Adolescents’ Openness to Include Refugee Peers in Their Leisure Time Activities

Abstract

Background: Against the backdrop of increasing refugee movements, the integration of refugees is becoming a more and more relevant topic for many European countries. As integration is a bidirectional process, the current study examines the openness of adolescents in Germany to include refugee peers from Syria in leisure time activities. Methods: Participants were presented with a hypothetical scenario in which they had to choose who of two peers (a German peer or a refugee peer from Syria) they would include in their activity. Additionally, we varied the relevant skills of the two protagonists to investigate the role of group functioning aspects for inclusion decisions. Three measures were applied: (1) adolescents’ own decisions, (2) what they expected their peer group to do, and (3) what they thought their peer group should do. Results: The findings demonstrate that the participants were generally very open to include refugees and that the protagonists’ skills were relevant for their decisions, though much more for the expected group decision than for adolescents’ own decisions. Reasoning analyses illustrate adolescents’ considerations for these decisions. Conclusion: This research helps to clarify the interplay of moral considerations and aspects of group functioning in adolescents’ inclusion decisions regarding refugee peers.

1. Introduction

In recent years, forced migration has increased considerably in all European countries. For instance, the Russian invasion of Ukraine in 2022 led many people to leave their homes and seek shelter in other European countries. Even before this, following the Syrian civil war that started in 2011, huge numbers of refugees moved to Europe and other parts of the world. Since the outbreak of the civil war in Syria, more than 6 million Syrian refugees have come to Europe (UNHCR 2024). Thus, refugees from Syria represent one of the biggest groups of refugees in many European countries (UNHCR 2024), including numerous children and adolescents (Eurostat 2021). As a result, the integration of refugees has gained relevance in many European countries. It is important to note that integration cannot be understood as a unidirectional process for which only refugees are responsible, as integration is rather a reciprocal process involving mutual adjustment between refugees and the members of the host society (Berry et al. 2006). It is essential for the members of the host society to be open to integration and inclusive to the incoming refugees (Berry 2011). In Germany, more than one third of Syrian refugees are underage (Mediendienst Integration 2024). Thus, if we want to understand the situation of refugee youth in a new host country, we need to focus on the openness of the local youth as well. The present study focuses on the openness of adolescents in Germany to include refugee peers from Syria into their peer activities. More specifically, we examine adolescents’ inclusion decisions regarding Syrian peers in hypothetical scenarios as well as their reasoning about these decisions.

1.1. Adolescents’ Openness Regarding Refugee Peers

To date, there is not much research on the openness of children or adolescents regarding refugees in Germany. However, the few existing studies indicate that adolescents in Germany are typically quite open regarding refugees (Albert et al. 2019; Beißert et al. 2020; Beißert and Mulvey 2022; Würbel and Kanngiesser 2023) and that they are more open than older groups of the German population (Kober and Kösemen 2019; Kösemen and Wieland 2022). This is in line with research from other countries. For instance, in a study from Turkey, primary school children were quite open and generally accepted refugee peers from Syria (Oncu and Yilmaz 2022). Also, research from Portugal showed that 80% of the participating adolescents were open to welcome refugees in general (Rapanta and Trovão 2020).

This openness of the adolescents in Germany is reflected in their attitudes as well as in their behavior, i.e., interactions and relationships with refugee peers (Andresen et al. 2021). In line with that, the majority of refugee youth in Germany reported in a study that they find it easy to make friends at school and that their peers at school seem to like them (Gambaro et al. 2020). These feelings of being welcomed may be driven by effective school-based programming to reduce prejudice and promote positive peer relationships with refugee and other newcomer youth (Crooks et al. 2022).

Research also demonstrates that the need to belong is key for refugee youth and that peers play a key role in refugee youth’s inclusion (Nakeyar et al. 2018). However, research on the openness of German adolescents regarding refugee peers is scarce and, as far as we know, there are only two studies that explicitly investigated adolescents’ inclusion decisions regarding refugee peers in Germany (Beißert et al. 2020; Beißert and Mulvey 2022). In these studies, vignettes were used to examine adolescents’ inclusion decisions regarding hypothetical peers from Syria. Both studies revealed that the participants were quite open to include refugee peers in their leisure time activities and that they also thought that including refugee peers is what ought to be done. However, both studies revealed that adolescents expected their peer groups to be less inclusive of Syrian refugee peers than they indicated themselves to be. Additionally, in one of the studies, the German language skills of a hypothetical refugee peer drove adolescents’ openness to include a refugee in leisure time activities (Beißert et al. 2020). More specifically, a refugee with good German skills was as likely to be included as a German peer. But, a refugee peer with poor German skills was less likely to be included.

