German Pre-service Teachers' Evaluations of and Reactions to Interethnic Social Exclusion Scenarios

This study investigated pre-service teachers' evaluations, reactions, and interventions with regard to interethnic exclusion scenarios in Germany. More specifically, we focused on pre-service teachers (N = 145, 99 female, Mage= 21.34) in the role of observers of exclusion among students. Using hypothetical scenarios in which either a German or a Turkish boy was excluded by other children of his class, we assessed teachers' evaluations of this exclusion behavior. This included evaluating how likely teachers were to intervene in the situation and what they would specifically do. The aim of this research was to examine whether the origin of an excluded student represents a relevant category for teachers' evaluations of and reactions to social exclusion. In addition, we aimed to determine whether teachers include aspects related to group functioning in their considerations. The analyses demonstrated that teachers generally reject social exclusion, with female participants rejecting exclusion even more than male participants. Further, participants evaluated the exclusion of a Turkish protagonist as more reprehensible than the exclusion of a German protagonist. Regarding the likelihood of intervention, the origin of the excluded person was only relevant for male participants; i.e., they were less likely to intervene when the excluded person was German than when the excluded person was Turkish. Analyses of teachers' reasoning revealed their strong focus on inclusion as a social norm, especially in cases of interethnic exclusion. That is, when participants reasoned about the exclusion of the Turkish protagonist, they referred to the social norm of inclusion much more than when talking about the German protagonist. In contrast, aspects related to group functioning were scarcely of importance. In terms of the specific actions that participants would undertake as a reaction to the exclusion situation, no differences related to the origin of the excluded person were found. Hence, the origin of the excluded person factored into both the evaluation of the exclusion and the likelihood of intervention, but once the decision to intervene was made, there were no differences in the specific actions. The results are discussed in light of practical implications and teacher training as well as in terms of implications for future research.


INTRODUCTION
The German educational system-as many others in Western Europe and the United States-has a student population with very heterogeneous cultural backgrounds. Germany has been an immigration country for at least half a century (Oltmer, 2017). As a result, the general population and, thus, the student body are characterized by considerable cultural diversity. Unfortunately, Germany has not been overly successful in achieving integration and educational equality so far. Even though some positive development has been noted in recent years, research has extensively demonstrated that students from ethnic minorities experience various disadvantages in the educational system (Müller and Ehmke, 2016;Weis et al., 2019). For instance, they are overrepresented in lower school tracks and underrepresented in higher school tracks (Baumert and Schümer, 2002;Kristen and Granato, 2007); they drop out of school more often (Rumberger, 1995); they are recommended for lower school tracks more often , and their academic achievement tends to be lower than that of their native peers (Walter, 2009;Klieme et al., 2010). While a seminal body of research has focused on achievement-related disparities, little is known about the social situation of immigrants in the educational system. For instance, do students from ethnic minorities face more social exclusion in peer interactions than their native peers? Are they socially well integrated into their peer group? And what roles do teachers play in this context?
While the study of interethnic friendships has a somewhat longer tradition (Reinders, 2004;Schacht et al., 2014), in recent years, research has started to investigate exclusion behavior among students in the context of interethnic group processes. As a result, research from Germany has recurrently shown that students are much more likely to choose peers of the same race as a friend than peers from another race (e.g., Kalter and Kruse, 2015;Schachner et al., 2016). This holds especially for close friendships (Winkler et al., 2011) and is particularly evident for children of Turkish origin (Schachner et al., 2016;Carol and Leszczensky, 2019). In addition, it has been shown that social exclusion often happens based on group memberships such as race or ethnicity (Killen and Stangor, 2001;Killen et al., 2010;Abrams and Killen, 2014;Hitti and Killen, 2015). Additionally, minority groups are especially likely to be confronted with stereotypical mindsets and behavior, which can also result in the exclusion of students from ethnic minorities (Killen et al., 2013). In line with these findings, belonging to an ethnic minority has been identified as a risk factor for exclusion among peers. Plenty and Jonsson (2017), for instance, investigated social exclusion among adolescents and found that students from ethnic minorities were rejected more than majority youth.
What remains unclear is the role of teachers in this context. In general, teachers can have a strong impact on their students' attitudes and behavior. For instance, they establish rules that indicate which behaviors are acceptable in class and which are not. They are important role models for their students, especially when it comes to ethnic topics (Evans, 1992). Thus, how they interact with their students is particularly important. With the way teachers behave, they transmit their attitudes and beliefs (Muntoni and Retelsdorf, 2020). Through their behavior they transmit both explicit and implicit messages about their attitudes related to the importance of inclusion and diversity in schools (Cooley et al., 2016). In this way, teachers' reactions and responses to interethnic social exclusion can have an impact on their students' attitudes and behavior. For instance, in a study that Verkuyten and Thijs (2002b) conducted in the Netherlands with students aged between 10 and 13 years old, youth from ethnic minorities reported less exclusion and dismissive behavior when they believed that they could confide in their teacher and when they believed that their teacher would take action if they told him or her that they had been treated unfairly. Thus, it can be assumed that teachers' commitment to addressing intergroup exclusion issues and their engagement in explicit discussions about the importance of inclusion have positive effects on their students' intergroup attitudes and behavioral tendencies related to inclusion or exclusion (Cooley et al., 2016). That is, the way teachers react to interethnic exclusion situations forms and impacts their students' attitudes and behavior in a subtle way. In this way, teachers can contribute to students' moral development and help them understand moral norms such as equality, fairness, and inclusion (Cooley et al., 2016). As teachers play such an important role in their class's social system, especially in the context of interethnic group processes, and as their behaviors can contribute to positive or negative intergroup dynamics, research is needed on teachers' responses to interethnic exclusion (i.e., social exclusion including students from different ethnicities) in order to better understand the social climate in classrooms.
Although there is only little research on teachers' reactions to exclusion, there are some studies that focus on teachers' responses to bullying which include exclusion behavior (Yoon and Kerber, 2003;Bauman and Del Rio, 2006;Shur, 2006). However, we are not aware of any research that has examined teachers' reactions to interethnic exclusion among students. Therefore, many open questions remain: How do teachers evaluate interethnic exclusion and how do they deal with it? Are their evaluations of exclusion biased depending on the ethnic origin of the excluded person? Do such biases influence their behavioral tendencies?

