The growing cultural heterogeneity in European societies as a consequence of continued immigration has sparked intense public and political debate around the central issue of how to support the integration and participation of immigrants in the economic, social, cultural, and political spheres of society (European Commission, 2020). A dominant integration policy, currently characterizing many European countries (Malik, 2015), emphasizes the importance of adopting the culture and language of the country of residence as much and as quickly as possible, hence, urging immigrants to assimilate into the majority society (Jaschke et al., 2021).

It is a firm belief, at least by majority group members and policymakers, that assimilation into the country of residence is the best option for equal opportunities, inclusion, and upward social mobility for immigrants (Kymlicka, 2012; Rumbaut, 2015). This is thought to hold in particular for the next generation, the children of immigrants, who participate in education or are about to enter the education system (Van Acker & Vanbeselaere, 2011). A strong emphasis on assimilation also implies that immigrant families should renounce their own heritage culture and language (Randeria & Karagiannis, 2020), which may put pressure on immigrant communities (Song, 2019). Indeed, studies indicate that assimilationist pressure can negatively impact immigrants wellbeing, feelings of belonging to the country of residence, and parental self-efficacy (Green et al., 2020; Igarashi, 2019; Kogan et al., 2018; Nguyen & Benet-Martínez, 2013), and may actually increase the distance between immigrant and majority communities (Brown & Zagefka, 2011; Guimond et al., 2014).

In contrast, several studies have found that a policy approach that strives for respectful bi-cultural integration is associated with favorable psychological outcomes and cognitive advantages for immigrants (Berry et al., 2006; Igarashi, 2023; Nguyen & Benet-Martínez, 2013; Phalet & Baysu, 2020). Consistent with these findings, research conducted with Turkish immigrants in Germany showed that an integration orientation was associated with less depressive symptoms, whereas marginalization was associated with more depressive symptoms (Morawa & Erim, 2014). However, evidence is not fully conclusive, as a recent meta-analysis showed that the main effects of integration on psychological and sociocultural adaptation, both cross-sectional and longitudinal, are small to very small in size and highly heterogenous (Bierwiaczonek & Kunst, 2021). This highlights the importance of putting acculturation in context and examining contextual moderators (e.g., integration policy climate in different countries, Yoon et al., 2020).

The impact of different national integration policies may hold in particular for immigrants with young children. First of all, immigrant parents face several cultural and linguistic choices while raising their children and often try to balance between the demands of the dominant society and their wish to maintain or pass on their cultural and linguistic heritage (Curdt-Christiansen, 2013; Song, 2019). Second, in their role as caregivers, they are connected to societies’ semi-public institutions, such as the early education and care and primary school systems their children attend. Professionals working in these school settings are among the most important representatives of the majority society for families with children in the (pre)school age and, as such, can fulfill a key role in connecting immigrants and ethnic-cultural minorities to society (Pastori et al., 2021). The beliefs, attitudes, and practices of professionals in education matter for the participation, feelings of belongingness, and upward social mobility of immigrant families (Pulinx et al., 2017; Romijn et al., 2021; Ünver & Nicaise, 2019). While previous studies mostly focused on immigrant adults and adolescents, the current study explicitly focuses on immigrant parents, as they lead the next generation into society.

The current study examines the ways in which immigrant parents engage in adapting to and integrating in the majority society, in light of the different integration policies in different countries. Rather than using an a priori categorization approach by defining the acculturation profiles in advance, the current study tries to do justice to the expected heterogeneity, by using a person-centered approach to identify acculturation profiles. Second, a key question of the current study is if, and in what way, the different national integration policies relate to parents’ own acculturation strategies; Do we find differences in the acculturation profiles in four European countries characterized by different integration policies: England, Germany, the Netherlands, and Norway? In the next sections, we will delve deeper into the concepts of acculturation strategies and different integration policies across Europe.

