Perceptions of Ukrainian and Other Refugees among Eighth-Graders in Slovenia: Characteristics of Students towards Inclusion of Refugee Students in Mainstream Schools

Abstract

Background: There is a lack of empirical data on the school-age population’s attitudes toward refugees. Despite this, the attention being paid to the integration of refugee students in schools worldwide is increasing. Objectives: First, the present study aims to explore the attitudes of eighth-grade students in Slovenia regarding refugees and to investigate whether there are differences in their attitudes based on the geographical areas from where refugees originate. Second, the study also explores the connections between attitudes and some characteristics of students. Methods: The present study is quantitative; sample: 3466 respondents (representative sample; year 2022). Descriptive statistics and binary logistic regression are used. Results: Most of the students would be willing to help refugees coming from any country. The students are least willing to help refugees from Syria and Afghanistan or Africa. A high proportion of students express fears about what refugees bring, though these fears are varied. Logistic regression shows that civic knowledge and positive attitudes towards immigrants are significantly and negatively related to students’ opinion that refugee children should study in separate schools, while student socio-economic status (SES) and own immigration status are not. Conclusions: The results call for different approaches to be maintained/introduced in schools for tolerance and reducing prejudices for (certain) groups of immigrants/refugees.

1. Introduction

Migration has always been part of human history (Boda et al. 2023); however, not all migrations are due to free will. Also, the proportion of (internally) displaced people and refugees has grown rapidly in recent years. Forced migration has become a global megatrend, and many refugees are school-aged (Boda et al. 2023). However, there is a need to explain and distinguish several terms first: (im)migrants, internally displaced people, refugees, and asylum-seekers. As defined in the Guiding Principles on Internal Displacement, internally displaced persons are “persons or groups of persons who have been forced or obliged to flee or to leave their homes or places of habitual residence, in particular as a result of or to avoid the effects of armed conflict, situations of generalized violence, violations of human rights or natural or human-made disasters, and who have not crossed an internationally recognized border” (UN n.d.). Asylum-seekers flee their homes and claim they are refugees, but their status has not yet been definitively determined (Concern Worldwide 2022). There is a difference between migrants and refugees—a migrant is someone who chooses to move (for, e.g., economic or educational reasons) and a refugee is someone who has been forced from their home (Eldridge 2022). Refugees, according to the Office of the United Nations High Commissioner for Refugees (UNHCR), are people who are “fleeing armed conflict or persecution” and “for whom denial of asylum has potentially deadly consequences” (Eldridge 2022). Refugees leave their home countries because it is dangerous for them to stay (Eldridge 2022). The distinction is an important one, as the UN Refugee Convention in 1951 outlined certain rights for people deemed refugees (e.g., refugees are protected from being deported/returned to situations that might threaten their lives; they have access to social services; and they are integrated into their new country’s society), whereas migrants have no such rights and are subjected to a country’s immigration laws and procedures and can be turned away or deported back to their homeland (Eldridge 2022).

The statistics show that, by the end of 2022, 108.4 million people around the globe were driven from their homes due to persecution, conflict, violence, human rights violations, and events seriously disturbing public order (UNHCR n.d.b). This represents an increase of 19 million people compared to the end of the year before, and it is the largest ever increase between years according to UNHCR’s statistics on forced displacement; more than 1 in every 74 people on Earth has been forced to flee (UNHCR n.d.b). Ongoing and new conflicts have driven forced displacement across the globe. The surge in refugees was predominantly driven by refugees fleeing armed conflicts; overall, 52% of all refugees and other people in need of international protection came from just three countries: the Syrian Arab Republic (6.5 million), Ukraine (5.7 million), and Afghanistan (5.7 million) (UNHCR n.d.b). In other words, by the end of 2022, Europe, including Türkiye, hosted more than one-third (36%) of all refugees globally (UNHCR n.d.a). The number of refugees in European countries has increased from 7 million at the end of 2021 to 12.4 million at the end of 2022, as millions of refugees from Ukraine sought safety in nearby countries (UNHCR n.d.a). As Xuereb (2023) claims, there is still limited empirical evidence examining people’s perceptions of refugees from different ethnicities or cultures. Furthermore, there is an even bigger lack of empirical evidence on young people’s (school-aged population) attitudes towards refugees.

As the enrollment of children from refugee backgrounds in schools across the globe rises, the attention being paid to their educational needs and experiences has increased (Cooc and Kim 2023). The UN Refugee Agency (UNHCR) revised its previous strategy and prepared the updated one, entitled “Refugee Education 2030: A Strategy for Refugee Inclusion”, a few years ago (UNHCR 2019). In this document, the authors highlighted that the inclusion of refugee children and youth1 in equitable quality education contributes to resilience and prepares them for participation in cohesive societies, where access to quality education enables them to learn, thrive, and develop their potential (UNHCR 2019). When we think about the integration of refugees into our school systems, we need to mention both structural and relational inclusion. While the term structural inclusion refers to access to institutions and services, relational inclusion means a sociocultural process related to identity development and an individual sense of belonging and connectedness (Cooc and Kim 2023). Monitoring both forms of inclusion is important to work toward educational equity for refugee students (Cooc and Kim 2023). However, statistics from the last UNHCR report from 2023 showed an alarming picture. It revealed that, by the end of 2022, more than half of the world’s 14.8 million school-aged refugee children (51%) were missing out on formal education, as data from 70 countries worldwide show (UNHCR 2023). This is a concerning trend, as it means that the number of school-aged refugees increased rapidly by nearly 50% from 10 million a year earlier, driven mostly by the full-scale invasion of Ukraine (UNHCR 2023). Also, there are varied enrolment rates by education level, with 38% enrolled in pre-primary level, 65% in primary, 41% in secondary, and just 6% in tertiary (UNHCR 2023). This shows that there is already a challenge in terms of the structural inclusion (access) of refugee children; thus, relational inclusion, which is even more complex, can be even more challenging.