Teachers' Evaluations of Students
A considerable part of a teacher's work involves evaluating students. In the context of achievement-related evaluations, it has been shown that teachers' judgments are often biased by irrelevant aspects related to their students (Glock, 2016;Holder and Kessels, 2017). Although teachers use relevant information about their students, their judgments or evaluations are often biased by information that, in fact, should be irrelevant for the respective judgment, such as a student's ethnicity (McCombs and Gay, 1988;Parks and Kennedy, 2007;Bonefeld et al., 2017). This has been shown for pre-service teachers as well as for inservice teachers Bonefeld and Dickhäuser, 2018;. Whereas, teachers' biased evaluations have been intensively investigated in the context of grading or other achievement-related aspects, it remains unclear whether ethnic origin is also relevant for teachers' judgments about social interactions, particularly in the context of social exclusion. Little is known about how teachers perceive, evaluate, and react to interethnic exclusion. As reactions to social exclusion might already be an important topic in teacher training, the current study investigates pre-service teachers' reactions to interethnic exclusion scenarios in Germany.
Although teachers' evaluations of their students are affected by the characteristics of the students, their own characteristics can also be important in this context (Südkamp et al., 2017). For example, a teacher's gender might be a relevant characteristic which has an impact on the evaluation of social exclusion. Previous research has shown that females (children, adolescents, and adults) tend to oppose exclusion more strongly than males (e.g., Killen and Stangor, 2001;Horn, 2003;Beißert et al., 2019). One reason for this could be gender-specific socialization. Typically, girls' socialization has a stronger focus on harmony, caring behavior, and the avoidance of interpersonal struggles (Cross and Madson, 1997;Zahn-Waxler, 2000). Further, parents have been shown to address the harmful consequences of aggressive behavior much more in the education of girls than that of boys, leading to enhanced empathy in girls compared to boys (Smetana, 1989). Based on these gender-specific socialization aspects, females might feel a stronger need to prevent exclusion and promote inclusion and, thus, might also be more likely to intervene in exclusion situations. In the current study, we want to examine whether these gender effects can also be found in teachers' reactions to social exclusion among students.
Further, a teacher's own immigration history might be particularly relevant for their evaluation of interethnic social exclusion. Prior research on teachers' attitudes has shown that teachers' own immigration history influences their evaluations regarding students of different ethnicities (Kleen et al., 2019). In terms of social exclusion, having an immigration history in the family might, for instance, enhance a teacher's empathy with an excluded student who is from an ethnic minority. Thus, research on interethnic exclusion should always take into account the participants' own immigration history.