Exploring Acculturation Strategies

Research on acculturation, defined as the process of cultural and psychological change that takes place as a result of contact between cultural groups and their individual members, has been strongly influenced by Berry’s two-dimensional Integration-Acculturation framework (Berry, 1997). In this framework, acculturation is defined as the interplay of two theoretically independent dimensions: immigrants’ preference for maintenance of the heritage culture and language (which can vary from low to high) and their preference for intercultural contact with members of the majority group and more broadly for participation in the host society (which can also vary from low to high; see Berry et al., 2006). Combined, these orientations result in four types of acculturation strategies: integration (comparatively strong preference for cultural maintenance and comparatively strong preference for intercultural contact and participation in the majority society), assimilation (orientation on the country’s culture and language together with a low preference for cultural maintenance and high preference for majority group contact and participation), separation (high preference for maintenance of the heritage culture together with a strong orientation on the cultural in-group and low majority group contact and societal participation), and marginalization (low preference for cultural maintenance together with limited connection to the in-group and low preference for majority group contact and participation in the majority society). In line with Berry (2022), we use the word strategies, as this reflects the different ways in which individuals and groups are actively engaged in the process of acculturation, usually resulting in different degrees of adaptation. These strategies go beyond mere preferences or attitudes as they reflect how, through which actions, individuals seek to promote the attainment of their integration goals in society.

Berry’s original model has been criticized and amended. For example, Bourhis et al. (1997) have proposed to focus on the degree of adoption of the host country’s culture rather than intercultural contact as second dimension to create a more univocal acculturation construct representing preferences or attitudes of immigrants (Matera et al., 2012). However, such a relatively minor change in operational definition can have a profound effect on the research results (e.g., different distributions of the participants across the acculturation profiles, Berry & Sabatier, 2011).

Others have argued that acculturation strategies may reflect the preferences of immigrants but also how members of the majority prefer newcomers to adapt to the country and what countries’ official integration policies aim at, which in turn shape the acculturation strategies of the immigrants. In this view, the acculturation of immigrants should be regarded as an interactive process involving both the migrant groups and the wider societal intergroup context, hence the Interactive Acculturation Model (Bourhis et al., 1997; Phalet & Baysu, 2020; Brown & Zagefka, 2011). Evidence in support of this model comes from two studies that related the national integration policies of 20 European countries (based on the MIPEX scores) to the national identification feelings of immigrants with similar immigration histories (Igarashi, 2019; Igarashi, 2023). In countries with comparatively tolerant multicultural policies, immigrants felt more belonging to the country and did not differ much from the majority in how strongly they identified with the country. In countries with a comparatively heavy emphasis on assimilation, immigrants felt less belonging and differed clearly from the majority in national identification (Igarashi, 2019). Furthermore, in countries with tolerant multicultural policies, most immigrants expressed positive feelings towards national and ethnic identifications (i.e., lean towards an integration acculturation strategy), while in a country with least tolerant multicultural policies, bicultural identification was difficult (Igarashi, 2023). Following up on these studies, the current study will examine the acculturation profiles of Turkish immigrants in four European countries with different integration policies.

Finally, the original Berry model has been criticized for classifying entire groups in a fixed number of categories while missing the diversity of identity construction within groups and the fluent boundaries between the distinguished categories (Howarth et al., 2014). The use of a priori classification methods, such as median splits, assumes that all four categories exist and are equally valid in the study population (Rudmin, 2003; Schwartz et al., 2010). However, given the large variation among immigrants and the highly varying circumstances under which they live, this assumption is disputable. Studies using empirical, data-driven ways of classifying individuals have shown that the four expected profiles are not always extracted or that different subtypes may exist (e.g., Stevens et al., 2004). This calls for a research approach that focuses on multiple dimensions of acculturation, is person-centered, and, while distinguishing profiles of acculturation, allows for fluent boundaries between profiles. The present study will use Latent Profile Analysis (LPA) for this purpose. LPA is a statistical technique that models the heterogeneity inherent in response patterns based on continuous dimensions of acculturation and detects latent profiles of immigrant parents with similar response patterns.

Different Integration Policies Across Europe

The current study includes four countries with a large Turkish immigrant community: England, Germany, the Netherlands, and Norway. These countries are characterized by different integration policies. England and the Netherlands have for long been known for their support to multicultural integration. In the Netherlands, however, since the beginning of the new millennium, the political discourse has shifted towards a stronger emphasis on assimilation (Entzinger, 2009; Maan et al., 2014; Malik, 2015). In contrast, although a similar shift in the national political discourse occurred in the UK, the multicultural policy framework has remained rather stable over the past decades (Malik, 2015; Mathieu, 2018). Germany, despite massive labor immigration from Turkey, was late to officially recognize that it had become a country of immigration. As a consequence, German integration policy has for long been characterized by the exclusion of immigrants from citizenship rights (Klusmeyer & Papademetriou, 2009). Cultural maintenance among immigrants was promoted in view of the expected remigration, while participation in society was discouraged, resulting in segregated parallel communities. Recently, state and national level policies changed into an assimilationist approach, and especially learning the German language is nowadays strongly promoted in early education programs and primary education (Heinemann, 2017). Of the four countries, Norway became a country of immigration most recently. The official integration policy emphasizes equality and multicultural integration, but in education, Norway is also known for its linguistic assimilation norms (Martiny et al., 2020).

