodemographic characteristics. Both groups stem from rather collectivistic cultures and migrated to countries where an individualistic culture is more prevalent (Hofstede, Hofstede, & Minkov, 2010). Thus, they might exhibit similar levels of discrepancies between their heritage and host cultures, which can arouse comparable levels of acculturative stress. Mexicans in the United States and Turkish-origin immigrants in Germany nowadays constitute one of the largest minority groups in the respective host countries, and they are at higher risk of social disadvantage which is perpetuated across generations (Fontenot, Semega, & Kollar, 2018; Robert Koch-Institut, 2015).

Social disadvantage is a major determinant of health disparities (Braveman, Cubbin, Egerter, Williams, & Pamuk, 2010), and Mexican and Turkish immigrants are indeed at increased risk for some physical and mental health adversities compared to the host population, including overweight/adiposity, unbeneficial health behaviors (e.g., unhealthy diet), and depressive and somatoform disorders (Robert Koch-Institut, 2015; Igde, Heinz, Schouler-Ocak, & Rössler, 2018; Vega, Rodriguez, & Gruskin, 2009; Velasco-Mondragon, Jimenez, Palladino-Davis, Davis, & Escamilla-Cejudo, 2016). Because some of these health adversities are known to be associated with acculturative stress as described above, the same observed link might explain health disparities among Turkish immigrants in Germany.

Study Aims

To sum up, research on measurement invariance of filter questions is scarce, and valid measures to assess acculturative stress in Turkish-origin immigrants in Germany are missing. Therefore, this project followed two aims: The aim of Study I was to test for measurement invariance of the filter questions between the original version of the MASI and a German translation that was adapted to immigrants from Turkey. In case of strong factorial invariance, comparisons of the MASI scores between Turkish and Mexican-origin respondents were permissible, but we assumed no group differences. The aim of Study II was to provide evidence of discriminant validity for scores of the German version regarding relationships with scores for acculturation and stress. We expected that the MASI scores based on the stress appraisal scale would be associated with perceived stress, and that the MASI scores based on the Occurrence of a stressor scale prompted by an affirmative response would be associated with different levels of acculturation (i.e., a stronger orientation towards the heritage culture would be associated with less acculturative stressors originating from contact with the heritage culture whereas a stronger orientation towards the host culture would be associated with less acculturative stress originating from contact with the host culture). Furthermore, we expected associations among MASI scores referring to stressors originating from contact with the heritage culture and among MASI scores referring to stressors originating from contact with the host culture.

Methods

Sample

N = 233 Turkish-origin first, second and third generation immigrants participated in an online study. The majority of the sample comprised second-generation immigrants, that is, they were born in Germany but at least one of their parents was born in Turkey. A smaller number were first-generation immigrants who were born in Turkey and migrated to Germany themselves, and only one participant was a third-generation immigrant, that is, the grandparents were born in Turkey and the respondent and parents were born in Germany. Sample characteristics of the Turkish group are presented in Table 1. This sample was used solely in study II; in study I, this sample was matched with the sample from the original MASI study from Rodriguez et al. (2002) that included N = 174 adults of Mexican origin residing in the greater metropolitan Los Angeles area (for further sample characteristics see Rodriguez et al., 2002). The majority of the Turkish sample was women, urban, and highly educated and therefore not representative of Turkish-origin immigrants in Germany especially in terms of educational attainment, albeit comparable to the Mexican sample, which had a similar selection bias. For example, 17% of the population in Germany have a university degree (Statistisches Bundesamt, 2018), while only 6.2% (first-generation) to 7.4% (second-generation) of Turkish-origin women and 13.7% (second-generation) to 14.3% (first-generation) of Turkish-origin men have a university degree (Schührer, 2018).

Table 1.

Sociodemographic characteristics of the Turkish sample (N = 233)

Frequency Mean (SD), range
Gender (female)
 Women 186 (80.87%)
 Men 44 (18.88 %)
 Missing 3 (0.25%)
Age 22.6 (4.84), 18-45
Generation status
 1st generation 17 (7.48%)
 2nd generation 212 (92.17%)
 3rd generation 1 (0.44%)
Language use
 Turkish 73 (32.20%)
 German 154 (67.80%)
Years of education 12.37 (1.28), 1-18
Highest educational degree
 Secondary school certificate or less 37 (15.91%)
 A-levels or comparable 138 (59.34%)
 Bachelor’s degree 36 (15.48%)
 Master’s degree 11 (4.73%)
 Completed job training 4 (1.72%)
 Other 1 (0.43%)
PSS score 2.04 (0.66), 0.30-3.80
FRACC orientation towards the heritage culture 3.95 (1.13), 0.10-6.00
FRACC orientation towards the host culture 3.78 (1.11), 1.10-6.00
Open in a new tab

Note. PSS = Perceived Stress Scale. FRACC = Frankfurt Acculturation Scale.

Our study was approved by the ethics committee of Humboldt-Universität zu Berlin and data collection was conducted in November 2017.

The online survey with the Turkish sample was conducted with the open source platform formr.org (Arslan & Tata, 2017). Participants stated consent to participate in the study. After completion of the questionnaires, the participants were informed about our aim to validate the German version of the MASI on participants of Turkish origin, and they received feedback on their responses to a personality short scale (Big Five Inventory-SOEP [Gerlitz & Schupp, 2005]). This personality scale was added simply for feedback purposes aimed at attracting more participants.

Measures

MASI.

