ATTITUDES TOWARDS MULTICULTURALISM 1

Attitudes towards Multiculturalism in : Measurement Invariance and Factor

Structure of the Multicultural Ideology Scale

Maria Stogianni1

Elke Murdock1

Jia He2, 3

Fons van de Vijver2, 4, 5, 6

1University of Luxembourg, Luxembourg

2Tilburg University, the

3DIPF Leibniz Institute for Research and Information in Education,

4Higher School of Economics, Russia

5North-West University, South Africa

6University of Queensland, Australia

This is a preprint. The paper is in press at the International Journal of Intercultural Relations

.

Please address correspondence to Maria Stogianni, email: [email protected]

This research was supported by a grant from the Fonds National de la Recherche, Luxembourg

[C16/SC/11337403/SWITCH/Murdock] ATTITUDES TOWARDS MULTICULTURALISM 2

Abstract

The present study examined the dimensionality and the measurement invariance of the

Multicultural Ideology Scale (MCI; Berry & Kalin, 1995), and mean differences across different cultural groups within the multilingual, multicultural context of Luxembourg. Luxembourg is a unique context to study attitudes towards diversity because 47% of the citizens are non-nationals, and minority and majority are increasingly difficult to define. Our sample included 1,488 participants from diverse ethnic backgrounds who completed the survey in German, French or

English. In contrast to previous findings, our analyses on responses to the MCI scale revealed a two-dimensional structure, distinguishing between positive and negative attitudes towards multiculturalism. The factor structure was partially invariant across ethnocultural groups:

Configural and metric invariance were established across natives and non-natives and different language versions. Scalar invariance was only established across gender groups. Natives and male participants reported more negative attitudes towards multiculturalism. We discuss the importance of assessing measurement invariance and provide recommendations to improve the assessment of psychological multiculturalism.

Keywords: diversity, multicultural ideology, attitudes, measurement invariance ATTITUDES TOWARDS MULTICULTURALISM 3

Attitudes towards Multiculturalism in Luxembourg: Measurement Invariance and Factor

Structure of the Multicultural Ideology Scale

Ethnic and cultural diversity is on the rise in most western European countries. The 21st century has in fact been called the age of migration (Castles & Miller, 2009) and this ongoing migration flow has resulted in major population changes worldwide. Due to increasing migration rates, most capital cities have been transformed into cosmopolitan hubs. Increases in migration flows have been also observed in smaller and less densely populated countries. Luxembourg, for example, has experienced great growth in its foreign-born population since the beginning of the

20th century (World Population Review, 2019). In the last 50 years, the total population nearly doubled. This growth is largely due to the foreign population percentage, which amounted to

18.4% in 1971 and 47.5% in 2019 (Statec, 2019). The total number and the demographic composition of the foreign population increased in this time period. Such demographic changes have particular implications for intergroup relations in modern societies. The boundaries between distinct cultural groups become permeable and people of different origins need to find ways of living together in new societies (Maehler & Schmidt-Denter, 2012). Both the autochthonous and the allochthonous populations grow up with multiple cultural influences and experience numerous opportunities for transnational contact. They gradually engage in a process known as acculturation. Acculturation refers to mutual cultural and psychological changes that occur when different cultural groups come into contact (Berry, 2016; Montreuil & Bourhis, 2001). It can affect multiple domains, such as attitudes, values, language, and behavior (Van Oudenhoven et al., 2006). As pointed out by Straub (2003) people become only aware of their cultural frames of reference in comparison with other cultural realities. A meeting of people of different cultural backgrounds also implies a meeting of different cultural frames of reference. In the past, ATTITUDES TOWARDS MULTICULTURALISM 4 acculturation research mainly focused on the immigrant perspective - the acculturation processes people face when they move into a relatively homogenous host society. Yet, migration patterns are changing. The diminishing size of native majority groups and the disappearance of a dominant culture in some countries are some examples (Van Oudenhoven & Ward, 2013).

Acculturation processes become increasingly complex and a critical question refers to how this diversity is experienced and managed by individuals living within these diverse societies.

A growing body of research investigates how people deal with increasing diversity and their attitudes towards multiculturalism and migration. Most of these studies report several factors associated with multiculturalism, including individual difference and contextual variables that shape multicultural attitudes (Verkuyten & Martinovic, 2006; Verkuyten, 2005a), advantages and disadvantages of living in culturally diverse societies (Stuart & Ward, 2019), and acculturation preferences of different ethnic groups (Arends-Toth & van de Vijver, 2003;

Verkuyten and Thijs, 2002). Understanding multicultural attitudes of majority and minority group members is particularly important in fostering positive intergroup relations but also in implementing policies that promote cultural diversity (Arends-Toth & van de Vijver, 2003).

However, conceptualizations of multiculturalism can vary across different cultural contexts; therefore, it is essential to assess the invariance of the construct and the measurement. Invariance assessment indicates whether and at which level data collected from different groups are comparable. It is necessary to gauge comparability of data representing diverse populations and lay a solid foundation for valid cross-cultural comparisons (Van de Vijver & He, 2017).

To this end, the present paper summarizes the current state of knowledge regarding the assessment of multicultural attitudes and provides empirical evidence from a unique social context that has grown increasingly diverse over the past years, the Grand Duchy of ATTITUDES TOWARDS MULTICULTURALISM 5

Luxembourg. Findings from Luxembourg can offer valuable insight regarding intergroup phenomena in super-diverse areas that have experienced major demographic shifts. Due to its multiethnic demographic composition, Luxembourg constitutes a natural laboratory to study intergroup relations in the context of diversity (Murdock, 2016). In this case, diversity does not only refer to ethnic diversity, but to the coexistence of various social groups with differential migration statuses, labour market experiences, linguistic and educational backgrounds, religious affiliations, age, and gender profiles (Meissner & Vertovec, 2015; Vertovec, 2007). The interplay of all these factors indicates that several dimensions can influence intergroup processes.

In the first section of the article, we introduce the concept of multiculturalism and present different definitions cited in the literature. We describe several dimensions of multiculturalism that contributed to the development of the Multicultural Ideology Scale (MCI; Berry & Kalin,

1995a), the first attitude scale specifically developed to measure the endorsement of multiculturalism. Then, we discuss methodological issues in the assessment of multicultural attitudes followed by an overview of relevant research findings.

The rest of the paper presents the empirical study conducted in Luxembourg. The study evaluates the applicability of the Multicultural Ideology Scale in a culturally diverse and multilingual environment. We examine the factor structure of the scale and its measurement invariance across three different language versions and ethnic groups. Previous studies tested the invariance of the scale among majority and minority group members living in the Netherlands

(Schalk-Soekar, 2007). In our study, the scale is tested among various cultural groups in a country where three national languages and English are widely spoken. Responses to different language versions are compared within the same cultural context and perceptions of multiculturalism are assessed in a highly heterogeneous society where foreign residents make up ATTITUDES TOWARDS MULTICULTURALISM 6 nearly half of the population. These findings will provide insight into the applicability of existing psychological measures in new migration contexts. A second aim is to assess intergroup attitudes and beliefs about diversity in a sample of adults with various ethnic backgrounds. We investigate similarities and differences in the endorsement of multicultural ideology between native

Luxembourgers and participants with migration background.

Multiculturalism - a multidimensional concept

Multiculturalism is a broad, multifaceted concept that can take various forms in different social contexts, organizations, and countries (Verkuyten, 2005a). In the literature, multiculturalism can refer to different aspects: demography, ideology with policy consequences, and the psychological dimension of multiculturalism (Berry et al., 1977; Van de Vijver et al.,

2008). The demographic aspect refers to the plural composition of a population. Multiculturalism may also describe a political ideology that promotes respect and support for cultural pluralism.