2. Literature Review

In this research review, we are focused on attitudes toward refugees (patterns and factors correlating with attitudes toward refugees), but also on the inclusion of refugee students in the classes (and the importance of social integration via intergroup relations).

2.1. Patterns of Attitudes towards Refugees

A study conducted in Switzerland in 2013 among Swiss citizens above the age of 18 showed that immigrants from distant cultures and those more likely to receive welfare benefits are perceived as more of a threat (Ruedin 2020). A study conducted in 2022 in the United Kingdom and Malta found that Ukrainian asylum seekers elicited higher positive emotions and attitudes towards helping, as well as lower negative emotions, classical prejudice, and conditional prejudice (such as when someone suggests that they do not have a problem with migrants/refugees, but are just concerned about their well-being or how they would fit in the host population; e.g., “They should return to their country once safe to do so”) compared to Syrian or Somali asylum seekers (Xuereb 2023). So far, the public response has been generally positive towards Ukrainian resettlement, with large shares of the European population being in favor of the policies implemented to host and support Ukrainian refugees (Tabasso 2022). Bansak et al. (2023) repeated conjoint experiments across 15 EU countries and Switzerland during the 2015–2016 and 2022 refugee crises, where they asked citizens to evaluate randomly varied profiles of asylum seekers, and showed that public preferences for asylum seekers with certain attributes have remained remarkably consistent; the authors discovered that overall support has, if anything, slightly increased over those two periods (Bansak et al. 2023). In 2022, Ukrainian asylum seekers were welcomed, with their demographic, religious, and displacement profile playing a larger role than their nationality, but this welcome did not diminish support for other marginalized refugee groups, such as Muslim refugees (Bansak et al. 2023).

A wide body of research has begun documenting patterns of attitudes toward refugees, with the results revealing pervasive and typically negative attitudes towards refugees (Cowling et al. 2019). What correlates with those attitudes? There are several groups of variables relevant when investigating attitudes toward refugees, and threat perception variables are among them, alongside demographic variables, emotions, and ideological dispositions (Cowling et al. 2019). The perceived threat of refugees has been identified as one of the most important predictors of prejudice in general and of attitudes towards refugees and migration in particular (Landmann et al. 2019). Using qualitative and quantitative methods, Landmann et al. (2019) identified six types of threats that the majority of group members in Germany experience in the face of refugee migration: symbolic threat (concerns about cultural differences), realistic threat (financial strain), safety threat (criminal acts), cohesion threat (conflicts within society), prejudice threat (increasing xenophobia, and altruistic threat (refugee care). The study found that various threat types induced negative emotions and correlated with negative attitudes towards refugees and backing for restrictive migration policies, yet, surprisingly, worries about refugee welfare strongly predicted support for such policies (altruistic threat paradox), possibly due to a deficiency in emotional reactions to this threat (Landmann et al. 2019). Meta-analyses by Cowling et al. (2019) highlighted threat perception as the primary predictor of negative attitudes towards refugees, albeit with caution due to the limited literature size, indicating that these perceptions, often fueled by media-induced false beliefs regarding refugees’ threat to national identity and security, seem to be the predominant driver of refugee-related prejudice.

However, there is a lack of evidence on young people (school-aged population)’s attitudes about refugees; empirical studies (especially those with representative samples) on their perceptions/attitudes toward refugees are limited (including in Slovenia).

2.2. Factors Correlating with Attitudes toward Refugees

Antecedents of attitudes towards refugees can be grouped into four larger categories: demographic correlates (levels of national identification, gender, age, political orientation, and education), ideological correlates (individual differences in ideological beliefs, roles of social justice principle—individuals’ orientations towards social justice, such as beliefs about a fair society), threat perception variables, and emotional variables (e.g., emotions such as empathy and intergroup anxiety may influence attitudes towards refugees) (Cowling et al. 2019). The results of meta-analyses published by Cowling et al. (2019) show that in the case of demographic factors, being male, religiously affiliated, highly nationally identified, politically conservative, and less educated were associated with negative attitudes towards refugees and asylum-seekers. Similarly, the same target population study as in the present study showed male eighth-graders in Slovenia, as well as those with lower proficiency levels of civic knowledge, holding fewer positive attitudes towards immigrants (Klemenčič Mirazchiyski 2023). Although age was typically found to correlate with prejudice, it was not systematically related to attitudes toward refugees (Cowling et al. 2019). Three ideological variables are moderately correlated with attitudes. Increases in RWA (right-wing authoritarianism, a threat-based cognitive motivational process promoting authoritarianism and societal uniformity) and SDO (social dominance orientation, a competition-based cognitive motivational process endorsing hierarchical social systems and intergroup inequality) were associated with more negative attitudes; conversely, endorsements of macro justice principles (the belief that a fair society equitably distributes resources amongst all) were linked to more positive attitudes (Cowling et al. 2019). Perceived threat has been identified as a key predictor of general prejudices (Landmann et al. 2019) and specifically of negative attitudes towards refugees (Cowling et al. 2019; Landmann et al. 2019). However, numerous studies have demonstrated that contact can reduce perceived threat, and these reductions in threat can mediate the relationship between intergroup contact and attitudes (Cameron et al. 2006, p. 362). This leads us to the importance of the inclusion of refugee students in mainstream classes/schools.