The original MASI (Rodriguez et al., 2002) is available in English and Spanish and was designed to assess acculturative stress in adults of Mexican origin living in the United States. The scale contains 25 items loading on four factors: “Spanish Competency Pressures” (F1, items 1-7), “English Competency Pressures” (F2, items 8-14), “Pressure to Acculturate” (F3, items 15-20), and “Pressure Against Acculturation” (F4, items 21-25). The items describe potentially stressful situations and respondents first indicate on a filter question whether the stated events occurred during the past three months (occurrence of stressor; “0”, not applicable, “1”, applicable) and if so, appraise how stressful those events were on a 5-point rating-scale (stress appraisal; “1”, not at all stressful, 5, “extremely stressful”). The dichotomous items on the occurrence of stressors were used in the measurement invariance analyses of Study I; the MASI factors based on the dichotomous as well as the metric scale were used for the validation analyses in Study II.

Translation of the MASI.

The translation procedure followed the recommendations of Banville, Desrosiers, and Genet-Volet (2000). The items were translated from English into German by a fluent German speaker with advanced English language skills (see Appendix). Another fluent German speaker checked the translation for content and linguistic correctness. In case of discrepancies, the items were corrected together by both translators. Afterwards the German items were translated back into English by a fluent English speaker. The same person compared the content of the back-translation to the original version. Of the 25 items, six items were identical, and 14 items were consistent in content. Another five items had slight discrepancies and went through the procedure of translation-back-translation again. Afterwards, all German items were considered to be invariant in content to the original items.

Because Turkey hosts numerous ethnic groups and this diversity is also reflected among residents who immigrated from Turkey to Germany, we avoided phrasings referring to the Turkish culture, such as “Turkish way of doing things” and used the term “culture of origin” instead. For the items on the heritage language we nonetheless referred to Turkish language as it constitutes the major language for most groups immigrating from Turkey.

Acculturation.

The Frankfurt Acculturation Scale (FRACC), a 20-item scale that assesses acculturation among Turkish-origin immigrants in Germany (Bongard, Kelava, Sabic, Aazami-Gilan, & Kim, 2007), was administered. It includes two subscales, “Orientation Towards the Heritage Culture” (α = .86) and “Orientation Towards the Host Culture” (α = .84). Respondents indicated their agreement to the statements on a 7-point rating-scale (“0”, absolutely not, “6”, absolutely).

Furthermore, we assessed language use, a commonly-used proxy measure of acculturation, by asking participants to name their most commonly used language (coded as 0 = more Turkish, 1 = more German).

Psychological distress.

Participants completed the German version of the 10-item Perceived Stress Scale (PSS; α = .84) (Klein et al., 2016) that assesses the experience of psychological distress during the past four weeks. Respondents indicated how often they experienced a number of situations on a 5-point rating-scale (“0”, never, “4”, very often).

Personality traits.

We administered the 15-item Big Five Inventory (BFI-S; Gerlitz & Schupp, 2005) and provided feedback as an incentive for participation. We did not use the Big Five scores for any analyses.

Sociodemographic factors.

Participants gave information on their age, gender, family status, place of residence, profession, job situation, number of school years completed, monthly net income, their highest educational degree, and their parents’ highest educational degree.

Data Analysis

Study I.

Confirmatory factor analyses (CFA) is an appropriate method for testing measurement invariance across groups (Byrne, 2008; Meredith, 1993). First, we conducted single-group CFAs in the Turkish and Mexican group to evaluate the four-factor-structure of the MASI. We modified the measurement model until an acceptable fit was achieved, while critically monitoring the theoretical plausibility of the added relations (Heene, Hilbert, Draxler, Ziegler & Bühner, 2011). The measurement model that had the best fit in both groups was then selected as a final model. It should be noted that this is no longer a confirmatory approach, however, only with fitting models are interpretations of parameters feasible. The second step comprised multi-group CFAs for testing measurement invariance whereby restrictions were made successively in order to test configural, weak factorial, and strong factorial invariance. These models were each compared by their fit indices for testing the assumption that a restriction does not deteriorate model fit. When the model fit deteriorated, the restrictions were deleted for certain items, which allowed for testing partial measurement invariance. Power analyses for the differences in RMSEA between the models of the multi-group analyses were conducted with the semTools package and indicated the effect size of the RMSEA (H0: RMSEA= .067; H1: RMSEA = .082).

Though standard practice, recent simulation studies indicated that blindfold reliance on the proposed cutoffs by Hu and Bentler led to a high failure to reject false, misspecified models (Greiff & Heene, 2017). The criteria proposed by Hu and Bentler (1999) emerged from CFA with intelligence tests but are hardly ever met in the assessment of other psychological constructs and therefore may not apply. Hopwood and Donellan (2010) showed that seven commonly used valid omnibus personality scales did not reach the cut-offs (e.g., CFI ranged between .61-.79). One simulation study indicated that the CFI was slightly above .80 when factor loadings were low in personality measures (Heene et al., 2011). We therefore set the CFI-cutoff to ≥.80, given that the MASI factor loadings were expected to be low. Furthermore, the importance of modelling the misfit of a model (e.g., correlated residuals) has been highlighted, otherwise parameter estimations are not trustworthy (Heene et al., 2011). We therefore allowed for model modifications at all steps in the CFA. Confirmatory factor analyses were conducted using the package lavaan (Rosseel, 2012) in the statistic software R (R Core Team, 2017). We used the maximum likelihood estimation with robust (Huber-White) standard errors (MLR) and full-information maximum likelihood (FIML) for handling missing data. The proportion of missing values was below 3% per MASI item that assesses the occurrence of stressors.

Study II.

