Do Ethnic-Racial Identity Dimensions Moderate the Relations of Outgroup Discrimination and Ingroup Marginalization to Self-esteem in Black and Latinx Undergraduates?

Guided by social identity and intergroup theory, we tested how two facets of ethnic-racial identity—felt typicality (perceived similarity to other ingroup members) and ingroup ties (felt closeness to other ingroup members)—potentially buffer the negative effects of outgroup discrimination and ingroup marginalization on self-esteem. Participants included 407 Latinx (65%) and Black (35%) undergraduates (Mage = 24.72 years, 79% women, 21% men) who completed an online survey. Our analyses yielded three key findings. First, both outgroup discrimination and ingroup marginalization predicted lower self-esteem; however, this association was significantly stronger with ingroup marginalization than outgroup discrimination. Second, the association between ingroup marginalization and self-esteem was reduced when ethnic-racial identity variables were controlled. Felt ethnic-racial typicality additionally moderated the association between ingroup marginalization and self-esteem—whereby the negative association was stronger when individuals felt higher ethnic-racial typicality. Our findings expand understanding of the impact of marginalization and discrimination from those within and outside of one’s ethnic-racial group, respectively. We also discuss the differing roles of ethnic-racial identity when experiencing outgroup discrimination and ingroup marginalization.

bias from within their ethnic-racial group such as pressures to conform to ingroup expectations and identity invalidation (Bergin & Cooks, 2002;Wilson, 2021). These ingroup experiences may create their own unique stressors, such as a decreased sense of belonging and feelings of shame (Cooper et al., 2015;Gonzalez, 2019;Murray et al., 2012;Wilson & Leaper, 2016).
We use the term ingroup marginalization to describe experiences with ingroup pressures and biases based on expected behaviors, interests, and physical appearance (Contrada et al., 2001). Much of the previous work on ethnicracial discrimination has focused on the effects of experiences from either one's outgroup or one's ingroup (e.g., Cobb et al., 2019;Harvey et al., 2005). To our knowledge, few studies have examined the roles of both ingroup marginalization and outgroup discrimination simultaneously. Our study contributes to the literature by investigating the effects of both ingroup marginalization and outgroup discrimination as well as the role of two separate dimensions of ethnicracial identity in buffering negative effects. Relatively little prior work has examined more than one facet of ethnic-racial 1 3 identity in relation to experiences with outgroup discrimination and ingroup marginalization.
In the following sections, we discuss the occurrence of ethnic-racial outgroup discrimination and ingroup marginalization and their relations to self-esteem among Black and Latinx groups. We then explore the role of ethnic-racial identity when confronting both of these forms of ethnicracial bias. Finally, we discuss our hypotheses regarding how ethnic-racial identity moderates the relationships between self-esteem and perceived outgroup discrimination and ingroup marginalization in our undergraduate sample.