Intergroup Processes and Social Exclusion
It is in the nature of humans to organize the social world into categories (Brewer, 2001). Just as we classify plants and animals into taxonomies based on their typical characteristics, we classify people into groups. Categories help us to simplify and organize our complex world. This process of classifying people into groups is called social categorization. Whenever we perform such categorizations, we also differentiate between ingroup (a group to which we psychologically identify as being a member) and outgroup (a group with which we do not identify). According to social identity theory, people define their social identity based on group memberships (Tajfel, 1978;Tajfel and Turner, 1979). Against this background and with a view to achieving a positive social identity, people desire to identify with and belong to social groups seen as superior to others (Tajfel, 1982). Group members compare their ingroup to outgroups and positively define their ingroup to maintain their status (Tajfel, 1974;Tajfel and Turner, 1986). This preference for the ingroup is called ingroup favoritism or ingroup bias; it results in preferences that favor or promote the ingroup's status, often at the expense of other groups . Given this general tendency to prefer ingroup members and depreciate outgroup members, it is not surprising that exclusion is often based on group memberships such as race or ethnicity, and that ingroup-outgroup processes affect the evaluation of exclusion situations (Dovidio et al., 2005;Hitti et al., 2011;Killen et al., 2013). Furthermore, when children and adolescents have to justify exclusion, they often cite reasons related to smooth group functioning (Hitti et al., 2011;Mulvey, 2016). Hence, it is not only "raw" ingroup favoritism that promotes the exclusion of outgroup members. In many instances, children and adolescents expect outgroup members to have a negative impact on the group functioning within their ingroup.
There is extensive research demonstrating the role of ingroupoutgroup processes and group functioning for exclusion among children and adolescents. However, it remains unclear whether teachers' evaluations of student exclusion are also affected by intergroup processes such as ingroup favoritism. Furthermore, it is unclear whether teachers, as observers, also consider group functioning aspects when evaluating interethnic social exclusion among their students.

Current Study
The current study investigated German pre-service teachers' reactions to interethnic exclusion scenarios. More precisely, we focused on pre-service teachers in the role of observers of exclusion among students and examined their evaluations of these situations as well as their hypothetical interventions. The study aimed to shed light on the question of whether teachers' evaluations of interethnic exclusion situations are biased by ingroup-outgroup processes based on ethnicity. More specifically, we analyzed whether the ethnic origin of an excluded student represents a relevant category for teachers' evaluations and reactions, and whether they include group functioning aspects in their considerations. In our study, we focused on students with Turkish roots because they are the largest ethnic minority in Germany (DESTATIS, 2016). Further, Turkish students are a very important group because research has shown that negative attitudes about Turkish people are widespread in Germany Glock and Karbach, 2015;. In order to examine the aforementioned issues, the current study used hypothetical exclusion scenarios in which the excluded protagonist was either a Turkish student or a German student. We assessed how pre-service teachers evaluated the exclusion scenario as well as how likely they would intervene in such a situation and how they would specifically react.

Evaluation of Exclusion
Given that the need to belong and be accepted by others is a fundamental human need (Baumeister and Leary, 1995), it is not surprising that children, adolescents (Killen and Rutland, 2011), and adults (Beißert et al., 2019) typically reject exclusion. Thus, we expected a strong general tendency to reject exclusion (rightskewed distribution) across both protagonists. However, given the importance of intergroup processes-more precisely ingroup favoritism-in social situations, we assumed that the participants would evaluate the exclusion of a German protagonist (ingroup member) as more reprehensible than the exclusion of a Turkish protagonist (outgroup member). Further, based on prior research with children and adolescents as well as on considerations related to gender-specific socialization, we expected females to generally reject exclusion more strongly-independently of the origin of the protagonist.