Discriminant validity analyses1 included examination of the relationships between the MASI factor mean scores and measures of acculturation and stress, as well as the relationships among the MASI factor mean scores. The PSS score was used as a measure of psychological distress. Measures of acculturation were the FRACC-subscales “Orientation Towards the Heritage Culture” and “Orientation Towards the Host Culture” and language use (Turkish vs. German) as a proxy-measure of acculturation. Pearson correlations between the MASI mean factor scores and the PSS, FRACC and language use were provided. Furthermore, regression analyses were conducted with PSS and FRACC subscales as the outcomes, the MASI mean factor scores as the main predictors, and age, gender and years of education as control variables. ΔR2 indicates the incremental amount of explained variance in the outcome variable by each MASI factor. All analyses were run separately for the MASI factors as based on the Occurrence of stressor items and as based on the stress appraisal items. The relationships among the MASI factors were depicted by intercorrelations. When discussing convergent and discriminant validity evidence, we refer to the definitions of Campbell and Fiske (1959) and also examine the idea of a nomological net proposed by Cronbach and Meehl (1955). This is in line with the editorial by Ziegler (2014) regarding the ABCs of test construction.

Study I - Results

Single Group Confirmatory Factor Analyses

We first fitted the four-factor model described in Rodriguez et al. (2002) separately to the data of the Mexican and the Turkish group (see ESM–Table 1). Due to bad model fit especially in the Turkish group, the models had to be modified, hence we added correlated residuals between items 3 and 5, 3 and 6, 5 and 9, 13 and 14, 10 and 13, 11 and 18, 13 and 14, 17 and 19, 19 and 21, and 21 and 23. These item pairs were either from the same factor each and/or were similar in content (e.g., item 5, “Since I don’t speak Spanish [Turkish] well, people have treated me rudely or unfairly” and item 9, “I have been discriminated against because I have difficulty speaking English [German]”). Adding these modifications improved the CFI, RMSEA, and SRMR. Descriptive statistics of the German items and the standardized factor loadings of the final measurement model for Turkish respondents are presented in Table 2. Most of the factor loadings were small in support of the cut-off of .80 for the CFI. All in all, these results indicate factorial validity for both language versions. At the same time, the correlated residuals implied that there was method variance in some of the items. A consequence of this could be that reliability estimates such as Cronbach’s alpha are positively biased. Therefore, McDonald’s Omega was utilized as an estimate of scale reliability, which indicated high reliability of all four scores for the Mexican sample (F1: ω = 0.91; F2: ω = 0.93; F3: ω = 0.88; F4: ω = 0.84) while for the Turkish sample the internal consistency estimates were acceptable (F1: ω = 0.74; F2: ω = 0.82; F3: ω = 0.68; F4: ω = 0.55).

Table 2.

Descriptive statistics and standardized factor loadings of the German MASI items

Items N Occurrence of Stressor (applicable) Stress Appraisal Mean (SD) Standardized factor loadings
Factor 1 – Turkish Competency Pressures
(1) 233 29 (12.5%) 2.36 (1.28) 0.578
(2) 232 20 (8.6 %) 2.53 (1.22) 0.656
(3) 232 23 (9.9 %) 2.91 (0.95) 0.658
(4) 233 12 (5.2 %) 2.25 (1.06) 0.453
(5) 232 12 (5.2 %) 3.50 (1.17) 0.386
(6) 232 23 (9.9 %) 2.83 (1.03) 0.396
(7) 232 15 (6.5 %) 2.87 (1.13) 0.371
Factor 2 – German Competency Pressures
(8) 232 3 (1.3 %) 2.00 (1.73) 0.426
(9) 232 13 (5.6 %) 3.23 (1.30) 0.578
(10) 231 16 (6.9 %) 3.67 (1.11) 0.828
(11) 232 21 (9.1 %) 3.15 (1.35) 0.505
(12) 232 21 (9.1 %) 3.30 (1.08) 0.459
(13) 232 8 (3.4 %) 3.14 (1.68) 0.708
(14) 233 24 (10.3 %) 2.70 (1.29) 0.345
Factor 3 – Pressure to Acculturate
(15) 232 139 (59.9 %) 3.38 (1.11) 0.668
(16) 233 209 (89.7 %) 3.82 (1.06) 0.420
(17) 231 76 (32.9 %) 3.05 (1.14) 0.309
(18) 233 65 (27.9 %) 3.08 (1.10) 0.490
(19) 231 92 (39.8 %) 3.26 (1.15) 0.358
(20) 230 113 (49.1 %) 3.6 (1.11) 0.373
(21) 231 83 (35.9 %) 3.31 (1.23) 0.528
Factor 4 – Pressure Against Acculturation
(22) 232 45 (19.4 %) 3.48 (1.28) 0.408
(23) 229 41 (17.9 %) 2.83 (1.28) 0.542
(24) 230 51 (22.2 %) 2.96 (1.19) 0.594
(25) 226 15 (6.6 %) 2.93 (1.21) 0.274
Open in a new tab

Multi-Group Confirmatory Factor Analyses

The results for the multi-group confirmatory factor models are shown in Table 3. The configural invariance model had an acceptable fit, but the test for weak factorial invariance deteriorated the model fit. Inspection of the loadings in the configural model revealed that item 8 (“I don’t speak English [German] or don’t speak it well.”) had the largest difference in loadings between both groups. This was expectable because respondents from the Mexican sample without English language proficiency had the option to choose the Spanish version of the questionnaire, whereas a Turkish version was not available and therefore Turkish respondents without sufficient German language skills were indirectly excluded from the survey. Hence, we established a partial weak factorial model where factor loadings were set equal for all item loadings except for item 8. The fit indices of this modified model were again acceptable and differences in fit indices indicated no substantial deterioration. Third, we tested for strong factorial invariance. The model fit deteriorated substantially and inspection of the intercepts was needed. A partial strong factorial invariance model where the intercepts were not set equal for the items 4, 6 (F1), 8, 12, 13, 14 (F2), 16, 17, 18 (F3), and 25 (F4) yielded an acceptable model fit without substantial deteriorations in comparison to the partial weak factorial invariance model. Because it is recommended that at least half of the items of a factor should be invariant (Steenkamp & Baumgartner, 1998; Vandenberg & Lance, 2000), we concluded partial strong factorial invariance for the factors F1 (5/7 items invariant), F3 (4/7 items invariant), and F4 (3/4 items invariant), whereas factor F2 (3/7 items invariant) did not meet this criterion. Taking a closer look at factor F2 reveals that items referring to language difficulty (e.g., item 8, “I don’t speak English or don’t speak it well”) were non-invariant while items referring to experiences of social pressure and exclusion due to language difficulties (e.g., item 9 “I have been discriminated against because I have difficulty speaking English.”) were invariant. RMSEA power for the model comparisons was around .99 in all multi-group models and therefore sufficiently high.