Ethnic-Racial Outgroup Discrimination and Ingroup Marginalization
Racism is "a system of dominance, power, and privilege based on racial-group designations, rooted in the historical oppression of a group defined or perceived by dominantgroup members as inferior, deviant, or undesirable" (Harrell, 2000, p. 43). According to critical race theory (Bell, 1987;Crenshaw, 1991), racism and discrimination are embedded in the fabrics of society, and it is a lens through which to scrutinize how the intersections of power, race, and other social identities influence injustices. For example, systemic racism and ethnic-racial outgroup discrimination can be seen in inequities throughout various sectors of society such as income inequalities (Manduca, 2018), exclusion from housing (Ross & Turner, 2005), disparate treatment in healthcare (Hall et al., 2021), biases in the criminal justice system (King & Johnson, 2016), and police harassment and surveillance (Fisher et al., 2000;Greene et al., 2006).
In addition to these larger systemic biases, ethnic-racial discrimination also occurs in everyday interpersonal interactions between ethnic-racial groups. These include intentionally hostile comments and actions as well as subtle microaggressions from outgroups (Sue et al., 2007). Perceptions of ethnic-racial outgroup discrimination are common among Black and Latinx groups. For example, a recent study among Latinx adults found that experiences with forms of discrimination, such as racist name-calling and assaults, were related to Spanish/English language proficiency and heritage (Cano et al., 2021). In addition, in their study with Black college students, Sellers and Shelton (2003) found that more than half of the participants reported experiencing discrimination at least once in the previous year. Although much of the research on discrimination has focused on intergroup bias, as described below, individuals may also experience marginalization from members within their own groups.
Ingroup marginalization describes a within-group system of prejudice that stems from and mirrors processes of privilege and exclusion traditionally observed with outgroup discrimination. It can be based on various factors including adherence to perceived group norms for behavior, language, speech styles, and phenotypic appearance or skin tone (Bergin & Cooks, 2002;Fordham & Ogbu, 1986;Hunter, 1998;Maddox, 2004;Uzogara, 2019;Wilson, 2021). It is important to note that we are not equating the societal structures that perpetuate outgroup racist discrimination with the processes underlying ingroup marginalization. Instead, as described below, we view ingroup marginalization as a product of racist discrimination; and we compare how experiences with both forms of bias affect Black and Latinx individuals. The genesis and propagation of ingroup judgments can be viewed through two lenses: (a) its sociohistorical context and (b) social identity and intergroup theory. We briefly discuss each in turn.
Legacies of slavery and colonialism among African American and Latinx communities provide an important sociohistorical context to consider when examining ingroup marginalization. For example, because many of the enslaved Africans with lighter skin tones were offspring of the slave owners, they often received preferential treatment. This included living in the home of the slave owner instead of on the field, receiving less violent treatment, less work-intensive tasks, and occasionally being released from slavery (Hunter, 2007). The hierarchical system that privileged those with lighter skin and European features over those with darker skin and West-African features persisted even after slavery was abolished and enabled lighter-skinned Black people to have greater social opportunities from both ingroup and outgroup members. Ingroup privileges are enacted through opportunities in education and employment (Hochschild & Weaver, 2007), evaluations of attractiveness (Hill, 2002), and social inclusion (Haywood, 2017), among others.
The colonization of Indigenous persons in the Americas by Europeans created a similar ingroup hierarchy within the Latinx community (Darity et al., 2010;Vigil & Lopez, 2004). Spanish colonizers spread and elevated cultural practices and features that reinforced European superiority such as having a lighter skin tone, speaking Spanish, clothing styles, and religious practices (Hunter, 2016). The legacy and internalization of colonial ideologies has been found to negatively impact ingroup peer dynamics (Barillas-Chón, 2010). For example, Gonzalez (2019) investigated attitudes and experiences with ingroup marginalization among mestizo/a Mexican (mixed with Spanish and Indigenous heritage) and Indigenous Oaxaqueño youth living in the US. These two groups share a common Mexican heritage, but a within-group hierarchy exists because of the mixed Spanish heritage of mestizos. Results indicated that participants of Indigenous Oaxaqueño heritage reported experiences with discrimination from their mestizo Mexican peers for cultural differences in language, dress styles, and skin color. Further, these experiences with ingroup marginalization were related 1 3 to conflicting feelings of shame, righteous indignation, and strengthening their cultural pride and identity.
In addition to sociohistorical explanations of ingroup marginalization, social identity and intergroup theory (e.g., Tajfel, 1982;Tajfel & Turner, 1979) offer an additional lens to examine how these biases are perpetuated. According to this theory, ingroup marginalization stems from ingroup members seeking to affirm and preserve ingroup identities and define clear boundaries between the ingroup and outgroup (e.g., Hogg & Rinella, 2018;Marques et al., 1998). Scholars have expanded on social identity theory to illuminate how typicality influences judgments of ingroup members. For example, Marques et al. (1998) found that students perceived ingroup members who adhered to group norms more positively than those who did not follow prescribed norms. This illustrates what Marques et al. (1998) called the black sheep effect, in which group members discriminate against other ingroup members whose behavior deviates from expectations. Ingroup marginalization may be experienced through social exclusion or through statements invalidating one's identity due to perceived deviation from normative ingroup behavior, such as "acting White" (Bergin & Cooks, 2002;Neal-Barnett et al., 2010) or "acting Spanish" (Carter, 2006). In Carter's (2006) study with Black and Latinx high school students, the "acting White" or "acting Spanish" accusations were described as a label to describe and ostracize peers based on behaviors such as their speech style, cultural preferences, and clothing style. Through the use of these labels, group identity is preserved by valuing ingroup normative behavior over behavior that is considered more typical of the dominant outgroup.