Diversity ideologies provide the framework for governmental policies about diversity (Cho et al.,

2018). Government or organizational policies maybe aimed at improving intercultural communication and integration of minority groups. Other examples of multicultural policies include efforts to establish equal rights and to improve the socioeconomic status of immigrant groups (Schalk-Soekar, 2007), promotion of multiculturalism in the school context (Schachner et al., 2019), allowance of dual citizenship, and financial support of cultural organizations (Berry &

Ward, 2016). The psychological aspect refers to attitudes and behaviors that support cultural diversity at the individual level (Celenk & Van de Vijver, 2014). These can involve support for and adjustment to a culturally diverse society, acceptance of minority ethnic cultures, anti- discrimination, and interactions between members of the host society and minority groups ATTITUDES TOWARDS MULTICULTURALISM 7

(Breugelmans & Van de Vijver, 2004). In studies assessing attitudes towards multiculturalism, psychological instruments focus on these dimensions.

Assessment Issues - The Multicultural Ideology Scale (MCI)

Attitudes towards multiculturalism have been assessed using quantitative and qualitative research methods. Self-report questionnaires are the most frequently used measures in the assessment of multiculturalism. The Multicultural Ideology Scale (MCI) is a popular attitude scale that was developed in 1974 in to measure attitudes towards multiculturalism (Berry et al., 1977). The first version was developed in line with the fourfold acculturation strategies framework (Berry, 1997). Endorsement of multiculturalism was considered to be a bipolar construct with support for integration of migrant groups on one end and support for the three other acculturation orientations (i.e. assimilation, separation, and marginalization) on the other end. The scale was later used in a survey designed by the Department of Multiculturalism and

Citizenship in Canada (Berry and Kalin, 1995b). It was tested in a large sample of 3,325

Canadian residents. Results from psychometric analyses including reliability checks, factor analysis, convergent and discriminant analysis, and scale intercorrelations fed into the final selection of the items (Berry & Kalin, 1995). This final questionnaire comprised 10 items (five positively worded and five negatively worded items) and was available in French and English. A higher scale score reflected a more positive attitude towards multiculturalism. Responses were recorded on a 7-point Likert scale. All items loaded on a single factor. The inter-item correlations were in the expected direction. Additional analyses revealed that the MCI scale correlated positively with a scale of attitudes towards specific multicultural programs and a scale assessing willingness to accept immigrant groups in Canada. It correlated negatively with ethnocentrism, authoritarianism, and discrimination (J. Berry, personal communication, ATTITUDES TOWARDS MULTICULTURALISM 8

September 16, 2019). In general, this attitude scale was tested in several studies assessing support for cultural diversity in Canada and acceptance of ethnic minority cultures (Berry &

Kalin, 1989). To date, adapted versions of the scale have been used in various countries dealing with diversity, including The Netherlands, Russia, Luxembourg, New Zealand, Latvia, Estonia,

Finland, Germany, Switzerland, Greece, Italy, Malta, Portugal, Spain, India, Hong Kong, and

Australia (Berry, 2017).

Arends-Toth and van de Vijver (2003) translated and adapted the scale to the Dutch context. The scale assessed support for having a culturally diverse society in the Netherlands. All items loaded high on a single factor and this unidimensional structure was confirmed in multiple factor analytic studies (Arends-Tóth & Van de Vijver, 2003; Schalk-Soekar, 2007; Van de Vijver et al., 2008). Reliability coefficients ranged from .67 to .86. Moreover, the questionnaire was invariant across different ethnic groups (Dutch, Turkish, and Moroccan nationals) and measurement occasions (Schalk-Soekar, 2007).

A revised version of this scale was later developed by Breugelmans and van de Vijver

(2004). This scale included 28 items covering the following domains: 1) attitudes towards diversity, 2) acculturation strategies employed by minority groups (assimilation vs. cultural maintenance), 3) acculturation preferences of majority group members, and 4) attitudes towards equal societal participation and interactions between immigrants and mainstream group members. Most scale items reflected the constructs described by the theoretical model of acculturation. New items, mainly covering the fourth domain, were developed based on the multicultural attitude survey (Ho, 1990) and the Quick Discrimination Index (QDI; Ponterotto et al., 1995). The following items are some examples (i. e. ‘I think that non-natives and mainstreamers should have equal rights’, ‘I feel uneasy when non-natives talk to one another in a ATTITUDES TOWARDS MULTICULTURALISM 9 language I do not understand’). This revised scale also yielded a one-factor solution, explaining

32% of the variance. Items belonging to the first domain had the highest loadings. Internal consistency of the measure was high (Cronbach’s α =. 96). Similar findings were observed in other studies using this revised attitude scale in the Netherlands and other countries. In a recent study assessing multicultural attitudes in New Zealand and the , Cronbach’s α were .89 and .93, respectively (Stuart & Ward, 2019). In these cited studies, the unidimensionality and the psychometric properties of the scale were ensured (Schalk‐ Soekar et al., 2008; Schalk‐ Soekar & Van de Vijver, 2008). However, in studies involving multigroup comparisons, the measurement invariance of the scale was not consistently tested or reported.

Measurement invariance should be tested on a larger, more representative sample of various cultural groups and different language versions.

Intergroup Attitudes to Multiculturalism

Cross-cultural research has predominantly focused on multiculturalism as attitude towards plurally composed societies (Celenk & Van de Vijver, 2014), both from a minority and majority group perspective. Evidence from Canada and European countries indicates that public attitudes towards living in a plurally composed society are relatively positive or neutral (Berry &

Kalin, 1995a; Schalk-Soekar et al., 2008) and unrelated to governmental policies aiming to promote diversity (Van de Vijver et al., 2008). More favorable attitudes towards ethnic out- groups have been observed in areas with larger shares of immigrants (Ward et al., 2011) and higher economic growth (Leong & Ward, 2006). Stronger endorsement of multicultural ideologies has been related to lower social distance towards outgroup members in the public sphere (Ng Tseung-Wong & Maykel Verkuyten, 2015). ATTITUDES TOWARDS MULTICULTURALISM 10

Support for multiculturalism can vary across different societal domains and ethnic groups of different status (Van de Vijver et al, 2008). One study in the Netherlands concluded that minority members express more positive attitudes toward multiculturalism but there is domain specificity: They prefer cultural maintenance in private domains and adoption of the mainstream culture in the public sphere (Arends-Toth & van de Vijver, 2003). In Germany, majority group members had favorable attitudes towards the acculturation strategies of integration and assimilation while minority members preferred integration (Pfafferott & Brown, 2006).

Discrepancies between acculturation preferences of majority and minority group members were related to lower levels of life satisfaction and negative intergroup attitudes. Other studies targeting majority group members showed that they display more positive attitudes towards culturally similar ethnic groups (Callens et al., 2019). They expect cultural adaption of immigrant groups to the host society in all life domains. For members of the mainstream culture, support for anti-discrimination and equal societal participation was higher than acceptance of minority cultures (Schalk-Soekar et al., 2008; Verkuyten, 2005b).

