WARNING CONCERNING COPYRIGHT RESTRICTIONS The copyright law of the United States (Title 17, United States Code) governs the making of photocopies or other reproduction of copyrighted materials. Under certain conditions specified in the law, libraries and archives are allowed to furnish a single photocopy or other reproduction for the sole purpose of private study, scholarship, or research. IT IS A VIOLATION OF COPYRIGHT LAW TO DISTRIBUTE OR REPRODUCE THIS MATERIAL ELECTRONICALLY OR IN PAPER FORMAT BEYOND A SINGLE COPY FOR YOUR OWN PERSONAL USE. JOURNAL OF RESEARCH ON ADOLESCENCE, ***(*), 1–16 Adolescents’ Retributive and Restorative Orientations in Response to Intergroup Harms in Schools Gabriel Velez Laura Pareja Conto , Angelica Restrepo, and Holly Recchia Marquette University Concordia University, Cecilia Wainryb University of Utah This mixed-methods study examined how adolescents understand and evaluate different ways to address intergroup harms in schools. In individual interviews, 77 adolescents (M age = 16.49 years; 39 girls, 38 boys) in Bogotá, Colombia, responded to hypothetical vignettes wherein a rival group at school engaged in a transgression against their group. Adolescents reported that students who were harmed should and would talk to school authorities, but also noted they would likely retaliate. In terms of teacher-sanctioned responses to harm, youth endorsed compensation most strongly, followed by apologies, and rated suspension least positively. Youths’ explanations for their endorsement of different disciplinary practices reflected varied concerns, including their perceptions of how justice is best achieved and how restoration could be attained. Key words: restorative justice – retributive justice – peer conflict – teacher–student relationships – intergroup harm Traditionally, schools in the Western world have responded to serious student misconduct with approaches that emphasize placing blame, punishing offenders, and reaffirming school authority (Okimoto et al., 2012; Zehr, 2002). Despite the pervasiveness of punitive discipline in schools, there is limited research supporting its effectiveness and growing scholarship underscoring the deleterious consequences of these practices (González, 2012). In particular, these approaches can negatively impact the school environment, disrupt children’s academic trajectories, and contribute to systemic injustices against historically marginalized students (Hinze-Pifer & Sartain, 2018). Recently, support has been growing for educational environments modeled after the principles of restorative justice. Broadly, restorative justice is based on the notion that everyone in the community is interconnected, and calls for responses to harm that involve dialogue between victims, perpetrators, and community members so as to facilitate reparation and prevent similar conflicts (Zehr, 2002). Although restorative justice models are youthcentered in theory, researchers have privileged the perspectives of teachers and school administrators, Requests for reprints should be sent to Holly Recchia, Department of Education, Concordia University, 1455 de Maisonneuve West, Montréal, QC H3G1M8, Canada. E-mail: holly.recchia@concordia.ca whereas less is known about how students understand and make sense of restorative approaches (Velez et al., 2020). To supplement traditional topdown approaches that prioritize adults’ views, it is necessary to document students’ own experiences, understandings, and evaluations of different approaches to addressing harm in their schools. This study thus sought to inform the implementation of restorative justice models in schools by adding to extant research in three ways: (1) by complementing the research on teachers’ perspectives with a careful examination of youths’ perspectives, (2) by drawing on a sample of nonEuropean or North American youth, and (3) by considering intergroup harms rather than dyadic harms. We were particularly interested in examining adolescents’ perspectives given their capacity to critically reflect on educational practices in their schools (e.g., Bell, 2020) and to draw meaningful insights as they reason about socio-moral transgressions (e.g., Recchia et al., 2020). Individuals’ perspectives on harm are grounded in the complex moral, pragmatic, and psychological understandings that emerge from their lived experiences (Smetana et al., 2014). That is, students’ prescriptive judgments of how conflicts with peers should be resolved may both differ from and be informed Ó 2022 Society for Research on Adolescence. DOI: 10.1111/jora.12785 2 PAREJA CONTO, RESTREPO, RECCHIA, VELEZ, AND WAINRYB by their descriptive expectations of what they believe will actually happen in their schools, and may reflect both retributive and restorative concerns. With these issues in mind, this mixed-methods study focused on adolescents’ reasoning about different approaches to address intergroup harm between rival peers at school. Given that many harms in schools involve transgressions between groups of students (Rutland & Killen, 2015), and students’ judgments about harm may vary depending on the group status of the perpetrator (Mulvey, 2016), we sought to illuminate youths’ perspectives on different approaches to addressing intergroup harms, including (1) their expectations and evaluations of the victimized group’s strategies in response to the harms, (2) their views regarding the role of teachers in addressing harms, and (3) their expectations and evaluations of different teacher-sanctioned responses. We also considered the different restorative and retributive concerns that are brought to bear when youth evaluate different approaches to addressing harm. Finally, after adolescents considered different ways to address the harms, we provided the space for them to describe their preferred approaches, which could include or not the strategies we had proposed. To these ends, we studied a community sample of adolescents attending two low-SES urban schools in Bogotá, Colombia. A focus on this population contributes to the extant literature by documenting experiences of harm in school among youth facing issues with poverty outside of a North American or European context. Colombian adolescents grow up amid social inequalities, corruption, and violence (Van Holstein, 2018). In a country-wide study of Colombian schools, the incidence of school aggression was particularly high in urban communities with higher levels of violence (Chaux et al., 2009). Similarly, while 22.1% of Colombian students report being a victim of bullying at least a few times per month, the incidence of bullying is even higher in schools serving students from disadvantaged backgrounds (OECD, 2017). This concentration of disadvantage can be further heightened in economically segregated cities such as Bogotá (Thibert & Osorio, 2014). Thus, we expected Colombian adolescents’ retributive and restorative concerns to be informed by these features of their neighborhood contexts and by how harms are addressed in their schools. Particularities of Intergroup Harm Intergroup harm refers to group-based transgressions by individuals from one group to individuals from another group (Goode & Smith, 2016). In this study, we developed hypothetical scenarios in which a rival peer group within the school had harmed the participants’ group. From an early age, children display psychological biases, such as ingroup favoritism, in relation to both naturally occurring groups (e.g., gender) and experimentally manipulated groups based on minimal differences (e.g., t-shirt color; Dunham et al., 2011). Categorizing individuals in groups can promote intergroup prejudice by accentuating between-group differences and minimizing within-group differences (Dovidio, 2013; Rutland & Killen, 2015). For instance, in a study on intergroup victimization, preadolescents reported liking their own group more when they were victimized by an outgroup, but they liked both groups equally when their ingroup was the victimizer (Gini, 2006). Similarly, youth attributed more blame to outgroup victimizers than ingroup victimizers. Thus, youth may judge harms committed by outgroup members more harshly. Importantly, however, most of the research on intergroup harm in schools is based within North American and European educational contexts, whereas these processes have been less often examined in other sociocultural milieus. Two Orientations to Justice As adolescents grapple with experiences of harm in their interpersonal relationships, their responses may be guided by both restorative and retributive concerns. Whereas retributive justice involves responding to harm in kind, and thus focuses on the proportionality of the punishment for a perpetrator in light of the initial offense, restorative justice is instead focused on restoring victims and building and repairing relationships (Okimoto et al., 2012; Zehr, 2002). Research with adults indicates that, in the context of severe harms, individuals more often prioritize retributive concerns motivated by desires to respond in kind (Darley et al., 2000; Gromet & Darley, 2009). However, individuals will endorse a combination of both retributive and restorative goals in situations where this option is provided (Gromet & Darley, 2006). Further, regardless of offense severity, individuals report greater concerns with restoring victims and promoting YOUTHS’ RETRIBUTIVE AND RESTORATIVE ORIENTATIONS communal values when victims’ or the community’s needs are made salient (Gromet & Darley, 2009). For instance, in addition to punishing perpetrators, individuals support both psychological (e.g., apologies) and material restoration (e.g., compensation) for victims. Notably, however, when harms are committed by outgroup members, preferred responses may be less often guided by a restorative orientation (Wenzel et al., 2010), as people may be less forgiving (and therefore less restoring) of outgroup members’ transgressions (Dovidio, 2013). Mirroring the adult literature, youths’ responses to peer conflict may reflect both retributive and restorative goals (McDonald & Asher, 2018; Wainryb et al., 2020). Youth (similar to adults) experience retributive desires in response to being hurt deeply by a peer, although they are also capable of containing or redirecting these desires in favor of more restorative goals (Recchia et al., 2019). Yet middle-class North American adolescents have been found to endorse low levels of retaliatory desires, even in the face of unambiguous harms (McDonald & Asher, 2018); youth predominantly report desires to maintain their relationships, regulate their emotions, and avoid getting hurt more. Similarly, youth who endorse relationship-oriented goals, aligned with a restorative orientation, support more prosocial and passive responses to peer conflict (Chung & Asher, 1996). When youth do endorse retaliation, they report greater feelings of anger, more internal attributions about the causes of harm, robust beliefs in responding to harm in kind, and desires to teach a lesson to the perpetrator (Ardila-Rey et al., 2009; McDonald & Asher, 2018; Recchia et al., 2019). The tendency to evaluate retribution favorably may be more common among youth exposed to violence and injustice (Guerra et al., 2003). In these environments, youths’ moral principles may be challenged when they are at odds with their experiences. For example, a study with Colombian adolescents displaced by armed conflict revealed a disconnection between youths’ prescriptive evaluations and their descriptive expectations (Posada & Wainryb, 2008); although all adolescents judged stealing and harming others as wrong, many still expected others to steal and harm. Further, displaced Colombian children endorse stealing from or harming others particularly in the context of revenge (Ardila-Rey et al., 2009; Posada & Wainryb, 2008), and Colombian youth with greater exposure to violence more often describe carrying out desires for revenge (Recchia et al., 2020). Yet, 3 despite their desires for revenge, Colombian children exposed to violence still report that it is possible to reconcile with transgressors (Ardila-Rey et al., 2009). In sum, youths’ restorative and retributive orientations are informed by their goals and interpretations of harm, and social-contextual factors such as exposure to violence and group membership. An important caveat of these findings is that they are based on adolescents’ reasoning about dyadic harms, while reasoning about intergroup harms may show some unique features, such as heightened desires for retribution (Gini, 2006). Adolescents’ Perceptions of Teachers’ Role in Addressing Peer Conflict While traditional punitive approaches emphasize the role of school authorities in responding to harm, restorative justice models challenge these hierarchical systems in favor of promoting more inclusive dialogues and equitable relationships (Llewellyn & Llewellyn, 2015; Zehr, 2002). Yet educators are nevertheless in a unique position to support restorative approaches and to help youth navigate their experiences of harm in schools. Consequently, in this study, we asked youth to evaluate different authority-mandated approaches to addressing harms (i.e., compensation, apology, and suspension). We selected compensation and apologies given the potential of these strategies to restore victims’ material and psychological loss, respectively, which may be in line with a restorative orientation to harm. As a counterpoint, we also explored youths’ perspectives on traditional school punishment (i.e., suspension). Overall, past research indicates that adolescents support teachers’ involvement in addressing harm in schools by, for instance, supporting victims, engaging with the implicated students, and involving parents (e.g., Frisén & Holmqvist, 2010). Yet less is known about the moral and pragmatic concerns that guide youths’ evaluations of school authorities’ different approaches. In one recent study, adolescents reported that power assertion (i.e., imposing punishments) was more effective than inductive discipline (i.e., encouraging empathy and perspective-taking) in addressing peer harms (Rote et al., 2020). Conversely, in another study, youth reported that teachers were more effective in addressing bullying when they responded with supportive-cooperative strategies (e.g., whole-school approaches, coordinating with parents), in comparison to supportive-individual 4 PAREJA CONTO, RESTREPO, RECCHIA, VELEZ, AND WAINRYB (e.g., talking to students, providing emotional support) and authoritarian-punitive strategies (e.g., punishing, threatening; Wachs et al., 2019). In this latter study, students also evaluated supportive strategies as more effective in the long term than authoritarian-punitive strategies. It is possible, however, that youth may advocate for more punitive responses from adults when victimizers are outgroup members (Gini, 2006). Although it is important to consider youths’ perceptions of the most effective teacher-sanctioned strategies for addressing peer harms, these descriptive judgments may differ from adolescents’ prescriptive evaluations about the fairness of these strategies. For instance, the same study by Rote et al. (2020) indicated that youth perceived inductive discipline as fairer than power assertion. Similarly, Black students who had been previously suspended in US schools reported that the disciplinary processes were unfair because punishments were excessive, and punitive practices did not allow for students to voice their perspectives (Bell, 2020). Overall, more research is needed to illuminate the variety of moral and practical concerns that guide youths’ evaluations of different authority-sanctioned responses to addressing harm in schools. The Current Study The main objective of this mixed-methods study was to examine Colombian adolescents’ reasoning about responses to intergroup harms in schools, as informed by retributive and restorative concerns. We conducted a series of quantitative and qualitative analyses of individual interviews with a sample of low-SES urban students. First, we examined youths’ evaluations of how they thought their group should respond, and their expectations regarding how their group would respond to hypothetical intergroup transgressions from rival peers at school. We expected youth to most often report that their group would retaliate in response to intergroup harm (Posada & Wainryb, 2008), but also to express that conflict should be solved by talking to authorities rather than by retaliating (Frisén & Holmqvist, 2010). Second, we asked adolescents to rate three teacher-sanctioned approaches to address the harms to examine their prescriptive evaluations and descriptive expectations of the likelihood of each approach. We expected that youth would evaluate compensation and apologies more positively than suspension (Wachs et al., 2019). To gain a richer understanding of adolescents’ reasoning about different teacher-sanctioned responses, we also investigated their justifications for prescriptive ratings in an exploratory way. This analysis focused on how their reasoning reflected varied retributive and restorative concerns. Finally, we explored adolescents’ perceptions of optimal responses to intergroup harms in their schools. These responses were analyzed qualitatively to further explain and contextualize findings from the quantitative analyses. METHOD Participants A total of 77 Colombian adolescents (39 girls, 38 boys) ranging from ages 14 to 19 years (M = 16.49, SD = .95) were recruited from grades 10 and 11 in two urban high schools in Bogotá, Colombia. This study was part of a larger investigation of youths’ moral development. For the specific questions forming the focus of the current paper, post hoc sensitivity analyses conducted in G*Power indicated that this sample size would enable us to detect medium to large effects (η2p > .06 to .08 with power of 80% at p < .05, depending on the degree of correction for nonsphericity); the magnitude of within-person differences reported in similar previous research is typically considerably larger than the effect sizes we were able to detect (e.g., Posada & Wainryb, 2008). Individual information regarding youths’ race/ethnicity was not collected; however, the sample was recruited in fairly homogenous neighborhoods in Bogotá (i.e., a largely White or Mestizo population; see Secretarı́a Distrital de Planeación, 2014). The sampling was guided by the country’s six-level socioeconomic stratification system in which neighborhoods are rated on a scale from 1 (low) to 6 (high) based on infrastructure and housing characteristics (Thibert & Osorio, 2014). We selected public schools serving communities in strata 2 and 3; most participants self-reported residing in these two strata (26% and 66%, respectively). The participating schools were located in neighborhoods facing increasing rates of crime and violence. Regarding family demographics, 65% of students reported living in single-parent households. Of those participants who knew their mother’s level of education (67%), youth reported that their mothers had attended elementary school (17%), high school (25%), or postsecondary school (25%). Most participants (89%) identified as Catholic. This study was approved by the Research Ethics Committee