Relations of Outgroup Discrimination and Ingroup Marginalization to Self-esteem
The effects of outgroup discrimination on individuals' psychological health have been well documented. For instance, in their meta-analysis on the effects of perceived discrimination, Schmitt et al. (2014) found that experiencing pervasive racism was related to decreased self-esteem and other indices of well-being. Furthermore, in their study with Black, Latinx, and Asian American youth, Greene et al. (2006) found that perceived discrimination from adults and peers was correlated with decreased self-esteem. Relatedly,  found that African American college students who reported more experiences with discrimination felt higher psychological distress. This relationship was particularly strong among those for whom race was a central aspect of their identity.
In addition to outgroup discrimination, empirical studies have suggested that ingroup marginalization can undermine individuals' self-esteem (Contrada et al., 2001;Durkee & Williams, 2015;Durkee et al., 2022;Murray et al., 2012). For instance, Small and Major (2019) found that those who contradicted ingroup expectations were perceived as untrustworthy and their racial identification was questioned by ingroup members. Furthermore, in their study with Latinx college students, Ojeda et al. (2012) found that pressure to adhere to ingroup norms was negatively related to wellbeing. Niemann et al. (1999) found similar results in their focus group study with Mexican-heritage participants living in Houston. Mexican immigrant participants reported experiences with discrimination and name-calling from Mexicanheritage persons who were born in the US or with multiple generations of family in the US. Further, they described feeling more stress when discrimination was felt from members of their own community than when experienced from others.
As explicated in social identity theory (Tajfel & Turner, 1986), evaluations from a valued ingroup may be more impactful to well-being than perceived discrimination from outgroups. Arneta and Hunt (2009) argued that when outgroup discrimination is felt, individuals may feel a sense of common fate with other ingroup members who may share similar experiences, which helps to protect self-esteem. However, when one is marginalized from ingroup members, this may disrupt their sense of belonging, which has important implications for self-esteem (Gailliot & Baumeister, 2007;Greenaway et al., 2016). Those who seek ingroup connections-but feel ostracized from ingroup members for violating ingroup norms for behavior or phenotype-may experience a lack of belonging and lower overall adjustment (Hagerty et al., 1996;Oyserman et al., 2006;Williams, 2009). Thus, because a sense of belonging is important for positive adjustment (Jetten et al., 2015), ingroup marginalization may have a greater impact on well-being and selfesteem than outgroup discrimination.

Ethnic-Racial Identity as a Buffer Against the Effects of Outgroup Discrimination and Ingroup Marginalization on Self-esteem
There is an extensive body of work examining the negative effects of perceived outgroup discrimination on selfesteem (Benner et al., 2018;Prelow et al., 2004;Schmitt et al., 2014). At the same time, prior studies have not found reliable evidence that overall self-esteem was lower among ethnic-racial minority groups. In one meta-analysis, ethnic-racial minorities either reported higher or comparable average rates of self-esteem relative to White samples (Twenge & Crocker, 2002). Further, studies suggest that strong ethnic-racial identities may enhance self-esteem among ethnic-racial minority groups (Lorenzo-Hernández & Ouellette, 1998;Parham & Helms, 1985). Understanding the role of ethnic-racial identity when experiencing outgroup discrimination or ingroup marginalization may help to clarify these patterns. Our study expands on previous work to investigate how different dimensions of ethnic-racial identity moderate the relationships of both outgroup discrimination and ingroup marginalization to esteem.

