Early adolescence is a time when youth begin to experience changes in ERI (12, 13). Studies have found that levels of ethnic affirmation increased during the transition into middle school for adolescents, especially African American and Latino youth (13). Other studies examining the longitudinal growth of ethnic identity have found that the majority of youth remain stable or increase in ERI levels over early and middle adolescence school (14–16). Research suggests these changes are influenced by the growing salience of ethnicity (16), highlighted by experiences and relationships in the school environment (17). In fact, studies have found a tendency for middle school youth to segregate themselves by their race and ethnicity compared to elementary school (18, 19). Thus, the emerging salience of ethnicity, along with the biological, psychological, and social changes during early adolescence, underscores the importance of examining ERI and its impact on adolescent wellbeing in this critical developmental period.

The extant literature suggests that racial/ethnic minority youth frequently have negative experiences with their peers at school. Studies have found that racial/ethnic minority students reported more perceived victimization than their non-Hispanic White (NHW) peers (20), specifically due to their ethnicity (21, 22). Research further demonstrates that African American and Hispanic adolescents report high rates of victimization from peers—87% of African American adolescents and 76% of Hispanic adolescents endorsing at least one discriminatory experience from their peers (23, 24). The differential treatment of youth based on their ethnicity exists in the school environment—as many as 46% of African American and 50% of Hispanic/Latino students have reported receiving lower grades based on their race/ethnicity (25), and disparities in the number of office referrals and suspensions for African American and Hispanic youth are well documented (26, 27). Furthermore, teachers have higher behavioral expectations of NHW youth than of minority youth (28). Racial/ethnic minority youth also perceive differing levels of support from their teachers based on their race and ethnicity, with NHW students reporting higher levels of perceived support from teachers than Chinese American and African American students (29, 30).

These findings are concerning, as negative peer and school experiences have detrimental effects on the identity development of racial/ethnic minority youth. Researchers have found that peer rejection and harassment based on racial/ethnic background are associated with negative ethnic identity beliefs (31, 32). Discrimination from peers is also associated with adolescents’ perceptions of their own ethnic group (25). Moreover, negative experiences in the school environment are associated with higher levels of psychological distress. For instance, adolescents who report high levels of peer victimization are at higher risk for depression, anxiety, and behavioral and adjustment problems (33–35). Student perceptions of low teacher support are also associated with more depressive symptoms for youth (36–38). Given the prevalence of these experiences, racial/ethnic minority youth are at increased risk for negative mental health outcomes compared to their NHW peers.

The overwhelming majority of this research has been conducted on monoracial minority youth. As such, the existing knowledge base inadvertently surmises that multiracial youth have similar ERI processes, and have similar outcomes as monoracial adolescents. While research with biracial adults suggest higher ERI is associated with better psychological outcomes (39), emerging research is demonstrating that developmental trajectories and outcomes may differ for multiracial youth. Multiracial adolescents reported higher rates of substance use and aggressive behaviors, and are at a higher risk for general health problems, academic concerns, and engaging in negative activities (40–42). In addition, studies have found that multiracial adolescents have higher rates of substance use in adulthood and are a higher risk for high lifetime substance use compared to their mono-racial/ethnic peers (41, 43). Multiracial youth were found to be more aware of race/ethnicity-related issues at an earlier age than monoracial youth, likely due to the awareness of these issues by the people with whom they interact and the potential feelings of marginality as a result of these (40). A study found that the majority of multiracial adolescents experienced a moderate decline in their ERI level over the early adolescent years, while the majority of minority adolescents experienced an increase (15). Furthermore, a strong sense of ERI was found to actually increase the likelihood of engaging in aggressive behaviors for multiracial adolescents instead of serving as a protective factor (40).

For multiracial individuals, developing ERI can be an especially complex and challenging process because they must navigate the integration of two or more racial backgrounds (44). Multiracial youth may experience more difficulties in their identity development due to various factors (45). For example, they may not receive the support they need to develop an achieved identity because their parents, who typically identify with one single racial/ethnic group may not have the necessary awareness and/or willingness to talk about issues related to their children’s ethnic identities (45). One study found that the majority of multiracial adolescents reported not talking about race at home (46). Studies on multiracial adults have found they often identity denial, which is defined as experiences in which one’s biracial identity is denied or questioned by others (47). Multiracial individuals often feel forced to choose just one racial group to identify with due to ways that they are being perceived by others based on physical appearance, social class, and other cultural and contextual factors (44). They also can experience marginalization and rejection from not just one side of their racial/ethnic background but from both sides, not belonging to or being fully accepted by either group. These findings are not generally relevant or encountered by monoracial minority youth, and suggest that the influences on ERI development for multiracial youth may differ for monoracial youth.

A surprisingly small number of studies have investigated the impact of the larger school context on ERI development and later psychological outcomes. Even fewer have examined how these relationships may differ for monoracial versus multiracial adolescents. This study examined the associations between negative peer experiences in schools, school environment, ethnic identity, and depression in a large sample of racial/ethnic minority early adolescents. Negative peer experiences in sixth grade was hypothesized to be associated with lower levels of ERI in middle school and higher levels of depression in ninth grade. Higher levels of ERI were hypothesized to be associated with lower levels of depression in ninth grade. ERI was hypothesized as a mediator in the relationship between negative peer experiences and school environment in sixth grade and depression in ninth grade.