The findings reflect different conceptualizations of multiculturalism among majority and minority group members. Explanations drawing on social categorization theory and integrated threat theory have been proposed to justify this variation in their acculturation preferences

(Arends-Toth & Van de Vijver, 2003; Van Osch & Bregeulmans, 2012). Endorsement of multicultural policies depends on the interests these ideologies serve. Multicultural ideologies are more beneficial for minority groups who can obtain a higher status in the society without losing their cultural identity. On the contrary, majority group members strive to maintain their cultural dominance and perceive some aspects of multiculturalism as threatening for their social status

(Van Oudenhoven et al., 1998). ATTITUDES TOWARDS MULTICULTURALISM 11

However, most of these studies were conducted over a decade ago and described situations in which there is a clear distinction of majority and minority groups. Since then, migration patterns have changed significantly, particularly in the European context. In most contemporary societies, diversity is not limited to interactions between majority and minority group members but refers to the coexistence of various groups with diverse historical and cultural characteristics (Meissner & Vertovec, 2015; Van Osch & Breugelmans, 2012).

Multicultural ideologies and their impact on intergroup relations are influenced by the sociohistorical context (Tseung-Wong & Verkuyten, 2018). Boundaries between minority and majority become blurred and there is no longer a coherent majority culture in some countries.

Native citizens live along with foreigners from diverse ethnic groups, including first- and second-generation immigrants, individuals of mixed cultural heritage, expatriate employees, international students, and refugees. It is therefore important to understand how these complex intercultural dynamics are associated with attitudes towards multiculturalism.

In the present study, we explore multicultural attitudes in Luxembourg, a country that hosts a large number of foreign residents. Several studies on multiculturalism have focused on the acculturation preferences of cultural minority groups (Schwarzenthal et al., 2018). However, the psychological processes underlying multicultural attitudes seem to differ among immigrant groups and host society members (Arends-Toth & Van de Vijver, 2003). The attitudes of host countries’ citizens have important implications for the integration of immigrant groups. We are particularly interested in the perspective of native citizens who grow up in culturally diverse settings. Native Luxembourgers live in a plurally composed society and are underrepresented in certain areas of the country. Following-up on previous studies (Murdock & Ferring, 2016), we ATTITUDES TOWARDS MULTICULTURALISM 12 investigate endorsement of multiculturalism within the context of the majority being in the minority position.

Multiculturalism in Luxembourg

With a foreign population percentage of 47.5% (Statec, 2019), Luxembourg looks like a cosmopolitan melting pot. According to Eurostat (2017), Luxembourg recorded high immigration rates during the past decade. It hosts a much higher proportion of foreigners in its population than most European countries (Eurostat, 2019). Luxembourg has a total population of

627,676 citizens (Worldometer, 2020). It is estimated that nearly 300,000 foreign residents are currently living in the country. Within the capital, home to a fifth of the country’s population, the percentage of the immigrant population rises to 70.63%, representing 164 nationalities (Ville de

Luxembourg, 2019).

The demographic composition of the immigrant population is constantly changing. New arrivals, originating from an increasingly wider range of countries live alongside second- and even third-generation immigrants. Commuters cross the border into trilingual Luxembourg on a daily basis. Interactions in various social contexts facilitate intercultural contact and switching between languages (Murdock, 2017). This fluidity and opportunity for intergroup contact is experienced as a source for professional and personal development. However, these shifting contexts can also present challenges for the newcomers who are being socialized in different cultures and for the native citizens who can find themselves in the minority in certain contexts

(Murdock & Ferring, 2016).

Given the rise in migratory movements, research into attitudes towards multiculturalism and diversity in Luxembourg has grown in importance. Previous studies in Luxembourg suggest that there is support for multiculturalism to some extent (Murdock, 2016). Both natives and ATTITUDES TOWARDS MULTICULTURALISM 13 foreigners appreciate benefits of living in a multicultural society, but they disagree regarding acculturation preferences in public and private life domains as was the case in other European countries (e.g., Arends-Toth & van de Vijver, 2003). In general, citizens with a migratory background evaluated multiculturalism more positively (Murdock, 2016). The principle – implementation gab (Yogeeswaran & Dasgupta, 2014) was also found for the Luxembourg native population: the idea of multiculturalism is endorsed, but some disagreement exists regarding the implementation in practice. Natives expressed more reluctant attitudes towards the integration and societal participation of immigrants (Murdock & Ferring, 2016). Specifically, natives expressed some disagreement with policies concerning voting rights and employment opportunities for minority groups.

In a recent study conducted in Luxembourg, Callens et al. (2019) targeted three groups

(natives, culturally similar immigrants from the neighboring countries, and culturally distant immigrants) to investigate their attitudes toward the integration of immigrants. Participants answered self-report questionnaires assessing multicultural attitudes, acculturation preferences, perceived immigrant-related threat, and frequency of contact with outgroup members. The scales assessing multicultural attitudes consisted of three items (i.e. focusing on minority group rights, cultural maintenance, and cultural exchange between natives and immigrants). Other dimensions of multiculturalism, such as attitudes towards integration policies and societal participation of immigrant groups were not covered adequately. Overall, the study found that immigrant related threat perceptions determine acculturation preferences of native citizens and predict lower endorsement of multiculturalism (Callens et al., 2019).

Gender Differences in Multicultural Attitudes ATTITUDES TOWARDS MULTICULTURALISM 14

There is controversial evidence regarding gender differences in the endorsement of multiculturalism. In some studies conducted in the Netherlands, women were more supportive of multiculturalism (Arends-Toth & Van de Vijver, 2003), compared to men who demonstrated greater prejudice against ethnic outgroups (Navarrete et al., 2010). In studies with Asian samples, men reported more negative attitudes and less awareness to racial injustice (Liu, 2002).

Gender difference in multicultural attitudes was absent in other studies conducted in the

Netherlands (Breugelmans & Van de Vijver, 2004; Ho, 1990; Stupar et al., 2014; Thijs &

Verkuyten, 2013) and in Balkan countries (Ballazhi & Ballazhi, 2017).

In Luxembourg, gender differences in multicultural attitudes are also inconclusive. In a study conducted with adult participants of mixed ethnic backgrounds, Murdock (2016) reported that female respondents expressed attitudes that are more positive. Subsequent studies involving adolescent participants and university students of different ethnic backgrounds did not reveal any gender differences with respect to multicultural attitudes. The mixed results thus far beg the question whether gender differences are salient in Luxembourg.

Different findings on the endorsement of multiculturalism between men and women could be attributed to cultural variables and particularly to differences in masculinity and femininity cultural values (Hofstede, 2001). Smaller gender differences are expected in countries such as the Netherlands that score high on feminine society and value equality and solidarity than countries with a high score on masculinity such as Luxembourg. It would be interesting to investigate whether gender differences are more salient in different samples in Luxembourg.

The present study

Building on previous findings, we further investigate attitudes towards multiculturalism of native Luxembourgers and immigrants living in Luxembourg. With data from a diverse ATTITUDES TOWARDS MULTICULTURALISM 15 sample, representing individuals of various cultural and socioeconomic backgrounds, we aim to assess the underlying structure and measurement invariance of the Multicultural Ideology Scale across different cultural groups and language versions. Lastly, we examine gender differences in the endorsement of multicultural ideology. We tested the following hypotheses:

1. The Multicultural Ideology Scale has a unidimensional structure.

2. Participants with a migration background have more positive attitudes towards

multiculturalism than native Luxembourgers.

Given the inconclusive findings of gender difference in the literature, we refrain from postulating specific hypotheses, but explore gender differences in support for multiculturalism.