Table 3.

Results of multi-group CFA

AIC BIC χ2 df CFI RMSEA SRMR ΔCFI ΔRMSEA ΔSRMR
Configural invariance 5772.6 6502.3 989.079*** 518 0.846 0.067 [CI 0.061-0.073] 0.075 - -
Weak factorial invariance 5843.3 6480.7 1049.561*** 541 0.833 0.068 [CI 0.062-0.074] 0.085 0.013 −0.001 −0.010
Partial weak factorial invariancea 5784.8 6426.2 1012.479*** 540 0.845 0.066 [CI 0.060-0.071] 0.081 0.001 0.001 −0.006
Strong factorial invariance 6041.4 6598.6 1280.871*** 561 0.764 0.079 [CI 0.074-0.085] 0.093 0.081 −0.013 −0.015
Partial strong factorial invarianceb 5797.9 6395.2 1047.202*** 551 0.837 0.067 [CI 0.061-0.072] 0.083 0.008 −0.001 −0.002
Open in a new tab

Note.

*

p < .05,

**

p < .01,

***

p < .001.

CFI and RMSEA are scaled. Δ represents the difference between the less restricted model and the more restricted model.

a

partialized for item 8,

b

partialized for items 4, 6, 8, 12, 13, 14, 16, 17, 18, and 25.

Group Differences in Latent Factor Means

Mean comparisons across groups were then permissible for factors F1, F3, and F4. Based on the partial strong factorial model, there were significant mean differences across all factors indicating that Turkish respondents scored significantly lower than Mexican respondents, and latent effect size d (Hancock, 2001) suggests that these differences were small (F3) to medium sized (F1, F4) (F1: d = -0.449, SE = 0.04, p < .001; F3: d = -0.155, SE = 0.04, p < .001; F4: d = -0.353, SE = 0.01, p < .001).

Study I – Discussion of Findings

In sum, we could replicate the four-factor structure of the MASI in the Turkish sample and showed partial strong factorial invariance for MASI factor F1, Turkish Competency Pressures, F3, Pressure to Acculturate, and F4, Pressure Against Acculturation, which allowed for group comparisons where the Turkish group scored lower than the Mexican group. For factor F2, German Competency Pressures, only partial weak factorial invariance was indicated.

As we could show partial strong factorial invariance across the English and German MASI, group comparisons for factors F1, F3, and F4 are permissible, and this is a relevant contribution to the tools of cross-cultural research. The Turkish group scored significantly lower in these factors, suggesting that they experience fewer potentially stressful situations in the context of acculturation, although we expected no group differences due to some cultural and sociodemographic similarities between the groups. However, these differences have to be interpreted in light of sample selection bias, and the unequal distribution of generation status (45% of the Mexican group vs. 8% of the Turkish group were first-generation immigrants) might carry weight here. In the study by Castillo et al. (2015), foreign-born respondents also scored significantly higher across factors F3 and F4 when compared to U.S.-born respondents.

It was necessary to modify the models during the CFAs in order to obtain a better model fit, and some items did not reach high levels of measurement invariance. However, we acknowledge that non-invariance is also a relevant finding (Greiff & Scherer, 2018), and as stated in the Result sections, most deviations were expectable. Some of the global-fit-indices of the CFA were in line with the proposed cut-off values by Hu and Bentler (1999) after model modifications, but the CFI was never >.90, the SRMR was merely on the border, and the chi-square value was significant in all models. However, when adapting other proposed cutoffs that take into account the type of test and size of factor loadings (Heene, Hilbert, Draxler, Ziegler, & Buehner, 2011; Hopwood & Donnellan, 2010) and a given sample size (Meade, Johnson, & Braddy, 2008), we could continue our analyses.

Study II – Results

Discriminant Validity Analyses

The Pearson correlations indicated that Occurrences of Pressure to Acculturate (F3) were related to a higher PSS score (r(222) = .26, p < .01), whereas the other MASI factors were unrelated to the PSS score (see ESM–Table 2). The correlations between the MASI factors and the acculturation measures emerged in the expected directions: Occurrences of Turkish Competency Pressures (F1) were significantly related to more German language use (r(225) = .18, p < .01), and less Orientation Towards the Heritage Culture (r(215) = −.33, p < .01), and vice versa, Occurrences of German Competency Pressures (F2) were related to more Turkish than German language use (r(225) = −.17, p < .01) and a lower Orientation Towards the Host Culture (r(215) = −.24, p < .01). Occurrences of Pressure to Acculturate (F3) were related to a significantly higher Orientation Towards the Heritage Culture (r(220) = .30, p < .01) and a lower Orientation Towards the Host Culture (r(220) = −.52, p < .01), whereas Occurrences of Pressure Against Acculturation (F4) were related to a significantly lower Orientation Towards the Heritage Culture (r(220) = −.17, p < .05).

All MASI factors as based on the stress appraisal items had a significant positive relationship with the PSS score (F1: r(67) = .24, p < .05; F2: r(53) = .38, p < .01; F3: r(208) = .19, p < .01; F4: r(91) = .29, p < .01). They were unrelated to the measures of acculturation, except for higher stress due to Pressure to Acculturate (F3) was related to a lower Orientation Towards the Host Culture (r(206) = −.26, p < .01).

