Promoting Prosocial Behavior Toward Refugees: Exploring the Empathy-Attitude-Action Model in Middle Childhood

Given the current global “refugee crisis,” there is a need for research investigating methods to promote positive intergroup relationships in host communities. This chapter will explore how empathy can be used to encourage host-society children to develop positive attitudes and engage in helping and prosocial behaviors toward refugees. The focus will be on middle childhood, a period when youth acquire advanced forms of perspective taking and an increased understanding of intergroup dynamics. This chapter extends prior research in adult populations by applying the empathy-attitudes-action model as an intervention for preparing children to engage positively toward incoming refugee children. This model outlines that inducing empathy toward a member of a stigmatized group may improve out-group attitudes as a whole, and in turn, increase prosocial behavior toward that collective out-group. Furthermore, the chapter recognizes the agency of children in establishing harmonious intergroup relations and in creating peaceful and inclusive societies. This chapter concludes with a call to action, advocating for the implementation of preventative programs that prepare children to engage meaningfully with their new peers.


Introduction
There are currently more than twenty-two million refugees in the world, over half of whom are school-aged children (UNHCR, 2018). Integrating these children into their new communities can be a challenging process that carries significant implications for both their personal well-being (Fazel, Reed, Panter-Brick, & Stein, 2012) and the establishment of long-term harmonious intergroup relationships (Esses, Hamilton, & Gaucher, 2017). Consequently, it is imperative that refugees enter welcoming and supportive environments. Children resettled in high-income countries tend to receive support that fulfills their basic material needs; however, there is a serious lack of provision to sustain their social-emotional well-being. One of the most important factors in promoting their social-emotional well-being is establishing supportive relationships with host-society children (Fazel, 2015). Within the broader context of peace psychology, positive intergroup relations in childhood are considered to be fundamental in the establishment of long-term peace (Christie, 2006). Despite this fact, investment in programs which prepare host-society children to engage meaningfully with their new peers is largely lacking. Although there is a widespread assumption that enhancing children's intergroup empathy could be an effective intervention strategy for promoting positive intergroup relations (Aboud & Levy, 2000;Nesdale, Griffith, Durkin, & Maass, 2005), this claim is largely unsubstantiated. This chapter aims to address this claim by outlining the empathyattitudes-action model (Batson, Chang, Orr, & Rowland, 2002) and demonstrating how it can be applied through a brief pilot study in Northern Ireland that focuses on helping recently arrived refugees in middle childhood. The chapter concludes with recommendations for future research and advocates for the implementation of interventions that prepare children to engage meaningfully with incoming refugees.

Children as Social Actors and Peacebuilders
Traditionally, little emphasis has been placed on the role that children can play in fostering positive and long-lasting intergroup relations. This is partly due to previous perceptions of children as either troublemakers, victims, or passive bystanders (Cummings, Merrilees, Taylor, & Mondi, 2017). This conceptual trap overlooks the role that children can play in building peace (for exception, see  and the reality that the relationships built during childhood serve as the building blocks for future generations (Abrams & Killen, 2014). Reflecting these truths, there has been a call to recognize children's participation as social actors or individuals who have agency and influence (IANYD, 2016;UNICEF, 1989). An increasing body of work has started to examine factors that promote young people's cooperation between groups in contexts where intergroup relations remain tense (McEvoy-Levy, 2006;Taylor et al., 2014). Encouraging children to engage in positive behaviors across group lines may prove particularly effective in reducing forms of episodic violence such as bullying or even in consolidating long-term peace.
Participation also requires that children have an effective voice in their own lives and in the community around them. There is a paucity of research that focuses on the perspective of refugee children themselves; however, existing research finds that refugee children place a high level of importance on their peer relations. Focus groups with newcomer pupils (including refugees) in Northern Ireland primary schools found that, before resettling, children's top concern was their ability to make friends in school (Kernaghan, 2015). In Britain, refugee and asylum-seeking pupils recommended schools implement preparatory programs which educate host-society children on potential refugee experiences prior to the arrival of refugee students (Fazel, 2015). As part of the move toward enhancing children's agency, it is necessary to explore the underlying processes involved in children's development of positive attitudes and prosocial behaviors toward refugees.