Further information on the four countries’ integration policies comes from a survey by Slot et al. (2018) among professionals in early education and care, primary education, and social youth work in ten European countries, including England, Germany, the Netherlands, and Norway. The professionals reported on their multicultural practices and multicultural beliefs. The study found large differences between the four countries of the present study. Professionals in England were most positive about multiculturalism and multilingualism and reported to implement more multicultural practices than the professionals in other countries. Professionals in Norway also reported above average positive multicultural beliefs and implementation of multicultural practices but were less positive about the value of multilingualism. Professionals in Germany and the Netherlands reported less positive multicultural beliefs and were especially less in favor of multilingualism, and they reported to implement multicultural practices the least. These findings suggest that national integration policies indeed impact the beliefs and practices of (early) education professionals who at the local level are in close contact with immigrant families with young children (Romijn et al., 2021). A question of the present study is if, and to what extent, Turkish immigrant families’ acculturation strategies also differ between the countries.

The Present Study

The aim of the present study is to examine the acculturation strategies of Turkish immigrant parents in England, Germany, the Netherlands, and Norway against the background of the different national integration policies of these countries. In contrast to previous research, we explicitly focus on parents as they fulfill a crucial role in the socialization of the next generation and as they have several interactions with the majority of society (i.e., via school and family support systems). Research has shown that there is a large heterogeneity in the acculturation preferences of immigrants with a Turkish background. On the one hand, Turkish immigrants are known to maintain their cultural values and customs in the country of residence to a larger extent than other immigrant groups (Backus, 2013), and to highly value maintenance of the Turkish language (Extra & Yağmur, 2010; Leseman et al., 2019). On the other hand, Turkish immigrants are reported to also adopt the culture and language of the country of residence, especially in the public domain (Arends-Tóth & Van de Vijver, 2004). Given this heterogeneity, the current study uses the earlier described technique of Latent Profile Analyses (LPA) to identify acculturation profiles. We include measures of cultural maintenance, contact with the majority group contact, and cultural adoption. While the first two measures are derived from Berry’s framework, the third one is adopted from Bourhis et al. (1997), who argued that cultural adoption might provide a better match with the cultural attitudes of cultural maintenance (Matera et al., 2012).

Based on studies by Berry (1997), Bourhis (1997) and others, we expected to find four multidimensional acculturation profiles: an integration profile, in which Turkish immigrant parents prefer to maintain their heritage culture, adopt the culture of the country of residence and desire to have contacts with the majority; an assimilation profile, indicating that parents do not prefer to maintain their heritage culture and only want to adopt to the culture of the country of residence, while favoring majority group contact; a separation profile, when parents desire to maintain their heritage culture without adopting the majority culture and without a wish for intercultural contact; and a marginalization profile, indicating there is both low preference for culture maintenance, adoption of the culture of the country and majority group contact. However, since we conduct data-driven analyses, other subtypes can also be found. Based on earlier studies on different integration policies and the influence on immigrants’ identification with society (e.g., Igarashi, 2019), we hypothesize that the sizes of the found acculturation profiles might differ per country as a consequence of its integration policy. For example, in countries where a multicultural integration climate is more predominant (e.g., England), we expect the integration profile to be larger than in countries with a less supporting integration climate. In countries where there is a strong assimilation tendency (e.g., Germany), we expect the assimilation profile of Turkish immigrant parents to be larger.

Method

Participants

The present study was part of a larger project conducted in ten European countries, the EU funded Inclusive Education and Social Support to Tackle Inequalities in Society (ISOTIS) project (www.isotis.org). In four countries out of the ten countries, parents with a Turkish immigration background were interviewed: 293 parents from England, 338 parents from Germany, 247 parents from the Netherlands, and 65 parents from Norway. Hence, in total, 943 parents with a Turkish immigration background (Mage = 37.99 years, SD = 5.81 years) were included. Interviewed parents had either a child in the three to six years age range who attended Early childhood Education and Care (ECEC) centers but did not start formal education yet (n = 400) or a child in the nine to twelve years age range who was in primary education (before entering middle school or secondary school; n = 543). See Table 1 for the descriptive statistics. The interviews were conducted with the primary caregiver of the child, in most cases the mother. In England and Norway, the proportion of first-generation immigrant parents (i.e., parents born in the country of origin) was higher than in Germany and the Netherlands. Turkish-Norwegian parents were higher educated than the participants in other countries, reflecting the differences in migration history in Europe.Footnote 1