at Concordia University and YOUTHS’ RETRIBUTIVE AND RESTORATIVE ORIENTATIONS local school administrations. Parents provided written informed consent, and youth provided written assent to participate. In appreciation for their participation, students received a cafeteria voucher. Procedure Individual semi-structured interviews were conducted in Spanish by well-trained graduate students in a private location at the schools. Each interview was conducted in one session of approximately 60 minutes; only components of the interview protocol relevant to the current study are described. Interviews were audiotaped and transcribed verbatim for analysis by native Spanish speakers. The vignettes and examples of coded responses provided below were translated for illustrative purposes and were verified by a second bilingual speaker. Hypothetical Vignettes Participants were presented with hypothetical vignettes describing two scenarios depicting intergroup harm involving rival peer groups, in a counterbalanced order. Pilot testing was conducted to develop vignettes that were ecologically valid for Colombian teenagers. In each scenario, participants were told that someone had seen a rival group in their school commit a transgression against their own group (see Table 1). Participants were prompted to imagine that this situation had actually happened. In previous research, youths’ responses to hypothetical vignettes are related to other informants’ evaluations of their behavior and to their responses to actual comparable events (see Rose & Asher, 1999; Turiel, 2008). After reading each vignette, the interviewer asked participants how the victims should and 5 would respond to the harm. Specifically, youth were asked “After the harm occurs, what do you think that [your team] should do, if anything?” then “Do you think that [your team] would actually do that? If not, what would [your team] do?” If participants did not mention involving school authorities, the interviewer prompted them to reflect on whether authorities should get involved: “Do you think that [your team] should reach out to the adults at your school?” Next, the interviewer presented participants with three teacher-sanctioned approaches to address the harms (in counterbalanced order): compensation, apology, and punishment. In responses depicting compensation, perpetrators were asked to repair the harm by returning the belongings and paying to replace broken items. In instances of apology, perpetrators were asked to apologize for the harm they caused and to express how sorry they felt. Lastly, in responses depicting punishment, perpetrators were suspended for three days from school. The presentation of each response was followed by a question assessing participants’ prescriptive evaluations (i.e., “Do you think that this is a good or not such a good way to handle the problem?”), and one assessing descriptive expectations —specifically, whether participants believed the approach was likely to occur (i.e., “How likely is ___ to actually happen?”). Youth rated each approach on six-point scales from not good at all (1) to really good (6), for prescriptive ratings, and very unlikely (1) to very likely (6), for descriptive ratings. Youth were also asked to explain their ratings for each question; the current analysis focused only on justifications for prescriptive ratings. Finally, after participants evaluated the three teacher-sanctioned responses we proposed, we asked them to describe the best approach to addressing each harm in a more open-ended way. TABLE 1 Intergroup Harm Vignettes Soccer championship Schoolyard Imagine that your school has two soccer teams. This year, your team is chosen to participate in an end-of-year tournament with other schools. You and your teammates are very excited because you think that you have a shot at the trophy. Two hours before the tournament begins, your equipment gets trashed and cannot be replaced on time. Witnesses say that the other team from your school is responsible for the vandalism. You and your teammates are very upset because your team has lost its chance to participate in this year’s tournament. Imagine that there has always been a rivalry with the other grade level at your school. One day, you and your friends are hanging out in the schoolyard. You leave your backpacks aside to go get some snacks. When you come back, you notice that your backpacks have been turned inside out. Your belongings are scattered across the schoolyard. The aggressors took money, notebooks, books, and other personal belongings, while other backpacks were ruined. Witnesses tell you and your friends that they saw a group from the other class doing the “empanada” [flipping the backpacks inside out] to your backpacks. 6 PAREJA CONTO, RESTREPO, RECCHIA, VELEZ, AND WAINRYB We asked this question at the end so that participants would have had the opportunity to reflect on different kinds of approaches that might be possible. Specifically, we asked youth: “If you were to come up with the best way to handle the situation, what do you think should happen? It can be a combination of things or something that I have not mentioned yet.” Coding and Reliability for Quantitative Analyses Participants’ responses were coded by two Spanish speakers (the first author and a second coder who was unaware of the hypotheses). As needed, a third collaborator was also consulted during the coding process. Coders first discussed the categories and their definitions, and then trained by jointly coding a subset of 10% of the narratives; interrater reliability was then established on an additional 21% of the narratives. Disagreements were resolved via discussion and consensus. After reliability was established, coders consulted with each other when they were unsure about how to code a specific response. We used Cohen’s kappas (κ) to calculate the degree of agreement between coders. First, we coded the strategies that youth described in response to open-ended questions concerning how victims should and would respond to the harms. Informed by previous research (e.g., Chung & Asher, 1996), adolescents’ responses were coded for the presence or absence of four types of strategies (κ = .95): talk to authorities, confront aggressors, retaliate, and lack of response. Talking to authorities involved appealing to authority figures such as teachers, referees, and coordinators at the school (e.g., “I would go talk to the professor in charge and then tell him what happened”). Confronting aggressors involved talking to the perpetrators, requesting apologies, or expressing emotions without using aggression (e.g., “Go talk to them, tell them like ‘why are they angry at us?’ or ‘why did they do that to us?’”). Retaliation was defined as seeking revenge or otherwise responding aggressively (e.g., “I would get very frustrated and seek revenge. I would do something that affects the other team.”). Finally, lack of response entailed moving on without doing anything to try to fix the problem (e.g., “Nothing. We would keep working to compete again next year.”). Since the strategy’s lack of response was only mentioned by six participants (4% of scenarios), this code was not considered in quantitative inferential analyses given concerns regarding the robustness of observed patterns in the context of floor effects. It was possible for adolescents to refer to multiple strategies within one response. We also coded the reasons that participants provided to justify their evaluative ratings for the three teacher-sanctioned responses to handle the conflict. The coding scheme for justifications was based on previous research (e.g., Gromet & Darley, 2009; Okimoto et al., 2012) and content analyses of 10% of the interviews. Participants’ responses were coded for the presence or absence of five overall categories of concerns, described below. We further specified whether the justification was used to support an approach (e.g., the strategy achieves retribution) or to criticize an approach (e.g., the strategy does not achieve retribution). Specifically, achieves retribution denoted endorsing an approach because it would punish or harm perpetrators, such as for the sole sake of retribution, or not endorsing an approach because it would not achieve these retributive goals (e.g., “It would be an ideal punishment because they would also come out losing. They would get a scolding.”; κs = .83–.85). Teaches a lesson denoted endorsing an approach because it would prevent future offenses and/or perpetrators would learn that what they did was wrong, or not endorsing an approach because perpetrators would not learn or would repeat the harm (e.g., “They would commit those acts again because they say like ‘I do it, they expel me three days, I come back, and everything stays the same.’”; κs = .92–.96). Fits the offense denoted endorsing an approach because it would be fair considering the perpetrators’ actions or it would be appropriate for the offense committed, or not endorsing an approach because it would not achieve these justice goals (“If I were to put myself in their shoes, I would do it because it is fair. Let’s say, if I damage something, I have to fix it because I was the one who did the damage.”; κs = .84). Benefits the victims denoted endorsing an approach because it would help victims or reduce the negative consequences of the harm, or not endorsing an approach because it would not benefit victims (e.g., “Because they cannot repair the harm like that. Because you can apologize and try to solve things, but the harm was already done.”; κs = .81–.91). Finally, repairs the relationship denoted endorsing an approach because it promotes reconciliation