Refugees as an Out-Group
A refugee is "someone who has been forced to flee his or her country because of persecution, war, or violence" (UNHCR, 2018, para. 1). After displacement, many refugees flee to neighboring countries, most commonly in refugee camps or urban settings. However, others resettle in a high-income country where the ethnic, religious, and cultural composition of the majority of inhabitants is usually vastly different from their own. Within these communities, they are an ethic out-group, that is, someone belonging to a different ethnic group (Schulz & Taylor, 2018). Moreover, refugees constitute a specific subset of the immigrant population who may be subject to particular stigmatizations, negative perceptions, or mixed emotional responses. These categorizations are critical when considering the attitudes and behaviors of host-society children toward refugees.
Ethnicity is a salient characteristic which influences children's development of out-group attitudes (Nesdale et al., 2005). By the age of 6 years old, children from the ethnically dominant group can recognize their ethnic group and may hold more positive in-group attitudes compared to other groups (Aboud, 1988). Children also demonstrate intergroup biases in their empathy and prosocial behavior toward others based upon group membership (Abrams, Van de Vyver, Pelletier, & Cameron, 2015). By the age of 7, these intergroup biases may turn to explicit prejudice and negative out-group-directed behaviors (Levy & Killen, 2008). As children enter into middle childhood, their attitudes toward ethnic groups become more nuanced and are influenced by an array of factors such as empathy, perception of out-group threat, and group norms (Nesdale et al., 2005). Given the complexity of attitude formation, it is important to consider the context in which it occurs. However, the majority of research concerning children's attitude formation toward ethnic out-groups is conducted within the context of more traditional majority/minority relations (e.g., White majority children and Black minority children in the United States). As a consequence, factors that lead to the development of positive attitudes and behaviors toward more novel out-groups, and specifically those which may be stigmatized or viewed as a threat by the broader society, are not well understood.
Children's attitudes toward ethnic out-group members are substantially shaped by the cultural contexts in which they occur (Pauker, Williams, & Steele, 2016). Thus, it is important to consider how refugees constitute a specific subset of the ethnic minority population and often face negative or mixed perceptions from the wider society. Attitudes toward refugees worldwide are worsening, with polling revealing that the majority of European adults hold negative views toward immigration (Dempster & Hargrave, 2017). While adults are more accepting of refugees than those classified as immigrants, compassion for refugees is often accompanied by anxiety (Dempster & Hargrave, 2017). In European countries, the majority of adults report feeling that refugees pose a threat to national security and experience anxiety over real world concerns when accepting refugees into their country (Wike, Stokes, & Simmons, 2016). Such "threat narratives" in the wider community shape children's attitudes toward refugees. Though the existence of intergroup biases and prejudice in childhood is well established, more specific understanding about how to promote positive out-group attitudes and behaviors toward refugees is sparse.

Previous Interventions
To date, two prominent studies have evaluated interventions targeted at improving attitudes and behaviors toward refugees. Cameron, Rutland, Brown, and Douch (2006) assessed the effectiveness of extended contact interventions in improving children's attitudes toward refugees and intended behaviors toward (hypothetical) refugee children (e.g., how much they would like to play with a refugee child or invite him/her to their house). Results demonstrated that extended contact was an effective intervention in improving attitudes toward refugees, but did not have a significant impact on children's behavioral intentions toward refugees. Thus, while benefits were seen for attitudes, those changes were not coupled with meaningful changes in behavior. Nesdale et al. (2005) note that "given that practical reconciliation between ethnic groups demands changes in behavior (e.g., a preparedness to interact, cooperate and share), there is a need for research that addresses the linkage between children's ethnic attitudes and behavior, in the context of their empathy toward ethnic minority group members" (p. 635). Turner and Brown (2008) evaluated "Friendship Project," a school-based intervention in the UK designed to improve children's (9-11 years old) attitudes toward refugees by combining multicultural curricula, anti-racist interventions, and an empathy-based component. In the empathy-based component, children were prompted to imagine children (similar to themselves) in situations similar to those faced by refugees. Children were then prompted to think about how they may feel in this situation and how refugees might feel. Results found that the program was effective in improving children's short-term attitudes, but not long-term attitudes, and that it had no effect on children's empathy toward refugees. Though this study evaluated an existing program aimed at promoting children's empathy toward refugees, it is important to note that the empathy component used was impersonal, focused on perspective taking, and used hypothetical circumstances. Furthermore, this study did not access children's behavioral intentions toward refugees and thus did not address the link between attitudes and behaviors within the context of empathy.
Interventions aimed at reducing prejudice toward more traditional ethnic minorities tend to utilize either multicultural curriculums or anti-racist programs. These programs are based on the assumption that prejudice stems from ignorance or misinformation regarding an out-group and providing correct information will reduce prejudice. However, such programs are not widely effective. A large meta-analysis examining 122 intervention-comparisons of programs around the world designed to reduce prejudice or promote positive intergroup attitudes in children and adolescents found only low to moderate intervention effects (Beelmann & Heinemann, 2014). Interventions that promoted empathy through contact and social-cognitive programs, however, were the most effective in promoting positive intergroup attitudes. Moreover, recent studies have found that the positive effects of such contact interventions are mediated by intergroup empathy (Vezzali, Hewstone, Capozza, Trifiletti, & Di Bernardo, 2017), suggesting that promoting empathy may bolster the effects of such interventions. Thus, exploring and evaluating interventions specifically aimed at promoting children's empathy toward refugees is a logical next step. Given the lack of research, testing models shown to be successful in adults could offer insight into how to promote positive attitudes and prosocial behaviors among children.