Table 1 Descriptive statistics per country
Full size table

Procedure

Parents were recruited in two to four urban or suburban areas per country with a moderate to high representation of the Turkish community. Parents were eligible if they were either first-generation immigrants (born in Turkey), second-generation immigrants (with their parents born in Turkey), or third-generation migrants who identified themselves as members of the Turkish community. An additional criterion for inclusion was applied: regarding the younger age group (3–6 years old), children had to be born in the host country, and regarding the older age group (9–12 years old), the child had to be living in the host country for at least five years. Recruitment strategies included approaching ECEC centers, primary schools, community centers, parent organizations, and mediating key persons who worked with the target group to establish contact with the parents. Exact response rates were difficult to determine due to the stepwise recruitment procedure and strict privacy protection rules in some countries, but overall response rates on the organization level ranged between 36 and 69% across the four countries (for more information, see Broekhuizen et al., 2018).

The structured interviews were conducted in person by trained interviewers with a Turkish immigrant background with a good command of Turkish, Kurdish, and the national languages, using an online questionnaire presented on a laptop. The questionnaire was available in all four countries’ national languages and in Turkish, and parents could switch between languages during the interview. For most questions, the interviewers read the question to the parent, the parent answered, and the interviewer entered the response. For more sensitive questions (e.g., regarding experienced discrimination, not reported in this study), parents could enter the answers themselves. The survey took 45 to 60 min to complete. Parents received a gift voucher of 5 to 10 € after participating in the interview. Data collection for the interviews ran from December 2017 to December 2018. The study was approved by the ethical committees of the research institutes involved in the study in each country. For more information about the interview study and training of the interviewers, see Broekhuizen et al. (2018).

Measures

Acculturation Attitudes

Cultural maintenance captured parents’ preference for maintaining the heritage culture. Parents were presented with two statements based on Zagefka et al. (2014) and asked to indicate to what extent they agreed with these statements on a scale ranging from disagree (1), slightly disagree (2), undecided (3), slightly agree (4), to agree (5). The two statements were: “I think it would be good if members of my group speak our original language often” and “I think it would be good if members of my group kept as much as possible our culture of origin and way of living.” The intercorrelation of the two items in the whole sample was r = 0.42 (ranging from r = 0.40 to r = 0.59 across the four countries). The final score was calculated as the mean of the answers to the two items. Higher scores indicated a stronger preference for cultural maintenance.

Cultural adoption indicated parents’ preference to adopt the culture of the host country. Parents were presented with two statements based on Zagefka et al. (2014) and asked to indicate their agreement with these statements on a similar answering scale as described above. The two items were: “I think it would be good if members of my group took on as much as possible of the [nationality] culture and way of living” and “I think it would be good if members of my group speak [national language] often.” As the intercorrelation of the two items was weak (r = 0.11 in the whole sample, ranging from r = 0.05 to r = 0.27 across the four countries), we decided to use only the item on cultural adoption and to exclude the item on language adoption from the analyses. Parents’ preference for language adoption was considered to be an ambiguous indicator of cultural adoption as language adoption can be motivated by the desire for better education and employment opportunities (Bezcioglu-Göktolga & Yağmur, 2018; Extra & Yağmur, 2010). Moreover, in a related study (Francot, 2021), we found that parents’ language choice at home was not related to their cultural preferences.

Preference for contact with the majority group represented parents’ wish to have intercultural contact with the majority group of the country. Parents were asked to indicate their agreement with two statements (based on Zagefka et al., 2011): “It is important to me that members of my group have friends with a [national] native background” and “It is important to me that members of my group spend some of their spare time with [nationality] native people.” A similar five-point answering scale was used as described above. The intercorrelation of the two items was r = 0.48 (ranging from r = 0.40 to r = 0.61 across the four countries). The final score was calculated as the mean of the answers to the two items. A higher score indicated that the parent found it more important for their community to have contact with the majority group.