2.3. Inclusion of Refugee Students in Classes—Social Integration via Intergroup Relations

School ethnic diversity may enhance the social integration of refugee students by fostering inter-ethnic relationships among the majority group, as increased intergroup contact typically reduces prejudice and aggression towards ethnic minorities, while also promoting more positive attitudes and friendships within diverse settings (Boda et al. 2023). In terms of the contact strategies for improving intergroup relations, the following research is especially relevant. Studies such as Cameron et al. (2006; in Tropp and Ramiah 2017) have explored the effectiveness of different group identity interventions among White British (in-group) elementary students aged 5 to 11 years. Over a six-week experiment/intervention, where students read stories involving relationships between in-group and refugee children, three models were tested: recategorization into a common in-group (e.g., where White British and refugee characters were represented as part of a common in-group in the stories), dual identity (highlighting both sub-group and common in-group identities), and decategorized identity (where group identities were de-emphasized in the stories and the individuating characteristics of refugee characters were emphasized instead). The study found the dual identity approach, which maintains subgroup identities while fostering a superordinate identity, most effectively improved attitudes towards refugees compared to other models and a control group (Tropp and Ramiah 2017). Thus, the authors argue that school interventions should both facilitate the development of an inclusive superordinate identity while also maintaining an emphasis on children’s subgroup identities (Tropp and Ramiah 2017, p. 360). Inclusive education within the context of refugee students is multidimensional; it should focus on interculturality rather than multiculturalism (Christani and Raftopoulos 2024). The terms multiculturalism and interculturalism represent distinct approaches to cultural diversity. Multiculturalism describes a society with multiple ethnic or cultural groups living side by side, often without significant interaction. In contrast, interculturalism involves communities that deeply understand and respect all cultures, emphasizing active intercultural communication which means to act appropriately when interacting with culturally different people (Dermol et al. 2023). While inclusive education should focus on interculturality, it should also put an emphasis on preserving the distinct cultural elements of ethnic groups and ensuring interaction among students with different cultural backgrounds (Christani and Raftopoulos 2024, pp. 70–71), but also other backgrounds. An emphasis on relational inclusion in schools is also beneficial for students who do not come from refugee backgrounds (Cooc and Kim 2023), which is very relevant for all students and the entire society. Relational inclusion or a sociocultural process is related to identity development and an individual sense of belonging and connectedness (Cooc and Kim 2023), which includes both an individual-level sense of belonging, or connectedness, as well as group-level social cohesion (Strang and Ager 2010; Dryden-Peterson 2018). Whereas structural integration such as “pure” access to (quality) education (Dryden-Peterson 2018) is an important prerequisite for relational integration, intergroup relations—having the opportunity to interact with others—seem to be essential when it comes to reducing prejudices and perceived threats as predictors of prejudices.

But first, we need to know what attitudes students already have regarding refugees and investigate whether there are differences in their attitudes regarding different geographical areas from where refugees originate. To better address and reduce prejudices towards refugees, we also need to know what the possible fears expressed by the students towards refugees are, and to explore potential connections between attitudes toward refugees and characteristics of students. By understanding and addressing students’ attitudes towards refugees, schools can create a more inclusive and equitable educational environment, benefiting all students (not only refugee students) and the wider community.

3. Statement of the Problem, Research Objectives, and Significance of the Study

In this section, the statement of the problem, research objective, and significance of the present study are described.

3.1. Statement of the Problem

In an era of mass migration with the proportion of displaced people rapidly growing, including refugees from different geographical areas, it is important to know and be aware of the attitudes of young people (school-aged students) at the level of compulsory education towards refugees, e.g., whether there is a general trend of accepting refugees among young people, how they see the country is coping with the issues, whether they are expressing more (in)tolerant opinions about the refugees, and what the main fears are about the refugees. The present study also seeks to understand the factors associated with attitudes towards refugees and immigrants, such as socio-economic background (SES), immigration status, and civic knowledge. This study is needed for a deeper understanding of a new era with more culturally and ethnically diverse societies in which we are living, as well as to better understand and support intergroup relations and integration. The present study’s focus is on the school-aged population (specifically, eighth-grade students), as the compulsory education period can contribute to solving challenges and gaps in attitudes, including those reducing intolerance towards others. It is also a very important period that can make an important contribution to a more inclusive education and inclusive society in the future for all.

3.2. Research Objectives

The research objectives for the present study are:

-

To explore the (overall) attitudes eighth-grade students in Slovenia have regarding refugees and to investigate whether there are differences in their attitudes regarding different geographical areas from where refugees originate;

-

To explore whether any fears related to refugees are reported and, if so, what those fears expressed by students towards refugees are;

-

To explore potential connections between attitudes toward refugees (especially regarding their opinion on the inclusion of refugee students in mainstream schools) and students’ socio-economic status (SES), immigration status, and proficiency levels in terms of civic knowledge.

3.3. Research Question

For the present study, three research questions were formed:

-

How do the attitudes of eighth-grade students in Slovenia differ based on the country/region from which refugees originate?

-

What types of fears do eighth-grade students in Slovenia express regarding refugees?

-

How do Slovenian eighth-grade students’ SES, immigration status, and civic knowledge correlate with their opinion that refugee students need to study in separate schools?

3.4. Significance of the Study

The present study can help us understand the differences in attitudes towards refugees from different geographical areas, those originating from closer geographical areas (i.e., within Europe) or those at the same time with past historical links (Slovenia being part of ex-Yugoslavia) versus those from further afield. It can also help gather information on the main fears that students have regarding refugees. This is the first national survey to explore these potential differences in attitudes in Slovenia. The present study is, therefore, significant nationally and internationally because it reveals some “profiles” of students with attitudes that do not support the inclusion of refugee students in mainstream schools. This research is likely to contribute significantly to the understanding of inclusion in educational settings, providing data that could help shape more effective and compassionate educational environments while also contributing to a better understanding of social cohesion.