Empathy-Attitudes-Action Model
The Empathy-Attitude-Action model (EAA) was developed to describe how empathy can promote prosocial behaviors across group lines (Batson et al., 2002). Empathy, defined as an "other-oriented emotional response congruent with the perceived welfare of another" (Batson et al., 2002(Batson et al., , p. 1856, is an important motivator for prosocial behaviors (Batson, 2010). The EAA model asserts that the link between empathy and prosocial behavior is not a direct link, but is instead mediated by outgroup attitudes. That is, inducing empathy toward one member of a stigmatized out-group (i.e., the empathy target) can improve attitudes toward that group to which that person belongs and, in turn, promotes prosocial behavior toward that group as a whole. A three-step process is described whereby (a) increasing empathy toward one member of an out-group may lead to an increased valuing of the individual's welfare, (b) this increased valuing is reflected through positive attitudes toward the out-group as a whole (Batson, Chang, Orr, & Rowland, 2002), and (c) these positive attitudes lead to an increase in motivation to help other members of that particular out-group, given that the need is salient to their group membership (Batson et al., 2002).
Evidence for a robust empathy-attitude-action model has been demonstrated in adult populations using a variety of stigmatized groups, such as individuals with AIDS, experiencing homelessness, or convicted of murder (Batson et al., 1997). Attitude change resulting from empathy induction appears to have relatively longlasting effects, with attitude improvements being evident at even one-to-two weeks after the empathy induction. These positive attitudes, in turn, promote a willingness to engage in prosocial behaviors toward other members of that group. For example, empathy induced by listening to the story of an individual addicted to heroin led to improved attitudes toward those struggling with substance abuse and, in turn, led to participants allocating resources to help those struggling with substance abuse (Batson et al., 2002). A simplified version of the EAA mediation model is depicted in Fig. 5.1.

Empathy
While there are many conceptualizations of empathy, Batson's definition refers to empathy as a state. That is, empathy which is experienced to varying degrees depending upon the specific situation. This is in contrast to trait empathy, which is considered a more stable personality variable in which individuals differ in their propensity to empathize with others (Davis, 1983). State-based empathetic responses are broadly thought to have two dimensions: cognitive and emotional, or affective. Affective empathy is a subjective state that is often thought to be the result of emotional cognition, or perspective taking (Batson et al., 1997). For example, listening to the story of an out-group member first prompts cognitive empathy by allowing the listener to put him/herself in that person's position (i.e., take the perspective of that person). This cognitive process then creates an affective empathetic response in the listener. This affective empathetic response then leads the listener to become more concerned with the welfare of the out-group member.
In this assertion, cognitive awareness of another's emotional state is necessary before other-oriented and parallel emotional responses can occur. This cognitive skill emerges around the age of five and becomes more advanced as children develop. By middle childhood, occurring between the ages of seven to eleven, children display more sophisticated forms of empathy, which impact their interpersonal prosocial behaviors (Abrams et al., 2015;Eisenberg, Eggum, & Di Giunta, 2010). By focusing on middle childhood, a period when children display advanced forms of empathy coupled with more nuanced understandings of group dynamics, we now explore each pathway of the empathy-attitude-action model among children.