Education Level

Parents’ education level represented the highest completed education level of the primary caregiver based on the International Standard Classification of Education (ISCED) levels (ISCED 2011), recoded into three levels with the following cut-off points: low = ISCED 0–2 (primary education, lower secondary education, or lower vocational training at most), medium = ISCED 3–5 (upper secondary, post-secondary non-tertiary, and short cycle tertiary education), and high = ISCED 6–8 (full tertiary education at the bachelor level or higher).

Generation of Immigration

Generation of immigration indicated whether the parent was a first, one-and-a-half, second- or third-generation immigrant. A parent was identified as a one-and-a-half generation immigrant when he or she was not born in the current country of residence, but moved to this county before the age of six, thus before formal primary education (cf. Rumbaut, 2004). For the present purpose, to facilitate interpretation, the variable was recoded into a dummy variable: first-generation immigrant (1) versus one-and-a-half, second- or third-generation immigrant (0).

Analysis Plan

Latent profile analysis (LPA) in Mplus 8.1 was conducted to identify parents’ acculturation profiles, using the dimensions cultural maintenance, cultural adoption, and preference for contact. LPA is a person-centered method suitable for continuous data (Muthén & Muthén, 1998). LPA models the heterogeneity inherent in response patterns and detects latent profiles, applying a Maximum Likelihood Approach. The statistical criteria applied were the Akaike Information Criterion (AIC) and the sample size adjusted Bayesian Information Criterion (SS Adj. BIC) (Nylund, 2007), with lower values indicating a better fit. An additional index of entropy was calculated to evaluate the homogeneity of the profiles, with values close to 1 indicating sufficient homogeneity of the profiles (Celeux & Soromenho, 1996). In addition, the parametric Bootstrapped Likelihood Ratio (BLRT) was consulted to determine if models that differed by one profile differed significantly from each other in model fit (Nylund, 2007). If the BLRT would be p < 0.05, the model with more profiles had a significantly better fit than the model with fewer profiles. To avoid the problem of local maxima, the analyses of each model were conducted with 1000 random sets of start values and 250 final stage optimizations to ensure that the best loglikelihood value was adequately replicated. Next to these statistical guidelines, also the interpretability of the profiles and the distribution of parents over the profiles were checked. With the aim to explain the found profiles more thoroughly, we also examine the education level and generation status of the immigrant parents in each profile. After conducting the LPA on the whole sample, the measurement invariance of the profiles across the four countries was examined using the multi-sample option of Mplus 8.1. First configural equivalence was established by identifying the same number and structure of profiles per country separately. Next metric and scalar invariance across countries was checked by imposing equality constraints across the countries for each profile (in line with Olivera-Aguilar and Rikoon [2018]).

Results

Latent Profile Analysis

Based on the three indicators of parents’ acculturation strategies, cultural maintenance, cultural adoption, and preference for contact with the majority group, using the standardized scores of these measures, Latent Profile Analysis (LPA) was conducted on the whole sample. Models with two to five latent profiles were compared. Table 2 shows the fit indices. Although the AIC and BIC values were lowest for a five profiles solution, the entropy was smaller than in the other models and the BLRT indicated that a five profile model did not significantly fit better to the data than the four profiles model. Therefore, based on statistical criteria as well as interpretability, we regarded the four profile solution as the best-fitting model.

Table 2 Fit indices and class proportions for the latent profile models
Full size table

Next, we investigated the four profiles model in detail and found that the profiles matched the typology of acculturation strategies as proposed by Berry (2017) and Bourhis (1997) rather well (see Fig. 1). Profile 1 was the smallest profile in the whole sample (8.3% of the parents fit this profile) and was characterized by below-average cultural maintenance, below-average cultural adoption, and slightly below-average preference for majority group contact. Therefore, this profile was typified as the “marginalization profile.” Profile 2 (15.2% of the whole sample) was characterized by below-average cultural maintenance, above-average cultural adoption, and above-average preference for majority group contact. Hence, this profile was characterized as the “assimilation profile.” Profile 3 (37.0% of the whole sample) showed above-average scores on all three acculturation variables: relatively high cultural maintenance, cultural adoption, and preference for majority group contact. Therefore, we labelled this profile the “integration profile.” Finally, Profile 4 was the largest profile in the whole sample (39.1%). This profile was characterized by the highest cultural maintenance scores and lowest cultural adoption and preference for majority group contact scores. This profile was regarded as the “separation profile.” See Table 3 for the descriptive statistics of the profiles. Finally, a closer examination (by assigning all parents to their most likely class membership) showed that these profiles did significantly differ in mean education level: In Profile 4, the “separation profile’, parents were in general lower educated (M = 1.72, SD = 0.72) than in the other profiles (F [3, 935] = 17.64, p < . 01). No significant differences between the other profiles in education level were found. Furthermore, using a Pearson chi-square test with a crosstabulation, we found no significant relation between the profiles and generation of migration (X2 [3, N = 937] = 7.27, p = 0.08).