Ethnic-Racial Identity
Ethnic-racial identity refers to how individuals make meaning of their ethnic-racial heritage and sense of belonging when thinking about themselves (Phinney, 1992;Rogers et al., 2020). Ethnic-racial identity is a lifelong process that includes both developmental (e.g., exploration) and content (e.g., attitudes) dimensions (Cheon & Yip, 2019;Rogers et al., 2020;Williams et al., 2020;Yip et al., 2019). Previous work has suggested that conceptualizing identity as a multidimensional construct illuminates important relations that might otherwise be obscured when considering identity as one composite construct (Rivas-Drake et al., 2014a). In the present study, we separately examined two facets: felt typicality (i.e., self-perceived similarity to other ingroup members) and ingroup ties (i.e., felt connection to others within ingroup) as introduced in previous measures of group identity (Cameron, 2004;Egan & Perry, 2001). In prior studies, each of these dimensions regarding ethnic-racial identity was positively associated with individuals' self-esteem (see Rivas-Drake et al., 2014b;Wilson & Leaper, 2016, for reviews). In addition, these two dimensions are aspects of identity that may influence how individuals are perceived by both ethnic-racial ingroup and outgroup members (Maddox, 2004). In the present study, we tested ethnic-racial typicality and ingroup ties in relation to adults' experiences with outgroup discrimination or ingroup marginalization and selfesteem in a sample of Black and Latinx adults. We elaborate below on the potential relations of both identity facets to experiences with outgroup discrimination or ingroup marginalization.

Ethnic-Racial Typicality
Previous research has found that those who appear more typical of their ingroup are more likely to experience outgroup discrimination (Maddox, 2004) and are perceived as more likely to be accepted among their ingroup (Wilson, 2021). Furthermore, previous studies have indicated that felt typicality among ethnic-racial minoritized persons was positively related to self-esteem and well-being (Hoffmann et al., 2020;Mitchell et al., 2018;Wilson & Leaper, 2016). Indeed, self-perceived ethnic-racial typicality may be protective when confronted with outgroup discrimination. For example, in their study with multiracial adults, Giamo et al. (2012) found that those who felt ethnic-racial discrimination from outgroups were more likely to feel similar to their ingroup and this similarity was related to higher well-being.
In contrast to outgroup discrimination, ingroup marginalization may be related to lower felt ethnic-racial typicality. Ingroup marginalization may make one feel less similar to their ingroup members which, in turn, would negatively affect self-esteem (Harvey et al., 2005). In addition, as highlighted in cognitive consistency or balanced identity theories (see Gawronski & Strack, 2012), persons who feel more typical of their ethnic-racial group may find ingroup marginalization is especially distressing given the discordance between their self-perceived identity and experience.

Ingroup Ties
When experiencing ethnic-racial outgroup discrimination, feeling a sense of connection and support from ingroup members may help mitigate the negative effects of these experiences (Kiang et al., 2008;Leach et al., 2008;Thibeault et al., 2018;Umaña-Taylor & Updegraff, 2007;Verkuyten & Thijs, 2006). Branscombe et al. (1999) theorized that experiencing outgroup discrimination may lead individuals to hold closer connections to their marginalized ingroup, which protects their overall well-being. In their study with Mexicanheritage youth, Baldwin-White et al. (2017) found that those who experienced outgroup discrimination were more likely to re-affirm connections to their ethnic-racial ingroup. Feeling a connection with ingroup members may protect against the negative effects of discrimination. In their review of the psychological effects of discrimination among Latinx adults, Andrade et al. (2021) found that discrimination was related to poorer mental health outcomes. However, having a sense of belonging with ingroup members helped to protect against these negative outcomes.
When experiencing ingroup marginalization, the role of ingroup ties in mitigating its negative effects is more complex. An individual might experience increased tension with group members when they do not feel accepted (Fordham, 1988;Steele & Spencer, 1992). In their study with Latinx college students, Castillo et al. (2007) tested the effects of ingroup marginalization among ethnic-racial minoritized college students. They investigated how pressures to conform to ethnic-group expectations and avoid "acting White" accusations among family, peers, and ethnic-group members relate to stress. Results indicated that ingroup marginalization from family members was related to increased stress and family conflict. Similarly, in their study with African American college students, Thompson et al. (2010) found that familial pressures to avoid the perception of "acting White" was related to increased stress. The "acting White" accusation has been considered a form of bullying (Davis et al., 2019), which is related to increased feelings of neglect and lower psychological health. Thus, although developing ingroup ties may be a helpful coping mechanism when confronted with outgroup discrimination, it might not be protective when experiencing ingroup marginalization because these ingroup relationships may be a distinct source of stress.