Empathy and Prosocial Behaviors
Prosocial behavior can be defined as a "voluntary behavior intended to benefit another" (Eisenberg et al., 2010, p.146) and is a superordinate category which includes behaviors such as sharing of resources, helping, and comforting (Schroeder & Graziana, 2015). Decades of research has identified empathy as a primary driving force for prosocial behavior, and increasing or inducing empathy has been linked to increased prosocial behavior across the lifespan (Eisenberg et al., 2010). For example, inducing empathy in children through the use of videos that portray others in need has been shown to increase children's willingness to share resources (Williams,  ). Yet, this research has traditionally been conducted in the interpersonal domain, which does not consider the role of group boundaries. Given that group membership may shape empathetic responses, there is a need to evaluate these processes within the intergroup domain (Sierksma & Thijs, 2017). Previous research by Abrams et al. (2015) found that when 5-to 10-year-old children were placed into minimal groups, those with higher levels of trait empathy were more willing to engage in out-group-directed prosocial behaviors. It follows that if trait empathy is predictive of out-group prosociality, increasing intergroup empathy might also promote prosocial behavior across group lines. In support of this assumption, Sierksma, Thijs, and Verkuyten (2015) found that in-group biases in children's helping intentions could be overcome by inducing empathetic understanding. In an experimental vignette, 8-to 13-year-old children were read a story about a child who was either part of their imagined friend group (an in-group member) or not (an out-group member). In the empathy induction condition, children where prompted to take the perspective of the child in need. Children in the noempathy condition listened to the same story, but were not prompted to take the perspective of the child in need. Children in the no-empathy condition gave more resources to the in-group member and fewer resources to the out-group member. However, those in the empathy induction manipulation gave equally to in-group and out-group members. This suggests that empathy induction has specific effects on children's behavior toward out-group members. These studies provide mounting evidence that inducing empathy toward an out-group member can promote prosocial behavior toward that group. Batson et al. (2002) demonstrated that, in adults, empathy induced for an out-group member in need may increase readiness to help that individual. Furthermore, this readiness to help may generalize to the out-group as a whole if the individual's need is related to their group membership. Developmentally, there is evidence that by middle childhood, children's growing sense of group dynamics and moral understanding could enable them to generalize the need of one person to the group as a whole. For instance, children have the capacity to understand the concept of groupbased need and inequality (Elenbaas & Killen, 2016). This understanding has been linked to prosocial behavior, as children may rectify injustices through their resource allocations (Elenbaas & Killen, 2016;Rutland & Killen, 2017). For example, 8-year-old children have been shown to give more resources to groups that have been historically disadvantaged within their society (Olson, Dweck, Spelke, & Banaji, 2011). Given that children understand and respond to group-based need, it is possible that inducing empathy for an out-group member could generalize into a willingness to help that group as a whole.

Empathy and Attitudes
The empathy-attitudes-action model proposes that attitudes mediate the relationship between empathy and prosocial behavior (Batson et al., 2002). The majority of research today has focused on trait empathy, rather than state-based empathetic responses which are tested in the pilot study below. Still, correlational studies between children's trait empathy and out-group attitudes suggest that empathy may play a vital role in shaping out-group attitudes. Nesdale et al. (2005) investigated the relationship between empathy in 5-to 12-year-old White Anglo-Australian children and their attitudes toward out-group members who were either of the same ethnicity (other Anglo-Australian children) or of a different ethnicity (Pacific Islander). Children were told they were taking part in a team drawing competition against one of these groups. In this study, children's trait empathy and attitudes toward the opposing team were then assessed. The results showed that empathy predicted liking for differently ethnic out-group members. However, liking for the same ethnicity out-group was not correlated with empathy. That is, empathetic children tended to express more positive attitudes toward out-group members, but not in-group members. This differentiation may have been attributed to children's understanding of social inequities and compassion for those that have less favorable circumstances. Thus, empathetic children may be more sensitive to intergroup inequalities and may hold more positive attitudes toward members of disadvantaged groups.
Building upon these results, Vezzali et al. (2017) conducted a correlational study among Italian immigrant and nonimmigrant primary school students (ages 8-11 years old), which investigated the links between trait empathy, attitudes, and contact. In keeping with the results of previous research (Nesdale et al., 2005), the study found that intergroup empathy mediated the relationship between contact and positive out-group attitudes. As such, children with higher levels of empathy had more positive out-group attitudes. These studies hold significant implications for supporting an empathy-attitude link in children, specifically toward minority outgroups. If higher levels of trait empathy are linked to more positive out-group attitudes, enhancing intergroup empathy should promote positive attitudes toward refugees. However, changing attitudes alone is not wholly beneficial, as they may not translate into meaningful action.