or a renewed consensus between victims and YOUTHS’ RETRIBUTIVE AND RESTORATIVE ORIENTATIONS perpetrators, or not endorsing an approach because it would not repair the relationship (e.g., “It is not going to fix the problem at its roots because [the perpetrators] will continue to hold a grudge towards us.”; κ = .74–.91). Qualitative Analyses Responses to the final open-ended questions concerning the best approach to addressing each harm were coded qualitatively. We followed an emergent explanatory sequential design wherein follow-up qualitative analyses were used to build on and contextualize quantitative findings (Creswell & Plano Clark, 2011). Specifically, we aimed to obtain a more in-depth view of how participants envisioned harms in schools should be handled. Following a phenomenological approach allowed us to explore addressing intergroup harm from the perspective of the participating youth (Creswell & Poth, 2017). Two coders (i.e., the first author and a second coder) read the transcripts multiple times to identify meaningful units of information until all relevant data had been extracted from this question. Open coding of participants’ responses allowed key concepts and distinct patterns to emerge (Strauss & Corbin, 1994). We initially identified specific codes in adolescents’ responses (e.g., “Invite them to have a peaceful dialogue so they share why they did it,” “A dialogue involving both grades,” and “Reach an agreement in which both teams are satisfied”). Then, we considered the similarities and differences between these codes to identify patterns in their reasoning. Disagreements between coders were resolved via discussions. The emerging codes were then compared, discussed, and clustered together into meaningful themes, a process that involved returning to the data (e.g., the three previous codes were grouped into the theme: Going beyond the harm with peaceful dialogues). After collaboratively developing distinct thematic categories, we calculated reliability based on the agreement between coders in identifying these patterns within the transcripts (κs > .81). RESULTS Statistical significance for quantitative analyses was assessed using two-tailed tests at p < .05. For each significant omnibus effect, effect sizes are reported as partial eta-squared (η2p ). When sphericity assumptions were violated, degrees of freedom were adjusted using the Huynh-Feldt correction when ϵ > .75 and the Greenhouse–Geisser 7 correction when ϵ < .75. We tested for differences in the endorsement of strategies across the two scenarios using a series of seven McNemar’s tests for responses to open-ended questions and six pairedsamples t-tests for ratings; of these 13 tests, only one effect was significant. As such, for parsimony, data were collapsed across the two scenarios for analysis. In addition, preliminary analyses did not reveal any significant bivariate correlations with age or point-biserial correlations with gender or school for any of the 13 previously mentioned variables. Thus, these factors were not considered further. There were no missing data for any of the variables included in the analyses below. Open-Ended Descriptions of How Victims Should and Would Respond to Harm We first examined youths’ open-ended responses regarding how victims should and would respond to the harms. To this end, we computed the proportion of times each response was endorsed; values ranged from 0 (i.e., never referenced) to 1 (i.e., referenced across both scenarios). We conducted a one-way repeated-measures MANOVA with the type of response (talk to authorities, confront aggressors, retaliate) as an independent variable, and the endorsement across should and would questions entered as dependent variables. The analysis revealed a multivariate main effect of response type, Wilk’s λ = .18, F(4, 73) = 85.60, p < .001, η2p = .82. Follow-up analyses revealed a univariate effect of response type for should, F(1.47, 111.49) = 86.43, p < .001, η2p = .53, and would questions, F(2, 152) = 13.70, p < .001, η2p = .15. Findings are presented in Table 2. Youth more often indicated that victims should talk to authorities, rather than retaliating (p < .001, d = 2.63) or confronting the aggressors (p < .001, d = 1.33). They also more often indicated that youth should confront aggressors than retaliate (p < .001, d = 0.92). In terms of youths’ expectations regarding what would happen, confrontation was described significantly less than both retaliation (p < .001, d = 0.90) and talking to authorities (p < .001, d = 0.74), which were reported at similar rates (p = 1.000, d = 0.12). When youth did not spontaneously indicate that victims should talk to authorities, we prompted them to consider whether authorities should be involved. Once prompted, adolescents overwhelmingly agreed that the victims should involve authorities in addressing the harm (M = .98, SE = .01). This finding supports the relevance of examining youths’ ratings of varied teacher-sanctioned 8 PAREJA CONTO, RESTREPO, RECCHIA, VELEZ, AND WAINRYB TABLE 2 Youths’ Open-Ended Descriptions of How Victims Should and Would Respond to Harm Should question Would question Type of response M SE M SE Talk to authorities Confrontation Retaliation .76 .31 .06 .04 .04 .02 .63 .35 .67 .04 .04 .04 Note. References to strategies were collapsed across the two scenarios ranging from 0 (i.e., not referenced for either scenario) to 1 (i.e., referenced in both scenarios). approaches to address peer harms, as described in the next section. Prescriptive and Descriptive Ratings of TeacherSanctioned Approaches To examine the ratings of the three teachersanctioned approaches across questions assessing prescriptive evaluations and descriptive likelihood expectations, we first conducted a one-way repeated-measures MANOVA with the type of teacher-sanctioned response (apology, compensation, suspension) as an independent variable, and the ratings across prescriptive and descriptive questions entered as dependent variables. The analysis revealed a multivariate main effect of teachersanctioned response, Wilk’s λ = .46, F(4, 73) = 21.36, p < .001, η2p = .54. Follow-up analyses revealed a univariate effect of response type for prescriptive ratings, F(1.86, 141.17) = 51.11, p < .001, η2p = .40, whereas the effect was not significant for descriptive ratings F (1.84, 140.23) = 2.08, p = .133, η2p = .03. Findings are presented in Table 3. Prescriptive TABLE 3 Ratings for Teacher-Sanctioned Responses to Harm ratings of compensation were more positive than for apologies (p = .002, d = 0.54) and suspension (p < .001, d = 1.51), and evaluations for apologies were also more positive than for suspension (p < .001, d = 0.96). Justifications for Prescriptive Evaluations of Teacher-Sanctioned Approaches To examine youths’ justifications for a given approach, we computed the proportion of times each justification was referenced for each approach; as such, values ranged from 0 (i.e., never referenced) to 1 (i.e., referenced across both scenarios). We conducted a one-way repeated-measures MANOVA with the type of teacher-sanctioned approach as the independent variable (apology, compensation, suspension) and the 10 possible justifications entered as dependent variables (see Table 4). The analysis revealed a multivariate main effect of teacher-sanctioned approach, Wilk’s λ = .06, F (20, 57) = 43.84, p < .001, η2p = .94. Follow-up analyses revealed univariate effects of approach type for all 10 subcategories of justifications (see Table 4). As compared to the other approaches, suspension was particularly favored as it would achieve retribution and teach a lesson to the perpetrators, but it was also sometimes criticized on the same bases (i.e., that it does not consistently satisfy retributive desires or teach a lesson). Suspension was also criticized relative to the other approaches in that it would not fit the offense and it would fail to repair the relationship. As noted above, compensation was generally viewed positively; in comparison to the other strategies, it was particularly favored as it would fit the offense and benefit victims. Finally, youth judged apologies would repair the relationships more than the other approaches; conversely, apologies tended to be criticized in that they would fail to benefit the victims. Youths’ Perspectives on Optimal Approaches to Addressing Harm Prescriptive ratings Descriptive ratings Teacher-sanctioned approach M SE M SE Suspension Compensation Apology 3.39 5.30 4.74 .18 .10 .14 3.47 3.56 3.87 .19 .13 .15 Note. Ratings were based on six-point Likert scales ranging from not good at all (1) to really good (6), for prescriptive ratings, and very unlikely (1) to very likely (6), for descriptive ratings. We conducted qualitative analyses to examine youths’ responses to the final open-ended question for each scenario. Specifically, we asked youth: “If you were to come up with the best way to handle the situation, what do you think should happen?” We identified six overall themes. Three of the themes reflected the authority-sanctioned responses that youth had considered earlier in the interview (i.e., forms of apology, compensation, and TABLE 4 Justifications for Prescriptive Ratings of Different Teacher-Sanctioned Approaches Compensation M (SE) Apology M (SE) Univariate effect for type of approach Achieves retribution .24 (.04)a .03 (.02)b .03 (.02)b F(1.53, 116.09) = 23.04, p < .001, η2p = .23 Does not achieve retribution .12 (.03)a .00 (.00)b .02 (.01)b F(1.34, 102.05) = 11.30, p < .001, η2p = .13 Teaches a lesson .51 (.05)a .32 (.04)b .28 (.04)b Does not teach a lesson .17 (.04)a .02 (.01)b .06 (.02)b F(1.88, 143.20) = 8.71, p < .001, η2p = .10 F(1.33, 101.48) = 12.18, p < .001, η2p = .14 Fits the offense .10 (.02)b .46 (.05)a .12 (.03)b F(1.86, 141.19) = 38.40, p < .001, η2p = .34 Does not fit the offense .67 (.05)a .22 (.03)b .26 (.04)b F(2, 152) = 43.66, p < .001, η2p = .36 Benefits the victim .03 (.02)c .80 (.03)a .25 (.04)b F(2, 152) = 164.48, p < .001, η2p = .68 Does not benefit the victim .14 (.03)b .19 (.03)b .33 (.04)a F(2, 152) = 9.59, p < .001, η2p = .11 Repairs the relationship Does not repair the relationship .02 (.01)c .13 (.03)b .40 (.05)a .10 (.02)a .06 (.02) .03 (.01)b F(1.45, 110.02) = 37.38, p < .001, η2p = .33 F(1.85, 140.82) = 3.57, p = .034, η2p = .04 Example “It is an exemplary punishment. . . one could say that it is equivalent to what they made us lose. . . And that is why it seems good to me. [. . .] [They lose] classes, grades. . . something important, for example, if they had math or physics that day. . . missing a class is already fatal.” “Because there are many who say ‘Oh, better for me. . .’ There are many who would not want to come to school so, of course, they are happy not to go to school for three days. So, it is better that they take away what they like the most, and that is the championship. . . if what they like the most, which is playing soccer, is taken away, then it will hurt.” “Like this, they are more scared to do it again. They will fear being suspended again, for more days, or that they will be expelled.” “People are going to take it like: ‘well, I did it, they suspended me for three days. . . yes, what I did was wrong’ but then later they think like ‘I did it, they suspended me and that’s it.’ Like it will not have the same importance as if they make them, for example, do an assignment to reflect or something like that.” “If they damaged their peers’ belongings, they have to replace it. . . It would be strange that they damaged my things and I had to pay for them. . . It would be like the most appropriate.” “It is not something academic, it is not something about our studies. Before it was academic, they [victims] had lost a notebook and it was already like three days away from school, but here it would be a better punishment to ban them [perpetrators] from another game, from another championship.” “There are people who financially do not have enough to buy it again. . . So, it seems good to me that they help with money so that the students of the other grade can buy the things they need.” “It is good to accept the mistake and everything, but they are not going to pay anything, they are not going to do anything. The damage is already done and they are not going to help.” “It would be like peace, no? So that there is no resentment between the two groups, so there are no other problems later, like now, it would be peace.” “If they suspend people, they will come back with more resentment. It will be resentment towards us because to them the scolding at home and everything will be our fault, so I think they will have more hatred toward us.” Note. Mean values are expressed as the proportionate use of a justification for a particular teacher-sanctioned approach across the two scenarios. Dissimilar alphabetic superscripts indicate significant pairwise differences in the use of a justification across approaches at p < .05 with a Bonferroni correction. No alphabetic superscripts indicate no significant pairwise differences. YOUTHS’ RETRIBUTIVE AND RESTORATIVE ORIENTATIONS Suspension M (SE) Type of justification 9 10 PAREJA CONTO, RESTREPO, RECCHIA, VELEZ, AND WAINRYB punishment), whereas three other themes referred to distinct approaches. Thus, the qualitative analysis further elucidated the concerns and reasoning underlying the responses that they had considered earlier but also revealed distinct approaches that adolescents in this sample particularly favored. This emphasizes the added value of also asking youth to reflect on their ideal solutions in an openended way. The names reported below are pseudonyms. Compensating victims for the harm caused to address material loss. Participants described the importance of compensating victims for the harm that had been caused. Consistent with the quantitative analyses, this was the most predominant theme. Most adolescents proposed that perpetrators should financially compensate victims. For instance, in light of the soccer equipment getting trashed, Carlos responded: “Obviously they have to pay us for the material damage that they caused.” Similarly, in response to the vignette in which belongings from the backpacks went missing, others proposed that perpetrators should help the victims to find their lost belongings or return what had been taken. Participants’ socioeconomic realities were present in their reasoning about compensating the harms. Certain youth, such as Clara, noted that some peers would not have money to buy their belongings again: “In my classroom, there are five people who do not have the financial means to buy notebooks, schoolbags, or anything like that so, yes, they [perpetrators] should help with that.” Simultaneously, other participants recognized that perpetrators might also not have enough money to pay back the lost belongings, so they offered creative solutions that would help to obtain compensation for the harms. For example, Alvaro proposed: “If they can’t pay it back, I think the best way to fix it would be to do a soccer match to collect funds or a raffle or a bazaar.” Apologies as a first step. A second theme focused on the need for perpetrators to apologize for the transgression, either following adults’ requests or voluntarily on their own. However, youth predominantly described apologies as an initial response that would need to be followed by other strategies. In particular, concerns with accountability often resulted in the endorsement of apologies in combination with compensation. For participants, material and symbolic compensation together had the potential to completely restore the harms. This perspective is evident in Luis’s remarks: [Apologies] because there would be a reconciliation between the teams. If they apologize, they will not do it again. . . Paying for the belongings would mean total reconciliation, half of it is peace between the teams and the other half is repairing their stuff. Some adolescents also proposed asking perpetrators to apologize to satisfy their initial retributive desires, underscoring the importance of asking youth to explain their reasoning for endorsing different approaches. For instance, Ricardo described apologies as “a punishment that wasn’t too severe.” Similarly, Fernanda observed that apologies could be used to “humiliate the person but not in a bad way.” Yet, regardless of whether apologies were endorsed to satisfy restorative or retributive desires, they were typically described in combination with other strategies. The need for punishment by excluding perpetrators. Participants described the need for punishments that excluded perpetrators from academic activities, including expulsion from school, suspension, removing students from class, or excluding them from soccer activities. Specifically, in the vignette about the soccer championship, some youth reasoned that the appropriate punishment should involve sports and not school, whereas others endorsed a more severe punishment that would impact both their academic and sports activities. For instance, Jairo suggested: “They should have the possibility to play [soccer] taken away forever.” Importantly, teenagers in Bogotá commonly play soccer during recess so this represented a severe punishment, as Jairo further explained: “That’s worse than losing, not being able to play anymore.” Consistent with adolescents’ lower prescriptive ratings of suspension, these punitive approaches were rarely mentioned as optimal strategies. Involving authority figures. The need for relevant authority figures to guide the process. Participants described the need to involve authorities such as teachers, parents, or school administrators. Although authority involvement was implicit in many of their proposed strategies (e.g., suspending perpetrators), in some cases, participants explicitly mentioned adults in their responses. Some described the role of teachers as fundamental given their position as authority figures in the school. For YOUTHS’ RETRIBUTIVE AND RESTORATIVE ORIENTATIONS instance, Clara, who previously mentioned the possible financial difficulties of victims, also believed that teachers’ involvement would be important to guarantee their compensation: “They should pay for what they did. . . Teachers should say ‘look guys, you did this’, that is, teachers should talk to them because if not they will not do it. . .” Other students, such as Lorenzo, suggested it would be best to follow the standard disciplinary procedure involving teachers and school administrators until the problem was fixed: “Following the normal procedure, if you can’t fix it with the teacher, it goes to the coordinator, and so on. . .” Youth also endorsed involving parents in the process. While some participants proposed home-school collaborations, others noted the unique role of parents in addressing harms and teaching moral lessons to their children. For example, José expected parents would “talk to them, correct them for what they did wrong, and teach them what is right.” Responses led by authority figures: How to extend beyond the usual processes. Although some participants endorsed adults’ typical approaches to conflict resolution within their schools, others noted ways in which these interventions could be improved upon or proposed their own alternative strategies. For example, although Ana