Present Study
A distinctive feature of the present study was our consideration of Black and Latinx adults' experiences with outgroup discrimination as well as ingroup marginalization. Most prior research on ethnic-racial discrimination in the US has addressed the effects of intergroup discrimination experienced from White European Americans (Fisher et al., 2000;Phinney et al., 2008;Tajfel, 1982). Scant research has investigated the effects of ingroup marginalization. Even less work has considered both types in the same study. Moreover, little work has examined more than one facet of ethnic-racial identity in this manner. The present study addresses these gaps by investigating the effects of ethnic-racial typicality and ingroup ties on both outgroup discrimination and ingroup marginalization among Black and Latinx adults. The following research questions and hypotheses were tested: (1) Do experiences with outgroup discrimination and ingroup marginalization differ in predicting selfesteem? Each of these experiences are known to predict lowered self-esteem (e.g., Schmitt et al., 2014;Small & Major, 2019). However, social rejection from ingroup members may be more harmful than negative perceptions from outgroup members (Armenta & Hunt, 2009). Accordingly, we expected the relation between ingroup marginalization and esteem would be stronger than the relation between outgroup discrimination and esteem (Hypothesis 1).
(2) Does ethnic-racial identity moderate the relationship between perceived outgroup discrimination and selfesteem? Social identity theory and related empirical work have demonstrated that ethnic-racial identity dimensions can be protective against the effects of discrimination on psychological well-being Tajfel & Turner, 1979). In line with this work, we hypothesized that felt typicality or ingroup ties would mitigate the negative relation between outgroup discrimination and esteem (Hypothesis 2). (3) Does ethnic-racial identity moderate the relationship between perceived ingroup marginalization and selfesteem? Empirical work has suggested that felt typicality and pressures from ingroup members to adhere to group expectations are related to increased stress and lower psychological health (e.g., Castillo et al., 2007). Thus, we hypothesized that felt typicality or ingroup ties would augment the negative relation between ingroup marginalization and esteem (Hypotheses 3).

Participants
The sample comprised undergraduates enrolled in psychology courses at a southwestern public university where the student body is diverse in ethnic-racial background and age. The mean age of students enrolled at the university is 26.7 years and the majority are part-time students (52%; Office of Data Analytics and Institutional Research, 2022). As described below, our sample reflected the diversity of the institution. Our initial sample included a total of 570 undergraduates. The self-reported ethnic-racial demographics of total participants included 46.5% Latinx (n = 265), 24.9% Black (n = 142), 11.2% White (n = 64), 8.9% Asian American (n = 51), and 8.4% Mixed/Other (n = 48). The current study focused on the experiences of Black and Latinx groups, and therefore the sample was restricted to the Black and Latinx participants. Thus, the effective sample included 407 undergraduates (79% women, 21% men) ranging in age from 17 to 61 (M = 24.72, SD = 7.86). Fifty-two percent of participants were between the ages of 17 and 22 and 38.5% were between the ages of 23 and 34. Sixty-five percent of participants (n = 265) were Latinx (80% women, 20% men); and 35% (n = 142) were Black (77% women; 23% men). Age was a missing variable for 1 Latina woman. In regression analyses described later, this missing value was substituted with the mean. See Table 1 for more information on the demographic background of our sample.