Attitudes to Action
The final component of the empathy-attitudes-action model is the link between out-group attitudes and an increased willingness to help out-group members. A considerable body of literature has explored this attitude-behavior relationship in adults and has found that attitudes generally serve as an antecedent for subsequent behavior (for a review see Ajzen & Fishbein, 2005). However, the nature of the behavior and the contexts in which it occurs are imperative to determining attitude-behavior consistency (Ajzen & Fishbein, 1980). In addition, cost-benefit considerations play a role when deciding whether or not to engage in a specific behavior. A person may hold a favorable view of another, but chose not to offer assistance if this comes at a cost to themselves or to others (Batson, 2011). Furthermore, the group membership of the recipient is also known to be a factor in the decision to engage in helping behaviors for both adults and children (Levy & Killen, 2008;Taylor & Hanna, 2018;Taylor et al., 2014). Studies aimed at improving children's attitudes toward out-group members have often found discrepancies between improvements in attitude and increases in prosocial behavior (Aboud et al., 2012).
As with the majority of research within the intergroup domain, the attitudebehavior relationship in this context has focused almost exclusively on the link between negative attitudes and behaviors. For instance, there is a wealth of research demonstrating that negative out-group attitudes in adults and children, such as prejudice, are associated with negative intergroup behaviors including aggression or violence (Genthner, Shuntich, & Bunting, 1975;Van Zomeren, Fischer, & Spears, 2007). A new line of thinking has proposed that the decision to engage in negative or positive behaviors may develop through different processes (Aboud, 2003). Thus, a shift in the research paradigm to focus specifically on the relationship between positive out-group attitudes and prosocial behaviors is necessary. However, much of the current literature addressing this call has been conducted using cross-sectional field studies and therefore has a limited capacity to explain and unravel causal patterns. For example, following an earthquake in Italy, the majority of children with more positive attitudes toward immigrants showed a greater willingness to meet and help immigrant victims (Vezzali, Cadamuro, Versari, Giovannini, & Trifiletti, 2015). By employing a longitudinal design, Taylor et al. (2014) found that positive outgroup attitudes were associated with later patterns of out-group prosocial behavior in young people living in a post-accord society. Taken together, there is converging evidence that promoting positive out-group attitudes could promote meaningful prosocial action.

Inducing Intergroup Empathy
Testing the empathy-attitudes-action model requires the successful induction of intergroup empathy. Narratives, presented through the use of books and storytelling, have previously been successful in inducing empathy (e.g., Sierksma et al., 2015). These approaches can be utilized in both experimental and applied contexts and are thought to induce empathy by allowing the individual to see a situation from another person's perspective.