4. Materials and Methods

The present study is quantitative. Details about the sample, instruments/data used, data collection and data preparation, and the analysis carried out are described below.

4.1. Sample, Data Collection

Similar to other IEA studies, the International Civic and Citizenship Education Study (ICCS) has a two-stage sampling approach. First, the schools are sampled (using the probability proportional to the size [PPS] method), followed by one eighth-grade class in each of the sampled schools. Schools were sampled from the entire population of schools in Slovenia. Within each sampled school, one intact class of students was sampled at random. The international target population coverage in Slovenia was 100% (Schulz et al. 2023a). The total number of sampled and participating schools was 168 and the number of sampled and participating students was 3466 (IEA and ACER 2024). The overall participation rate of the sampled Slovenian eighth-graders was 95% (Schulz et al. 2023a). The overall exclusion rate (i.e., schools, within-sample, and student exclusions) in Slovenia for ICCS 2023 was 3.9%, which is within the limits set by the IEA (Schulz et al. 2023a). The sample is representative on the level of the population (Klemenčič Mirazchiyski 2023, pp. 33–34). The additional data used for the present study were collected with an additional national questionnaire regarding attitudes towards refugees, administered to the same students as the International Civic and Citizenship Education Study (ICCS 2022).

Data collection occurred from 11 April 2022 to 22 June 2022, involving all participating 3466 eighth-grade students in Slovenia. However, some of the students did not fill in the questionnaire due to absences in the session. However, these represented just 15 students from the entire sample. The sample characteristics for the ICCS main study and the additional questionnaire conducted in Slovenia is presented in

Table 1. Sample sizes for 8th-grade students in Slovenia taking the international questionnaire and the national additional questionnaire.

Sample	Gender	Sampled Cases	Population Estimate	(SE)	Percentages	(SE)	Average Age	(SE)
International questionnaire	Boy	1755	10,621.24	(189.69)	50.59	(0.76)	13.96	(0.01)
International questionnaire	Girl	1711	10,375.20	(166.85)	49.41	(0.76)	13.90	(0.01)
Total	-	3466	20,996.44	(158.92)	100.00	(0.00)	13.93	(0.01)
National questionnaire	Boy	1751	10,592.04	(191.59)	50.67	(0.77)	13.96	(0.01)
National questionnaire	Girl	1700	10,313.94	(173.24)	49.33	(0.77)	13.90	(0.01)
Total	-	3451	20,905.98	(168.23)	100.00	(0.00)	13.93	(0.01)

below. As the table shows, the differences in students taking both instruments is very small both for the international questionnaire and the national addition on attitudes towards refugees. Thus, for both instruments, the collected data are representative at the national level.

4.2. Data Preparation and Analysis

In our analyses, we used data from three instruments. The first two are part of the International Civic and Citizenship Education Study (ICCS) 2022 instruments (data available: IEA and ACER 2024), namely, a regular international questionnaire (for students’ family SES, and immigration status, as well as measuring positive attitudes towards immigrants) and an ICCS cognitive test (for results regarding civic knowledge scale). The following scales from the international ICCS 2022 database (IEA and ACER 2024) were used: students’ family SES, students’ immigration status, students’ positive attitudes towards immigrants, and students’ civic knowledge scale. The decision to use these three scales in our analyses is of national relevance, as the trends in positive attitudes towards immigrants in the ICCS cycle 2022 has fallen significantly compared to previous cycles. Civic knowledge is often significantly associated with different attitudes (for more details, see Klemenčič Mirazchiyski 2023). In addition, we would like to check whether a student’s own immigration status is associated with their opinion that refugee children should study in separate schools. The methodology for constructing the used scales and indices in ICCS 2022 followed the same methodology as in the previous cycles (2009 and 2016).

To measure and report on socio-economic background (students’ family SES), the following variables were used: highest parental occupational status, highest parental education, and number of books at home (Klemenčič Mirazchiyski 2023, p. 74), and a composite variable (scale in the international ICCS 2022 database, IEA and ACER 2024) was derived. The highest parental education status had five categories (did not complete ISCED 2, ISCED 2, ISCED 3, ISCED 4 or 5, and ISCED 6, 7, or 8). The first two categories (did not complete ISCED2 and ISCED 2) were collapsed with the final categories matching the following educational levels: lower-secondary or below, upper-secondary, tertiary non-university, and university education (Schulz and Friedman 2018). The variable on the number of books at home originally had six categories, with the highest being “more than 200 books,” but was collapsed to five categories, merging the last two: 0 to 10 books; 11 to 25 books; 26 to 100 books; more than 100 books. The values of the three variables (highest parental occupation, highest parental education, and the number of books at home) were converted to standardized z-scores (N[0, 1]) and the socio-economic background index was composed using these three variables in a principal component analysis (PCA) for each weighted national sample. The final index consists of the factor scores for the first principal component and is a standardized z-score (N[0, 1]) (Schulz and Friedman 2018).

The student’s family immigration background is a simple index. The students were asked to indicate in which country they and each of their parents or guardians were born. For the purpose of the analysis, each student and each parent of that student was classified as either “born in the country of the test” (in our case, Slovenia) or “not born in the test country” (Klemenčič Mirazchiyski 2023, p. 77). The data from these three questions (student’s, mother’s, and father’s place of birth) were used to categorize the immigration background into three distinct categories: (1) students and at least one parent were born in the country; (2) students born in the country, but parents born abroad; and (3) students and their parents born abroad (Schulz and Friedman 2018). For the analysis in this paper, the values for the categories of the immigration background index were reversed, so that students and parents born in the country becomes the highest category and students and parents born abroad become the lowest. The percentages in population for the separate categories are as follows: students and parents born abroad—4.49%, students born in the country of the test, but parents born abroad—16.34%, and at least one parent born in the country of the test—79.17%. That is, 21% of the students had an immigrant background.