The 593 adolescents comprised of 51.5% males (n = 305) and 48.5% females (n = 287). The mean age at sixth grade was 11.90 years, and the ethnic composition was as follows: 35.9% NHW adolescents (n = 213), 18% Latino adolescents (n = 107), 15.2% African American adolescents (n = 90), 8.9% Asian/Pacific Islander American (n = 53), 2.5% American Indian/Native American (n = 15), and 19.2% multiracial adolescents (n = 114). In order to address the research questions for the present study, participants who self-identified as a person of color with one race (e.g., only African American, only Latino) were combined into one group labeled monoracial to allow for comparison with multiracial adolescents. Monoracial adolescents represented 44.5% (n = 264) of the adolescents in this study.

All study variables were measured using the Child and Family Center Youth Questionnaire (49, 50), which included a range of items that assess for adolescent social, emotional, and behavioral wellbeing.

The latent construct of negative peer relationships was measured using three measures drawn from the larger Child and Family Youth Questionnaire, which assessed for youth’s peer and social skills (50). These questions have been widely validated in studies with diverse populations of youth (52, 53). Perception of peers is a seven-item youth self-report measure was used to assess adolescents’ perceptions of the peers at their school. Each item ranged on a scale from 1 to 5 and had a descriptive word used to describe the peers. Items were unfair–fair, mean–nice, cold–warm, unfriendly–friendly, bad–good, cruel–kind, and dishonest–honest, and higher scores indicated more-favorable perceptions of other students. Wave 1 (Grade 6) of this measure was included as part of the latent variable for negative peer relationships. This measure was reliable (α = 0.91). Problems with peers was assessed using this four-item measure on the frequency of problems that adolescents had with other students in the past month. On a 5-point scale ranging from never or almost never to always or almost always, youth rated their level of agreement to questions such as “I had a problem with other students,” “Students called me names, swore at me, or said mean things to me,” and “A student hit, pushed, or fought me.” Wave 1 was included in the analysis and partially comprised the latent variable of negative peer relationships. Reliability of the measure was α = 0.77 at Wave 1. Teased by peers was assessed using a five-item measure that determined the frequency of teasing experienced by youth at their school. On a scale of 1–5, adolescents rated their level of agreement ranging from never or almost never to always or almost always on items such as “I was teased by kids at school for no reason” and “I was ignored by kids I wanted to hang out with.” The measure in sixth grade (Wave 1) was included as the latent variable measuring negative peer relationships. Reliability of this measure was α = 0.78 at Wave 1.

The latent construct of school environment was measured using three measures assessing youth perception of their teachers, their school climate, and school safety. Perception of teachers included seven items and asked about adolescents’ perceptions of their teachers at school. Youth used a scale of 1–5 to assess the following qualities: unfair–fair, mean–nice, cold–warm, unfriendly–friendly, bad–good, cruel–kind, and dishonest–honest, with higher scores indicating more-favorable perceptions of teachers. This measure was reliable in this sample (α = 0.91). School climate was a nine-item measure that evaluated adolescents’ positive experiences at school, including availability of teachers, amount of praise received, and opportunities to be involved in extracurricular activities. Youth rated their experiences on a 5-point Likert scale ranging from never to almost never to always or almost always. The reliability of the measure at Wave 1 was 0.73. School safety perception asked youth to report how safe they perceived their school to be in sixth grade (Wave 1) was used as part of the school environment latent variable. Adolescents rated their level of agreement from strongly agree to strongly disagree on seven items, such as “I feel safe in the school hallways,” “I feel safe in the school restroom,” “I feel safe in the classroom,” and “I feel safe outside the school.” This measure was reliable (α = 0.88).

A multigroup, path analysis was conducted using structural equation modeling (SEM) to determine the mediational role of youth ethnic identity in the relationship among negative peer relationships and the school environment in middle school and their impact on depression symptoms in ninth grade (see Figures 1–3). Direct and indirect effects were examined. Adolescent gender and baseline depression symptoms were covariates in the analyses. Multigroup analysis (MGA) was used to determine if these relationships differed by racial group (NHW, monoracial, multiracial). All analyses were conducted using AMOS software version 28. Missing data was imputed using multiple imputation with the Bayesian approach (54). This approach has been found to achieve better and more reliable results (55). The following indices were used to assess model fit: comparative fit index (CFI) (56), with values greater than 0.95 indicating good model fit and root mean square error of approximation (RMSEA) (56), with values less than 0.08 indicating reasonable model fit. A model was determined to fit well if these criteria were met. A statistical significance level of 0.05 was used to evaluate the statistical significance of individual model parameters (e.g., factor loadings, correlations).

Racial group differences were determined using MGA (57). First, an unconditional model was examined, which allowed for all parameters to be freely estimated across groups. Next, a constrained model was examined—this model constrained all parameters at the measurement level (e.g., path coefficients) to be equal across racial groups. The chi-square difference test was used to determine if the unconstrained and constrained models differed significantly – non-significance indicates no differences in parameters between groups, whereas significance indicates a lack of equality in the constrained model and differences across groups.

The model demonstrated relatively good fit, χ²(30) = 114.37, p < 0.001, CFI = 0.92, RMSEA = 0.07, SRMR = 0.07. To determine if the model path coefficients differed by racial group, the chi-square difference between the measurement and unconditional models were examined – this test was statistically significant (p < 0.05), indicating the path coefficients in the model were not equal across racial groups. Thus, the direct and indirect paths were examined by racial group and discussed separately.