Pilot Study
To test the empathy-attitude-action model in middle childhood in response to the refugee crisis in Europe, a pilot study was developed which induced empathy toward Syrian refugee children (Glen, 2017). Syrian refugees were selected as the target group given the United Kingdom's announcement to expand the Vulnerable Persons Relocation scheme to resettle at least 2,000 Syrian refugees in Northern Ireland (NI) and because they would represent an ethnic out-group in the NI host society.
Ninety-two children between 8 and 11 years old participated in an intervention using a realistic refugee scenario in schools across NI and were randomly assigned to an information-only (control) or a story (intervention) condition. Across both conditions, participants were first given a child-friendly definition of a refugee. Next, children were introduced to gender-matched (hypothetical) child-Mohammad or Ayesha (M/A)-and told that he/she was a refugee from Syria who was moving to NI and would soon become a pupil at their school. In the informationonly condition, participants moved on to the dependent variable questions in a oneon-one interaction with a trained researcher, while in the story condition, children heard an adapted version of a published children's story book based on a true account of two Syrian children's journey to resettle in Europe. The story was "read" to children using a pre-recorded audio to ensure that each participant heard the story in an identical manner. Children then completed the dependent variable tasks, which included an empathy manipulation check, out-group attitudes toward refugees, a resource allocation task, and realistic helping opportunity directed toward M/A. Empathy was assessed through self-report using a one-item question derived from Williams et al. (2014). Children were asked either "How did you feel while listening to the story" (story condition) or "How did you feel when learning about refugees" (information-only condition) and responses were recorded using a four-point Likerttype facial affective scale (FAC), with responses ranging from awful to brilliant. Outgroup attitudes were assessed using a three-question scale created by Nesdale et al. (2010). Children were asked (1) "How much do you like Syrian refugee children," (2) "How much would you trust Syrian refugee children," and (3) "How much would you like to play with Syrian refugee children." Responses were recorded using a four-point Likert-type scale and summed to form a composite score. Consistent with Batson's model, prosocial behavior toward refugee children as a generalized out-group was assessed through a resource allocation task. Children allocated seven one-pound coins between three groups, helping either Syrian refugee children, animals, or nature.
Extending Batson's model, realistic helping intentions toward the empathy target were assessed using a two-question item from Vezzali et al. (2015). Children were asked "How much time do you want to help M/A during lunch" and "How much time do you want to help M/A after school." Responses were recorded using a fivepoint Likert-type scale of progressively shaded in clocks ranging from none of the time to all of the time and were aggregated to form a composite score. This extension allowed us to differentiate the role of empathy and out-group attitudes in both prosocial behavior toward the empathy target and prosocial behavior toward the empathy target's group.
The pilot study found that empathy was successfully induced in a school setting through a brief narrative intervention. That is, children in the story condition reported significantly more empathy than those in the information-only condition, echoing Batson's finding that exposure to a realistic narrative can be an effective method for inducing empathy toward a member of an out-group. It should be noted, however, that those in the information-only condition also reported relatively high levels of empathy; that is, children empathized readily with refugees, even when provided only with brief, factual information. This finding is encouraging as intergroup biases in empathy are well documented and reveal that, typically, adults and children do not readily empathize with those they perceive to be dissimilar to themselves (Chiao & Mathur, 2010;Vanman, 2016). Furthermore, other studies using more hypothetical and impersonal means of promoting empathy toward refugees failed to produce similar results (Turner & Brown, 2008). Taken together, this finding provides evidence for future studies to utilize realistic narratives as a means of empathy induction toward refugees specifically.
Furthermore, although children in both conditions were willing to provide direct assistance to M/A, children in the story condition reported greater realistic helping intentions toward the target. Regarding the other two dependent variables, there were no significant differences in children's out-group attitudes or the number of resources allocated to refugees. For each of these outcomes, children in both conditions expressed positive attitudes toward refugees and consistently allocated resources in favor of refugees in general. According to Batson, choosing to allocate resources in favor of the empathy target's group suggests a higher valuing of the out-group's welfare over that of competing groups. This is reflective of developmental research which finds that by the age of 8 years old, children's helping behaviors are significantly influenced by the perceived need and merit of the recipient (Moore, 2009;Schmidt, Svetlova, Johe, & Tomasello, 2016). Correlations for the dependent variables can be found in Table 5.1.
Next, two mediation analyses were run to test the EAA model. There was a significant link from attitudes to realistic helping intentions; however, empathy was not directly related to out-group attitudes nor to either type of prosocial behavior. Thus, the pilot study did not find conclusive evidence for the EAA model. However, within the context of empathy induction, children's out-group attitudes were highly predictive of their willingness to engage in meaningful, realistic helping behaviors toward the incoming refugee pupil. This corroborates previous research which found that young people with more positive out-group attitudes engage in more prosocial behavior toward out-group members (Taylor et al., 2014;Vezzali et al., 2015). Given the focus on specific, realistic helping intentions, this finding may provide a fairly reliable representation of actual future helping behavior (Ajzen & Fishbein, 2005). Furthermore, the intentions assessed were costly and required sacrifices in the form of both time and effort. As cost-benefit considerations are important in the decision to engage in helping behaviors (Batson, 2011), this is a particularly encouraging finding. As the first study of EAA in middle childhood, the successful induction of empathy and identification of the attitude-behavior link are promising directions as we begin to unravel the relations within this model. However, several limitations of the pilot study should be acknowledged and addressed in future research.