Positive attitudes towards immigrants is a scale (from ICCS 2022 international database: IEA and ACER 2024) composed from the following questionnaire items: Immigrant children should have the same opportunities for education that other children in the country have; immigrants who live in a country for several years should have the opportunity to vote in elections; immigrants should have the opportunity to continue their own customs and lifestyle; immigrants should have the same rights that everyone else in the country has; immigrants bring many cultural, social, and economic benefits to [country of test] (Schulz et al. 2023a, p. 301). In this study, the country of test was Slovenia (Klemenčič Mirazchiyski, p. 130). The items had four response categories: strongly agree, agree, disagree, and strongly disagree. The values for the response categories were reversed and the scale was produced using item response theory (IRT). As the response categories are more than two, the scaling used partial credit model (PCM). The international item parameters were derived using equally weighted countries’ datasets. The final student scores for this scale were derived using weighted likelihood estimate (WLE). The obtained scores were z-standardized scores (N[0, 1]), which were transformed to a more convenient metric (N[50, 10]).

Like in all previous cycles, ICCS 2022 uses a large number of test items to measure civic knowledge reliably. As it is not possible for all students to take all test items, ICCS uses a matrix-sampling design where items are distributed in multiple blocks and each block contains a unique set of items that do not appear in any other block. Some of the countries preferred the paper version of the test, but others (like Slovenia) preferred the electronic one, which had 14 item blocks distributed in 14 booklets. Each of the 14 electronic booklets contains unique combination of 3 of the 14 item blocks (Schulz et al. 2023b). While convenient, this rotated design, where no student takes all items, restricts the use of classical test theory (CTT). Instead, all cycles of ICCS use more complex methodology using IRT and the so called “plausible values” (PVs) methodology. In a first step, the item parameters are estimated. For the dichotomously scored (i.e., multiple-choice) items, a one-parameter IRT model was used. For items with more than two scoring categories (i.e., constructed-response items with partial and full credit), a partial credit model was used. The data were weighted equally in all countries’ samples. In a second step, the student scores were generated using the PVs methodology. The estimated item parameters were used to score the data using the test item responses and the information from the background questionnaires. To reduce the large number of variables from the background questionnaires, PCA was used to extract the components that account for 99% of the variance in these variables. PVs were drawn from the marginal posterior of the latent distribution at the individual level. The estimation of these individual scores is based on the item response conditional on the population model, which also includes a regression on the background variables used in this conditioning. As the PVs are actually imputed scores, instead of drawing just one score, five of them are derived for each student (Gebhardt and Schulz 2018).

The ICCS 2022 test items (as in previous cycles) were designed to establish a framework for deriving a scale of civic knowledge, since ICCS 2016 comprised four levels of proficiency (Schulz et al. 2023b). The proficiency-level descriptions synthesize item descriptors within each level, outlining a hierarchy of civic knowledge characterized by progressively sophisticated content knowledge and cognitive processes (Schulz et al. 2023b, p. 101). As the scale was derived empirically, rather than based on a specific model of cognition, the ascending levels reflect increasingly complex content and cognitive processes, as evidenced by performance (Schulz et al. 2023b).

The additional (national) questionnaire asks a series of questions related to refugees in Europe and Slovenia. The first question asks students “Which refugee would you be willing to help?”: (1) all refugees; (2) from Afghanistan; (3) from Africa; (4) from Ukraine; (5) from other EU countries; (6) from former Yugoslavia (e.g., Bosnia, Serbia); and (7) nobody. The students were asked to choose only one of the available options.

The second question is a set of subquestions under a common stem (“How should Slovenia solve the issue with refugees?”) where students can choose more than one answer (all that apply): (1) should accept the refugees; (2) should maintain the same policy as other EU countries; (3) should strengthen its border (e.g., build a wall); and (4) the problem with refugees does not exist for us.

The third question is also a set of subquestions under a common stem (“What are the risks for Slovenia that refugees can bring?”) where students can select all that apply: (1) terrorism can be carried out; (2) refugees can impose their faith and traditions on the local people; (3) refugees may take away employment from the local people; (4) refugees can infect local people with hazardous diseases; (5) the crime rates will increase; and (6) refugees do not endanger the safety of Slovenia.

The fourth question also includes a set of separate questions under a common stem and is identical to the third question, with the only difference being that the questions are not about Slovenia, but about the EU. The common stem is “What risks could refugees bring to the EU?”. This way, the questionnaire distinguishes between students’ opinions on their perception of the hazards refugees could bring to their own country and the EU as a whole.

These additional national questions were presented to the students with a single checkbox. If the box was checked, the response was marked as affirmative (i.e., agreeing with the statement). If left blank, the value was coded as a missing value.

The last question from the additional questionnaire was from a set of questions with a common stem, “How much do you agree with the following statements about the refugee students?”, and it is a simple statement: “Refugee children should learn in the separate schools”. This statement had four response categories, from “Strongly agree” to “Strongly disagree”.

The percentage of missing values in the variables used in the analysis is presented in

Table 2. Variable names, questions, and percentage of missing values for the variables used in the analysis.