Despite the conceptual overlap of the constructs, all of these correlations were rather small or medium sized, indicating discriminant validity (i.e., measures discriminate as constructs are dissimilar) rather than convergent validity (i.e., measures converge on the same construct as constructs are similar) of the MASI factors (Campbell & Fiske, 1959).

Regression analyses were run with the PSS score, FRACC orientation towards the heritage culture and FRACC orientation towards the host culture as the criteria with singular MASI factors as the predictors adjusted for age, gender, and years of education (see ESM–Tables 3, 4 and 5). The same relationships emerged as in the bivariate correlations. As indicated by ΔR2, the MASI factors as based on the occurrence of stressors explained greater variance in acculturation (e.g., 10.6% by F1 in FRACC Orientation Towards the Heritage Culture and 25.6% by F3 in FRACC Orientation Towards the Host Culture) than in perceived stress where Factors F1, F2, and F4 explained below 1% of the variance and F3 around 7%. In contrast, MASI factors as based on the stress appraisal scale explained greater variance in perceived stress (e.g., 15.2% by F3) than in acculturation where the incremental variance was maximum 3% in FRACC Orientation Towards the Heritage Culture and FRACC Orientation Towards the Host Culture.

There were significant positive intercorrelations between the factors that described pressures from the heritage culture, i.e., F1, Turkish Competency Pressures and F4, Pressure Against Acculturation (occurrence of stressor: r(231) = .32, p < .01; stress appraisal: r(40) = .62, p < .01), and between the factors on pressures from the host culture, i.e., F2, German Competency Pressures and F3, Pressure to Acculturate, (occurrence of stressor: r(231) =.23, p < .01; stress appraisal: r(54) = .35, p < .01) (see ESM–Table 2).

Study II – Discussion of Findings

In sum, the analyses provide evidence for discriminant validity of the MASI scores. The stress appraisal of acculturative stressors was mostly related to psychological distress whereas the occurrences of acculturative stressors were more related to different levels of acculturation in the expected directions, and matching factors of the MASI correlated among each other. The small correlations between the MASI factors and the PSS highlighted that the MASI indeed measures stress, but also something beyond what is captured by the PSS which is presumably the acculturative aspect of the stress. To experience certain stressors thus seems to be a matter of acculturation – being more oriented towards the heritage culture and greater usage of the Turkish language involves less Turkish competency stress and less pressure against acculturation, and vice versa being more oriented towards the host culture and greater usage of the German language involves naturally less German competency stress and less pressure to acculturate. The intercorrelations among the factors F1, Turkish Competency Pressures and F4, Pressure Against Acculturation, and among F2, German Competency Pressures and F3, Pressure to Acculturate suggest that the emergence of language competency stress is not isolated from cultural pressures.

General Discussion

Implications

There have been numerous studies on measurement invariance in the cross-cultural context to warrant between-group comparability of scales (e.g., Frenzel, Thrash, Pekrun, & Goetz, 2007; Malham & Saucier, 2014), but the specific case of filter questions has been understudied in measurement invariance testing so far. To our knowledge, this is among the first studies to test measurement invariance of filter questions. We find this to be a relevant endeavor because filter questions precede another scale of interest, and measurement variance in filter questions may therefore also impact the assessment of the subsequent scale. Although the model fit indicated a good match of the ascribed items to the factors overall, the model modifications indicated that some items might not be unidimensional. These items might possibly tap into other constructs or subfacets not yet operationalized. We therefore used factor mean scores (not sum scores), because these are controlled for the correlated residuals.

However, to this end, measurement invariance of the MASI stress appraisal scale between the Mexican- and Turkish-origin group were not tested, therefore, group comparisons based on the stress appraisal scale are not permissible. In the Turkish sample, the number of observations was extremely reduced on the stress appraisal scale as many acculturative stressors were reported as inapplicable. Hence investigating measurement invariance for the stress appraisal scale would require further analyses with a much larger sample. However, these limitations do not restrict the usefulness of the German translation of the MASI for its application in research. As can be followed from the validity analyses, the MASI factors are associated with psychological distress and levels of acculturation and explained variance in these constructs. Therefore, we suggest using the German MASI in research with Turkish-origin immigrants.

Limitations and Outlook

The results of our studies are promising, but some further limitations should be considered. As stated above, measurement invariance of the stress appraisal scale needs to be tested in a larger sample with a higher share of first-generation immigrants to conclude full measurement invariance. In Study I, modifications to the models were necessary at all stages of the CFA to improve the model fit indices. This shows that some items might not be unidimensional and furthermore, the correlated residuals need to be replicated. Factor F2, German Competency Pressures, turned out to be problematic for comprehensible reasons as outlined above. A Turkish translation of the inventory that can be applied to Turkish-origin immigrants with greater proficiency in Turkish than German (i.e., first-generation in particular) is therefore needed and is currently being developed by our research group to test for strong factorial invariance for this factor. The Turkish sample was biased and consisted mainly of female, urban, second-generation immigrant university students. Future studies should also investigate measurement invariance across gender, generation status, education, different ethnic groups who originate from Turkey, as well as other immigrant groups in Germany and other countries in more diverse samples. The MASI could be also translated into other languages.

Taken together, we can now provide a measure to conduct group-comparisons with Mexican-origin immigrants regarding the occurrences of acculturative stressors and to assess acculturative stress in Turkish-origin immigrants in Germany. Future studies could now apply the German MASI to analyze its relationship to health disparities in Turkish-origin immigrants and thereby contribute knowledge on the underlying psychological and cultural adaptation processes and the role of stress in this context.

Supplementary Material

Supplemental tables

All ESM-tables can be found in the document MASI_ESM.docx.

ESM-Table 1. Model fit indices of single-group CFA.

