Exploring Multisystemic Resilience among Youth of Color Exposed to Direct and Indirect Violence

Objectives: Youth of color (YOC) are more likely to experience direct and indirect violence compared to non-Hispanic White youth. Although the negative consequences of violence exposure are well-established, less is understood about factors that may facilitate resilience. Methods: The current study utilized hierarchical linear regression modeling to examine the associations between resilience and psychological health, social support, and school engagement among YOC in the United States who have experienced violence. Participants included 75 YOC (M age = 9.39; SD = 1.56; 84.1% Black or African American, 9.3% Biracial or Multiracial, 5.3% Hispanic, 1.3% Indigenous) who were directly or indirectly exposed to violence in the past year. They were recruited from a family justice center in the southern United States. Results: The final model was significant (F(5,74) = 21.71; p < .001) and accounted for 57% of the variance in resilience. Specifically, lower levels of direct violence exposure ( β = -.18; p = .002), greater social support from friends and nonparental adults ( β = .35; p < .001), and more school engagement ( β = .47; p < .001) were related to higher resilience.


Introduction
Youth exposure to violence is pervasive worldwide. The World Health Organization (2020) defines "violence" as the threatened or actual use of physical force or power against someone, which actually or likely results in injury, harm, or death. Youth may be exposed to violence directly or indirectly. Specifically, direct violence includes acts of aggression perpetrated against the person directly, while indirect violence includes witnessing or hearing about acts of violence (Finkelhor et al., 2015). Approximately one billion individuals under the age of 18 experience direct violence each year across Asia, North America, and Africa (Hillis et al., 2016). Moreover, approximately 49% of youth in the United States experience indirect forms of violence (Finkelhor et al., 2015). Reflecting larger systems of inequities, youth of color (YOC) in the United States are 1.5 times more likely than non-Hispanic White youth to experience violence (Slopen et al., 2016). The negative consequences of violence exposure on short-and long-term health are well-established (Jackson & Deye, 2015), but less is known about the assets that may be associated with resilience following exposure to violence, particularly among YOC (Henderson et al., 2016). Calls to action have highlighted the need for research on the assets of YOC that help them navigate adversity (Henderson et al., 2016). The current study employed a strengthsbased approach to explore how psychological health, social support, and school engagement were related to resilience among YOC exposed to direct and/or indirect violence.

Violence Exposure Among YOC
In the United States, 55% of youth are exposed to direct violence via physical assault, 40% via property victimization, 26% via maltreatment, and 10% via sexual victimization (Finkelhor et al., 2015). Regarding indirect violence exposure, 39% of youth have witnessed violence in their home or community, and 10% experienced other forms of indirect violence, such as hearing or learning about family assault or community violence, household theft, school violence, or threat of attack (Finkelhor et al., 2015). Previous research has identified a higher prevalence of direct and indirect forms of violence among YOC in the United States compared to non-Hispanic White youth (Antunes & Ahlin, 2021;Slopen et al., 2016). Specifically, YOC have a higher likelihood of experiencing abuse (Wilson et al., 2015) and peer victimization (Jackman et al., 2020). YOC may also be more likely to witness community violence and violence between their caregivers (Slopen et al., 2016). Elevated rates of violence among minoritized racial/ethnic groups in the United States are driven by a plethora of factors, including inequitable access to economic opportunities, education, housing, and healthcare, as well as the impact of the criminal justice system (Frazer et al., 2018).
A growing body of research has examined the consequences of violence exposure among YOC. For example, Hsieh and colleagues (2021) found that exposure to direct and indirect violence among African American youth may increase psychological and physiological stress. Previous literature has typically combined direct and indirect violence, assessing it cumulatively (Antunes & Ahlin, 2021;Frazer et al., 2018;Hsieh et al., 2021;Slopen et al., 2016). However, some literature suggests that direct and indirect violence may differentially affect youths' functioning (Elsaesser, 2018;Zeringue, 2019;Zimmerman & Posick, 2016). For example, in addition to the cumulative effects related to direct and indirect violence exposure, Fleckman and colleagues (2016) also found that direct violence was distinctly related to greater externalizing behaviors. Further, Shukla and Wiesner (2015) found that indirect and direct violence were independently associated with depressive symptoms among YOC when analyzed concurrently. Thus, research examining differential effects of direct and indirect violence could provide a more comprehensive understanding of how these experiences may manifest among YOC.