Method

Participants

A sample of 1488 respondents (50.2% female) participated in a series of studies exploring attitudes towards multiculturalism. The majority of them were young adults (n =

1,040) with an average age of 32.48 (SD = 12.80) and 407 were adolescents attending at two different schools in Luxembourg (Mage = 15.52, SD = .90). Forty-one respondents did not provide their age. Our sample comprised 671 native Luxembourgers and 817 first and second generation immigrants, predominantly from European countries. The sample breakdown by each language version is presented in Table 1. Most of the immigrants were born in Luxembourg (50.7%) and some obtained the citizenship through birth (7%) or naturalization (7.8%). With respect to the number of nationalities mentioned, 18.1% of the participants listed two or more nationalities. Participation in the study was voluntary and informed consent was obtained.

Measures ATTITUDES TOWARDS MULTICULTURALISM 16

Participants completed a demographic questionnaire and the Multicultural Ideology Scale

(Berry & Kalin, 1995a), that assesses attitudes towards a culturally plural society. As noted above, the original MCI scale consists of 10 items (e.g. “A society that has a variety of cultural groups is more able to tackle new problems as they occur”). One item (“ A society made up of many cultural groups has greater difficulty retaining its national identity compared to a society made up of one or two cultural groups”) was removed after the pilot study because it did not make sense for the Luxembourgish context. Responses were recorded on a 7-point Likert scale

(1 = totally disagree, 7 = totally agree).

Procedure

The present study was part of a larger project assessing attitudes about diversity, intergroup relations, and personal values. The data collection took place between the years 2012-

2014. Participants selected either a paper (67.2%) or an online version of the survey (32.8%).

High-school students completed the survey in classroom settings and their teachers collected the data. Adults were recruited through social media, personal networks, and a company intranet.

The questionnaire was available in three languages: German, French, and English.

Questionnaires were translated from English to French and German using a translation back- translation procedure. Respondents could choose which language version they would like to complete. Demographic information of these groups of participants are provided in Table 1.

--- Insert Table 1 here---

Results

Data Analysis Strategies

We first performed an Exploratory Factor Analysis to investigate the factor structure of the MCI scale in our sample. Then, we proceeded with Exploratory Structural Equation ATTITUDES TOWARDS MULTICULTURALISM 17

Modeling (ESEM) for different grouping variables (i.e., language, gender, and immigration status), to establish the measurement invariance of the scale. These findings formed the basis for mean comparisons between different language versions and members of different cultural and gender groups.

Exploratory Factor Analysis

Initially, an Exploratory Factor Analysis was performed on the nine MCI scale items in the pooled sample. A two-factor solution showed a much better fit to the data based on the scree plot and an examination of the eigenvalues. The first two factors had eigenvalues of 3.51 and

1.17, explaining 39% and 13% of the variance, respectively. Given that the two factors were correlated, an oblique rotation was performed. The first component included items that reflected positive attitudes towards multiculturalism (e.g. ‘It is good that many different groups with different cultural backgrounds live in Luxembourg’) and the second component items that reflected negative attitudes (e.g. ‘The unity of the country is weakened by non-Luxembourgers’).

Most items loaded substantially on only one component. The factor loadings from this exploratory analysis for the pooled sample are presented in the first two columns in Table 2.

Therefore, two subscales were created: the Positive Attitudes and the Negative Attitudes.

--- Insert Table 2 here---

Measurement Invariance

To validate the structure and metrics of the two-factor model for the MCI, invariance of measurements across groups was assessed using Exploratory Structural Equation Modeling

(ESEM) with Geomin rotation (Asparouhov & Muthén, 2009). This method was selected to overcome the limitations associated with conventional Multigroup Confirmatory Factor Analysis

(CFA) in the case of large samples and presence of cross-loadings. Furthermore, research has ATTITUDES TOWARDS MULTICULTURALISM 18 shown that ESEM incorporates the flexibility of exploratory and the testing rigor of confirmatory factor analysis, whilst also providing more accurate parameter estimates and a better understanding of the construct under investigation (Xiao et al., 2019).

Multigroup measurement invariance analyses were conducted with the Mplus statistical package (Version 8.3; Muthen & Muthen, 2017) and estimated with the robust maximum likelihood estimator (MLR). Separate analyses were performed for three different groupings: migration status (Luxembourgish nationals and immigrants), gender (female and male participants), and language version (English, French, and German). The overall model consisted of the two-factor model with positively worded items loading on the positive attitude factor and the negatively worded items loading on the negative attitude factor, respectively. Meanwhile, the negatively worded items were also allowed to load on the positive attitude factor with a close-to- zero loading, and the same goes for the positively worded items with close-to-zero loadings on the negative attitude factor. For each grouping variable, three different levels of invariance were tested: the configural model, which tests the invariance of the overall factor structure, the metric model for the invariance of factor loadings, and the scalar model for the invariance of item intercepts across groups.

Model fit was evaluated by the Comparative Fit Index (CFI) (acceptable above .90), Root

Mean Square Error of Approximation (RMSEA) and Standardized Root Mean Square Residual

SRMR (acceptable below .08) (Cheung & Rensvold, 2002). The acceptance of a more restrictive model is based on the change of CFI and RMSEA values. The changes with .01 in ΔCFI and

ΔRMSEA from a less to a more restrictive model indicated an acceptance to the more restricted model. Chi-square statistics were not used given their sensitivity to large sample sizes, as in this study. The goodness-of-fit indexes for each model are displayed in Table 3. ATTITUDES TOWARDS MULTICULTURALISM 19

--- Insert Table 3 here---

As shown in Table 3, across the three language versions (English, French, and German), we found support for the configural and metric invariance, indicating that the factor structure and the factor loadings were equivalent across different language versions. The scalar invariance model was not accepted (drop of CFI from the metric to the scalar model was .063 and that of

RMSEA .022). Similar conclusions were reached for comparisons between native

Luxembourgish participants and immigrants, where configural and metric invariance were accepted. Scalar invariance was marginally accepted, with the drop of CFI at .015 and that of

RMSEA .005 from the metric to the scalar invariance model. Between the two gender groups, scalar invariance of the MCI scale was established. In this comparison, observed changes in CFI and RMSEA were marginal (all within .01) from the less to the more restricted model. We thus concluded that the construct of multiculturalism has the same structure, metric and measurement origins for female and male respondents.

To sum up, the two factors of MCI measure the same constructs with metric invariance across all three grouping variables. Hypothesis 1 on the dimensionality of the scale is not supported. Scalar invariance was only reached between the two gender groups, thus mean scores can be compared between the gender groups. Scalar invariance was not established for the different language versions and it was marginal for the native-immigrant comparisons.

Therefore, mean group differences should be interpreted with caution in these two cases.

Group comparisons

We then ran the ESEM with the pooled sample and extracted the factor scores of the two subscales reflecting positive and negative attitudes towards multiculturalism. Higher scores indicated a higher level of agreement with the positive and the negative aspects of ATTITUDES TOWARDS MULTICULTURALISM 20 multiculturalism, respectively. The factor loadings in the ESEM model are provided in the last two columns in Table 2, which largely correspond to the exploratory factor analysis well. The two factors were negatively correlated r = -.79, p < .01. These factor scores were then used for the group mean comparisons.

We performed a between-groups MANOVA to investigate differences in attitudes towards multiculturalism, as measured by the two MCI subscales. We used the factor scores of the two MCI subscales as dependent variables. Language version of the questionnaire, gender and migration status were the independent variables. Multivariate normality tests based on

Mahalanobis distances indicated that five participants had an unusual combination of scores across the two dependent variables. These participants were excluded from the analysis as multivariate outliers, leaving a sample of 1,483 respondents. Participants who completed the three language versions differed significantly on their overall multiculturalism attitudes, F(4,

2 2,900) = 28.115, p < .001; Wilks’ Lambda = .927, ηp = .037. Moreover, there was a statistically significant difference between natives and immigrants, F(2, 1,450) = 33.701, p < .001; Wilks’

2 Lambda = .956, ηp = .044. We also found a significant multivariate main effect in relation to

2 gender, F(2, 1,450) = 10.994, p < .001; Wilks’ Lambda = .985, ηp = .015.