Likelihood of Intervention
It was an open question as to whether participants would differ in their reactions to the exclusion scenario depending on the respective protagonist (German vs. Turkish). For instance, were they more or less likely to intervene in situations with one protagonist or with the other? And when they decided to intervene or not, would their considerations differ depending on the excluded person?

Specific Actions
Our ultimate objective was to explore how exactly pre-service teachers would react to the exclusion situation and whether their specific actions would differ based on the origin of the excluded protagonist.

Participants
The study included 145 pre-service teachers (99 female, M age = 21.34, SD = 2.13) from various school tracks who were students at a university in the southwest of Germany. Within this sample, 58% of the participants had completed a school teaching internship as a mandatory part of their program. Sixteen of the participants had an immigration history in their family (i.e., at least one parent was born in a country other than Germany), but all participants were born in Germany. Three participants were excluded from the analyses because they had a Turkish background and, thus, the outgroup manipulation in the scenarios would not have worked for them. Participation was voluntary and informed consent was obtained from all participants.

Design and Procedures
The study was conducted in a research lab at the participants' university. The participants were recruited personally on campus. Additionally, flyers and posters advertising the study were distributed on campus. After arriving in the lab, the participants were seated in front of a computer screen. They were informed that they were participating in a study about social issues in school. Before the assessments started, they were informed of their data protection rights and learned that participation in the study was anonymous and voluntary, and that there were no disadvantages if they decided not to participate or leave the study early without completing it. Participants had to confirm that they understood the information and were willing to participate in the study. The study took approximately 10 min per person, and the participants were given a cupcake as an incentive for participation.
The participants completed a computer-based survey including a questionnaire collecting demographical information and were then presented with a hypothetical scenario in which one student was excluded from a learning group by his classmates. The excluded protagonist had either a typical German or a typical Turkish name. The names used in the scenario had been pretested in a former study by Bonefeld and Dickhäuser (2018). The exact wording of the scenario was as follows: "While packing up after class (grade 7 1 ), you observe some students making an appointment to study together. Max/Murat would like to join the learning group. The other students tell him that he can't join." The study used a between-subjects design, and the participants were randomly assigned to the experimental conditions (71 were assigned the version with the German protagonist, 74 the version with the Turkish protagonist).

Measures
The participants' evaluations of the exclusion scenario were assessed with a scale consisting of three items on a seven-point Likert-type scale. The participants were asked to indicate how (1) not okay/okay, (2) unfair/fair, and (3) unjustifiable/justifiable they evaluated the scenario. A score was created indicating a participant's evaluation of the exclusion based on these three items (Cronbach's alpha = 0.84). High numbers indicate high acceptability of exclusion; low numbers indicate strong rejection of exclusion.
Additionally, participants were asked how likely it was that they would intervene in the situation if it happened in their class. This was also assessed using a seven-point Likert-type scale (very unlikely to very likely). Further, they were asked to justify their decision and to indicate what specifically they would have done (open-ended questions).

Coding of Open-Ended Questions
The coding systems for the open-ended questions (justification of likelihood of intervention, specific actions) were inductively developed from the surveys themselves (see Tables 1, 2 for an overview and examples). To prevent a loss of important information, coders were allowed to code up to three relevant justifications for each statement (if necessary). Coding was completed by two independent coders. On the basis of 25% of the interviews, interrater reliability was high, with Cohen's kappa = 0.85 for the justifications of the likelihood of intervention, and kappa = 0.95 for the specific actions. We included the mostused categories (all of which were used by more than 10% of the participants) in the analyses here.

Data Analyses
Univariate ANOVAs were used to test for differences in the evaluation of exclusion and the likelihood of intervention between the two different experimental conditions (German protagonist vs. Turkish protagonist). Repeated-measures ANOVAs were used for analyses on the justifications of the decisions and on the specific actions. In order to test for

Category
Example N Need for information "I would intervene in order to find out why they didn't let him join the group."