A Multisystemic Approach to Resilience
Much of the literature examining violence exposure, particularly among YOC, is deficit-based and has focused on maladaptive outcomes, such that even studies of "resilience" often operationalize the construct as an absence of psychopathology (Chancy et al., 2022;Huang et al., 2020;Wright & Wachs, 2019). A deficit-based framework can contribute to harmful stereotypes by placing individual responsibility upon YOC for experiences that are often, at least in part, due to systematic inequities. Alternatively, a strengths-based framework presents key opportunities to better understand the capacities and capabilities of YOC (Henderson et al., 2016). For YOC who are often experiencing multiple types of violence, it is imperative to identify pathways of resilience and underscore factors that bolster positive functioning following adversity to inform culturally relevant prevention and intervention efforts.
Stemming from Bronfenbrenner's (1979) ecological theory of development, the multisystemic model of resilience offers a strengths-based framework that conceptualizes resilience across individual, relational, and contextual domains (Ungar, 2021). This model highlights the importance of factors across the social ecology and frames resilience 3 3 in terms of interactions between individuals and their environment that promote adaptive functioning (Ungar, 2021). Notably, resilience can be defined in a myriad of ways and conceptualized as a predictor, moderator/mediator, or outcome. For the current study, resilience was viewed as an outcome and defined through the multisystemic model of resilience as the sum of one's access to, and engagement with, social ecological resources (Ungar, 2021). In line with this conceptualization, the Child and Youth Resilience Measure-Revised (CYRM-R; Liebenberg et al., 2012) was used to assess resilience in the current study. The CYRM-R was developed with diverse youth across cultures and contexts; it has demonstrated reliability and validity with YOC in North America (Liebenberg et al., 2012). The current study explored three factors that may promote multisystemic resilience among YOC in the United States: psychological health, social support, and school engagement.

Psychological Health
Psychological health can be understood as one's ability to achieve goals within internal and external environments (Preedy & Watson, 2010). Psychological health, which includes feeling positive emotions such as happiness and satisfaction with life, has been linked to resilience in the face of adversity (KIDSCREEN Group Europe, 2006;Picardi et al., 2012). Notably, the literature suggests that YOC may experience greater psychological health than non-Hispanic White youth, due to their connection with their ethnic identity (Berry et al., 2006). A small body of work has examined psychological health among YOC exposed to violence (Kaslow & Thompson, 2008;Storch et al., 2003). Kaslow and Thompson (2008) found that youth who experienced higher levels of direct and indirect violence reported lower levels of psychological health, such as higher internalizing and externalizing problems and greater stress. This study identified a direct relationship between violence exposure and psychological health among YOC (Kaslow & Thompson, 2008), but to our knowledge no research has examined how psychological health may be related to resilience among YOC exposed to violence.

Social Support
Social support, or the emotional comfort provided by others, is one of the most robust predictors of resilience in youth exposed to violence (Yule et al., 2019). Among YOC, social support may be a valuable resource for marginalized communities in ameliorating difficulties that commonly follow exposure to direct and indirect violence (Chancy et al., 2022;Shoemaker et al., 2022). Social support received from family members, peers, teachers, and neighbors has been linked to fewer internalizing and externalizing behavior problems in YOC exposed to adversity and/or violence (Huang et al., 2020;Wright & Wachs, 2019). Importantly for YOC, social support may transcend the immediate family, as extended family members and adults within the community are often involved in child-rearing (Henderson et al., 2016). Williams and Bryan (2013) utilized qualitative methods to identify positive themes regarding the importance of friends, family members, neighbors, and other nonparental adults in providing encouragement, feedback, and support when YOC were faced with adversities. Still, quantitative research is needed to further assess how peer and adult support networks contribute to resilience among YOC exposed to violence.