Procedure
This study was reviewed and approved by the Institutional Review Board (IRB). Participants were recruited online from the university's psychology participant pool. They gave their informed consent before completing the online survey for course credit. The survey included multidimensional measures of identity for ethnicity-race and gender, felt ethnic-racial and gender discrimination, and self-esteem. The present study focused on the analysis of ethnic-racial identity and ethnic-racial discrimination.

Measures
The following scales were analyzed for the present study. Unless otherwise indicated, items for each scale were rated on a 5-point Likert scale (1 = disagree strongly to 5 = agree strongly). Alpha coefficients of internal consistency were satisfactory for all scales (see Table 2).

Multidimensional Measures of Identity
We used Wilson and Leaper's (2016) multidimensional measure of ethnic-racial identity. Their measure consisted of facets of ethnic-racial identity including centrality, ingroup affect, ingroup ties, felt typicality, and felt conformity pressure. Identity dimensions correlated with expected outcomes of well-being, including self-esteem. Sub-scales utilized in the present analyses included ethnic-racial typicality (6 items; e.g., "I feel like I'm just like all the other people within my ethnic group"; α = 0.70) and ingroup ties (4 items; e.g., "I have a lot in common with other people within my ethnic group"; α = 0.74).

Reported Ethnic-Racial Outgroup Discrimination
Self-reported ethnic-racial outgroup discrimination was measured using an adapted version of Schmitt et al.'s (2002) measure of gender discrimination. This measure includes three subscales that were adapted to assess ethnic-racial discrimination. Recent studies have similarly adapted Schmitt et al.'s (2002) measure to assess discrimination based on various social identities (e.g., Norman & Chen, 2020;Saleem & Ramasubramanian, 2019). For example, in support of theoretical expectations, Norman and Chen (2020) found that their adapted measure of ethnic-racial discrimination predicted racial-ethnic identity-and thereby provided evidence for its' construct validity. For the current analyses, we focused on participants' responses to past experiences with ethnic-racial discrimination to measure outgroup discrimination (6 items; e.g., "Prejudice against my ethnic group has affected me personally"; α = 0.90). Because our measure of outgroup discrimination was an adapted measure, we ran an exploratory factor analysis to evaluate its structural validity. We entered all 6 items of the outgroup discrimination measure in an exploratory factor analysis using principal axis factor extraction with direct oblimin rotation. The analysis suggested a 1-factor model and revealed satisfactory factor loadings all above 0.71. See Supplementary Materials for all question items and factor loadings.

Reported Ethnic-Racial Ingroup Marginalization
We created a measure of self-reported ingroup marginalization to evaluate participants' experiences of being targeted for appearances or behaviors that do not conform to ethnic-racial ingroup norms. The development of the 8-item scale was based on a review of the literature on ingroup marginalization and the "acting White" accusation (e.g., Bergin & Cooks, 2002;Carter, 2006). We included items to reflect ingroup marginalization experiences reported by participants in empirical research on this topic including behavioral expectations (e.g., "I have been criticized by people within my own ethnic group for the way I speak"), speech styles (e.g., "I have been criticized by people within my own ethnic group for the way I speak"), and skin tone (e.g., "I have felt discriminated against from other people within my ethnic group for my skin tone [e.g., having light or dark skin]"). Graduate and undergraduate students in the second author's lab reviewed the items for understanding, clarity, and grammar. See Supplementary Materials for all question items. The measure yielded satisfactory internal consistency (α = 0.88).
As described in the Results, the measure was negatively related to self-esteem, which provides support to its' theoretical construct validity. We also ran an additional exploratory factor analysis to test the structural validity of our new ingroup marginalization measure. We entered all 8 items in the exploratory factor analysis using principal axis factor with direct oblimin rotation. The analysis revealed a 1-factor model and indicated satisfactory factor loadings above 0.60. See Supplementary Materials for factor loadings for each item.