These findings might be indicators that adolescents, on the one hand, have generally open attitudes regarding refugees (from Syria) but that they, on the other hand, include aspects related to group dynamics and group functioning into their considerations. In the current study, we want to extend prior research on adolescents’ inclusion decisions by explicitly adding an aspect that might be relevant for smooth group functioning. More specifically, we focus on and experimentally vary the skills that are relevant for a specific group activity.

1.2. Theoretical Framework

As a theoretical framework, we used the Social Reasoning Development perspective (Rutland et al. 2010; Rutland and Killen 2015). This framework is based on social domain theory (Smetana et al. 2014; Turiel 1983) and social identity theory (Tajfel and Turner 1976, 1986) and posits that when making social decisions, individuals balance moral considerations with the interests of the social groups they belong to (Rutland et al. 2010; Rutland and Killen 2015). Thus, as adolescents are making decisions about including refugee peers whose group-relevant skills vary, they may attend to moral principles but may also attend to group goals and functioning.

In line with this theoretical framework, past research has documented that on the one hand, moral principles influence adolescents’ inclusion decisions. That is, adolescents generally reject social exclusion, evaluating it as morally unacceptable (Killen and Rutland 2011), and consider aspects of fairness and harm in their reasoning when making inclusion decisions (Killen et al. 2001; Killen and Smetana 2015). On the other hand, aspects related to smooth group functioning have been shown to influence adolescents’ inclusion decisions as well (Rutland and Killen 2015). For instance, adolescents are more likely to accept exclusion when it helps to protect the group and its norms (Mulvey 2016) but often reference smooth group functioning when evaluating exclusion (Hitti et al. 2011). Thus, aspects such as group functioning or loyalty to the group can, at times, bear greater weight than moral considerations (Rutland and Killen 2015).

In this context, making moral decisions about inclusion of refugee peers also hinges on intergroup processes, given that German adolescents may perceive refugee peers to be outgroup members. Adolescents often prefer including an ingroup member over an outgroup member justifying these decisions with reasons related to group functioning and group loyalty (Killen and Smetana 2015; van Bommel et al. 2021). With increasing age, children and adolescents care more about social groups, effective group functioning, and existing norms and expectations that are related to the group (Killen and Stangor 2001; Turiel 1983). Hence, when it comes to inclusion or exclusion in inter-ethnic contexts, both adolescents’ own moral principles as well as group norms are quite essential and may influence personal decisions as well as behavior of adolescents (Killen et al. 2017; McGuire et al. 2018). Group pressure is a strong issue for adolescents and their expectations about their groups’ attitudes can have a strong impact on their own decisions and behavior (Brown 2013). Thus, group norms are critically important for shaping group inclusivity (McGuire et al. 2015; Nesdale et al. 2010) and adolescents’ individual decisions may be influenced by their perceptions of their peers’ attitudes (Mulvey et al. 2014).

Therefore, it is conceivable that adolescents would like to be inclusive (based on their moral considerations), but the expectations of their peer group’s desires around inclusion may inhibit them from engaging in inclusive behavior. That is, group processes and moral principles exist simultaneously when adolescents make inclusion decisions in intergroup contexts (Killen and Rutland 2011). In the current research, the aim is to better understand the interplay of moral considerations and aspects of group functioning in adolescents’ inclusion decisions regarding refugee peers.

1.3. Current Study

The current study focuses on adolescents’ inclusion decisions regarding refugee peers. More specifically, using hypothetical scenarios, we examined how open youth in Germany are to including refugee peers in their leisure time activities. To shed more light on the aspect of group functioning, we included a variation in the level of relevant skills for the respective leisure time activity in our study design, as variation in skills may suggest greater or less ability to aid the group in accomplishing their goals. Specifically, we presented high- and low-skills conditions to participants. In the high-skills condition, the target refugee peer was described as “very good” at the activity, and in the low-skills condition, the target refugee peer was described as not having carried out the activity before.