22
Children's autonomy "It's the children's' choice who to include or not"

20
Group functioning "If they don't like him, studying together wouldn't work well."

11
Empathy for the victim "I would intervene because the excluded person would be very sad."

17
Social norm of inclusion "Because exclusion is generally not ok!", "Everybody should get a chance to participate."

65
Class-oriented perspective "It is better for the sense of community to impede exclusion."  differences between male and female teachers, the variable gender was included in all analyses. As participants' own ethnic background might influence their responses and reactions, their family immigration history was included in all analyses as a control variable.

Evaluation of Exclusion
As expected, we found a general tendency to reject exclusion across both protagonists, i.e., a right-skewed distribution on the evaluation scale with a skewness of 1.43 (SE = 0.20), a mean of 1.95 (SD = 1.08), mode = 1.0, and median = 1.67. See Figure 1, for the distribution of the means of the evaluation scale. In order to test for differences in the evaluation of exclusion between the two protagonists, a 2 (protagonist: German, Turkish) × 2 (gender: male, female) univariate ANOVA was conducted with the participants' immigration history as a covariate. The results revealed a main effect of the protagonist, F(1, 141) = 19.72, p < 0.001, η p 2 = 0.12; see Figure 2. In contrast to our expectations, the participants evaluated the exclusion of the Turkish protagonist as more reprehensible (M = 1.64, SD = 0.78) than the exclusion of the German protagonist (M = 2.28, SD = 1.24). Further, as expected, a main effect of the participants' gender was found, F(1, 141) = 14.71, p = 0.001, η p 2 = 0.10, revealing that female participants rejected exclusion generally more strongly (M = 1.76, SD = 0.94) than male participants (M = 2.36, SD = 1.23). No effects of participants' immigration history were found. No interaction effects were found.

Likelihood of Intervention
In a next step, we analyzed differences in the participants' likelihood to intervene in such a situation by conducting a 2 (protagonist: German, Turkish) × 2 (gender: male, female) univariate ANOVA with participants' immigration history as a covariate. No main effects were revealed. However, a significant interaction of participants' gender and the protagonist was revealed, F(1, 140) = 4.39, p = 0.038, η p 2 = 0.03. Specifically, male participants were less likely to intervene in the condition with the German protagonist than in the condition with the Turkish protagonist (p = 0.024), whereas female participants did not differ between the two conditions (p = 0.710); see Figure 2.

Justification of Likelihood of Intervention
Analyses were conducted on the participants' reasoning based on the proportional use of the targeted justification codes (all of which were used by more than 10% of the participants). The resulting codes were: "social norm of inclusion, " "empathy for the victim, " "need for information, " and "children's autonomy." ANOVAs provide appropriate frameworks for performing repeated-measures reasoning analyses because they are robust to the problem of empty cells, whereas other data analytical procedures require cumbersome data manipulation to adjust for empty cells [see Posada and Wainryb (2008), for a more extensive explanation and justification of this data analytical approach].
In order to test for differences in participants' justifications based on the origin of the protagonist, a 2 (protagonist: German, Turkish) × 4 (justification: social norm of inclusion, empathy for the victim, need for information, children's autonomy,) × 2 (gender: male, female) ANOVA was run with repeated measures on the factor "justification" and with participants' immigration history as a covariate.
We found a main effect of justification F(2.50, 342.96) = 12.23, p < 0.001, η p 2 = 0.08. The Greenhouse-Geisser adjustment method was used to correct violations of sphericity. The analyses revealed that justifications referring to the social norm of inclusion were used much more frequently than any other type of justification (ps ≤ 0.001). This main effect was qualified by a  significant interaction effect of justification and the protagonist, F(2.50, 342.96) = 3.02, p = 0.039, η p 2 = 0.02. This meant that this justification was used more often than any of the other types of justifications only when referring to the Turkish protagonist, p < 0.001. When the participants referred to the German protagonist, there were no differences in the use of justifications. Additionally, comparisons revealed that this type of justification was used considerably more often when referring