Variable	Question	Percent Missing
V1	Which refugees would you be willing to help?	12.99
V2a	How should Slovenia solve the issue regarding refugees?/Should accept refugees	36.2
V2b	How should Slovenia solve the issue regarding refugees?/Should maintain the same policy as other EU countries	56.67
V2c	How should Slovenia solve the issue regarding refugees?/Should strengthen its border (e.g., build a wall)	82.26
V2d	How should Slovenia solve the issue regarding refugees?/The problem with refugees does not exist for us	90.98
V3a	What risks could refugees bring to Slovenia?/Terrorism can be carried out	46.98
V3b	What risks could refugees bring to Slovenia?/Refugees can impose their faith and traditions on the local people	60.21
V3c	What risks could refugees bring to Slovenia?/Refugees may take away employment from the local people	71.55
V3d	What risks could refugees bring to Slovenia?/Refugees can infect local people with hazardous diseases	67.55
V3e	What risks could refugees bring to Slovenia?/The crime rates will increase	62.78
V3f	What risks could refugees bring to Slovenia?/Refugees do not endanger the safety of Slovenia	74.61
V4a	What risks could refugees bring to the EU?/Terrorism can be carried out	41.86
V4b	What risks could refugees bring to the EU?/Refugees can impose their faith and traditions on the local people in EU	63.33
V4c	What risks could refugees bring to the EU?/Refugees may take away employment from the local people in EU	70.12
V4d	What risks could refugees bring to the EU?/Refugees can infect local people in EU with hazardous diseases	69.26
V4e	What risks could refugees bring to the EU?/The crime rates will increase	60.85
V4f	What risks could refugees bring to the EU?/Refugees do not endanger the safety of EU	77.37
V6c	How much do you agree with the following statements about the refugee students?/Refugee children should learn in the separate schools	3.67
S_IMMPOS	Students’ positive attitudes toward immigrants scale	1.59
S_IMMIG	Immigration status index	3.45

below. Again, please note that for the variables used in the descriptive statistics (i.e., V1, V2a, V2b, V2c, V2d, V3a, V3b, V3c, V3d, V3e, V3f, V4a, V4b, V4c, V4d, V4e, and V4f), only the answers from students agreeing with the statements were recorded. If a student declined to answer, their answer was set to a missing code and not counted as disagreeing.

Just like with other international large-scale assessments, the analysis of ICCS data do not comply with the usual statistical procedures. Any analysis needs to use the provided sampling weights. The complex sampling design in ICCS requires different techniques for estimating the sampling variance. ICCS uses jack-knifing repeated replication (JRR) due to the different probabilities of selection. As each student score is represented as five PVs, the analysis needs to be repeated with each PV and then the results are averaged. The imputation variance stemming from the five PVs also needs to be estimated. The final standard error for any estimate using ICCS data needs to account for both the sampling and imputation variance (Cortes and Atasever 2024). This requires using specialized software. All data processing (converting the SPSS into RData sets and recoding of variables) and analyses (descriptive statistics and binary logistic regression) were performed through the R Analyzer for Large-Scale Assessments (RALSA) package (Mirazchiyski 2021). RALSA is a package for the R statistical software and programming language (The Comprehensive R Archive Network n.d.). At the time of conducting the analyses for this publication, RALSA version 1.4.5 and R version 4.4.0 were used. The downloaded SPSS files in the international database of ICCS (IEA and ACER 2024) were converted to RData files using the RALSA package (Mirazchiyski 2021) to be able to analyze them. The outcome variable for the logistic regression (“Refugee children should learn in separate schools”) had four distinct categories (from “Strongly agree” to “Strongly disagree”) and were recorded by collapsing the categories into “Disagree” and “Agree”.

5. Results

First, we start by presenting the results of the descriptive statistics, where we asked students which refugees they would be willing to help (regarding refugees from different geographical areas), how Slovenia should solve issues with refugees, and then their attitudes about which risks refugees bring to Slovenia and to the European Union (EU). Second, we will present the results regarding attitudes towards immigrants (for details, see previous section) in relation to civic knowledge, and, later, the results for the statement “Refugee children should study in separate schools” based on different characteristics of students and civic knowledge of those students.

The results for the first question (“To which refugees you would be willing to help?”) are presented in Figure 1. Most of the eighth-grade students in Slovenia (69.4%) would be willing to help refugees coming from any country. The second most frequent response was that students would help refugees from Ukraine (9.3%), the third—no one (8.1%), the fourth—from EU countries (5.6%), and the fifth—from former Yugoslavia (5.1%). Slovenian students are least willing to help refugees from Africa (1.8%) and Syria and Afghanistan (0.9%).

The results from the analysis of how students think Slovenia should solve the issue regarding refugees are presented in Figure 2.

Most of the students (63.8%) are of the opinion (Figure 2) that Slovenia should host refugees. A total of 43.3% are of the opinion that Slovenia should adhere to EU policies in this regard. A total of 17.7% think Slovenia should strengthen its borders. Just 9% think that Slovenia does not have issues with refugees.

As for the dangers refugees could bring to Slovenia (Figure 3), more than half (53%) of the students think that terrorism could be carried out, 39.8% think refugees will impose their faith and traditions on the locals, 37.2% think crime rates will increase, 32.5% think refugees could bring hazardous diseases, and 28% think refugees could take jobs from the local population. A quarter (25.4%) think that refugees are not a safety threat. The same question was asked to students concerning the EU (see Figure 4).

The results (Figure 4) show that, in general, Slovenian students are more of the opinion that the potential issues are of more concern for the EU than for their country (Slovenia). A total of 58.1% think that terrorism could be carried out, 39.1% that crime rates will increase, 36.7% that refugees could impose their faith and traditions on the local population, and 29.9% think that there will be risks to the job security of the local population. Just 22.6% think that there are no risks associated with refugees in the EU.

The association between civic knowledge and positive attitudes towards immigrants is presented in the following two figures. First, the percentages of students reaching each of the ICCS benchmarks were computed (Figure 5). The benchmarks (proficiency levels) are defined with scores, e.g., scores between 311 and below 395 are level D, from 395 to below 479 is level C, etc. The highest level is proficiency level A, which means scores at or above 563. The averages of the students’ positive attitudes toward immigrants are presented in Figure 6.