Further, adolescents’ decisions may not only be based on their own individual attitudes, norms, or values, but group norms may also play into adolescents’ decisions (Killen et al. 2017; McGuire et al. 2018), suggesting the importance of assessing what adolescents anticipate what their peer group would do. Youth may struggle with social decisions in situations in which group norms conflict (Rizzo et al. 2018). Thus, it is important to examine how much adolescents’ individual decisions are in line with what they believe their peers would decide. The present study not only focused on adolescents’ own decisions but also considered adolescents’ expectations about their peer group’s behaviors. Finally, in addition to adolescents’ own decisions and their expected group decisions, we were interested in what they thought would be the right thing to do. To examine this, we did not only ask them about what they thought the group would do but also about what they thought the group should carry out.

Based on the considerations described above as well as on prior research, our hypotheses were the following:

• Adolescents expect that their group would be less inclusive of the Syrian peer than they themselves would be and than they thought their group should be.
• Adolescents’ reasoning about the group decision will include more references to socialconventional aspects and group functioning (compared to the reasoning about their own decision and about what the group should do).
• Adolescents’ reasoning about their own decision and about what the group should do will include more moral reasoning (compared to the reasoning about what the group would carry out).
• A protagonist with high skills will be more likely included than one with low skills. This should be true for adolescents’ own decisions and their expected group decisions; skill should be less likely to be associated with expectations about what the group should do.
• If there were differences in the skill level of the German and the Syrian peer, there should be more reasoning about group functioning than in conditions in which the skills of the protagonists were the same.

Besides adolescents’ inclusion decisions and reasoning, we were interested in what adolescents in Germany knew about (Syrian) refugees as strong stereotypical beliefs might have affected their openness regarding refugee peers as well.

2. Materials and Methods

2.1. Participants

This study was conducted in the year 2019 at a high school in the south-west of Germany (Federal State of Rhineland-Palatinate). All in all, 475 children and adolescents between 9 and 17 years old participated in the study. A total of 13 surveys had to be excluded from the analyses because the students had Syrian roots and thus the ingroup–outgroup manipulation would not have been clear for them (n = 4) or because they had not seriously answered the questions (n = 9). So, the final sample included 462 adolescents (M_age = 13.29 years; SD = 1.80; 268 female; 194 male) attending grades 5 to 10. In this sample, 39% of the participants had a migration history in the family (i.e., they themselves or at least one parent was born in a country other than Germany).

2.2. Design and Procedures

This study was approved by the Supervisory and Service Directorate of the Federal State of Rhineland-Palatinate. Participation in the study was voluntary. Informed assent was obtained from all participants. The parents of all participants provided informed consent. Under the guidance of a trained research assistant, all participants completed a paper–pencil questionnaire in class. Before the questionnaires were handed out to the participants, the research assistant explained the voluntary and anonymous natures of the study once again and pointed out that there were no disadvantages if some adolescents decided not to participate or leave the study early without completing it.

When the survey started, participants had to answer some demographic questions (age, grade, gender, migration history in the family). Afterwards, they were presented with a hypothetical scenario. At the end of the survey, participants’ (stereotypical) knowledge about refugees in general and Syrian refugees specifically was assessed. Once the participants had finished the surveys, they were provided with a debriefing regarding the background of the study, could ask questions, and could talk with the research assistant about the aims and the background of the study.

2.3. Material and Measures

To assess adolescents’ inclusion decisions, we used a hypothetical scenario in which the participants had to indicate which of two additional peers they would be more likely to include in a leisure time activity (playing video games). Those additional peers were both described as new in class: one moved here from another German town and the other one was a refugee that had fled with their family from Syria. Additionally, the skill level related to the activity (i.e., playing video games) of these two protagonists was varied. The scenario started for all participants as follows (the names of the protagonists matched the gender of the participants):

Imagine you have a group of friends at school. You usually spend recess and much of your free time together. The following situation refers to this group.

Now, imagine you and your group are planning to play video games this afternoon. You can only invite one more person. But there are two boys/girls who would like to join your group: Lukas/Laura and Rami/Shata. Both are new at your school. Lukas/Laura moved here from Bremen, he/she is German. Rami/Shata came to Germany with his/her family as a refugee from Syria.

This text was followed by a sentence in which we varied the skill level of the protagonists by indicating the experiences they had with video gaming. This led to four conditions (see

Table 1. Overview of experimental conditions.