School Engagement
School engagement may promote resilience among youth. Specifically, school engagement is a multifaceted construct that encompasses behavioral, cognitive, and emotional components (Fredricks et al., 2005). These components include involvement in academic, social, and extracurricular activities, investment in learning, and positive and/or negative responses to one's school environment (Fredricks et al., 2005). Research has identified the potential benefits of school engagement in enhancing positive functioning, including academic success and reduced aggressive behaviors among YOC affected by violence (DiClemente et al., 2018;Elsaesser et al., 2016;Irvin, 2012). Despite this knowledge, there is a dearth of research on school engagement and resilience among violence-exposed YOC. Research suggests that indirect violence exposure can negatively impact school engagement among YOC, although some YOC maintain school engagement despite exposure to violence (Daly et al., 2009). Findings have also shown that engagement in school-based activities may provide supervision and structure that minimize youth violence exposure (Ceballo et al., 2021). The current study expanded on previous work by exploring how school engagement may be associated with multisystemic resilience among YOC who have been exposed to direct and indirect forms of violence.

The Current Study
Much of the available literature examining the effects of direct and indirect violence exposure has been limited by a deficit-based lens and lack of consideration for the distinct experiences of YOC. The current study focused on YOC and their strengths by investigating how psychological health, social support, and school engagement were related to multisystemic resilience among YOC exposed to violence. In this case, all primary variables were self-reported by YOC, whereas the caregivers of YOC participants provided some demographic information. It was hypothesized that: (1) less frequent experiences of direct and indirect violence would be associated with higher resilience and, (2) greater psychological health, higher social support, and greater school engagement would each be related to higher resilience.

Method Participants
Participants were 75 YOC aged 7-12 years who had been exposed to direct or indirect violence in the past year. Participants were recruited from a family justice center (FJC) in the southern United States. The FJC is a community organization that provides access to coordinated services, such as legal support, mental health care, housing, and other social services for individuals who have experienced violence. Approximately half of the participants self-identified as girls, and the remaining youth self-identified as boys. Most youths identified as Black or African American (84.1%), followed by Biracial or Multiracial (9.3%), Hispanic (5.3%), and Indigenous (1.3%). Most participants were living below the federal poverty line for a family of three, with 65.4% of families having an annual household income below $20,001 USD. Detailed sociodemographic information on participants is provided in Table 1.

Procedures
After receiving IRB approval from the first author's institution (University of Memphis), caregiver-child dyads were recruited from the FJC as part of a larger study exploring the effects of participating in a weeklong summer camp for youth exposed to adversity. The current study utilized baseline data from this project; therefore, none of the participants had engaged in project-related services at the time of assessment. To participate in the study, families needed to fluently speak English and have a child aged 7-12 years old. After being informed about the nature and purpose of the study, eligible and interested families were scheduled for an assessment. Before completing the assessment, caregivers provided written consent and permission for their child to participate. Youth provided written assent. Trained study staff met individually with each caregiver and child in separate, private rooms. Interviewers read all items aloud to participants, who followed along with the survey and provided verbal responses to each question. Interviewers recorded participant responses in an online database. Each interview lasted approximately 45 minutes. The child of families who were allocated to the summer camp (i.e., intervention) condition participated in a free weeklong overnight summer camp, whereas families allocated to the control condition were compensated with a $25.00 USD gift card for completing the interview and offered an opportunity to join a waitlist for their child to attend camp the following summer.