Fig. 1
figure 1

Standardized scores of the acculturation profiles

Full size image
Table 3 Unstandardized means and standard deviations for acculturation profiles
Full size table

Measurement Invariance

As a next step, the measurement invariance of the acculturation profiles across the four countries was examined. First, to test configural invariance, latent profile solutions were estimated separately for each country to determine the optimal number of profiles for each country. We identified the same number of four profiles across the countries. Examination of the structure of the profiles revealed that the profiles for Turkish parents in England, Germany, and Norway were similar and could be interpreted and labeled in the same way as for the whole sample. The Turkish parents in the Netherlands showed similar Profiles 1 (marginalization), 2 (assimilation) and 4 (separation). Profile 3, however, differed from the other countries. The structure did not indicate an integration profile. Instead, Profile 3 in the Netherlands was characterized by below-average cultural maintenance and above-average cultural adoption, but below-average preference for majority group contact. We will return to this deviant pattern in the “Discussion”.

Finally, we tested the scalar invariance in a multi-sample analysis by constraining all means and variances of the profile indicators to be equal across the four countries. Comparison of the fully constrained and the unconstrained model using Likelihood Ratio Testing revealed a significantly worse fit of the constrained model (χ2(36) = 117.28, p < 0.01), indicating that scalar measurement invariance could not be established. After freeing the third profile in the Dutch sample, partial measurement invariance could be established (χ2(33) = 20.55, p = 0.94), indicating that all means and variances of the four profiles were invariant across the four countries, except for Profile 3 in the Netherlands.

Table 4 displays the distribution of the parents over the four profiles per country. Both in England, Germany, and the Netherlands, the marginalization profile was the smallest and the separation profile the largest profile. In Norway, the separation profile was the smallest, and the assimilation profile the largest. Note that the small sample size in Norway makes the results less reliable. There are differences between the countries in the distributions of the profiles, a chi-square test confirmed that the countries and profiles are associated, χ2(9) = 56.14, p < 0.01. Conducting Pairwise Z-tests, we see that the separation profile in Germany is significantly larger than the separation profile in England, and vice versa, the integration profile in England is significantly larger than the integration profile in Germany (and the deviant profile in the Netherlands). The assimilation profile in the Netherlands is significantly larger than the assimilation profile in England and Germany. Keeping the small sample size of Norway in mind, we see that the marginalization profile in Norway is significantly larger and the separation profile is significantly smaller than in the other three countries.

Table 4 Sample sizes of the acculturation profiles per country
Full size table

Discussion

The increasing diversity in Europe, both between and within immigrant groups, calls for more insight in the acculturation strategies of immigrants themselves and see how these might differ across countries. In the current study, we explicitly focused on immigrant parents, as their acculturation strategies will likely influence the upbringing and socialization of the next generation.

Integration policies in many countries currently urge immigrant families to adopt the country’s national culture and language, while often ignoring or even denying the value of their heritage culture. Underlying this is the strong, yet unproven belief that only assimilation will promote integration and upward social mobility (Song, 2019). As earlier described, in some European countries, this assimilation pressure is more present than in other countries. As immigrants differ in their own preferences for acculturation and in their susceptibility for the assimilation pressure from their environment (both on local and national level), we expected to find a heterogeneous pattern of acculturation strategies within and between countries. The current study focused on immigrant parents with a Turkish background who have settled in four European countries with different national integration policies: England, Germany, the Netherlands, and Norway. We conducted latent profile analysis (LPA) to identify the acculturation strategies, using theory-based indicators of acculturation following Berry (1997, 2017; Berry et al., 2006) and Bourhis et al. (1997).