Condition		Wording
A	Syrian peer skilled, German peer skilled	Both are very good in playing Wii.
B	Syrian peer not skilled, German peer not skilled	Both have never played Wii before.
C	Syrian peer skilled, German peer not skilled	Rami/Shata is very good in playing Wii. Lukas/Laura has never played Wii before.
D	Syrian peer not skilled, German peer skilled	Lukas/Laura is very good in playing Wii. Rami/Shata has never played Wii before.

). That is, there were two conditions in which both protagonists had the same skill level (condition A and B). And, there were two conditions with differences in the skill levels; in one, the German peer was more skilled than the Syrian peer (condition D), and in the other one, the Syrian protagonist was more skilled than the German protagonist (condition C).

After reading the scenario, the participants were asked to answer the following three questions for each protagonist separately: (1) How likely is it that you would choose xxx? (own decision). (2) What do you think, how likely is it that your group would choose xxx? (expected group decision). (3) Do you think your group should choose xxx? (prescriptive group decision).

Each of these measures was displayed on an individual page, which included the questions regarding both protagonists. The questions were presented in the same order for all participants (as described above). The participants answered all questions on six-point Likert-type scales. Specifically, for the first two questions (own decision, expected group decision), the scale ranged from 1 = very unlikely to 6 = very likely. For the third question (prescriptive group decision), the scale ranged from 1 = not at all to 6 = definitely. Additionally, for each question, participants were asked to provide justifications for their decisions (open-ended question: Why?).

Previous research on social inclusion has mostly employed a forced-choice paradigm, requiring participants to choose between two peers. (e.g., Hitti and Killen 2015; Mulvey et al. 2014; Toppe et al. 2020). In contrast, in the current study, the participants were asked to rate the likelihood of inclusion for both a refugee and local peer separately. With this study design, a more detailed picture of their evaluations and decisions could be assessed.

2.4. Coding of Open-Ended Answers

2.4.1. Knowledge About Refugees

Participants were asked what they knew about refugees in general and, in a second question, what they knew about Syrian refugees specifically. A coding system was created that was applied for both questions. The system included four categories: (1) correct answers (e.g., “they had to leave their country because of war”, (2) stereotypical answers (e.g., “they come here to spread violence”; “most of them are criminals”), (3) no knowledge (e.g., “I don’t know anything about refugees”), and (4) undifferentiated (answers that did not respond to the question, e.g., “it doesn’t matter where refugees come from, that does not change anything”). Answers were coded by two independent coders. Interrater reliability was high (Cohen’s kappa = .90).

2.4.2. Justifications for Inclusion Decisions

To code participants’ answers to the open-ended questions (e.g., the justifications of their decisions), a coding system was developed based on prior research (Beißert et al. 2020; Beißert and Mulvey 2022) that was extended by additional categories inductively developed from the surveys themselves (see

Table 2. Coding system and frequencies of usage.

	Own Decision		Expected Group Decision		Prescriptive Group Decision
	German	Syrian	German	Syrian	German	Syrian
MORAL DOMAIN
Moral “because there should be fairness”	103	102	42	37	132	130
Prosocial “because I want to help her find friends”	165	165	73	95	164	189
SOCIETAL DOMAIN
Group fit “it’s easier to play with someone who knows our culture”	8	3	20	20	11	5
Language “if she doesn’t know German, we can’t explain her the game.”	97	98	103	93	47	41
Skills “I’d choose her because she is better in playing Wii.”	66	52	74	62	30	23
Origin “I’d choose her because she is from Germany.”	19	16	79	67	21	36
Hostility and stereotypes “refugees don’t belong here”, “refugees will steal our stuff”	13	21	58	73	8	18
PSYCHOLOGICAL DOMAIN
Autonomy “because I want to get to know her”	17	17	8	18	10	18
Psychological information “if she is nice and friendly why should I not choose her”	83	78	56	52	52	46
Other	11	4	10	4	16	6

for the coding system and examples; see Appendix A for an overview of the frequencies of code use). The coders were allowed to code up to three relevant justifications per statement. A value of 1.0 was used if the participant assigned one code. If two codes were used, each was assigned a value of 0.5. If they used three codes, each of them received a value of 0.33. Two independent coders completed the coding. Based on 30% of the interview, interrater reliability was very high, with Cohen’s kappa = .98.