Univariate testing indicated that participants responded differently in both subscales of the MCI questionnaire using a Bonferroni adjusted alpha level of .017. There was a significant difference in positive attitudes towards multiculturalism as a function of language F(2, 1,451)

2 =13.721, p < .001; ηp = .019. Specifically, Tukey HSD post-hoc comparisons indicated that participants who completed the English language version scored higher (M = 0.257, SE = 0.066) than those who completed the German version (M = -0.09, SE = 0.027) on the positive attitudes.

These French speaking respondents also had a higher score M = 0.116 (SE = 0.056) than the ATTITUDES TOWARDS MULTICULTURALISM 21

German speaking respondents on the positive attitudes. A significant effect of language was

2 observed for the negative attitudes subscale, F(2,1451) =49.012, p < .001; ηp = .063. Post-hoc comparisons using the Tukey HSD test showed that the German speaking participants reported more negative attitudes (M = 0.156, SE = 0.024) than those who completed the English (M = -

0.37, SE = 0.059) and the French version (M = -0.27, SE = 0.05). There were no significant differences between the French and the English language versions on either subscale.

Subsequent univariate analyses revealed a significant main effect for migration status, confirming our hypothesis 2. The results of pairwise comparisons demonstrate that natives reported more negative attitudes M = 0.02 (SE = 0.041) than immigrants (M = -0.344, SE =

2 0.029) [(F(1, 1,451) = 67.111, p < .001; ηp = .044)], and had a lower score (M = -0.07, SE =

0.046) than immigrants (M = 0.258 SE = 0.032) on the positive attitudes [(F(1, 1, 451) = 43.914,

2 p < .001; ηp = .029]. Significant univariate effects were obtained for gender on both subscales.

Men had a higher score (M = -0.07, SE = 0.034) than women (M = -0.254, SE = 0.035) on the

2 negative attitudes [F(1, 1,451) = 20.588, p < .001; ηp = .014] and a lower score (M = -0.001, SE

= 0.037) than women (M = 0.19, SE = 0.039) on the positive attitudes [F(1, 1,451) = 17.901, p <

2 .001; ηp = .012].

Discussion

The study investigated the attitude towards multiculturalism among different population groups in Luxembourg with the MCI scale. The main findings included that (1) A two-factor structure of the MCI is supported across gender, language and immigration status; (2) Metric invariance is reached for all grouping variables, but scalar invariance is only reached between the two gender groups; (3) Females in comparison to males, and immigrants in comparison to mainstreamers report more positive attitudes toward multiculturalism. ATTITUDES TOWARDS MULTICULTURALISM 22

The Dimensionality and Measurement Invariance of MCI

The unidimensional structure of the MCI scale was not confirmed in our sample. In the literature, the MCI scale has been used as a bipolar unidimensional scale with higher scores indicating support for diversity and low scores indicating negative evaluation of diversity and endorsement of assimilation. In our study, both exploratory factor analysis and exploratory structural equation modeling revealed and supported a two-factor solution, distinguishing between positive aspects of multiculturalism (such as cultural exchanges between different ethnic groups) and negative aspects of living in a culturally diverse society (disadvantages concerning security and heterogeneity issues that affect the host society). The two factors were inversely correlated. This distinction between positive and negative aspects of multiculturalism give us a more detailed view on cross-cultural similarities and differences in the endorsement of multiculturalism. Similar findings have been reported in the literature. Hahn and colleagues

(2015) observed that positive diversity ideologies like multiculturalism and negative diversity ideologies like assimilation tap into different dimensions. In their study, a multi-factor CFA model showed a better fit to the data than the one factor model of diversity ideologies.

We also examined the factor structures and the measurement invariance of the scale across different groups of respondents. To date, attitudes towards multiculturalism among majority and minority group members have been assessed with the same instruments but there is a lack of empirical support for the measurement invariance of these questionnaires in different cultural contexts. In the present study, invariance of the MCI was tested across different gender, languages and cultural groups. We found support for the full comparability of the scale across the two genders, but across the language versions and the immigration status, only metric invariance was accepted. This is in agreement with another recent study conducted in ATTITUDES TOWARDS MULTICULTURALISM 23

Luxembourg (Callens et al., 2019), where scalar invariance was not established for the measures assessing multicultural attitudes. The psychological aspects of multiculturalism can be interpreted in different ways by respondents of different cultural backgrounds, or these respondents may have different styles to present themselves when responding to Likert-scale questions (He & van de Vijver, 2013). Lack of scalar invariance can also be attributed to different language translations of the items assessing support for multiculturalism. Language is connected to one’s cultural background and culture can greatly influence perceptions and attitudes to multiculturalism. Therefore, these empirical findings leave room for alternative interpretations of the group mean differences as the scores may be confounded with differences in other cultural and individual characteristics. Future studies should always test the applicability and the invariance of the measure in various cultural contexts. Measurement invariance is an important aspect of psychometric testing in order to ensure comparability of the data and deal with methodological challenges in cross-cultural research.

The Group Differences on MCI

Regarding group differences in the endorsement of multicultural ideology, we observed significant differences in multicultural attitudes for all grouping variables but as previously mentioned, these findings should be interpreted with caution when there is lack of scalar invariance. In line with our expectations and previous studies (Arends-Toth & van de Vijver,

2003), natives demonstrated more negative attitudes towards practices that support multiculturalism and more favorable attitudes towards assimilation of immigrant groups in the host society.

This finding is also evident when looking at the different language versions. Participants who selected the German version of the questionnaire had more negative attitudes than those ATTITUDES TOWARDS MULTICULTURALISM 24 who selected the two other language versions. The majority of the German speaking respondents in our sample were native Luxembourgish citizens while French and English speaking respondents had a more diverse cultural background. Many of these French and English speakers were first or second-generation immigrants and dual-citizenship holders who do not identify strongly with one culture only. Attitudes of native citizens and individuals with a mixed cultural background clearly diverge depending on their interests and their personal experiences. These findings have significant implications. Differences in attitudes towards multiculturalism between natives and outgroup members can lead to conflictual relationships and disagreements regarding the rights of immigrant groups who come to live in the society of Luxembourg. The possibility for immigrants groups to maintain their ethnic culture is always an issue of debate. Natives support multiculturalism as an ideology but expect immigrants’ adaption to the host society.

Between the gender groups, women reported more positive attitudes than men towards living in Luxembourg. This finding is in line with one study conducted in the Netherlands

(Arends-Toth & van de Vijver, 2003). Future studies should further investigate gender differences in multicultural attitudes. Such studies should employ a diverse methodology and consider qualitative approaches to understand gender differences in thinking styles regarding tolerance and multiculturalism.