From Figure 5, we see that the percentages of eighth-grade students in Slovenia reaching the last three civic knowledge benchmarks (proficiency levels), which means more complex content in the domain of civic and citizenship education as well as more complex cognitive processes, are much larger than those reaching the first benchmark or below. Most of the students achieve higher benchmark levels of civic knowledge.

The averages of students’ positive attitudes towards immigrants according to each performance level are presented in Figure 6. As the figure shows, the higher the performance level students reach, the more positive the attitudes toward immigrants they tend to have. The differences in the averages between each performance level are statistically significant. The only exception is for the students performing at the lowest level, where the standard errors are rather large due to the small number of cases in this group.

Table 3. Binary logistic regression results for the statement “Refugee children should study in separate schools”.

Estimate	Civic Knowledge	Immigration Status	Positive Attitudes towards Immigrants	SES
Coefficients	−0.56	−0.08	−0.26	−0.02
(SE)	(0.06)	(0.04)	(0.05)	(0.04)
Wald	−9.7	−1.76	−5.26	−0.49
p	<0.001	0.079	<0.001	0.621
Odds Ratio	0.57	0.92	0.77	0.98
(SE)	(0.03)	(0.04)	(0.04)	(0.04)
Wald lower 95 CI	−0.68	−0.17	−0.36	−0.11
Wald upper 95 CI	−0.45	0.01	−0.17	0.06
Odds lower 95 CI	0.51	0.85	0.7	0.9
Odds upper 95 CI	0.64	1.01	0.85	1.07

represents the standardized coefficients for the binary logistic regression where the data for the statement “Refugee children should study in separate schools” were used as the dependent variable. The original response categories (1—Strongly agree; 2—Agree; 3—Disagree; and 4—Strongly disagree) were dichotomized into two categories: 0—Disagree; and 1—Agree. As the table shows, student immigration status is not significantly related to the opinion that refugee students should study in separate schools (β = −0.08, p = 0.079). SES is also not significantly related to the opinion of whether refugee students should study in separate schools (β = −0.02, p = 0.621). Civic knowledge and positive attitudes towards migrants, however, are significantly related to the outcome. In both cases, the coefficients are negative (β = −0.56 and β = −0.26, respectively) and statistically significant (p < 0.001 in both cases); i.e., students with lower civic knowledge and those with less positive attitudes towards immigrants tend to be more of the opinion that refugee students should study in separate schools. This is also supported by the odds ratios, which are below 1 (0.57 and 0.77, respectively), showing that supporting the statement that refugee students should study in separate schools is less likely to occur as civic knowledge and positive attitudes towards immigrants occur. That is, having higher levels of civic knowledge and more positive attitudes toward immigrants have a negative effect on support for educating refugee students in separate schools, while SES and immigrant status are unrelated.

For each unit increase in civic knowledge, the odds of supporting the notion that refugee students should study in separate schools decrease by a factor of 0.57, or by 43% (1 − 0.57 = 0.43). For each unit increase in positive attitudes towards immigrants, the odds of supporting the notion that refugee students should study in separate schools decrease by a factor of 0.77, or by 23% (1 − 0.77 = 0.23). The confidence intervals for both the civic knowledge and positive attitudes odds ratios do not go above the value of 1, which means that with both predictors increasing, the odds of supporting the notion that refugee students should study in separate school decrease.

The model fit statistics are presented in

Table 4. The model fit statistics.

Statistic	Estimate	(SE)
Null Deviance (−2LL)	21,723.35	(421.41)
Deviance (−2LL)	20,245.82	(428.76)
DF Null	3154	(191.11)
DF Residual	3150	(191.11)
AIC	20,343.91	(432.42)
BIC	20,374.19	(432.45)
Chi-Square	1477.53	(198.65)
R2 (Hosmer and Lemeshow)	0.07	(0.01)
R2 (Cox and Snell)	0.37	(0.05)
R2 (Nagelkerke)	0.37	(0.05)

. It shows that the amount of explained variance (Cox and Snell) is 37%. This value is quite strong and demonstrates that a large part of the variance in the outcome variable (agreement or disagreement that refugee children should study in separate schools) and the predictors are strongly associated with the outcome.

6. Discussion

The majority of positive reports of providing help to refugees are in line with previous studies on the adult population, e.g., Bansak et al. (2023). In 2022, Ukrainian asylum seekers were welcomed into European countries, with their demographic, religious, and displacement profiles playing a more significant role than their nationality; this welcome did not reduce the support for other marginalized refugee groups, such as Muslim refugees (Bansak et al. 2023). However, we also see that when asking students about refugees from different geographical areas, such as Ukraine, other EU countries, and ex-Yugoslavia, versus more distanced geographical areas, the results are in line with other studies. Namely, people are generally more inclined to help those who are perceived as psychologically close; empathy with Ukrainians can be a relevant process intervening in the effect of dispositional prosociality and European identification on helping intentions as well. Another relevant factor might also be identity fusion (framed as perceptual “closeness” with an individual or a group that motivates personally costly altruistic behaviors) with Ukrainians (Politi et al. 2023).

However, we can also be concerned from the results of our study about perceived threats as predictors of prejudices—both in general and also those towards refugees and migration (Landmann et al. 2019)—as a large proportion of the students expressed fears about different threats refugees may bring to a country (Slovenia) and the EU, such as security, symbolic (cultural), economic, and other threats. The relevant question still arises—do we have problems with prejudices based on the different geographical areas from which refugees originate? How can we overcome this?