Limitations and Future Research
First, the empathy measure used in the pilot study may not have adequately captured the level or type of empathy experienced by participants. While there are many forms of empathy, the pilot study only assessed affective empathy as this reflected Batson's original study in adults. However, Batson's study used multidimensional assessment of affective empathy that required an advanced vocabulary and ability to express complex emotions and, thus, was not appropriate for the current population. It is possible that the child-friendly, single-item assessment used in the current study did not adequately capture the complexity of affective empathy.
Future research, therefore, should use more comprehensive measures which account for multiple forms of empathy. For instance, measures which differentiate between affective and cognitive state empathy, such as the personal reactivity and empathetic concern subscales of the Interpersonal Reactivity Index (Davis, 1983), could be modified to assess out-group-directed empathy. Furthermore, future research may try to assess the degree to which trait empathy is a precursor for responding to message-induced state empathy in middle childhood. Capturing the complexity and multidimensional experience of empathy may provide a clearer picture of its role in attitude formation and out-group-directed prosocial behavior in the future.
Second, the resource allocation task may not have provoked adequate competition between groups. Factors such as fairness, recipient's need, and group loyalty play a role in children's resource allocation (Rutland & Killen, 2017), and future research should continue to explore these factors. For example, allocation benefiting refugees could be placed in direct competition with other children, such as those from the in-group, other ethnic groups, or with a serious illness.
Third, this study was limited in our ability to assess the directionality of effects. For example, future studies should assess out-group attitudes at multiple time points, such as before and after empathy induction. Moreover, while there is mounting evidence that trait and state empathy are precursors for positive out-group attitudes (Batson et al., 1997;Vezzali et al., 2017), positive out-group attitudes may be a precursor for out-group empathy. More complex designs may allow for future researchers to unravel the dynamic and multidimensional relation between empathy and attitudes. Fourth, although random assignment was used with the empathy induction, the pilot study did not control for children's previous quantity or quality of contact with refugees or other ethnic groups. This is an important limitation as Northern Ireland is historically a homogenous society in which children may have had little experience with members of other ethnic groups. Contact is known to play a role in the formation of out-group attitudes (Aboud, Mendelson, & Purdy, 2003) and, thus, future research should assess both the quantity and quality of children's previous contact with diverse out-groups. At the same time, contact interventions should consider incorporating empathy-inducing techniques, such as narratives, to enhance the impact on positive intergroup relations (Vezzali et al., 2017).
Finally, research should strive to incorporate the views of refugee children themselves and investigate how intergroup processes may differ between majority and minority group children. Intergroup relations are by no means one-way, and research would benefit from adopting a more dynamic approach.

Call to Action
It is imperative that research not only advance theoretical understanding but also inform policy and program implementation. As highlighted in this chapter, children may actively contribute to improving intergroup relations. Although the empowerment and facilitation of children as social actors is gaining traction in research, the rhetoric of participation far outweighs its implementation.
Interventions to promote episodic peacebuilding and consolidate long-term peace often focused on fostering positive intergroup encounters (Christie, 2006). However, such efforts are often reactionary rather than preventative, intervening only after negative group dynamics have taken root. At this point, hostile group dynamics are hard to change. The relatively recent influx of refugees into highincome countries offers an opportunity to not only prevent the development of negative intergroup dynamics, but also to prepare children to build long-lasting, meaningful relationships with their new peers.
New, empirically driven models on promoting cross-group friendships place a focus on making children 'contact ready' and preparing them to engage in meaningful, sustained relationships with out-group members (Turner & Cameron, 2016). Such models advocate for socio-cognitive capacity building interventions which enhance children's empathy. On a broader scale, programs which aim to promote trait empathy and helping behaviors from a young age should be more widely implemented. For example, Roots of Empathy (Gordon, 2005) has shown some promise in cultivating children's empathy. Future research should extend these types of interventions to focus specifically on empathy in intergroup contexts.
Interventions aimed specially on preparing children to welcome refugees should consider using a socio-cognitive approach to enhance empathy toward refugees.
As demonstrated by our pilot study, exposing children to realistic narratives of refugees' experiences can be a useful form of empathy induction. Such empathyinducing narratives could be combined with extended or imagined contact (Cameron et al., 2006;Turner & Cameron, 2016). For example, narratives could be extended to incorporate stories of friendship between refugee and host-society children. Such interventions would be relatively low-cost, logistically easier to organize than interventions involving face-to-face contact, and administered by para-professionals. Schools may be key sites to facilitate such preparatory interventions as they are often the first point of contact between host-society and incoming refugee children. Although schools are often overloaded with new agendas and policies, the interventions outlined above could be incorporated into existing curriculum and administered by teachers in the classroom.
By investing in interventions that prepare host-society children to welcome their new peers in meaningful ways, we are responding to the voices of refugee children. Preventative programs are an important long-term investment that can aid in addressing the multifaceted nature of integrating refugees into new communities (Fazel & Betancourt, 2017). Although integrating refugees into a new society is challenging, host-society children can play an integral role in establishing an environment that supports the well-being of refugees and fosters peace for generations to come.
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.