Results
We first describe preliminary analyses that were conducted to check for bivariate correlations and group differences in the investigated measures. We next summarize the results from the hierarchical regressions performed to test our hypotheses. Prior to conducting all statistical tests, the measures for self-esteem, ingroup marginalization, outgroup discrimination, and ethnic-racial identity facets were converted to standardized z scores. (For ease of interpretation, the nonstandardized scores are provided in the text and in Table 2 with the descriptive statistics.).

Descriptive Statistics and Bivariate Correlations
Descriptive statistics and bivariate zero-order correlations among the key variables are presented in Table 2. In the bivariate correlations, participants' age was positively correlated with self-esteem, ingroup ties, and reported outgroup discrimination. Self-esteem was positively correlated with typicality and ingroup ties; and it was negatively related to ingroup marginalization and outgroup discrimination. Ingroup marginalization was significantly associated with both identity measures, whereas outgroup discrimination was significantly correlated only with ingroup ties. Ingroup marginalization and outgroup discrimination were moderately associated.
For exploratory purposes, we additionally ran these correlations separately for the Black and Latinx samples. We subsequently performed comparison analyses by converting correlation coefficients for each ethnic-racial group to Fisher z values, and then testing differences between the Z values (see Supplementary materials). One significant group difference was indicated in the associations between self-esteem and the predictor variables: Self-esteem and age were more strongly related among the Latinx than the Black participants (z = 2.30, p < 0.05). In addition, there was a significant group difference in the correlation between two predictor variables. A negative association between ingroup ties and outgroup discrimination occurred only among the Black participants (z = 2.22, p < 0.05).

Comparing Relations of Ingroup Marginalization and Outgroup Discrimination to Self-esteem
In our first hypothesis, we predicted a stronger negative correlation between ingroup marginalization and self-esteem than between outgroup discrimination and self-esteem. In support of our hypothesis, a correlation slope analysis (Steiger, 1980)

Testing for Moderation in Hierarchical Regression Analysis
To test whether identity measures moderated any associations between ingroup marginalization and outgroup discrimination to self-esteem, we performed a hierarchical regression analysis. We controlled for age and gender by entering them in the first step. We considered this advisable given the wide age range of our sample, the significant age difference between the Black and Latinx samples, and the positive correlation between age and self-esteem. In the second step, experiences with ingroup marginalization and outgroup discrimination were added. The two identity dimensions-ethnic-racial typicality and ingroup tieswere entered in the third step. In the third step, Typicality x Ingroup Marginalization, Typicality x Outgroup Discrimination, Ingroup Ties x Ingroup Marginalization, and Ingroup Ties x Outgroup Discrimination interactions were entered to consider possible moderation. Finally, because preliminary analyses revealed the association between ingroup ties and outgroup discrimination differed for Black and Latinx participants, we entered a 3-way Ethnic-Racial Group x Ingroup Ties x Outgroup Discrimination interaction in a fifth step for exploratory purposes. The latter step did not add significantly to the model, F(1, 394) = 1.15, p = 0.284.
As summarized in Table 3, each step in the regression added significantly to the model. Collinearity diagnostics revealed low VIF values (VIF < 3) for all entered predictors at each step. At Step 2, both ingroup marginalization and outgroup discrimination were negatively related to selfesteem. When typicality and ingroup ties were entered in the third step, both identity measures were significantly and positively associated with self-esteem. However, outgroup discrimination remained significant, while ingroup marginalization was not significant. Finally, when the interaction Table 3 Hierarchical regression with self-esteem Collinearity diagnostics revealed low VIF values for all predictors at each step (VIF < 2 in Steps 1-3 and VIF < 3 in Step 4). Because preliminary analyses revealed the association between ingroup ties and outgroup discrimination differed for Black and Latinx participants, we entered a 3-way Ethnic-Racial Group x Ingroup Ties x Outgroup Discrimination interaction in a fourth step for exploratory purposes. The latter step did not add significantly to the model, F(1, 394) = 1.15, p = .284 N = 407. IGM = Ingroup marginalization, OGD = Outgroup discrimination, Typ = Typicality, IG Ties = Ingroup ties *p < .05, ** p < .01, *** p < .001 Step 1 Step 2 Step 3 Step 4 To test the interaction, we compared the simple slopes of participants who scored high (+ 1 SD), average, and low (− 1 SD) on typicality (while controlling for age). The analyses indicated that ingroup marginalization was negatively related to self-esteem particularly among those who scored high in ethnic-racial typicality, t(404) = − 3.97, p < 0.001, or average in ethnic-racial typicality, t(404) = -3 .39, p = 0.001. In contrast, ingroup marginalization appeared unrelated to esteem among those low in ethnic-racial typicality, t(404) = − 1.01, p = 0.313. As depicted in Fig. 1, the association was especially strong for the high typicality group.