Measures
Demographics. Caregivers provided information about their annual household income and their child's race and ethnicity. Youth provided information regarding their age and gender.
Resilience. The CYRM-R (Liebenberg et al., 2012) is a 17-item measure of youth's multisystemic resilience. Youth under age 10 completed an age-appropriate version of the measure using simplified language and a 3-point Exploring Multisystemic Resilience among Youth of Color Exposed to Direct and Indirect Violence 5 5 5 scale: "No", "Sometimes", or "Always" (Resilience Research Centre, 2018). Participants aged 10-12 completed a version of the CYRM-R using standard language and a 5-point scale ranging from 1 ("Not at all") to 5 ("A lot"). To merge data from younger and older participants, measure scores from younger participants were scaled up at the item level to match the 5-point scale used by older participants, as recommended by measure developers. Furthermore, as per the measure developer guidelines, 17 items that overlap across the two versions of the measure were summed to create a total score ranging from 17-85, with higher scores indicating higher levels of resilience (Liebenberg et al., 2012). In the current study, Cronbach's alpha was .70.
Direct and Indirect Violence Exposure. The Juvenile Victimization Questionnaire -2nd Revision Reduced Item Version, Youth Past-Year Form (JVQR2-RIV) is a 12-item measure assessing exposure to property victimization, physical assault, peer victimization, sexual victimization, and witnessing/indirect victimization (Finkelhor et al., 2005). Youth reported past-year violence exposure by responding to items on a dichotomous "Yes" or "No" scale. Affirmative responses were summed to generate two total scores: direct violence exposure (9 items), and indirect violence exposure (3 items). The JVQR2-RIV has been used with racially and socioeconomically diverse youth in the United States (Finkelhor et al., 2005). As a child's exposure to one form of adversity does not necessarily relate to their exposure to other forms of adversity (Gray et al., 2004), internal consistency was not calculated for this measure.
Psychological Health. Youth completed the KIDSCREEN-27, a 27-item measure of youth's subjective health, well-being, and quality of life across five subscales (KIDSCREEN Group Europe, 2006). The current study utilized the Psychological Well-Being subscale, comprised of seven items that examine the presence of positive emotions, absence of negative emotions, and satisfaction with life over the past week. Items were scored on two 5-point scales (i.e., "Not at all" to "Extremely"; "Never" to "Always"). A total psychological health score was computed by summing responses to each item and converting the summed score to a t-score, per measure developer recommendations. Higher scores reflect greater psychological health. The subscale has demonstrated adequate internal consistency (α = .84) and discriminatory power (KIDSCREEN Group Europe, 2006). The measure was developed as part of cross-cultural collaboration and has shown evidence of cross-cultural equivalence (KIDSCREEN Group Europe, 2006;Robitail et al., 2007). In the current study, Cronbach's alpha was .71.
Social Support. The Social Support -Friends and Adults (SS-FA) measure Zimet et al., 1988) is comprised of six items that assess support from friends and nonparental adults. Participants responded on a 4point scale, ranging from 1 ("Not true about me") to 4 ("Mostly true about me"). Items were summed to generate a total score ranging from 6-24, with higher scores indicating higher levels of support . The SS-FA was developed using simplified language/wording intended to be accessible to a range of audiences. The SS-FA demonstrated adequate internal consistency (α = .90) among Latinx adolescents (Nunez, 2019). Further, the SS-FA was adapted from Zimet's (1988) Multidimensional Scale of Perceived Social Support, which demonstrated adequate reliability and validity with racially/ethnically diverse youth from urban communities in the United States (e.g., Canty-Mitchell & Zimet, 2000). For the current study, Cronbach's alpha was .67. School Engagement. The 19-item School Engagement Measure (SEM) assesses three domains of school engagement: behavioral (5 items), cognitive (8 items), and emotional (6 items; Fredricks et al., 2005). Responses are provided on a 5-point scale from 1 ("Not at all true") to 5 ("Very true"). Items were summed to create a total score ranging from 19-95. The SEM has demonstrated adequate internal consistency (α = .67-.88) and concurrent validity (Fredricks et al., 2005). The SEM was initially developed for use with youth living in urban areas who primarily identified as Black or Hispanic and were typically from low-income households (Fredricks et al., 2005). It has since been used with other ethnically diverse United States elementary school students from low-income households (Goldschmidt, 2008). In the current study, Cronbach's alpha was .92.

Data Analytic Plan
Preliminary analyses assessed skewness, kurtosis, outliers, and multicollinearity to determine whether the variables of interest met key statistical assumptions (Tabachnick & Fidell, 2019); no concerns were identified. Missingness patterns were analyzed, and the percentage of missing data on study variables ranged from 0.0% to 2.0%. Little's MCAR was not significant (p > .05), suggesting that data were missing completely at random and appropriate for multiple imputation (Rubin et al., 2007). A power analysis conducted in G*Power 3.1 indicated that a sample size of 75 would be sufficient to detect a medium effect with 95% statistical power (Faul et al., 2009). Hierarchical linear regression modeling was completed in IBM SPSS version 27.0 to examine associations between youth's resilience and their violence exposure, psychological health, social support, and school engagement. In the first block of the model, associations between resilience and youth's violence exposure (i.e., direct and indirect) were assessed. The second block added individual, relational, and contextual factors (i.e., psychological health, social support, and school engagement), exploring how these resources were associated with youth's resilience while accounting for direct and indirect violence exposure. Results from correlations and t-tests indicated that demographic variables (i.e., child age, child gender, child racial identity, and family income) were not significantly associated with resilience. Therefore, these variables were not included as covariates in the regression analyses.