Acculturation Profiles

The LPA resulted in a model with four profiles that fitted the data best in the total sample as well as in the samples per country. The four profiles showed metric invariance across England, Germany, and Norway, and regarding three of the four profiles also the Netherlands. This largely confirms the construct of acculturation and the models proposed by Berry (1997) and Bourhis et al. (1997) across the four countries for the current Turkish immigrants sample; one profile in the Netherlands deviated. The profile characterizing the largest number of parents in the whole sample was labelled the separation profile. This was also the largest profile in England, Germany, and the Netherlands, but not in Norway. This profile was characterized by a relatively strong preference for cultural maintenance and a relatively weak preference for cultural adoption and majority group contact. The second largest profile in the whole sample was labelled the integration profile. The profile was also second largest in England and relatively large in Norway, but not in Germany, while in the Netherlands a different profile was found, as will be discussed below. Parents assigned to the integration profile expressed a relatively strong preference for maintaining their own culture while also preferring to adopt the culture of the country of residence and to have contact with the majority community. The third largest profile in the whole sample, and also in the sample in England, but the second largest profile in Germany and the Netherlands, and the largest profile in Norway (but note the small sample size in Norway), was the assimilation profile, characterized by a below-average preference for cultural maintenance, and above-average preference for cultural adoption and majority group contact. Finally, a small group of parents in the whole sample showed a marginalization profile, characterized by low cultural maintenance, low cultural adoption, and below-average preference for cultural contact. In England, Germany, and the Netherlands, this was the smallest profile. In Norway, this profile was quite large, but due to the small sample size of Norway, this finding is less reliable. Important to note is that the marginalization profile did not reveal severe marginalization, as parents with this profile reported only slightly below-average preference for majority group contact. Further examination showed that parents with this profile were not lower educated than parents with other acculturation profiles. Therefore, it might be that in the present study, this profile represents a group of immigrant parents who are still searching how to combine both cultural frames and how to adapt to the country of residence, rather than being immigrant residents who have secluded themselves from the majority group, as was found in previous research (Berry et al., 2006; Del Pilar & Udasco, 2004; Schwartz et al., 2007).

Remarkably, the integration profile could not be established for the Turkish parents in the Netherlands. Instead, an unexpected profile was found showing aspects of assimilation but in a marginalized or separated way. Parents assigned to this profile had a relatively weak preference for maintaining their culture and rather wanted to adopt the majority culture, but they expressed a relatively weak desire for contact with the majority group. A possible explanation for this partly assimilationist, partly separationist profile (characterizing 18.2% of the Dutch sample) might be the relatively high residential and school segregation in the Netherlands (Boterman, 2013, 2019) in combination with the reported strong intragroup ties of the Turkish community (Backus, 2013). This together may satisfy the need for contact and social support and decrease the need for contact with the majority group, without necessarily having a strong orientation on maintaining their own culture. Note that this profile might not have been found if we had used a priori classification methods, which shows the importance of using person-centered methods to model the heterogeneous data.

Acculturation Profiles as Related to National Policies

In the current study, we focused on four countries that differ in national integration policies. We could replicate the profiles in each country, with the exception of the integration profile in the Netherlands. However, across the four countries, the profiles differed in size, which might point to the influence of national policies. In England, where a multicultural policy is still predominant, at least at the level of local professionals in early childhood education and care and in primary education (Slot et al., 2018), the integration strategy was significantly more prevalent among parents (27.30%) than in Germany (19.88%), where an assimilationist policy currently predominates (Heinemann, 2017). Also in Norway, in line with the official multicultural integration policy, the integration profile pertained to a larger proportion of parents than in Germany. However, in Norway, the assimilation profile was the biggest profile (31.25%), which may reflect the reported strong emphasis on language assimilation in the Norwegian early and primary education system, despite the official embrace of multiculturalism (Martiny et al., 2020; Romijn et al., 2021; Slot et al., 2018). The separation profile was the largest profile (44.51%) in Germany, where, until recently, exclusionary and separationist policies predominated (Heinemann, 2017).

Finally, the Netherlands may be an example of the complex interplay of national integration policies and other social policies, and how this interplay may create uncertainty and polarization within immigrant communities (Crul et al., 2012). Dutch integration policy used to promote multiculturalism until the end of the previous millennium, and then shifted to a predominant assimilation policy which is also reflected in the beliefs and practices of early childhood and primary education professionals at the local level (Romijn et al., 2021; Slot et al., 2018). At the same time, despite the emphasis on assimilation, residential and (pre)school segregation based on the constitutional freedom of school choice is rather strong in the Netherlands (Boterman, 2013; Inspectorate of Education, 2018), also in an international perspective (Ladd & Fiske, 2011), which may influence the majority group contact dimension of acculturation. In this complex policy context, the present study revealed for the Netherlands both a relatively large assimilation profile (significantly larger than in Germany and England) and a relatively large separation profile, in addition to an unexpected profile that was a mix of assimilation and separation strategies, while a clear integration profile was absent. Altogether, the present results provide tentative support for the view that integration policies and other social policies of a country (e.g., regarding urban planning and school choice) influence the acculturation preferences of immigrant families living in those countries (in line with earlier studies conducted by Igarashi [2019 and 2023]).