The literature shows that less educated people are associated with negative attitudes towards refugees and asylum-seekers (Cowling et al. 2019). Our findings, if we take civic knowledge performance as a proxy for more/less educated people, are in line with previous research on the adult population. Just as in previous studies, such as in the year 2016 (Klemenčič Mirazchiyski et al. 2019), this study shows that Slovenian eighth-graders with higher civic knowledge performance (higher benchmark levels) showed more positive attitudes towards immigrants. This is also a very similar result to that in other European countries. For example, eighth-graders with higher scores on the civic knowledge scale reflected greater endorsement of equal opportunities for immigrants in Belgium (Flemish), Bulgaria, Croatia, Denmark, Estonia, Finland, Italy, Latvia, Lithuania, Malta, the Netherlands, Norway (ninth-graders), and Sweden (Losito et al. 2018), which refers to students (on average) from all participating countries (and educational systems) in the ICCS 2016 European regional module. In the 2022 cycle of the ICCS, we can observe a similar trend throughout and outside Europe.2 Besides Slovenia, eighth-graders (on average) from all other participating countries (and educational systems) in the study showed similar results. This means that eighth-grade students from Bulgaria, Chinese Taipei, Colombia, Croatia, Cyprus, Estonia, France, Italy, Latvia, Lithuania, Malta, the Netherlands, Norway (ninth-graders), Poland, Romania, Serbia, the Slovak Republic, Spain and, Sweden (Klemenčič Mirazchiyski 2023; Schulz et al. 2023a).3

The present study shows that immigration backgrounds are not related to the statement that refugee students should study in separate schools. As Christani and Raftopoulos (2024, p. 70) emphasized, fostering awareness and a sense of acceptance among the entire student/parental population, different practices of inclusion (such as engaging refugee families in the school’s everyday policy, organizing school activities that reflect the traditions, customs, and values of the students’ countries of origin, and hosting events to support refugee families and preserve their individual identities) are essential to integrating refugee families into the greater society. These practices not only promote inclusion, but also strengthen the national consciousness of refugees (Christani and Raftopoulos 2024). As such, this is why we think it is essential to be aware of the attitudes of the school-aged population (those with or without immigrant backgrounds, and other characteristics), as well as the knowledge proficiency they show.

The last two decades have presented several challenges to Europe and the world, including the 2008 economic crisis, the 2015 migration flows, Brexit in 2020, the COVID-19 pandemic from 2020 onwards, and the full-scale invasion of Ukraine in 2022 (Trunk et al. 2023). These events may, at least to some extent, influence young people’s perceptions (Trunk et al. 2023) of the societies in which they live.

Increasing cultural diversity is a fact faced by almost all European countries, especially due to increasing migration (Dermol et al. 2021, p. 232), including refugees. The present study showed a generally high proportion of eighth-grade students in Slovenia who are willing to help refugees, although we see some discrepancies when it comes to refugees from Arabic and African regions. Indeed, from the opinion of students, we see a high proportion of them indicating that refugees can impose their faith and traditions as a risk to Slovenia and the EU. However, we see the highest proportion of students being afraid of refugees because of terrorism and only roughly a quarter of students are of the opinion that the refugees are not a safety risk for the country (Slovenia) and the EU. All of these results need to bring attention to school education, as this can indicate certain issues with prejudices and stereotypes around people from different cultural, ethnic, and other backgrounds. In this case, teachers have a special role to play in battling prejudice. However, we must also be aware of what Wiseman and Galegher (2019, p. 75) claim in that “traditional classroom teachers are often unprepared to meet the needs of refugee and asylum-seeking youth who settle in their communities and attend mainstream schools”.

These factors all call for different approaches to be introduced in our schools for increasing tolerance and reducing prejudices towards certain groups of migrants/refugees. Not only this, this research also calls for special attention to be paid to intergroup relations within schools (and society), including for emotional and social competencies to be strengthened, as the literature shows (e.g., Cowling et al. 2019; Xuereb 2023) that emotions are in correlation with general prejudices as well as those towards refugees.

7. Conclusions

The European educational systems should foster a new culture of acceptance and respect for diversity (Christani and Raftopoulos 2024, p. 71). This new culture should not only exist in Slovenia and not just in Europe, but globally, as mass migration and the increase in refugees are essentially a constant across the world. It is essential to develop new and improved strategies that support the positive development of immigrant and refugee youth across their families, schools, workplaces, and communities; this support is vital not only for the individuals and families involved, but also for society as a whole (Juang et al. 2018). Therefore, social integration within schools is crucial not only for refugee (or immigrant) students, but for all groups of students. However, migration experiences are diverse, and immigrant and refugee youth are not a homogeneous “at risk” group (Juang et al. 2018), which means that educators would have to take this into account as well. This effort would not only be a language barrier matter, but also multidimensional, leading towards true interculturality by respecting intergroup relations. Our study also showed that education (possibly) matters, which is a good result, as those students reaching higher proficiency levels in civic knowledge show less support for the idea that refugee students should study in separate schools and, therefore, do not support segregation in this respect.

This paper is not without limitations. The first limitation is the construction of the questionnaire. This has two essential points: first, almost all of the questions were designed in such a way that the student responses were only coded as valid values if they agreed to the statements, and if they disagreed or skipped answering, their responses were coded as missing values. This impedes the use of more complex analyses and, thus, the analyses were limited mainly to descriptive statistics. The second limitation is from the perspective of content asked in the questionnaire. The literature shows that although students may have positive attitudes toward refugee students, this still does not ensure social acceptance of refugee students (Aktan 2023). This is very important for a future research direction if we would like to obtain answers not only on general attitudes towards the refugees (and refugee students) among students, but also to investigate the relational inclusion of refugee students in mainstream schools. This aspect is not covered in this paper, but it would be of added value for future research, as it would also be valuable to have data from different time points on different cohorts of students.