Conclusions

Our findings contribute to understanding intergroup dynamics and differences in multicultural attitudes among various cultural groups living in Luxembourg, a country where immigrants make half of the population. In addition, the results provide evidence regarding the factor structure and the applicability of the MCI scale in a unique, multicultural context. Some limitations of the measure should be addressed in future studies. The Multicultural Ideology ATTITUDES TOWARDS MULTICULTURALISM 25 questionnaire includes a limited number of items and focuses on specific domains relevant to multicultural attitudes. Most items investigate acculturation preferences such as cultural maintenance and cultural adaption of immigrant groups. Given that support for multiculturalism varies across different life domains, existing scales should be refined to cover different attitudinal and behavioral dimensions of multiculturalism. Attitude measures should include items addressing the following domains: societal participation of immigrant groups, interactions between in-group-outgroup members, and actions that promote support for cultural diversity in real life settings. A mixed-methods approach that combines evidence from self-report questionnaires, interviews, and experimental research should be used to understand the dynamics of attitudes to multiculturalism. Evidence from different attitude measures will broaden our knowledge regarding the cognitive and affective components that contribute to multicultural attitudes and enable us create effective interventions to improve intergroup relations. ATTITUDES TOWARDS MULTICULTURALISM 26

Acknowledgements

We would like to thank co-autor Prof. Fons van de Vijver for his advice and dedicated supervision. He was involved in the project from the beginning until his sad passing in June

2019. The final submission was reviewed and agreed upon by all other co-authors. We would also like to thank Saskia Schalk-Soekar for providing information regarding previous unpublished studies on the topic, late Prof. Dieter Ferring and Philipp Scischka for their contribution at the earlier stages of the project, Prof. Georges Steffgen and Prof. John Berry for their helpful comments on the manuscript.

Declaration of conflicting interests The Authors declare that there is no conflict of interest. ATTITUDES TOWARDS MULTICULTURALISM 27

References

Arends-Toth, J. V. & Van de Vijver, F. J. R. (2003). Multiculturalism and acculturation: views

of Dutch and Turkish–Dutch. European Journal of Social Psychology, 33(2), 249–266.

https://doi-org/10.1002/ejsp.143

Asparouhov, T., & Muthén, B. (2009). Exploratory structural equation modeling. Structural

Equation Modeling, 16(3), 397-438. https://doi-org/10.1080/10705510903008204

Ballazhi, S. O., & Ballazhi, S. (2017). Attitudes Toward Multiculturalism and Assimilation and

Contact with Members of Outgroup. European Journal of Social Science Education and

Research, 4(6), 308-313. http://dx.doi.org/10.26417/ejser.v11i2.p308-313

Berry, J. (1997). Immigration, Acculturation, and Adaptation. Applied Psychology: An

International Review, 46(1), 5-68. https://doi.org/10.1111/j.1464-0597.1997.tb01087.x

Berry, J. (2016). Diversity and equity. Cross Cultural and Strategic Management, 23(3), 413-

430. https://doi.org/10.1108/CCSM-03-2016-0085

Berry, J. W., & Kalin, R. (1989). Attitude Survey Research on Multiculturalism: Design

Specifications. Report to Multiculturalism and Citizenship Canada (pp. 136).

Berry, J. W., & Kalin, R. (1995a). Ethnic and civic self-identity in Canada: Analyses of 1974 and

1991 national surveys. Canadian Ethnic Studies, 27(2), 1.

https://search.proquest.com/openview/0f0f1373b52243cf688ab1ce831f1f9f/1?cbl=1817

20&pq-origsite=gscholar

Berry, J. W., & Kalin, R. (1995b). Multicultural and ethnic attitudes in Canada: An overview of

the 1991 national survey. Canadian Journal of Behavioural Science. 27(3), 301-320.

https://doi.org/10.1037/0008-400X.27.3.301 ATTITUDES TOWARDS MULTICULTURALISM 28

Berry, J.W., Kalin, R., and Taylor, D. (1977). Multiculturalism and Ethnic Attitudes in Canada.

Ministry of Supply and Services.

Berry, J. W., & Ward, C. (2016). Multiculturalism. In D. L. Sam & J. W. Berry (Eds.), The

Cambridge handbook of acculturation psychology (pp. 441-463). Cambridge University

Press.

Breugelmans, S. M., & Van De Vijver, F. J. (2004). Antecedents and components of majority

attitudes toward multiculturalism in the Netherlands. Applied Psychology, 53(3), 400-

422. https://doi.org/10.1111/j.1464-0597.2004.00177.x

Brown, T. A. (2014). Confirmatory factor analysis for applied research. Guilford.

Callens, M. S., Meuleman, B., & Marie, V. (2019). Contact, Perceived Threat, and Attitudes

Toward Assimilation and Multiculturalism: Evidence From a Majority and Minority

Perspective in Luxembourg. Journal of Cross-Cultural Psychology, 50(2), 285-310.

https://doi-org/10.1177/0022022118817656

Castles, S., & Miller, M. J. (2009). The age of migration: International population

movements in the modern State(3rd ed.). Palgrave Macmillan.

Celenk, O., & Van de Vijver, F. J. (2014). Assessment of psychological acculturation and

multiculturalism: An overview of measures in the public domain. In V. Benet-Martinez &

Y. Y. Hong (Eds.), The Oxford handbook of multicultural identity (pp. 205-226). Oxford

University Press.

Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing

measurement invariance. Structural Equation Modeling, 9(2), 233-255.

https://doi-org/10.1207/S15328007SEM0902_5 ATTITUDES TOWARDS MULTICULTURALISM 29

Cho, J., Tadmor, C. T., & Morris, M. W. (2018). Are all diversity ideologies creatively equal?

The diverging consequences of colorblindness, multiculturalism, and

polyculturalism. Journal of Cross-Cultural Psychology, 49(9), 1376-1401.

https://doi-org/10.1177/0022022118793528

European Commission (2017). Ten trends shaping migration.

https://ec.europa.eu/epsc/sites/epsc/files/epsc_-_10_trends_shaping_migration_-_web.pdf

Eurostat (2017). Migrant integration statistics. Eurostat Statistical Books (Rev. Ed.). https:// ec.europa.eu/eurostat/web/products-statistical-books/-/KS-05-17-100

Eurostat (2019). Migration and migrant population statistics.

https://ec.europa.eu/eurostat/statistics- explained/index.php? title=Migration_and_migrant_population_statistics

He, J., & van de Vijver, F. J. R. (2013). A general response style factor: Evidence from a multi

ethnic study in the Netherlands. Personality and Individual Differences, 55(7), 794-800.

https://doi.org/10.1016/j.paid.2013.06.017

Hofstede, G. (2001). Culture's consequences: comparing values, behaviors, institutions, and

organizations across nations. Sage.

Ho, R. (1990). Multiculturalism in Australia: A survey of attitudes. Human Relations, 43(3),

259-272. https://doi-org/10.1177/001872679004300304

Leong, C. H., & Ward, C. (2006). Cultural values and attitudes toward immigrants and

multiculturalism: The case of the Eurobarometer survey on racism and

xenophobia. International Journal of Intercultural Relations, 30(6), 799-810. https:// doi.org/10.1016/j.ijintrel.2006.07.001 ATTITUDES TOWARDS MULTICULTURALISM 30

Liu, W. M. (2002). Exploring the lives of Asian American men: Racial identity, male role norms,

gender role conflict, and prejudicial attitudes. Psychology of Men & Masculinity, 3(2),

107-118. https://doi.org/10.1037/1524-9220.3.2.107

Maehler, D., & Schmidt-Denter, U. (2012). Migrationsforschung in Deutschland: Leitfaden und

Messinstrumente zur Erfassung psychologischer Konstrukte. [Migration research in

Germany: Guidelines and measurement instruments for the assessment of psychological

constructs]. Springer-VS.