believed the commitments made between teachers and students would be effective, she also thought students’ behavior could be followed more closely to make sure they abide by the agreements: “There should be a stronger follow-up [. . .] Like monitoring the student’s behavior in the classroom, with the teachers and everything.” In addition, some participants suggested that adults should be involved as impartial third parties moderating a conversation between the groups, while others advocated for a more active role from authority figures. This active involvement would sometimes entail punishing perpetrators; however, adolescents also proposed varied strategies that adults could employ to promote understanding, cooperation, and empathy between the groups. For example, in the context of the soccer vignette, youth suggested that teachers should organize friendly matches or mix the teams to promote mutual understanding and overcome group divides. Going beyond the harm with peaceful dialogues. Participants also advocated for peaceful and constructive conversations between victims and perpetrators. The stated aims of these dialogues included to better understand why they 11 committed the harm, to help the perpetrators understand the consequences of the harm, to understand both sides of the conflict, to reach a consensus, or to otherwise end the conflict. For instance, in response to the schoolyard vignette in which there was an ongoing rivalry with the other grade level, Pablo suggested: “Clarifying the situation that happened previously to cause this. . . because if it continues to occur it means that there is something behind that which is causing this resentment or this impulse to do something against someone.” Youth varied in their beliefs about who should be involved in these dialogues, with some endorsing authority figures’ involvement in group discussions. For example, Liliana observed: “My classmates would first discuss among themselves because my grade is close-knit. . . then, all together, they would talk to the other grade. . . and there should be an impartial adult in the middle.” However, others endorsed dialogues involving peers alone, as explained by Raquel, in response to the vignette about the soccer equipment getting trashed: “They are supposed to be teens, right? Well, they already have the capacity to reason, so they could discuss among themselves with no need for adult intervention.” Giving back to the community by facilitating learning opportunities. A final theme involved actions whereby the perpetrators would give back to the community at large. Some adolescents suggested that the perpetrators could make posters or give presentations related to the harms. For example, Angelica proposed: “I would make them do a campaign. . . like a campaign about respect towards others and their belongings.” Some youth noted that these initiatives could serve to prevent similar future transgressions. In this sense, these campaigns would not be necessarily restricted to those directly impacted by the harm but rather more geared towards supporting the school community as a whole. For instance, Juan proposed that “[The perpetrators] should go around the school explaining to other students how bad it is to employ those tactics against other peers.” This tendency to educate the community seemed to be informed by youths’ previous experiences in their schools. Some, such as Cristina, also saw this type of community service as an opportunity for self-reflection: “it seems better to me that they do a presentation about responsibility, care. . . so they can teach other classmates, and like that they can reflect.” 12 PAREJA CONTO, RESTREPO, RECCHIA, VELEZ, AND WAINRYB DISCUSSION This study examined how low-SES urban Colombian youth reason about competing approaches to addressing intergroup harms between rival peers at school. In particular, we explored the strategies that adolescents thought their group should and would use to respond to the harms, and their prescriptive evaluations of and descriptive likelihood expectations about three teacher-sanctioned approaches. By asking youth to reflect on their reasons for endorsing different teacher-mandated responses, we also gained new insight into the retributive and restorative concerns that guide their judgments. Finally, our qualitative findings document adolescents’ perceptions of optimal responses to intergroup harms in their schools to further incorporate participants’ voices and lived experiences. These results are considered in light of the salience of violence in these youths’ neighborhoods (Chaux et al., 2009), which provides important context for both their prescriptive and descriptive judgments. How Adolescents Think that Teens Should and Would Respond to Intergroup Harm As expected, our results revealed that youth often reported that victims would seek revenge or otherwise respond aggressively after being harmed. In part, this expectation may be linked to adolescents’ understandings of intergroup relations, and their predictions about the likelihood of hostile responses to intergroup provocations (Dovidio, 2013). Simultaneously, however, they recognized that this strategy was not consistent with their prescriptive evaluations of how harms should ideally be addressed. This pattern illustrates how, despite their expectations of retaliation, adolescents still developed generalized moral concepts about how people should behave. These findings are consistent with previous research documenting a gap, among displaced Colombian adolescents, between youths’ understandings of what is and what ought to be vis-a-vis the harms they encounter in their environments (Ardila-Rey et al., 2009; Posada & Wainryb, 2008). Thus, it seems that when youth are exposed to higher levels of violence, their prescriptive evaluations of how harms should be addressed come into conflict with their expectations of the likelihood of retaliation. In this way, our findings support Yeager et al.’s (2018) observation that traditional interventions with youth tend to overemphasize knowledge transmission at the expense of pondering and taking account of the underlying motives that encourage youth to retaliate. Specifically, given that youth in our study appear to know and judge that retaliation is wrong, interventions aimed at preventing revenge and aggression might need to not only leverage youths’ judgments about how harm ought to be addressed but also acknowledge and explore their complex lived experiences. Regarding other strategies, adolescents also frequently reported victims should and would seek support from authority figures in their schools to address the harms. Similarly, when discussing optimal responses, youth endorsed the involvement of teachers, parents, and school administrators. These findings are in line with previous research suggesting that adolescents support teachers’ involvement in the resolution of severe conflict (e.g., bullying; Frisén & Holmqvist, 2010). In addition, youth may have endorsed adults’ involvement because conflicts between groups can more easily escalate into violence than dyadic conflicts (Rutland & Killen, 2015). Yet, although adolescents’ endorsement of seeking help from authorities was consistent with the seriousness of the harm and the potential risks following intergroup harm, questions remain about their belief in adolescent victims’ capacity to respond to harm without engaging in retaliation. That is, despite reporting that confrontation would be preferable to retaliation, they did not often report that victims would use this approach, and they endorsed confrontation less often than the involvement of authorities. In line with this finding, Reimer (2019) suggested that an overreliance on teachers’ support to address peer conflicts may reflect students’ lack of confidence in their ability to solve conflicts constructively by themselves. This concern was also evident in adolescents’ responses as they noted that confronting aggressors might escalate into retaliation. Their proposed optimal responses further emphasized this tension between confronting perpetrators directly and involving adults. While some youth advocated for autonomous constructive dialogues between victims and perpetrators, others endorsed involving adults as mediators of these conversations. It is also possible that some youth struggled to imagine approaches to harm that did not require adults’ involvement given the predominance of top-down disciplinary practices in Colombian schools (Ardila-Rey et al., 2009; Bustamante et al., 2021). Thus, both cultural norms about deference to authorities and the dynamics of intergroup conflict may have influenced adolescents’ endorsement of involving adults. Future studies might further investigate YOUTHS’ RETRIBUTIVE AND RESTORATIVE ORIENTATIONS youths’ reasoning about confronting aggressors, in terms of the potential risks and benefits associated with responding assertively to intergroup harm. Youths’ Views on Different Forms of Authority Involvement in Addressing Harms We also sought to illuminate how adolescents envisioned and judged different forms of authority involvement. Their descriptive expectations regarding the likelihood of compensation, apologies, and suspension did not differ significantly. Nonetheless, youth clearly endorsed some approaches over others: compensation was evaluated most positively, followed by apologies, and then suspension. Our qualitative analyses echoed these findings regarding adolescents’ endorsement of compensation and the involvement of authority figures in addressing harms at school. Adolescents proposed varied strategies, often involving the