ESM- Table 2. Means, standard deviations, and bivariate correlations between the MASI variables and related constructs in the Turkish group.

ESM-Table 3. Regression analyses with psychological distress as the criterion.

ESM-Table 4. Regression analyses with FRACC Orientation towards the heritage culture as the criterion.

ESM-Table 5. Regression analyses with FRACC Orientation towards the host culture as the criterion.

Acknowledgments

The preparation of this manuscript was supported by German Research Foundation (DFG) grant EN 851/2-1 and National Institutes of Health (NIH) grant R01 MD-010738 to Sonja Entringer.

Appendix

MASI items in German language and adapted to German-based Turkish-origin immigrants

Factor 1: “Turkish Competency Pressures”
Item 1: “Ich spreche Türkisch nicht oder nicht gut.”
Item 2: “Ich fühle mich unwohl, wenn ich unter Leuten bin, die nur Türkisch sprechen.”
Item 3: “Ich empfinde Druck, Türkisch zu lernen”
Item 4: “Es fällt mir schwer, andere zu verstehen, wenn sie Türkisch sprechen.”
Item 5: “Weil ich nicht gut Türkisch spreche, wurde ich unfreundlich oder unfair behandelt.”
Item 6: “Es stört mich, wenn Leute davon ausgehen, dass ich Türkisch spreche.”
Item 7: “Ich wurde diskriminiert, weil ich Schwierigkeiten habe, Türkisch zu sprechen.”
Factor 2: “German Competency Pressures”
Item 8: “Ich spreche Deutsch nicht oder nicht gut.”
Item 9: “Ich wurde diskriminiert, weil ich Schwierigkeiten habe, Deutsch zu sprechen.”
Item 10: “Weil ich nicht gut Deutsch spreche, wurde ich unfreundlich oder unfair behandelt.”
Item 11: “Ich empfinde Druck, Deutsch zu lernen.”
Item 12: “Es stört mich, dass ich Deutsch mit Akzent spreche.”
Item 13: “Es fällt mir schwer, andere zu verstehen, wenn sie Deutsch sprechen.”
Item 14: “Ich fühle mich unwohl, wenn ich unter Leuten bin, die nur Deutsch sprechen können.”
Factor 3: “Pressure to Acculturate”
Item 15: “Es stört mich, wenn Leute mich dazu drängen, mich an die deutsche Art und Weise, Dinge zu tun, anzupassen.”
Item 16: “Es stört mich, wenn Leute die Werte meiner Herkunftskultur nicht respektieren.”
Item 17: “Aufgrund meines kulturellen Hintergrunds fällt es mir schwer, mich bei den Deutschen anzupassen.”
Item 18: “Ich fühle mich unwohl, wenn andere erwarten, dass ich die deutsche Art und Weise, Dinge zu tun, kenne.”
Item 19: “Ich fühle mich von Deutschen nicht akzeptiert.”
Item 20: “Ich fühle mich unwohl, wenn ich zwischen der deutschen Art und Weise, Dinge zu tun, und der meiner Herkunftskultur wählen muss.”
Item 21: “Leute behandeln mich geringschätzig, wenn ich türkische Bräuche praktiziere.”
Factor 4: “Pressure Against Acculturation”
Item 22: “Ich hatte Konflikte mit anderen, weil ich deutsche Bräuche denen meiner Herkunftskultur vorziehe.”
Item 23: “Leute behandeln mich geringschätzig, wenn ich deutsche Bräuche praktiziere.”
Item 24: “Ich fühle mich unwohl, wenn andere erwarten, dass ich die Art und Weise, Dinge zu tun wie in meiner Herkunftskultur üblich, kenne.”
Item 25: “Ich fühle mich unwohl, da meine Familienmitglieder die Art und Weise, Dinge zu tun wie in unserer Herkunftskultur üblich, nicht kennen.”
Open in a new tab

Footnotes

1

When discussing convergent and discriminant validity evidence, we refer to the definitions of Campbell & Fiske (1959) and also examine the idea of a nomological net proposed by Cronbach & Meehl (1955). This is in line with the editorial by Ziegler (2014) regarding the ABCs of test construction.