Meissner, F., & Vertovec, S. (2015). Comparing super-diversity. Ethnic and Racial Studies,

38(4), 541-555. https://doi-org/10.1080/01419870.2015.980295

Montreuil, A., & Bourhis, R.Y. (2001). Majority acculturations toward “valued” and “devalued”

immigrants. Journal of Cross-Cultural Psychology, 32(6), 698–719.

https://doi-org/10.1177/0022022101032006004

Murdock, E. (2016). Multiculturalism, identity, and difference: Experiences of culture contact.

Palgrave Macmillan.

Murdock, E. (2017). Identity and its construal: Learning from Luxembourg. Integrative

Psychological and Behavioral Science, 51(2), 261-278.

https://doi-org/10.1007/s12124-017-9385-7

Murdock, E., & Ferring, D. (2016). Attitude towards multiculturalism – majority in minority

perspective. In C. Roland-Lévy, P. Denoux, B. Voyer, P. Boski, & W. K. Gabrenya Jr.

(Eds.), Unity, diversity and culture (pp. 83 - 88 ).

https://www.iaccp.org/iaccp_publications/unity-diversity-and-culture/

Muthén, L.K, & Muthén, B.O. (2017). Mplus User’s Guide (8th Edition). Muthén & Muthén. ATTITUDES TOWARDS MULTICULTURALISM 31

Navarrete, C. D., McDonald, M. M., Molina, L. E., & Sidanius, J. (2010). Prejudice at the nexus

of race and gender: an outgroup male target hypothesis. Journal of Personality and

Social Psychology, 98(6), 933-945. https://doi.org/10.1037/a0017931

Ng Tseung-Wong, C., & Verkuyten, M. (2015). “I’d rather we be neighbours than lovers”: The

two-sidedness of multiculturalism. Group Processes & Intergroup Relations, 18(4), 437

453. https://doi-org/10.1177/1368430214546068

Ng Tseung-Wong, C., & Verkuyten, M. (2018). Diversity ideologies and intergroup attitudes:

When multiculturalism is beneficial for majority group members. Group Processes and

Intergroup Relations, 21(2), 336-350.

https://doi-org/10.1177/1368430216663021

Pfafferott, I., & Brown, R. (2006). Acculturation preferences of majority and minority

adolescents in Germany in the context of society and family. International Journal of

Intercultural Relations, 30(6), 703-717. https://doi.org/10.1016/j.ijintrel.2006.03.005

Ponterotto, J.G., Burkard, A., Rieger, B.P., D’Onofrio, A., Dubuisson, A., Heenehan, M.,…

Sax, G. (1995). Development and initial validation of the Quick Discrimination Index

(QDI). Educational and Psychological Measurement, 55(6), 1016–1031.

https://doi-org/10.1177/0013164495055006011

Rienzo, C. & Carlos Vargas-Silva. (2019, October 4). Migrants in the UK: An overview. The

Migration Observatory at the University of Oxford.

https://migrationobservatory.ox.ac.uk/resources/briefings/migrants-in-the-uk-an-

overview/

Schachner, M. K., Schwarzenthal, M., van de Vijver, F. J., & Noack, P. (2019). How all students

can belong and achieve: Effects of the cultural diversity climate amongst students of ATTITUDES TOWARDS MULTICULTURALISM 32

immigrant and nonimmigrant background in Germany. Journal of Educational

Psychology, 111(4), 703-716. https://doi.org/10.1037/edu0000303

Schalk-Soekar, S. R. G. (2007). Multiculturalism: A stable concept with many ideological and

political aspects [Unpublished doctoral dissertation]. Department of Psychology, Tilburg

University.

Schalk‐ Soekar, S. R., Breugelmans, S. M., & Van De Vijver, F. J. (2008). Support for

multiculturalism in The Netherlands. International Social Science Journal, 59(192), 269-

281. https://doi-org/10.1111/j.1468-2451.2009.00698.x

Schalk‐ Soekar, S. R., & Van de Vijver, F. J. (2008). The concept of multiculturalism: A study

among Dutch majority members. Journal of Applied Social Psychology, 38(8), 2152-

2178. https://doi.org/10.1111/j.1559-1816.2008.00385.x

Schwarzenthal, M., Schachner, M. K., van de Vijver, F. J., & Juang, L. P. (2018). Equal but

different: Effects of equality/inclusion and cultural pluralism on intergroup outcomes in

multiethnic classrooms. Cultural Diversity and Ethnic Minority Psychology, 24(2), 260-

271. https://doi.org/10.1037/cdp0000173

Statec (2019). Evolution of total Luxembourgish and foreign population [Infographic].

Statistiques.public.lu.

https://statistiques.public.lu/stat/TableViewer/tableViewHTML.aspx?ReportId=12858&

F_Langage=eng&MainTheme=2&FldrName=1

Straub, J. (2003). Psychologie und die Kulturen in einer globalisierten Welt. [Psychology and

Cultures in a Globalized World]. In A. Thomas (Ed.), Kulturvergleichende Psychologie

(543-566). Hogrefe. ATTITUDES TOWARDS MULTICULTURALISM 33

Stuart, J., & Ward, C. (2019). Exploring everyday experiences of cultural diversity: The

construction, validation, and application of the normative multiculturalism

scale. European Journal of Social Psychology, 49(2), 313-332.

https://doi.org/10.1002/ejsp.2542

Stupar, S., van de Vijver, F. J., Te Lindert, A., & Fontaine, J. R. (2014). Multicultural attitudes

mediate the relation between personality and perceived ethnic outgroup distance in the

Netherlands. International Journal of Intercultural Relations, 38, 24-35.

https://doi.org/10.1016/j.ijintrel.2013.05.002

Thijs, J., & Verkuyten, M. (2013). Multiculturalism in the classroom: Ethnic attitudes and

classmates’ beliefs. International Journal of Intercultural Relations, 37(2), 176-187.

https://doi.org/10.1016/j.ijintrel.2012.04.012

Van de Vijver, F. J. R., Breugelmans, S. M., & Schalk-Soekar, S. (2008). Multiculturalism:

Construct validity and stability. International Journal of Intercultural Relations, 32(2),

93-104. https://doi.org/10.1016/j.ijintrel.2007.11.001

Van de Vijver, F. J. R., & He, J. (2017). Equivalence in research on positive development of

minority children: Methodological approaches. In N. J. Cabrera , B. Leyendecker (Series

Eds.), Handbook on positive development of minority children and youth (pp. 53-66).

Springer.

Van Osch, Y. M., & Breugelmans, S. M. (2012). Perceived intergroup difference as an

organizing principle of intercultural attitudes and acculturation attitudes. Journal of

Cross-Cultural Psychology, 43(5), 801-821. https://doi.org/10.1177/0022022111407688 ATTITUDES TOWARDS MULTICULTURALISM 34

Van Oudenhoven, J. P., Prins, K. S., & Buunk, B. P. (1998). Attitudes of minority and majority

members towards adaptation of immigrants. European Journal of Social Psychology,

28(6), 995–1013.

https://doi.org/10.1002/(SICI)1099-0992(1998110)28:6<995::AID-EJSP908>3.0.CO;2-8

Van Oudenhoven, J. P., Ward, C., & Masgoret, A. M. (2006). Patterns of relations between

immigrants and host societies. International Journal of Intercultural Relations, 30(6),

637-651. https://doi.org/10.1016/j.ijintrel.2006.09.001

Van Oudenhoven, J. P., & Ward, C. (2013). Fading majority cultures: The implications of

transnationalism and demographic changes for immigrant acculturation. Journal of

Community & Applied Social Psychology, 23(2), 81-97.

https://doi.org/10.1002/casp.2132

Verkuyten, M., & Thijs, J. (2002). Multiculturalism among minority and majority adolescents in

the Netherlands. International Journal of Intercultural Relations, 26(1), 91-108.