Specific Actions
We also asked the participants what exactly they would have done if they had intervened in the exclusion situation. To analyze these specific interventions, we conducted analyses on the proportional use of the mentioned actions (all of which were referred to by more than 10% of the participants). The resulting categories were "ask for reasons, " "help to find inclusionoriented solution, " "explain norm of inclusion, " "find alternative solution for excluded student, " and "class-based intervention." In order to test for differences in the participants' specific actions, a 2 (protagonist: German, Turkish) × 5 (action: ask for reasons, help to find inclusion-oriented solution, explain norm of inclusion, find alternative solution for excluded student, classbased intervention) × 2 (gender: male, female) ANOVA was run with repeated measures on the factor "action" and with the participants' immigration history as a covariate. There was a main effect of action, F(3.28, 449.22) = 17.78, p < 0.001, η p 2 = 0.12. The Greenhouse-Geisser adjustment method was used to correct violations of sphericity. More specifically, participants stated they would ask for reasons (M = 0.37, SD = 0.38) and try to find an alternative solution for the excluded student (M = 0.24, SD = 0.38) more than they would help to find an inclusion-oriented solution (M = 0.13, SD = 0.28), explain the norm of inclusion (M = 0.06, SD = 0.19), or aim for a class-based intervention (M = 0.08, SD = 0.26), all ps < 0.05.
There were neither main effects of the protagonist or of the participants' immigration history nor any interaction effects.