Results
Descriptive statistics and correlations among study variables are provided in Table 2. All youth had been exposed to at least one form of direct or indirect violence. Specifically, 90.0% had experienced at least one instance of direct violence (M = 2.32, SD = 1.56), and 81.0% had been exposed to at least one instance of indirect violence (M = 1.52, SD = 1.05). The most common types of direct violence were: physical assault (72.0%, M = 1.24, SD = 1.06), psychological aggression (56.0%, M = 0.67, SD = 0.66), and property theft (40.0%, M = 0.40, SD = 0.49). The most common types of indirect violence were: witnessing parental intimate partner violence (54.7%, M = 0.73, SD = 0.76), hearing violence outside the home in one's neighborhood or community (e.g., hearing gunshots; 48.0%, M = 0.48, SD = 0.51), and seeing a person being attacked with a weapon (29.3%, M = 0.29, SD = 0.46). Combining direct and indirect violence exposure, 24.0% of youth had been exposed to six or more types of violence (M = 3.80; SD = 2.29; Range = 1-10), indicating the presence of substantial polyvictimization (Finkelhor et al., 2015). Findings from the hierarchical linear regression model are presented in Table 3. Direct and indirect violence were entered into the first block, and the model was significant (F(2, 77) = 5.21, p = .009), accounting for 10.0% of the variance in resilience. Partially consistent with the first hypothesis, less direct violence exposure (β = -0.38, p = .002) was associated with more resilience; however, indirect violence exposure did not emerge as significant. The second block added resources representing youth's psychological health, social support, and school engagement. The model was significant (F(5, 74) = 21.71, p < .001) and accounted for 57.0% of the variance in resilience. More social support from friends and nonparental adults (β = 0.35, p < .001) and greater school engagement (β = 0.47, p < .001) were significantly related to higher resilience, consistent with the second hypothesis. However, direct violence exposure, indirect violence exposure, and psychological health were not significantly associated with resilience. The effect size of the final model (Cohen's ƒ 2 = 0.84) is considered large (Selya et al., 2013).