3. Results

To analyze participants’ inclusion decisions and their reasoning, repeated-measures ANOVAs were used. For the inclusion decisions, preliminary analyses showed that there were no effects based on the participants’ gender and no systematic effects of age related to our research questions. Thus, these variables were not included as factors in the analysis. However, age was included as a covariate to show effects above and beyond age. For reasoning analyses, age was also included as a covariate and we tested only two-way interactions (for instance, between type of reasoning and measure).

3.1. Inclusion Decisions

To test for differences in inclusion decisions for the two protagonists and the three measures, a 4 (condition: German skilled and Syrian skilled, German skilled and Syrian not skilled, German not skilled and Syrian not skilled, German not skilled and Syrian skilled) × 2 (protagonist: German, Syrian) × 3 (measure: own inclusion decision, expected group decision, prescriptive group decision) ANOVA was conducted with repeated measures on the last two factors with age as a covariate. The Huynh–Feldt adjustment was used to correct violations of sphericity.

The analysis revealed an interaction effect of protagonist and measure (F (1.56, 620.63) = 11.73; p < 0.001; ηp² = 0.03). Pairwise comparisons revealed that for their own decision and the prescriptive group decision, participants were more inclusive of the Syrian protagonist than of the German one, whereas in the expected group decision, participants expected their group to be more inclusive of the German protagonist (see Figure 1).

However, additionally, there was a significant three-way interaction of protagonist, measure, and condition (F (4.67, 620.63) = 9.48; p < 0.001; ηp² = 0.07). Pairwise comparisons revealed that the patterns varied between the different conditions. For their own decision, in the condition where the German was skilled and the Syrian was not skilled, participants were more inclusive of the Syrian protagonist than of the German protagonist (p = 0.006). In all other conditions, the differences between the Syrian and the German peer were not significant (p > 0.05). For the expected group decision and the prescriptive group decision, there were no differences between the two protagonists in the condition where the Syrian was skilled and the German was not skilled (p > 0.05). In all other conditions, participants expected the group to be more inclusive of the German protagonist than of the Syrian one (ps < 0.001); however, they indicated that they thought that the group should be more inclusive of the Syrian protagonist (condition A: p = 0.006; condition B: p = 0.039; condition D: p < 0.001). See Figure 2 for an overview of the adolescents’ inclusion decisions on all three measures by experimental condition.

3.2. Reasoning Analyses

Appendix A gives an overview of the frequencies of code use by the three different measures.

To test for differences in adolescents’ reasoning, a 4 (condition: German skilled and Syrian skilled, German skilled and Syrian not skilled, German not skilled and Syrian not skilled, German not skilled and Syrian skilled) × 2 (protagonist: German, Syrian) × 3 (measure: own inclusion decision, expected group decision, prescriptive group decision) × 9 (code: moral, prosocial, group fit, language, skills, origin, autonomy, psychological information, hostility/stereotypes) ANOVA was conducted with repeated measures for the last three factors with age as a covariate. The Greenhouse–Geisser adjustment was used to correct violations of sphericity.

The findings demonstrated an interaction effect of code and measure (F (11.26, 4558.52) = 3.73; p < 0.001; ηp² = 0.01). Pairwise comparisons revealed the following results. Participants used more moral and prosocial reasoning in their own decision and the prescriptive group decision than in the expected group decision, and moral reasoning was used even more in the prescriptive group decision than in their own decision. Reasoning related to group fit, language, origin, and skills was used more in the expected group decision than in the prescriptive group decision, and group fit and origin were also used more than they were in adolescents’ own decision. Language and skills reasoning were used more and origin-related reasoning less in their own decision than in the prescriptive group decision. Reasoning including aspects of hostility and stereotypes was used more in the expected group decision than in the prescriptive group decision and than in adolescents’ own decisions. Justifications referring to psychological information were used more in adolescents’ own decisions than in the expected group decision and than in the prescriptive group decision. See

Table 3. Means and standard errors of reasoning by three different measures.