References

  1. Arslan RC, & Tata CS (2017). formr.org survey software (Version v0.16.14). 10.5281/zenodo.1000787 [DOI] [Google Scholar]
  2. Bai J (2016). Development and validation of the Acculturative Stress Scale for Chinese college students in the United States (ASSCS). Psychological Assessment, 28, 443–447. 10.1037/pas0000198 [DOI] [PubMed] [Google Scholar]
  3. Banville D, Desrosiers P, & Genet-Volet Y (2000). Translating questionnaires and inventories using a cross-cultural translation technique. Journal of Teaching in Physical Education, 19, 374–387. 10.1123/jtpe.19.3.374 [DOI] [Google Scholar]
  4. Basáñez T, Dennis JM, Crano WD, Stacy AW, & Unger JB (2014). Measuring acculturation gap conflicts among Hispanics: Implications for psychosocial and academic adjustment. Journal of Family Issues, 35, 1727–1753. https://doi-org.ccl.idm.oclc.org/10.1177/0192513X13477379 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Berry JW (1997). Immigration, acculturation, and adaptation. Applied Psychology, 46, 5–34. 10.1111/j.1464-0597.1997.tb01087.x [DOI] [Google Scholar]
  6. Berry JW, Kim U, Minde T, & Mok D (1987). Comparative studies of acculturative stress. The International Migration Review, 21, 491–511. 10.1177/019791838702100303 [DOI] [Google Scholar]
  7. Bongard S, Kelava A, Sabic M, Aazami-Gilan D, & Kim YB (2007). Akkulturation und gesundheitliche Beschwerden bei drei Migrantenstichproben in Deutschland [Acculturation and health complaints in three migrant samples in Germany]. In Eschenbeck J, Heim-Dreger U, Kohlmann CW (Eds.), Beiträge zur Gesundheitspsychologie, p. 53. Schwäbisch-Gmünd, Germany: Pädagogische Hochschule 752 Schwäbisch Gmünd. [Google Scholar]
  8. Braveman PA, Cubbin C, Egerter S, Williams DR, & Pamuk E (2010). Socioeconomic disparities in health in the United States: What the patterns tell us. American Journal of Public Health, 100, 186–196. 10.2105/AJPH.2009.166082 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Byrne BM (2008). Testing for multigroup equivalence of a measuring instrument: A walk through the process. Psicothema, 20, 872–882. [PubMed] [Google Scholar]
  10. Campbell DT, & Fiske DW (1959). Convergent and discriminant validation by the multitrait-multimethod matrix. Psychological Bulletin, 56, 81–105. 10.1037/h0046016 [DOI] [PubMed] [Google Scholar]
  11. Castillo LG, Cano MA, Yoon M, Jung E, Brown EJ, Zamboanga BL, … & Whitbourne SK. (2015). Factor structure and factorial invariance of the Multidimensional Acculturative Stress Inventory. Psychological Assessment, 27, 915–924. https://doi-org.ccl.idm.oclc.org/10.1037/pas0000095 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Cervantes RC, Fisher DG, Padilla AM, & Napper LE (2016). The Hispanic Stress Inventory Version 2: Improving the assessment of acculturation stress. Psychological Assessment, 28, 509–522. 10.1037/pas0000200 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Driscoll MW, & Torres L (2013). Acculturative stress and Latino depression: The mediating role of behavioral and cognitive resources. Cultural Diversity and Ethnic Minority Psychology, 19, 373–382. https://doi-org.ccl.idm.oclc.org/10.1037/a0032821 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Finch BK, Hummer RA, Kolody B, & Vega WA (2001). The role of discrimination and acculturative stress in the physical health of Mexican-origin adults. Hispanic Journal of Behavioral Sciences, 23, 399–429. 10.1177/0739986301234004 [DOI] [Google Scholar]
  15. Finch BK, & Vega WA (2003). Acculturation stress, social support, and self-rated health among Latinos in California. Journal of Immigrant Health, 5, 109–117. 10.1023/A:1023987717921 [DOI] [PubMed] [Google Scholar]
  16. Fontenot K, Semega J, & Kollar M (2018). Income and poverty in the United States: 2017. (Current Population Reports, P60-263). Retrieved from: https://www.census.gov/library/publications/2018/demo/p60-263.html [Google Scholar]
  17. Frenzel AC, Thrash TM, Pekrun R, & Goetz T (2007). Achievement emotions in Germany and China: A cross-cultural validation of the Academic Emotions Questionnaire—Mathematics. Journal of Cross-Cultural Psychology, 38, 302–309. 10.1177/0022022107300276 [DOI] [Google Scholar]
  18. Gerlitz J-Y, & Schupp J (2005). Zur Erhebung der Big-Five-basierten Persönlichkeitsmerkmale im SOEP [On surveying the Big Five-based personality traits in SOEP]. DIW Research Notes 4, Berlin: German Institute for Economic Research-. [Google Scholar]
  19. Greiff S, & Heene M (2017). Why psychological assessment needs to start worrying about model fit. European Journal of Psychological Assessment, 33, 313–317. 10.1027/1015-5759/a000450 [DOI] [Google Scholar]
  20. Greiff S, & Scherer R (2018). Still comparing apples with oranges? Some thoughts on the principles and practices of measurement invariance testing. European Journal of Psychological Assessment, 34, 141–144. 10.1027/1015-5759/a000487 [DOI] [Google Scholar]
  21. Haasen C, Levit O, Gelbert A, Foroutan N, Norovjav A, Sinaa M, & Demiralay C (2007). Zusammenhang zwischen psychischer Befindlichkeit und Akkulturation bei Migranten [Relationship between mental distress and acculturation among migrants]. Psychiatrische Praxis, 34, 339–342. 10.1055/s-2007-971014 [DOI] [PubMed] [Google Scholar]
  22. Heene M, Hilbert S, Draxler C, Ziegler M, & Buehner M (2011). Masking misfit in confirmatory factor analysis by increasing unique variances: A cautionary note on the usefulness of cutoff values of fit indices. Psychological Methods, 16, 319–336. 10.1037/a0024917 [DOI] [PubMed] [Google Scholar]
  23. Hofstede G, Hofstede GJ, & Minkov M (2010). Cultures and organizations: Software of the mind intercultural cooperation and its importance for survival. New York, NY: McGraw-Hill. [Google Scholar]
  24. Hopwood CJ, & Donnellan MB (2010). How should the internal structure of personality inventories be evaluated? Personality and Social Psychology Review, 14, 332–346. 10.1177/1088868310361240 [DOI] [PubMed] [Google Scholar]