Limitations and Strengths

The most important limitation of the present study concerns the representativeness of the samples. The samples in the four countries were purposive, focusing on two to four (sub)urban areas in these countries. Within these urban areas, sample recruitment focused on neighborhoods with a substantial representation of Turkish immigrant families. Therefore, the country samples cannot be considered representative for the entire Turkish immigrant populations in these countries. Although the present samples represent relevant variation within the Turkish immigrant populations in these countries, caution is warranted when generalizing the findings. Second, the sample size in Norway was small, especially compared to the sample sizes of the other countries, resulting in less reliable estimates of the acculturation profiles in Norway and limiting the possibilities for a more detailed analysis of the heterogeneity within and between the country samples.

One of the strong points of this study is the use of a data-driven approach, such as Latent Profile Analysis, to identify the acculturation profiles. This is preferred over classifying individuals as high or low in categories, using a priori cut-off points such as the midpoint of a range (Jang et al., 2017; Schwartz & Zamboanga, 2008). We also included multiple dimensions to get a more complex overview of the concept of acculturation, integrating the models proposed by both Berry (with the dimensions cultural maintenance versus majority group contact; 1997) and by Bourhis (with the dimensions cultural maintenance versus cultural adoption; 1997). Correlational analyses confirmed that these dimensions are only moderately related to each other (cf. Van Acker & Vanbeselaere, 2011), thus these dimensions constitute different aspects of parents’ acculturation preferences and should all be taken into account when examining acculturation strategies. In future research, using the latent profile approach, more dimensions could be included, such as the practices, values, and identification with the heritage culture and the receiving culture to more fully capture acculturation processes (Schwartz et al., 2010). Furthermore, it should be acknowledged that acculturation is a difficult concept to operationalize and assess (see Celenk & Van de Vijver [2011] for an overview of acculturation instruments in research). In the current study, we assessed acculturation with the short but validated measurements based on Zagefka et al. (2014). However, future studies could apply more elaborate measures of acculturation strategies, such as the Acculturation Index by Ward and Kennedy (1999) or take the complex relation between language (preference) and cultural orientations into account (see Doucerain et al., 2017).

Using a data-driven approach, we found nearly complete metric measurement invariance across the countries, with the exception of one profile in the Netherlands. This implies we can obtain accurate comparisons between the countries (Kwok, 2023), and draw, at least tentative, conclusions about the impact of the different (integration and other social) policies on the acculturation strategies of Turkish immigrant parents.

Finally, a particular strong point of the current study was the successful strategy of reaching out to often difficult-to-reach groups, immigrant parents with young children, who are infrequently studied, thereby giving voice to these groups in personal and culturally sensitive interviews. Given the important role of parents, future studies should focus on how these acculturation strategies relate to their education and socialization practices at home (Francot et al., 2024).

Conclusions and Implications

To return to the central issues of this study, we summarize the main findings that may inform practitioners, local and national policymakers, and politicians aiming at the integration and upward social mobility of immigrant families. The sampled Turkish immigrant parents in England, Germany, the Netherlands, and Norway showed different acculturation profiles in line with the Interactive Acculturation Model, and the present study provides tentative support that national integration policies may influence the distribution of parents over acculturation profiles. The integration profile, generally regarded as most favorable based on several studies (e.g., Berry et al., 2006; Igarashi, 2023; Nguyen & Benet-Martínez, 2013), was more prevalent in countries with an official multicultural orientation, while the assumingly less favorable profiles of assimilation and separation were more prevalent in countries with an emphasis on assimilation in combination with segregation tendencies that followed from past or current social policies. This might suggest that other national integration policies than the widely supported assimilation policy would be more beneficial for integration.

In addition, we found indications that other social policies can influence parents’ acculturation profile as well. For example, urban planning (related to residential segregation) and freedom of school choice (related to school segregation), are typically policies that as such are not part of the national integration model, but through either limiting or enhancing the opportunities for intergroup interactions, these policies may influence the attitudes of immigrant parents towards the majority society, their wish for intercultural contact and, thereby, indirectly impact on the intercultural socialization of the next generation.