	Own Decision M (SE)	Expected Group Decision M (SE)	Prescriptive Group Decision M (SE)
Moral	0.22 (0.02) a,c	0.09 (0.01) a	0.30 (0.02) c
Prosocial	0.35 (0.02) b	0.18 (0.02) b	0.39 (0.02)
Group fit	0.04 (0.01) e	0.15 (0.02) d,e	0.03 (0.01) d
Language	0.01 (0.01) j	0.04 (0.01) f	0.02 (0.01) f,j
Skills	0.04 (0.01) k	0.17 (0.02) h	0.07 (0.01) h,k
Origin	0.21 (0.02) i	0.23 (0.02) g	0.11 (0.01) g,i
Autonomy	0.04 (0.01)	0.03 (0.01)	0.04 (0.01)
Psychological information	0.19 (0.02) n	0.12 (0.01) n,o	0.12 (0.01) o
Hostility/stereotypes	0.13 (0.01) l	0.15 (0.02) l,m	0.06 (0.01) m

^{a,b,c,e,f,g,h,j,k,l,m,n,o} p < 0.001; ^d p < 0.01; ⁱ p < 0.05.

for means and p-values.

Further, there was an interaction effect of code and condition (F (15.21, 2053.91) = 9.00; p < 0.001; ηp² = 0.06). Pairwise comparisons revealed interesting patterns in the use of the various categories: In the conditions in which both protagonists were skilled (condition A) or both were not skilled (condition B), prosocial reasoning and reasoning referring to the protagonists’ skills were used less than in the conditions in which the skill levels of the protagonists differed (condition C and D). Additionally, in the conditions in which both protagonists were skilled (condition A) or both were not skilled (condition B), moral reasoning and reasoning including stereotypes or hostility were used more than in the condition in which the Syrian peer was not skilled and the German peer was skilled (D). Further, the categories of origin and psychological information were used more in the condition where both protagonists were skilled (condition A) than in the two conditions with different skill levels (condition C and D). Finally, language-related reasoning was used more in the two conditions in which the protagonists had the same level of skills (A, B) than in the condition in which the Syrian peer was skilled and the German peer was not skilled (C). Additionally, in the condition where both peers were not skilled (B), language-related reasoning was used more than in the condition in which the German peer was skilled and the Syrian peer was not (D). There were no differences between the four conditions for group fit- or autonomy-related reasoning. See

Table 4. Means and standard errors of reasoning by the four different conditions.

	Syrian Peer Skilled, German Peer Skilled (Condition A) M (SE)	Syrian Peer Not Skilled, German Peer Not Skilled (Condition B) M (SE)	Syrian Peer Skilled, German Peer Not Skilled (Condition C) M (SE)	Syrian Peer Not Skilled, German Peer Skilled (Condition D) M (SE)
Moral	0.25 (0.03) i	0.22 (0.03) k	0.20 (0.02)	0.14 (0.03) i,k
Prosocial	0.19 (0.03) a,c	0.24 (0.03) b,d	0.35 (0.03) a,b	0.44 (0.03) c,d
Group fit	0.04 (0.01)	0.02 (0.01)	0.03 (0.01)	0.02 (0.01)
Language	0.20 (0.03) q	0.25 (0.03) r,s	0.11 (0.03) q,r	0.15 (0.03) s
Skills	0.01 (0.02) e,f	0.02 (0.02) g,h	0.20 (0.02) e,g	0.23 (0.02) f,h
Origin	0.13 (0.02) m,n	0.10 (0.02)	0.07 (0.01) m	0.07 (0.01) n
Autonomy	0.04 (0.01)	0.03 (0.01)	0.04 (0.01)	0.02 (0.01)
Psychological information	0.21 (0.03) o,p	0.14 (0.02)	0.08 (0.02) o	0.13 (0.02) p
Hostility/stereotypes	0.09 (0.02) a,b,j	0.10 (0.01) l	0.06 (0.01) a	0.05 (0.01) b,j,l

^{a,c,d,e,f,g,h,o,r} p < 0.001; ^i,l,m,n p < 0.01; ^b,j,k,p,q,s p < 0.05.

for means and p-values.

3.3. Knowledge About Refugees

At the end of the survey, participants were asked what they knew about refugees in general and what they knew about Syrian refugees specifically. Most of the adolescents had accurate knowledge about Syrian refugees (80%) and refugees in general (82.5%). Only a small percentage of the participants stated that they did not have any knowledge about Syrian refugees (8%) and about refugees in general (4%). However, there was still a certain number of adolescents giving stereotypic statements about Syrian refugees (11%) and about refugees in general (12.4%). The remaining one percent of the participants gave undifferentiated answers.