support of adults (including parents), to promote cooperation and empathy between the teams or in the broader school community. Particularly, youth advocated for peaceful dialogues to delve deeper into the causes and consequences of harm and to resolve conflict. In this sense, our findings are in line with past work suggesting that adolescents favor restorative approaches to discipline over punitive strategies (e.g., Wachs et al., 2019). Further, youth in our study were seeking comprehensive responses that addressed the needs of all involved parties, while also going beyond the harm to reflect on how these relationships were embedded within a larger community (Llewellyn & Llewellyn, 2015). In this way, our qualitative analyses revealed an orientation towards giving back to their communities in the aftermath of harm that was not otherwise evident in the data. In addition to examining adolescents’ evaluations, we also considered the justifications they provided to support their ratings. Overall, adolescents reasoned that compensation would be a desirable response because it would fit the offense and benefit victims, whereas they endorsed apologies to repair the relationship. These concerns are in line with a restorative orientation to harms; rather than prioritizing retributive aims, adolescents were oriented towards redressing harms, restoring victims, and repairing relationships (Gromet & Darley, 2009; Okimoto et al., 2012; Zehr, 2002). Interestingly, even though the harms were committed by a rival group, adolescents still favored restorative goals; this finding diverges from previous research by Wenzel et al. (2010). 13 Similarly, youths’ perspectives on optimal responses included apologies as a first step to repair the emotional damage, but further illuminated their views that apologies would be preferable in conjunction with some sort of material compensation. These responses illustrate how adolescents’ preferred approaches to harm are likely to vary across situations and types of harm; that is, adolescents may believe that in some cases symbolic reparations are not enough. In this way, adolescents’ consideration of varied facets of situations underlines the inadequacy of traditional punitive practices in schools that promote one-size-fits-all or “zero tolerance” responses to harm (Okimoto et al., 2012; Zehr, 2002). This concern with punitive practices was further illustrated when adolescents criticized suspension because it would fail to benefit victims or repair relationships. When adolescents did endorse suspension in response to the harm, they tended to justify their ratings by noting that it would provide a learning opportunity for perpetrators. Yet, although not to the same extent, adolescents also believed that compensation and/or apologies could serve this function. Their desire to ensure that perpetrators learned from these approaches was typically motivated by pragmatic concerns with deterrence to prevent future harm by the perpetrators or other members of the school community, or the belief that perpetrators could be transformed from this experience. Conversely, some adolescents noted that suspensions would not be effective in providing a learning opportunity to the perpetrators. Despite adolescents’ predominant orientation towards restoration, some also reported retributive aims, particularly in relation to suspension. Additionally, when reflecting on optimal responses, youth proposed different strategies with the stated aim of punishing the perpetrators, such as expulsion, suspension, exclusion from sports activities, and even apologies. Our findings thus highlight that, although youth believed that victims should not respond aggressively to their peers’ provocation, they recognized that authority-mediated responses to harm in schools provide other pathways to achieve desires for retribution. This juxtaposition challenges perspectives on restoration and retribution as contrasting orientations to justice; adolescents in this study and participants in previous research have reported endorsing a mixture of both retributive and restorative goals (e.g., ArdilaRey et al., 2009; Gromet & Darley, 2006). In the case of suspension and their other proposed strategies to exclude perpetrators, adolescents were 14 PAREJA CONTO, RESTREPO, RECCHIA, VELEZ, AND WAINRYB motivated by an affinity to responding to harm in kind with punishments that would negatively impact the perpetrators (Darley et al., 2000). However, similar to their concerns about whether suspension would be effective in providing a learning opportunity to the perpetrators, most adolescents reported this strategy would not fit the offense because it would not be fair, and others doubted that suspension would even achieve retribution. In this way, our findings mirror studies in which adolescents described punitive practices as unfair (Bell, 2020; Rote et al., 2020). Further, contrary to Rote et al. (2020) but consistent with Wachs et al. (2019), many youths in our study were uncertain regarding the effectiveness of punitive responses. Limitations The results of this study are based on a community sample recruited from two schools in Bogotá, Colombia, and thus this investigation contributes to a growing literature on adolescents’ experiences in schools beyond a North American or European context. As is the case in most studies, the current sample may not be representative of Colombian youth with markedly different demographic characteristics. Additional work is needed to further unpack how socio-ecological contexts inform adolescents’ retributive and restorative orientations to harm. For instance, it would be useful to examine these questions in samples of Colombian adolescents within different regions (e.g., those more and less directly affected by political violence) or from different socioeconomic backgrounds. Furthermore, asking participants to describe the best approach(es) to addressing the harms did not allow us to consistently distinguish moral concerns (e.g., what is fair) from pragmatic concerns (i.e., what works). This challenge has also been discussed in the school discipline literature, as an overemphasis on efficacy may result in the use of disciplinary practices that violate students’ moral rights (Tilson & Oxley, 2020). For instance, although expelling a student from school may prevent future disruptions, the interests of the school community should not override what is best for individual students (e.g., their right to moral and academic education). Thus, it will be important for future research to further disentangle these considerations in documenting youths’ perspectives on the best approaches to addressing harm. In addition, since we used self-report measures, social desirability may have influenced our results, particularly with respect to youths’ endorsement of retaliation. Finally, it is common practice to conduct member checking in qualitative studies; because we conducted our qualitative analyses long after data collection had been completed, this was not possible. SUMMARY AND CONCLUSIONS Despite these limitations, our findings make various contributions to the literature on adolescents’ moral development, justice orientations, and experiences of disciplinary practices in their schools. Our results suggest that adolescents recognize the importance of involving authority figures in the constructive resolution of intergroup harm. In addition, our findings demonstrate that adolescents in low-SES urban schools exhibit a restorative orientation to harm as they endorse peaceful dialogues with perpetrators, advocate for strategies that benefit the larger school community, and favor the symbolic and material reparation of victims. Nonetheless, participants’ responses to vignettes also sometimes reflected an endorsement of punitive or retributive aims. In this respect, our findings suggest that it is necessary to move beyond a straightforward contrast between retributive and restorative orientations, to capture a more nuanced perspective on victims’ competing desires in response to harms. This study aimed to center youth voices by documenting their justice orientations and preferred approaches in response to hypothetical peer harms. Our findings have implications for policy and practice in schools. In particular, they underscore the need for schools to support young people in deconstructing and reflecting on their retributive desires in response to peer transgressions. In addition, the results from this study can inform the implementation of restorative justice models in ways that fit youths’ needs and take their perspectives as a starting point for effective and equitable prevention and intervention strategies. For instance, this study highlighted that adolescents hold competing retributive and restorative concerns in response to peer harms; thus, youth-oriented implementations of restorative models ought to create a safe space to acknowledge and explore their varied concerns and how these are guided by interpretations and motivations in the aftermath of being deeply hurt. In this respect, a useful direction for future research is to more thoroughly investigate how to support adolescents’ relationship-oriented responses to harm as they coordinate their YOUTHS’ RETRIBUTIVE AND RESTORATIVE ORIENTATIONS competing desires for retribution and restoration. Overall, this study highlighted youths’ preferences regarding different approaches to addressing intergroup harms and the concerns that guide their reasoning. Centering youths’ concerns may facilitate the development of approaches to harm that are responsive to their needs. ACKNOWLEDGMENTS We thank the participating adolescents and schools, Angelica Alvarez for her assistance with data collection, and Sandra Silva for her help with the coding. This work was supported by graduate fellowships from the Social Sciences and Humanities Research Council of Canada to the first and second authors, and funding from the Concordia University Research Chairs program to the third author. REFERENCES Ardila-Rey, A., Killen, M., & Brenick, A. (2009). Moral reasoning in violent contexts: Displaced and nondisplaced Colombian children’s evaluations of moral transgressions, retaliation, and reconciliation. Social Development, 18(1), 181–209. https://doi.org/10.1111/j. 1467-9507.2008.00483.x Bell, C. (2020). “Maybe if they let us tell the story I wouldn’t have gotten suspended”: Understanding Black students’ and parents’ perceptions of school discipline. Children and Youth Services Review, 110, 1–11. https://doi.org/10.1016/j.childyouth.2020.104757 Bustamante, A., Maldonado-Carreño, C., González, O. L., Navia, M. A., Restrepo, E. C., & Torres, D. F. (2021). ‘Who said we play to hit each other?’ Early childhood teachers’ responses to children’s transgressions. Journal of Moral Education, 50(2), 233–249. https://doi.org/10. 1080/03057240.2019.1708710 Chaux, E., Molano, A., & Podlesky, P. (2009). Socioeconomic, socio-political and socio emotional variables explaining school bullying: A country-wide multilevel analysis. Aggressive Behaviour, 35(6), 520–529. https:// doi.org/10.1002/ab.20320 Chung, T. Y., & Asher, S. R. (1996). Children’s goals and strategies in peer conflict situations. Merrill-Palmer Quarterly, 42(1), 125–147. https://www.jstor.org/ stable/23090523 Creswell, J. W., & Plano Clark, V. L. (2011). Designing and conducting mixed methods research (2nd ed.). SAGE Publications. Creswell, J. W., & Poth, C. N. (2017). Qualitative inquiry & research design: Choosing among five approaches. (4th ed.). SAGE Publications. Darley, J. M., Carlsmith, K. M., & Robinson, P. H. (2000). Incapacitation and just deserts as motives for 15 punishment. Law and Human Behavior, 24(6), 659–683. https://doi.org/10.1023/A:1005552203727 Dovidio, J. F. (2013). Bridging intragroup processes and intergroup relations: Needing the twain to meet. British Journal of Social Psychology, 52(1), 1–24. https://doi. org/10.1111/bjso.12026 Dunham, Y., Baron, A. S., & Carey, S. (2011). Consequences of ‘minimal’ group affiliations in children. Child Development, 82(3), 793–811. https://doi.org/10. 1111/j.1467-8624.2011.01577.x Frisén, A., & Holmqvist, K. (2010). Adolescents’ own suggestions for bullying interventions at age 13 and 16. Scandinavian Journal of Psychology, 51(2), 123–131. https://doi.org/10.1111/j.1467-9450.2009.00733.x Gini, G. (2006). Bullying as a social process: The role of group membership in students’ perception of intergroup aggression at school. Journal of School Psychology, 44, 51–65. https://doi.org/10.1016/j.jsp.2005.12.002 González, T. (2012). Keeping kids in schools: Restorative justice, punitive discipline, and the school to prison pipeline. Journal of Law and Education, 41(2), 281–335. Goode, C., & Smith, H. J. (2016). Retribution or restoration: Symbolic justice concerns shape how victim group members react to intergroup transgressions. Current Opinion in Psychology, 2016(11), 105–109. https://doi.org/10.1016/j.copsyc.2016.07.003 Gromet, D. M., & Darley, J. M. (2006). Restoration and retribution: How including retributive components affects the acceptability of restorative justice procedures. Social Justice Research, 19(4), 395–432. https:// doi.org/10.1007/s11211-006-0023-7 Gromet, D. M., & Darley, J. M. (2009). Punishment and beyond: Achieving justice through the satisfaction of multiple goals. Law and Society Review, 43(1), 1–38. https://doi.org/10.1111/j.1540-5893.2009.00365.x Guerra, N. G., Huesmann, L. R., & Spindler, A. (2003). Community violence exposure, social cognition, and aggression among urban elementary school children. Child Development, 74(5), 1561–1576. https://doi.org/ 10.1111/1467-8624.00623 Hinze-Pifer, R., & Sartain, L. (2018). Rethinking universal suspension for severe student behavior. Peabody Journal of Education, 93(2), 228–243. https://doi.org/10.1080/ 0161956X.2018.1435051 Llewellyn, J. J., & Llewellyn, K. (2015). A restorative approach to learning: Relational theory as feminist pedagogy in universities. In T. Penny Light, J. Nicholas, & R. Bondy (Eds.), Feminist pedagogy in higher education: Critical theory and practice (pp. 11–31). Wilfrid Laurier University Press. McDonald, K. L., & Asher, S. R. (2018). Pacifists and revenge-seekers in response to unambiguous peer provocation. Journal of Youth and Adolescence, 47(9), 1907–1925. https://doi.org/10.1007/s10964-017-0767-4 Mulvey, K. L. (2016). Evaluations of moral and conventional intergroup transgressions. British Journal of Developmental Psychology, 34(4), 489–501. https://doi. org/10.1111/bjdp.12145 16 PAREJA CONTO, RESTREPO, RECCHIA, VELEZ, AND WAINRYB OECD. (2017). PISA 2015 results (volume III): Students’ well-being. OECD Publishing. https://doi.org/10.1787/ 9789264273856-en Okimoto, T. G., Wenzel, M., & Feather, N. T. (2012). Retribution and restoration as general orientations towards justice. European Journal of Personality, 26, 255– 275. https://doi.org/10.1002/per.831 Posada, R., & Wainryb, C. (2008). Moral development in a violent society: Colombian children’s judgments in the context of survival and revenge. Child Development, 79(4), 882–898. https://doi.org/10.1111/j.1467-8624. 2008.01165.x Recchia, H. E., Wainryb, C., & Pasupathi, M. (2019). “I wanted to hurt her”: Children’s and adolescents’ experiences of desiring and seeking revenge in their own peer conflicts. Social Development, 28(4), 840–853. https://doi.org/10.1111/sode.12370 Recchia, H. E., Wainryb, C., & Posada, R. (2020). The juxtaposition of revenge and forgiveness in peer conflict experiences of youth exposed to violence. Journal of Research on Adolescence, 30(4), 956–969. https://doi. org/10.1111/jora.12573 Reimer, K. E. (2019). Relationships of control and relationships of engagement: How educator intentions intersect with student experiences of restorative justice. Journal of Peace Education, 16(1), 49–77. https://doi. org/10.1080/17400201.2018.1472070 Rose, A. J., & Asher, S. R. (1999). Children’s goals and strategies in response to conflicts within a friendship. Developmental Psychology, 35(1), 69–79. https://doi.org/ 10.1037/0012-1649.35.1.69 Rote, W., Corona, A., Moore, L., Patrick, R., & Flak, S. (2020). Adolescent perceptions of inductive discipline as a response to peer aggression: Variation by socialization agent and individual characteristics. Social Development, 30, 428–448. https://doi.org/10.1111/ sode.1249 Rutland, A., & Killen, M. (2015). A developmental science approach to reducing prejudice and social exclusion: Intergroup processes, social-cognitive development, and moral reasoning. Social Issues and Policy Review, 9(1), 121–154. https://doi.org/10.1111/sipr. 12012 Secretarı́a Distrital de Planeación. (2014). Rostros y rastros: Afrodescendientes e Indı́genas en Bogotá. https://www.sdp.gov.co/sites/default/files/voces_ afro_indi_2014.pdf Smetana, J. G., Jambon, M., & Ball, C. (2014). The social domain approach to children’s moral and social judgments. In M. Killen & J. Smetana (Eds.), Handbook of moral development (2nd ed., pp. 23–45). Psychology Press. Strauss, A., & Corbin, J. (1994). Grounded theory methodology: An overview. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (pp. 273– 285). SAGE Publications. Thibert, J., & Osorio, G. A. (2014). Urban segregation and metropolitics in Latin America: The case of Bogotá, Colombia. International Journal of Urban and Regional Research, 38(4), 1319–1343. https://doi.org/10.1111/ 1468-2427.12021 Tilson, J., & Oxley, L. (2020). Children’s moral rights and UK school exclusions. Theory and Research in Education, 18(1), 40–58. https://doi.org/10.1177/1477878520912509 Turiel, E. (2008). Thought about actions in social domains: Morality, social conventions, and social interactions. Cognitive Development, 23(1), 136–154. https:// doi.org/10.1016/j.cogdev.2007.04.001 Van Holstein, E. (2018). Experiences of participatory planning in contexts of inequality: A qualitative study of urban renewal projects in Colombia. Planning Theory & Practice, 19, 39–57. https://doi.org/10.1080/ 14649357.2017.1406981 Velez, G., Hahn, M., Recchia, H., & Wainryb, C. (2020). Rethinking responses to youth rebellion: Recent growth and development of restorative practices in schools. Current Opinion in Psychology, 35, 36–40. https://doi.org/10.1016/j.copsyc.2020.02.011 Wachs, S., Bilz, L., Niproschke, S., & Schubarth, W. (2019). Bullying intervention in schools: A multilevel analysis of teachers’ success in handling bullying from the students’ perspective. Journal of Early Adolescence, 39(5), 642–668. https://doi.org/10.1177/ 0272431618780423 Wainryb, C., Recchia, H., Faulconbridge, O., & Pasupathi, M. (2020). To err is human: Forgiveness across childhood and adolescence. Social Development, 29(2), 509– 525. https://doi.org/10.1111/sode.12413 Wenzel, M., Okimoto, T. G., Feather, N. T., & Platow, M. J. (2010). Justice through consensus: Shared identity and the preference for a restorative notion of justice. European Journal of Social Psychology, 40(6), 909–930. https://doi.org/10.1002/ejsp.657 Yeager, D. S., Dahl, R. E., & Dweck, C. S. (2018). Why interventions to influence adolescent behavior often fail but could succeed. Perspectives on Psychological Science, 13(1), 101–122. https://doi.org/10.1177/ 1745691617722620 Zehr, H. (2002). Little book of restorative justice. Good Books.