DISCUSSION
The current study investigated pre-service teachers' reactions to interethnic exclusion scenarios in Germany. More specifically, we focused on pre-service teachers in the role of observers of exclusion among students. Using hypothetical scenarios in which either a German or a Turkish boy was excluded by other children in his class, we assessed teachers' evaluations of this exclusion behavior as well as the likelihood that they would intervene in the situation, and the specific action they would take. The aim of this research was to examine whether the origin of an excluded student represents a relevant category for teachers when responding to social exclusion.
Generally and regardless of the origin of the protagonists, we found a strong tendency to reject exclusion, with female participants rejecting exclusion even more strongly than male participants. However, the origin of the excluded student represented a relevant category for participants' evaluations of the exclusion scenario. Interestingly, the effect confounded our expectations: Contrary to our anticipations, the participants evaluated the exclusion of a Turkish protagonist as more reprehensible than the exclusion of a German protagonist. Therefore, origin affected teachers' evaluations, but in contrast to our expectations, not in the sense of ingroup bias (i.e., the tendency to favor ingroup members over outgroup members).
Regarding the likelihood of intervention, the origin of the excluded person was only relevant for male participants. They were less likely to intervene when the excluded person was German than when the excluded person was Turkish. For female participants, there was no difference; i.e., they were very likely to intervene independently of the origin of the protagonist.
The origin of the protagonist was also relevant for teachers' justifications of their decisions to intervene or not. Namely, an interesting interaction was found between the origin of the excluded student and the type of justification. When the participants reasoned about the exclusion of the Turkish protagonist, they referred to a general social norm of inclusion much more than when talking about the German protagonist. In addition, when speaking about the Turkish protagonist, they referred to the social norm of inclusion more often than to any other reason. In other words, the general norm of social inclusion seems to be particularly salient when ethnic minorities are involved and an important issue for teachers. Interestingly, female participants named the social norm of inclusion even more often than male participants-for both protagonist groups.
In line with our expectations, female pre-service teachers seem to value social inclusion even more than their male counterparts. They focus more strongly on social inclusion as a general norm, reject exclusion more strongly, and are very likely to intervene in exclusion situations in order to promote inclusion. Extending previous research that typically focused on exclusion in symmetrical relationships between peers, the current data showed that these gender differences also hold for pre-service teachers as observers of student exclusion. As females have been found to be more inclusive than males at various ages and in very different kinds of relationships or situations, this seems to be something deeply rooted in the minds of girls and women. This might indicate that these differences are based on gender-specific socialization. As a consequence, a possible way of enhancing males' inclusivity would be a stronger focus on inclusion and community during their early education and socialization.
In terms of the specific actions potentially taken by participants as a reaction to the exclusion situation, we found no differences related to the origin of the excluded person. Regardless of the origin, the vast majority of the participants would have asked the group to name the reasons why it was excluding the student, and also a large number of the participants would have tried to find an inclusion-oriented solution or an alternative solution for the excluded student. Thus, origin mattered for both the evaluation of the exclusion situation and the likelihood of intervention, but once the decision to intervene was made, there were no differences in the specific actions.
All in all, we found evidence that the origin of an excluded person is a relevant category for teachers' reactions to social exclusion. The origin of the excluded student affected teachers' evaluations of the exclusion scenario and, for males, also their likelihood to intervene or not. Additionally, their underlying considerations differed based on the origin of the excluded person. However, although these differences were significant, it is important to emphasize that our participants predominantly rejected exclusion for both protagonists, and the differences were not very big. Further, the teachers in our study did not exhibit ingroup bias. In contrast, they evaluated the exclusion of the student from an ethnic minority (outgroup member) as even more reprehensible than the exclusion of a German protagonist and referred to the value of social inclusion more often when talking about the minority student.
We found it highly encouraging that pre-service teachers generally reject interethnic exclusion and show no discrimination of children with Turkish origin. Moreover, pre-service teachers reject the exclusion of a Turkish protagonist even more than for a German student. This could be due to the official educational mission and protection mandates regarding minorities, which include the integration of ethnic minorities at school (Ungern-Sternberg, 2008). Given the huge heterogeneity in German schools, this mandate might be particularly salient in teachers' minds. This is also in line with the finding that our participants referred to the general norm of inclusion much more often when reasoning about the exclusion of a Turkish student. Further, this might indicate that pre-service teachers are sensitive to the fact that exclusion based on origin is an issue in schools, and that it is important to promote inclusion as a general norm in class. Due to the strong human need to belong, social exclusion can have severe implications for health and well-being (Buhs and Ladd, 2001;Rutland and Killen, 2015). For students from ethnic minorities, exclusion can have a particularly strong impact (Ward et al., 2001;Verkuyten and Thijs, 2002a). This makes it even worse that social exclusion among children and adolescents is often based on group memberships such as ethnicity (Killen and Stangor, 2001;Abrams and Killen, 2014). Thus, it is very promising that the pre-service teachers in our study attached great importance to the norm of social inclusion in class and did not succumb to ingroup bias.