A t-test was conducted to test whether participants expressing stereotypes made different individual inclusion decisions regarding the Syrian refugee than those not expressing stereotypes. This analysis revealed that adolescents with stereotypical views (M = 3.44; SD = 1.41) were less inclusive regarding the Syrian peer than adolescents without stereotypical answers (M = 3.93; SD = 1.35; t(1.53) = 2.39; p = 0.009).

4. Discussion

The present study examined the openness of youths in Germany to include refugee peers from Syria in their leisure time activities. Namely, adolescents’ inclusion decisions and reasoning were assessed using hypothetical scenarios. More specifically, we assessed adolescents’ own inclusion decisions as well as what they would expect their peer group to decide and what they think their peer group should carry out. The aim of this research was to better understand the interplay of moral considerations and aspects of group functioning in adolescents’ inclusion decisions regarding refugee peers.

4.1. Inclusion Decisions

In the first step, when looking at the data across all four experimental conditions, we could replicate prior findings by Beißert and Mulvey (2022) demonstrating that the adolescents thought that they themselves would be more inclusive toward a Syrian peer than a German peer and that their peer group should be more inclusive toward a Syrian peer than a German peer. In contrast to their own decisions and to what they think their peer group should do, adolescents expected their group to be more inclusive of a German peer than of a Syrian peer.

However, when considering the experimental manipulation of the relevant skills, these patterns changed. More specifically, the pattern described above was only found if the German peer was skilled and the Syrian peer was not. Interestingly, and in contrast to our expectations, adolescents were more likely to choose the Syrian peer than the German peer in this condition (own decision). Based on prior findings demonstrating that aspects related to smooth group functioning influence adolescents’ inclusion decisions (Hitti et al. 2011; Rutland and Killen 2015), we had expected that especially in this condition adolescents would more likely choose the German peer to increase the functionality of the group. However, the reasoning analyses revealed that many adolescents in our study considered that skills might come from experience. Many justified their choice by saying that the Syrian peer had never had the chance to play video games before and thus they chose him/her to provide the opportunity. In the other conditions, there were no differences in the likelihood of choosing one peer or the other. That is, skills—or rather experiences—regarding a certain activity made a difference for adolescents’ inclusion decisions, even though the influence seemed to be different than what we expected it to be.

For the expected group decision, the findings are in line with our considerations related to group dynamics. On the one hand, in three of the four conditions, the participants expected their peer groups to prefer the German peer over the Syrian peer. However, in the condition where the Syrian peer was skilled and the German one was not, there was no difference between the two peers. That is, when the Syrian peer was skilled, this increased their likelihood of being chosen and decreased the preference for the German peer. This is in line with prior research showing that adolescents often prefer including an ingroup member over an outgroup member to enhance group functioning (Killen and Smetana 2015; van Bommel et al. 2021). However, in the condition in which the Syrian peer was more skilled than the German peer, smooth group functioning could be achieved by choosing the outgroup member, and thus, the likelihood of choosing the German peer decreased.

For the prescriptive group decision, we found a similar impact of the skill level. Whereas in all other conditions the adolescents thought that the group should choose the Syrian peer, in the condition where the Syrian peer was skilled and the German peer was not, there was no difference in the recommendation whom of the two to choose. That is, when a Syrian peer had no obvious disadvantage compared to the German peer, the moral obligation to prefer the Syrian over the German peer vanished.

So, as hypothesized, the variation in the skill level of the protagonists mattered for the participants’ inclusion decision—though in a different way than what we had expected. Namely, participants were more likely to include the Syrian peer than the German peer only in the condition where the German protagonist was skilled and the Syrian protagonist was not skilled. However, the reasoning analyses demonstrated that our manipulation of the skill level led participants to consider experience and express a desire to provide a new opportunity. In this scenario, participants often stated that they would include the Syrian peer because they had never had the chance to play video games or because they wanted to provide them with the opportunity to play video games as they had never had the chance before. In future research, it would be interesting to also vary the stakes of the gameplay; it may be that skill alone (and not creating opportunities for new experiences) would matter if the video game play was competitive and there were prizes to be awarded. Prior research does suggest that adolescents are especially likely to consider group functioning in competitive situations (McGuire et al. 2018, 2019).