https://doi.org/10.1016/S0147-1767(01)00039-6

Verkuyten, M. (2005a). identification and group evaluation among minority and

majority groups: Testing the multiculturalism hypothesis. Journal of Personality and

Social Psychology, 88(1), 121-138. http://dx.doi.org/10.1037/0022-3514.88.1.121

Verkuyten, M. J. A. M. (2005b). Immigration discourses and their impact on multiculturalism: A

discursive and experimental study. British Journal of Social Psychology, 44(2), 223-240.

https://doi.org/10.1348/014466604X23482

Verkuyten, M., & Martinovic, B. (2006). Understanding multicultural attitudes: The role of

group status, identification, friendships, and justifying ideologies. International Journal

of Intercultural Relations, 30(1), 1-18. https://doi.org/10.1016/j.ijintrel.2005.05.015 ATTITUDES TOWARDS MULTICULTURALISM 35

Vertovec, S. (2007). Super-diversity and its implications. Ethnic and Racial Studies, 30(6), 1024-

1054. https://doi.org/10.1080/01419870701599465

Ville de Luxembourg (2019). Etat de la population. [City of Luxembourg. State of the

population]. https://www.vdl.lu/en/city/a-glance/facts-and-figures

Ward, C., Masgoret, A.-M. & Vauclair, M. (2011). Attitudes toward immigrants and

immigrant experiences: Predictive models based on regional characteristics. IMSED

Research, Department of Labour.

https://www.researchgate.net/publication/305682932_Research_Report_Attitudes_towa

d_immigrants_and_immigrant_experiences_Predictive_models_based_on_regional_cha

acteristics

Worldometer (n.d.). Luxembourg population

https://www.worldometers.info/world-population/luxembourg-population/

World Population Review (2019). Luxembourg population 2019.

http://worldpopulationreview.com/countries/luxembourg-population/

Xiao, Y., Liu, H., & Hau, K. T. (2019). A comparison of CFA, ESEM, and BSEM in test

structure analysis. Structural Equation Modeling: A Multidisciplinary Journal, 26(5),

665-677. https://doi.org/10.1080/10705511.2018.1562928

Yogeeswaran, K., & Dasgupta, N. (2014). The devil is in the details: Abstract versus concrete

construals of multiculturalism differentially impact intergroup relations. Journal of

Personality and Social Psychology, 106(5), 772-789. https://doi.org/10.1037/a0035830 ATTITUDES TOWARDS MULTICULTURALISM 36

Table 1

Demographic characteristics and item scores across all language versions of the MIS subscales

Total German French English

N 1488 1036 261 191

Female (%) 50.2% 51.1% 48.3% 48.2%

Age (M, (SD)) 27.71 (13.27) 29.09 (13.23) 26.12 (14.43) 22.37 (9.90)

Cronbach’s α .80 .81 .71 .66

Item Scores (M, (SD))

PA1. It is good that many different 5.38 (1.64) 5.17 (1.70) 5.78 (1.41) 5.95 (1.27) groups with different cultural backgrounds live in Luxembourg.

PA2. Ethnic minorities should preserve 4.57 (1.67) 4.47 (1.69) 4.66 (1.59) 4.97 (1.61) their ethnic heritage in Luxembourg

NA3. It would be best if all people forget 2.43 (1.68) 2.53 (1.67) 2.01 (1.57) 2.51 (1.77) their background as soon as possible.

PA4. A society that has a variety of 4.34 (1.70) 4.23 (1.8) 4.32 (1.71) 4.98 (1.54) cultural groups is more able to tackle new problems as they occur.

NA5. The unity of the country is 3.65 (1.90) 3.94 (1.87) 3.06 (1.84) 2.88 (1.76) weakened by non-Luxembourgers.

NA6. If immigrants want to keep 3.70 (1.99) 4.00 (1.95) 2.95 (2.02) 3.09 (1.73) their own cultures they should keep to themselves. ATTITUDES TOWARDS MULTICULTURALISM 37

Total German French English

PA7. Native Luxembourgers should do 4.12 (1.79) 4.06 (1.83) 4.19 (1.73) 4.38 (1.65) more to learn about the customs and traditions of the other cultural groups.

PA8. Immigrant parents must encourage 4.07 (1.71) 3.78 (1.65) 4.79 (1.67) 4.7 (1.65) their children to retain the culture and

traditions of their homeland.

NA9. Immigrants to Luxembourg 4.97 (1.84) 5.45 (1.56) 4.37 (1.95) 3.18 (1.78) should change their behavior to be more like the Luxembourgish people. ATTITUDES TOWARDS MULTICULTURALISM 38

Table 2 EFA and ESEM Factor loadings for the Multicultural Ideology Scale (MIS) components

Positive Attitudes Negative Attitudes Positive Attitudes Negative Attitudes

(EFA Model) (ESEM Model)

PA1. It is good that many different groups with .644 .250 .655 -.142 different cultural backgrounds live in Luxembourg.

PA2. Ethnic minorities should preserve their .758 .069 .802 .068

Ethnic heritage in Luxembourg.

PA4. A society that has a variety of cultural groups .693 .033 .620 .004 is more able to tackle new problems as they occur.

PA7. Native Luxembourgers should do more to .785 -.183 .595 .103 learn about the customs and traditions of the other cultural groups.

PA8. Immigrant parents must encourage their .623 .079 .493 -.076 children to retain the culture and traditions of their homeland.

NA3. It would be best if all people forget their .095 .601 -.127 .438 background as soon as possible.

NA5. The unity of the country is weakened -.046 .754 .043 .631 by non-Luxembourgers.

NA6. If immigrants want to keep their own cultures .029 .712 .006 .627 they should keep to themselves.

NA9. Immigrants to Luxembourg should change their -.013 .706 -.003 .564 ATTITUDES TOWARDS MULTICULTURALISM 39 behavior to be more like the Luxembourgish people.

Table 3

Summary of Goodness-of-Fit Statistics and Information Criteria for all ESEM Models

Model and Description χ2 Δχ2 CFI ΔCFI RMSEA ΔRMSEA SRMR AIC BIC (df) (Δdf) Language Configural Invariance 159.054*** .966 .060 .030 49696.89 50253.935

(57)

Metric Invariance 251.635*** 92.581*** .945 -.021 .063 .003 .056 49733.471 50141.97

(85) (28)

Scalar Invariance 452.136*** 200.500*** .882 -.063 .085 .022 .074 49905.971 50240.198

(99) (14) Migration status

Configural Invariance 130.067*** .968 .057 .027 49900.958 50272.321

(38)

Metric Invariance 170.660*** 40.592** .959 -.009 .055 -.002 .040 49913.55 50210.64

(52) (14)

Scalar Invariance 219.007*** 48.348*** .944 -.015 .060 .005 .045 49947.897 50207.852

(59) (7) Gender

Configural Invariance 97.299*** .979 .046 .023 49268.243 49638.324

(38)

Metric Invariance 117.046*** 19.748 .977 -.002 .041 -.005 .033 49259.991 49556.056

(52) (14)

Scalar Invariance 139.26*** 22.214** .972 -.005 .043 .002 .033 49268.205 49527.261

(59) (7)

Note. df = Degrees of freedom; CFI = Comparative fit index; RMSEA = Root-mean-square error of approximation; SRMR

= Standardized root mean square residual; AIC = Akaike information criterion; BIC = Bayesian information criterion;

***p<.001; **p<.01; *p<.05. ATTITUDES TOWARDS MULTICULTURALISM 40