Discussion and Implications
The current study utilized a strengths-based approach to examine how direct and indirect violence exposure, psychological health, social support, and school engagement were related to multisystemic resilience. Participants included school-aged YOC from primarily low-income households in the United States, most of whom had endured multiple instances of direct and indirect violence. In contextualizing the high rates of direct and indirect violence in this sample, it is important to consider that these youths were recruited from an FJC that serves individuals experiencing adversity and other hardships, so their exposure to violence may have been uniquely elevated as compared to other YOC. Given the adversities faced by many YOC, as well as co-occurring challenges related to poverty and the intergenerational and ongoing effects of discrimination (Henderson et al., 2016), investigating modifiable factors that support resilience among this population is critical. Such work extends beyond a deficit-based framework that has contributed to the systematic marginalization and over-pathologizing of communities of color. Results from the current study illustrate the importance of social support and school engagement among YOC, as these factors were significantly related to higher multisystemic resilience in this sample.
The first hypothesis proposed that less frequent experiences of direct and indirect violence exposure would be associated with higher resilience, which was partially supported by the results. When these facets of violence were assessed in the first block of the regression model, only direct violence exposure was significantly linked to resilience, such that less direct violence was associated with higher resilience. Previous literature has examined rates and effects of cumulative violence exposure among YOC without accounting for variability based on types of violence (e.g., Antunes & Ahlin, 2021;Frazer et al., 2018;Hsieh et al., 2021;Slopen et al., 2016). Few studies have analyzed the potential for direct and indirect violence to have differential relations to resilience. Despite significant overlap and high prevalence of both direct and indirect violence exposure among YOC in the current study, results suggest that direct violence exposure may be more influential on YOC resilience compared to indirect violence exposure. This is consistent with previous literature that has linked direct, but not indirect, violence with increased negative mental health problems among primarily YOC from low-income backgrounds (Javdani et al., 2014).
Notably, current study findings extend previous research by focusing on the associations between types of violence exposure and resilience, which has rarely been examined among YOC. It may be that the natural intensity of direct violence, such as injury or threat to one's life, has disproportionately affected YOC's resilience. Surprisingly, when psychological health, social support, and school engagement were included in the model, direct violence exposure was no longer significantly related to resilience. This finding highlights the powerful contribution of protective factors to resilience, given that the negative impact of direct violence on resilience was no longer significant once these resources were included in the model. Additionally, this finding emphasizes the capabilities of violence-exposed YOC to utilize social support and school engagement as resources that may facilitate resilience.
Indirect violence exposure should not be discounted, as studies have demonstrated significant associations between indirect violence and internalizing and externalizing behavior problems (Elsaesser, 2018;Zeringue, 2019). In the current study, indirect victimization was measured using the JVQR2-RIV, which has only three items regarding experiences of indirect violence. Specifically, the following instances of indirect violence were assessed: seeing violence between caregivers, seeing someone being attacked on purpose with a weapon, and seeing or hearing gunshots, bombs going off, or street riots. It may be the case that inquiring about a wider variety of indirect violent experiences would have provided additional context. Further, differentiating the youth's relation to the person (e.g., family, friend, acquaintance, stranger; Zeringue, 2019) who was seen or heard being involved in violence could further elucidate potential relations between indirect violence exposure and resilience. In addition to methodological considerations, it is pertinent to consider desensitization theory (Gaylord-Harden et al., 2016). This theory posits that youth who are chronically exposed to violence may normalize such experiences through the reduction of emotional responses to violence exposure (Gaylord-Harden et al., 2016), which may contribute to a null association with resilience.
The second hypothesis investigated the differential effects of psychological health, social support, and school engagement on resilience while accounting for experiences of violence. Contrary to expectations, psychological health was not significantly related to resilience. The current study took a novel approach to conceptualize psychological health as the cumulative presence of positive emotions and satisfaction with life, whereas previous research investigating psychological health has examined the absence of negative emotions (e.g., Kaslow & Thompson, 2008). It is possible that assessing a broader spectrum of emotional experiences (e.g., both positive and negative versus one or the other) would offer varying results concerning resilience. There may also be individual factors other than psychological health that could be salient for YOC, such as hope and connection to one's ethnic identity (Berry et al., 2006;Mullin, 2019). Hope has been argued as essential in building children's resilience (Mullin, 2019), whereas the strength of ethnic identity may provide a connection with one's heritage and culture, thereby enhancing resilience (Berry et al., 2006).
In line with what was hypothesized, more social support from friends and nonparental adults was significantly associated with higher resilience. Extant research has identified social support as a critical factor for promoting resilience following violence exposure among youth (Yule et al., 2019), and previous studies have linked social support to reduced internalizing and externalizing behaviors among violence-exposed YOC (Huang et al., 2020;Wright & Wachs, 2019). The current study adds to these findings by establishing a direct link between social support and multisystemic resilience among violence-exposed YOC. Moreover, we utilized a broad conceptualization of social support by accounting for friends, as well as nonparental adults (e.g., extended family members and adults within the community). Given that extended family networks and nonfamilial adults are often essential to the nurturing and upbringing of YOC, this approach to measuring social support is particularly relevant among this population (Henderson et al., 2016;Williams & Bryan, 2013). Thus, research that explores multiple facets of social support (i.e., peers, nonparental adults, caregivers) would provide valuable information regarding how to facilitate social support for YOC exposed to violence.
As expected, greater school engagement was significantly linked to higher resilience. Compared to social support, school engagement has received far less attention in the literature, particularly concerning resilience among YOC. One study linked behavioral and emotional aspects of school engagement to higher achievement and less aggression among YOC (Irvin, 2012), whereas another highlighted the possibility of cognitive school engagement in reducing violence exposure (Ceballo et al., 2021). Although these studies suggest an association between school engagement and violence/aggression, the current study offers novel information regarding the importance of school engagement for YOC exposed to violence by identifying a cumulative, positive relation between resilience and behavioral, emotional, and cognitive school engagement. Schools provide meaningful chances to cultivate important cognitive, social, and psychological developmental skills through academic and extracurricular involvement within a safe and supportive environment (Ceballo et al., 2021;DiClemente et al., 2018), which can offer opportunities for youth to build resilience. Specifically, supporting participation in academic and extracurricular activities, facilitating positive experiences with academic content and interpersonal relations, and fostering a safe space to ask for help and assistance with challenging topics may promote resilience among YOC (Fredricks et al., 2005).