4.2. Reasoning

The reasoning enabled a deeper understanding of adolescents’ underlying motives for their inclusion decisions. Interestingly, in terms of reasoning, there were differences between the three measures. Namely, reasons from the moral domain were used mostly for the adolescents’ own decision and for the prescriptive group decision. That is, these two decisions seemed to be driven mainly by moral considerations. In contrast, considerations from the societal domain, more specifically aspects related to smooth group functioning such as group fit, origin, language, and skills, were used more in the expected group decision than in the prescriptive group decision, and some of them were also used more than they were in the own decisions. That is, the expected group decision seemed to be driven mainly by aspects related to group functioning. This is in line with the theoretical framework of the Social Reasoning Development perspective, positing that when making social decisions, individuals balance moral considerations with the interests of the social groups they belong to (Rutland et al. 2010; Rutland and Killen 2015). That is, for inclusion or exclusion decisions in inter-ethnic contexts, both adolescents’ own moral principles as well as group norms are relevant and may impact adolescents’ decisions and behavior (Killen et al. 2017; McGuire et al. 2018).

Further, hostility and stereotypes were mainly relevant for the expected group decision and hardly used to justify the own decision or the prescriptive group decision. That is, adolescents did not base their individual decisions on hostility or stereotypes, but they expected their peer groups to do so. This could be understood as a risk factor because group pressure is a strong issue for adolescents and their expectations about their groups’ attitudes can affect their own decisions and behavior (Brown 2013). So, this could be a starting point for prevention work at schools.

Further, it was striking that reasoning including stereotypes or hostility was (across all measures) used more often in conditions in which both protagonists had the same skill level. Thus, it may be that if there is no “objective” reason to base the inclusion decision on, stereotypes might become more relevant. This assumption is supported by the finding that in conditions with varying skills, adolescents would include skill-related reasons more frequently in their considerations than in conditions in which both protagonists had the same skill level. So, if there is factual information to base the decision on (e.g., the skills of the protagonist), adolescents will use this information. But, if there is a lack of information, they more strongly rely on stereotypes.

And, finally, it was interesting to see that prosocial reasoning was used much more frequently in conditions with varying skills. This can be explained by the fact that the participants conceptualized our variation in the skills, rather, as experiences, so in the conditions in which one of the peers was lacking these experiences, they wanted to provide opportunities.

4.3. Knowledge About Refugees

Encouragingly, the vast majority of the adolescents had accurate knowledge about (Syrian) refugees and only very few participants stated that they did not know anything about refugees. However, there was still a small but not negligible proportion of adolescents in the study who made statements characterized by common stereotypes. And, these stereotypes seemed to shape the adolescents’ behavioral intentions for the inclusion of a Syrian refugee peer as adolescents with stereotypical answers were less inclusive regarding the refugee peer than adolescents not expressing stereotypes in their answers. This means that there is still a need for awareness-raising and prevention work in schools.

4.4. Strengths and Limitations

The current study adds to our understanding of how adolescents consider different factors when making inclusion decisions. Additionally, the study provides nuanced insight into their underlying reasons for their decisions, showing both that adolescents are aware of limited opportunities refugees may have and demonstrate a desire to rectify these opportunity gaps and that they understand that their peers may be driven more by concerns about group functioning than about creating opportunities. However, this study does have limitations. First, the assessment of the survey took place in class. As completing this study in a group setting may have impacted responses, future research should try different approaches such as a survey taken alone.

Second, as the scenarios were hypothetical, it is not clear whether adolescents would be as inclusive as they indicate in real life. Additionally, we were unable to assess participants’ intergroup contact with Syrian refugees, but it may be that those with more prior interaction with Syrian refugees would respond differently than those with low contact.

Finally, our reasoning analyses focused only on two-way interactions to ensure that we were adequately powered, but future research might delve more deeply into how reasoning may differ on multiple axes.

5. Conclusions

In sum, this research suggests that adolescents are generally open to including refugee peers into their peer activities. However, when making inclusion decisions, they carefully consider both their group’s best interests and the experiences of those who seek to join their group for social activities. The findings extend prior research on individual and group evaluations (Mulvey et al. 2014) to a new context, clarifying that when considering the inclusion of Syrian refugee adolescents and German peers, adolescents make sophisticated decisions. While there may be differences found in different contexts, these results suggest that interventions to promote the inclusion of refugee peers should focus on helping adolescents to recognize that their peers generally want to include refugee peers and to increase opportunities for refugee peers to have the chance to engage in leisure activities alongside their new peers.