However, the question arises: What makes interethnic social exclusion different from other issues that teachers have to evaluate or react to? As described above, in Germany, children with an immigration history are disadvantaged in educational settings in many ways. Especially when performing achievementrelated evaluations (e.g., grading), teachers seem to be biased by their students' ethnic origin (Bonefeld et al., 2017;Bonefeld and Dickhäuser, 2018). One possible explanation for this discrepancy between our findings regarding social exclusion and the findings of many prior studies on teacher bias in achievement evaluations is provided by the dual process models of information processing and judgment formation (Brewer, 1988;Fiske and Neuberg, 1990;Fiske et al., 1999). Such models assume that we typically process information via two routes when we make judgments about other persons: a more automatic route where we rely on obvious-but often irrelevant-cues such as social categories (in our context, for instance, the ethnic origin of a student) and a more controlled, integrating route where we review all information that might be relevant for the judgment. The automatic route is typically used for judgments related to routine tasks whereas the controlled route is typically used for non-routine activities or judgments that are considered as particularly important. In prior research it was often assumed that achievement-related judgments are, to a great extent, made via the automatized route (Glock and Krolak-Schwerdt, 2013). This is plausible because achievement-related judgments represent routine tasks in teachers' daily lives. Dealing with (interethnic) social exclusion in class, on the other hand, is something less routine and an issue that we would expect to be considered as very important given the severe consequences it can have. When integrating all relevant information in order to make a judgment using the controlled route, teachers make more accurate judgments (i.e., judgments that are less biased by social categories such as ethnicity). Therefore, this could be one reason why social exclusion is less affected by the ethnic origin of students.
Another interesting finding of this study is that group functioning aspects played only a very minor role in pre-service teachers' considerations. Previous research with children and adolescents has shown group functioning to be an important justification for interethnic exclusion (Hitti et al., 2011;Mulvey, 2016). On the other hand, in our study, only a very small number of participants (8%) mentioned aspects related to group functioning. However, teachers might not have had enough information to name group functioning as an underlying motive in the presented situations. Many of the participants stated that they would intervene in order to ask for reasons. This indicates that pre-service teachers do not want to judge the situation superficially. They want to understand the underlying motives why the student is excluded. Of course, exclusion is always harmful for the excluded person. However, some reasons might be more appropriate justifications for excluding a certain person in a certain situation than others. Smooth group functioning might be one of the more appropriate reasons to justify social exclusion. Also, prior negative behavior of a student or interpersonal struggles might be rather valid justifications for excluding someone from a certain situation, without representing a strong moral transgression in contrast to exclusion solely based on someone's origin. Many participants in our study stated explicitly that it was hard for them to evaluate the scenario and to say whether they would intervene or not, because they were provided with so little background information regarding the situation or the relationship of the protagonists. Obviously, teachers want to understand the situation more completely when making decisions about how to react appropriately. In real life, teachers typically have more knowledge about their students and know who is friend with whom, etc. This might limit the external validity of our results. However, our approach has clear benefits: The experimental approach allowed us to analyze the sole effect of the protagonists' origin, isolated from other aspects which in real life might bias teachers' evaluations or reactions. Additionally, the situations described in our scenarios are not too disconnected from everyday life. Although teachers typically know their students, they do not have knowledge about all things that happen in class and are often supposed to make instant evaluations or decisions without having additional information. However, further research should address this issue and systematically vary the background information about the situation and the excluded student.
One important restriction of our study is that our data were collected using hypothetical scenarios and self-reports which might be biased by social desirability. However, Turiel (2008) demonstrated that reasoning and evaluations by children and adolescents in hypothetical situations correspond to those in reallife situations and thus are comparable. In addition, research using a similar paradigm as the current study demonstrated that, for children, self-reports correspond with actual behavior (Mulvey et al., 2018). However, this has not been proven for adults yet. Therefore, it would be of interest for future research to connect self-report data with behavioral observations in order to determine the extent to which self-reports correspond with actual behavior. As social desirability is especially relevant regarding explicit measures of intergroup attitudes (Nesdale and Durkin, 1998;Rutland et al., 2005), it would also be interesting to see whether implicit measures of intergroup attitudes reveal different results.
Moreover, future research should also focus on in-service teachers because they already work with students and have an impact on their development and behavior. It would be interesting to see whether studies with in-service teachers replicate our findings or lead to different results. However, both research with pre-service teachers and in-service teachers can produce essential findings and have important implications for teacher training with the objective of placing well-trained staff in schools right from the start of their careers.
Further, it would be very interesting to compare different contexts of exclusion or to focus on additional ethnicities. Prior research has demonstrated that the context of exclusion (e.g., leisure time activities vs. achievement-related activities) affects judgments about peer exclusion (Horn, 2003;Tenenbaum et al., 2018;Beißert et al., 2019). But does this also hold for teachers when they evaluate social exclusion among students? And will the current findings be replicated for protagonists from other ethnicities or with a different immigration status, such as refugees? Are aspects related to group functioning possibly more relevant in some contexts or for some target persons than for others? Future studies should address these issues and systematically investigate teachers' reactions to social exclusion and include different methodical paradigms, different types of teachers, different target groups, and different contexts.
Nevertheless, in summary, our study revealed important findings regarding teachers' reactions to social exclusion in interethnic interactions. Encouragingly, pre-service teachers do not seem to underlie ingroup bias when evaluating interethnic social exclusion. In contrast they value the social norm of inclusion even more when the excluded student is an outgroup member. These findings imply that the inclusion of ethnic minorities in class be promoted by teachers. Teachers are important role models for their students, especially when it comes to ethnic topics. Thus, it is a very positive sign that they resist ingroup bias and try to establish inclusive class norms. With this, they have the potential to contribute to their students' moral development by promoting equality and social inclusion as a norm in class. Accordingly, future research should also focus on the development and thorough evaluation of prevention and intervention programs based on the current findings. Such programs should aim at raising students' and teachers' awareness of exclusion of ethnic minority students in the classroom, and teachers should be trained to successfully contribute to students' moral development and help them understand and internalize moral norms such as equality, fairness, and inclusion.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.