Strengths and Limitations
This study has numerous strengths, including its theoretical grounding in an approach that allows for simultaneous examination of multiple resources that may be related to resilience among YOC exposed to violence. This approach showcases the assets that youth and families of color hold, which have historically been neglected in the literature. Still, these results should be considered in the context of important limitations. The sample size was relatively small, which limits the complexity of analyses that could be conducted. Relying on self-report exposes the study to potential biases (e.g., social desirability, response bias). The use of cross-sectional data prevents assumptions about temporality or directionality. Furthermore, limits to generalizability across YOC are present because most youth self-identified as Black or African American. Generalizability is also limited given participant recruitment through the FJC, as YOC recruited from this agency may have a higher prevalence of violence exposure compared to YOC more broadly. Finally, the current study did not include sexual and gender minority youth. Specifically, binary male and female response options were provided alongside an "other" text entry option. Including a transgender or nonbinary option would have been valuable, as it reduces the stigmatization of gender identities outside the male-female binary. Further, no item assessed sexuality. Including the additional gender response options and an item pertaining to sexual identity would have offered meaningful opportunities to explore resilience among youth with intersectional identities.

Future Directions
Future research should examine additional assets, such as emotion regulation, familial social support, and neighborhood characteristics. Future work should also assess racial and ethnic groups separately, as between-group differences in the mechanisms of multisystemic resilience may arise. It is imperative to acknowledge and account for the historical and current heterogeneous experiences of youth from different racial and ethnic backgrounds. Understanding individuals' level of identification and engagement with their racial and/or ethnic group, as well as experiences of individual and systemic discrimination and oppression, would provide a more comprehensive understanding of how best to support YOC. A longitudinal examination of these factors would offer valuable insight into the directionality of relations over time and across developmental periods. Future researchers should aim to collect data from multiple informants (e.g., parents, teachers) in addition to youth, as this may facilitate a more thorough understanding of resilience and how various resources concurrently or independently play a role in positive development. Finally, future research should assess gender while including transgender or nonbinary options, in addition to an item regarding sexuality.

Clinical and Policy Implications
Results from the current study offer several clinical and policy implications. Mental health practitioners working with YOC should consider screening for violent experiences with an emphasis on discerning direct from indirect exposure, considering the particularly harmful nature of direct violence exposure. To connect YOC with accessible resources and supports, clinicians should screen for strengths and resources, such as social support and school engagement. Such an assessment could provide information regarding youth's current resilience and opportunities to enhance or strengthen relational and contextual resources. Screening for strengths and resources also provides a less stigmatizing approach to assessment, which is critical for YOC who often face compounding challenges and discrimination. Families and school personnel may facilitate youth's social support by establishing consistent, safe relationships characterized by listening, empathizing with, and supporting YOC. Further, school personnel can promote youth's school engagement by guiding YOC in their academic endeavors and encouraging familial participation in school-related activities through open and flexible communication.

Conclusions
The present study used a strengths-based, theoretically informed approach to explore how psychological health, social support, and school engagement were associated with resilience among YOC exposed to violence. Results underscore the importance of social support and school engagement, which are two mutable resources that could be enhanced via intervention. This study adds valuable insight to the literature regarding resilience among YOC, a topic that has often been neglected in previous research. Resilience in urban African American adolescents: The protective enhancing effects of neighborhood, family, and school cohesion following violence exposure.