  25. Hovey JD (2000). Acculturative stress, depression, and suicidal ideation in Mexican immigrants. Cultural Diversity and Ethnic Minority Psychology, 6, 134–151. 10.1037/1099-9809.6.2.134 [DOI] [PubMed] [Google Scholar]
  26. Hovey J, & King C (1996). Acculturative stress, depression, and suicidal ideation among immigrant and second-generation Latino adolescents. Journal of the American Academy of Child & Adolescent Psychiatry, 35, 1183–92. 10.1023/A:1009556802759 [DOI] [PubMed] [Google Scholar]
  27. Hovey JD, & Magaña C (2000). Acculturative stress, anxiety, and depression among Mexican immigrant farmworkers in the Midwest United States. Journal of Immigrant Health, 2, 119–131. 10.1023/A:1009556802759 [DOI] [PubMed] [Google Scholar]
  28. Hu L, & Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 1–55. 10.1080/10705519909540118 [DOI] [Google Scholar]
  29. Igde E, Heinz A, Schouler-Ocak M, & Rössler W (2018). Depressive und somatoforme Störungen bei türkeistämmigen Personen in Deutschland [Depressive and somatoform disorders in persons of Turkish origin in Germany]. Der Nervenarzt, 90, 25–34. 10.1007/s00115-018-0602-5 [DOI] [PubMed] [Google Scholar]
  30. Klein EM, Brähler E, Dreier M, Reinecke L, Müller KW, Schmutzer G, … Beutel ME (2016). The German version of the Perceived Stress Scale – Psychometric characteristics in a representative German community sample. BMC Psychiatry, 16, 159–169. 10.1186/s12888-016-0875-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Meade A, Johnson EC, & Braddy PW (2008). Power and sensitivity of alternative fit indices in tests of measurement invariance. Journal of Applied Psychology, 93, 568–592. 10.1037/0021-9010.93.3.568 [DOI] [PubMed] [Google Scholar]
  32. Meredith W (1993). Measurement invariance, factor analysis and factorial invariance. Psychometrika, 58, 525–543. 10.1007/BF02294825 [DOI] [Google Scholar]
  33. Milfont TL, & Fischer R (2010). Testing measurement invariance across groups: Applications in cross-cultural research. International Journal of Psychological Research, 3, 111–121. 10.21500/20112084.857 [DOI] [Google Scholar]
  34. R Core Team. (2017). R: A language and environment for statistical computing. R Foundation for Statistical Computing. Vienna, Austria. [Google Scholar]
  35. Robert Koch-Institut. (2015). Gesundheitsberichterstattung des Bundes. Gesundheit in Deutschland. Robert Koch-Institut (Hrsg) (2015) Gesundheit in Deutschland. [Health in Germany.] Gesundheitsberichterstattung des Bundes. Gemeinsam getragen von RKI und Destatis. RKI, Berlin. Berlin: Robert-Koch-Institut. 10.25646/3173 [DOI] [Google Scholar]
  36. Rodriguez N, Myers HF, Mira CB, Flores T, & Garcia-Hernandez L (2002). Development of the Multidimensional Acculturative Stress Inventory for adults of Mexican origin. Psychological Assessment, 14, 451–461. 10.1037//1040-3590.14.4.451 [DOI] [PubMed] [Google Scholar]
  37. Rosseel Y (2012). An R package for structural equation modeling and more. Journal of Statistical Software, 48, 1–36. 10.18637/jss.v048.i02 [DOI] [Google Scholar]
  38. Sandhu DS, & Asrabadi BR (1994). Development of an acculturative stress scale for international students: Preliminary findings. Psychological Reports, 75, 435–448. 10.2466/pr0.1994.75.1.435 [DOI] [PubMed] [Google Scholar]
  39. Schührer S (2018). Türkeistämmige Personen in Deutschland: Erkenntnisse aus der Repräsentativuntersuchung “Ausgewählte Migrantengruppen in Deutschland 2015” (RAM) [Turkish-origin persons in Germany. Findings from the representative study “Selected migrant groups in Germany 2015” (RAM).]. Bundesamt für Migration und Flüchtlinge: Nürnberg. [Google Scholar]
  40. Schwartz SJ, Zamboanga BL, Rodriguez L, & Wang SC (2007). The structure of cultural identity in an ethnically diverse sample of emerging adults. Basic and Applied Social Psychology, 29, 159–173. https://doi-org.ccl.idm.oclc.org/10.1080/01973530701332229 [Google Scholar]
  41. Statistisches Bundesamt (2018). Bildungsstand der Bevölkerung. Ergebnisse des Mikrozensus 2017. [Educational Status of the Population. Results of the Microcensus 2017.] Retrieved from https://www.destatis.de/DE/Themen/Gesellschaft-Umwelt/Bildung-Forschung-Kultur/Bildungsstand/Publikationen/Downloads-Bildungsstand/bildungsstand-bevoelkerung-5210002177004.pdf?__blob=publicationFile [Google Scholar]
  42. Steenkamp JEM, & Baumgartner H (1998). Assessing measurement invariance in cross‐national consumer research. Journal of Consumer Research, 25, 78–90. 10.1086/209528 [DOI] [Google Scholar]
  43. Vandenberg RJ, & Lance CE (2000). A review and synthesis of the measurement invariance literature: Suggestions, practices, and recommendations for organizational research. Organizational Research Methods, 3, 4–70. 10.1177/109442810031002 [DOI] [Google Scholar]
  44. Vega WA, Rodriguez MA, & Gruskin E (2009). Health disparities in the Latino population. Epidemiologic Reviews, 31, 99–112. 10.1093/epirev/mxp008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Velasco-Mondragon E, Jimenez A, Palladino-Davis AG, Davis D, & Escamilla-Cejudo JA (2016). Hispanic health in the USA: A scoping review of the literature. Public Health Reviews, 37, 1–27. 10.1186/s40985-016-0043-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Ziegler M, & Bensch D (2013). Lost in translation: Thoughts regarding the translation of existing psychological measures into other languages. European Journal of Psychological Assessment, 29, 81–83. 10.1027/1015-5759/a000167 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental tables

All ESM-tables can be found in the document MASI_ESM.docx.

ESM-Table 1. Model fit indices of single-group CFA.

ESM- Table 2. Means, standard deviations, and bivariate correlations between the MASI variables and related constructs in the Turkish group.

ESM-Table 3. Regression analyses with psychological distress as the criterion.

ESM-Table 4. Regression analyses with FRACC Orientation towards the heritage culture as the criterion.

ESM-Table 5. Regression analyses with FRACC Orientation towards the host culture as the criterion.

NIHMS1608573-supplement-Supplemental_tables.docx (37.1KB, docx)

RESOURCES