Summary
In support of Hypothesis 1, we found that the correlation between ingroup marginalization and self-esteem was significantly stronger than the correlation between outgroup discrimination and self-esteem. In partial support of Hypothesis 2, felt ethnic-racial typicality moderated the negative association between ingroup marginalization and self-esteemwhereby the association was stronger when individuals felt higher ethnic-racial typicality. However, ingroup ties did not moderate this association. There was also indication that the ethnic-racial identity measures mitigated the potential impact of ingroup marginalization on self-esteem inasmuch that ingroup marginalization was no longer significant once these identity measures were entered into the model. Contrary to Hypothesis 3, neither of the ethnic-racial identity measures moderated the association between outgroup discrimination and esteem.

Discussion
Our study investigated the relations among ethnic-racial identity, outgroup discrimination, ingroup marginalization, and self-esteem in Black and Latinx undergraduates. We were guided by social identity theory and intergroup theory to test our hypotheses that ethnic-racial typicality and ingroup ties would moderate the relations of outgroup discrimination and ingroup marginalization to self-esteem. In our subsequent discussion, we first consider the relative relations of outgroup discrimination and ingroup marginalization to self-esteem. We next address the extent that ethnic-racial identity moderated the relations of outgroup discrimination and ingroup marginalization to self-esteem. Then, we review some of the notable patterns among study variables indicated in our preliminary analyses. In the final section, we consider our study's limitations and offer recommendations for future research.

Comparing Relations of Ingroup Marginalization and Outgroup Discrimination to Self-esteem
One of the notable features of our study was its consideration of both outgroup discrimination and ingroup marginalization. Consistent with prior work and our hypotheses, both of these reported experiences were negatively related to selfesteem (e.g., Schmitt et al., 2014;Small & Major, 2019). However, in support of our first hypothesis, ingroup marginalization was more strongly related to lower self-esteem than was outgroup discrimination. This result affirms prior work finding individuals reported more distress after experiencing bias from those within their own group compared to discrimination from outgroup members (Niemann et al., 1999). Consistent with this premise, bioecological systems theory (Bronfenbrenner & Morris, 2006) and related empirical work suggests that spheres of influence nearer to the individual have a more readily detectable effect on outcomes than influences that are more distant (Benner et al., 2018). That is, the institutionalization of systemic racism in the macrosystem is manifested in the microsystems of people's everyday lives (Rogers et al., 2021).
By extension, one interpretation from our findings is that Black and Latinx groups might feel prepared to encounter discrimination from those outside of their group; at the same time, they may expect to seek comfort from ingroup peers and family for protection against the negative effects of those experiences (Neblett et al., 2012). However, when marginalization is felt within one's own group, this may be particularly hurtful or isolating. Indeed, this marginalization occurs from those one might expect support due to shared heritage or background (Greenaway et al., 2016). These findings underscore the importance of addressing ingroup

Outgroup Discrimination
We expected that outgroup discrimination would predict lower self-esteem and that ethnic-racial typicality or ingroup ties would mitigate this association. In support of our hypothesis, perceiving outgroup discrimination was related to lower self-esteem. However, contrary to our predictions, neither of the ethnic-racial identity variables moderated the relationship between outgroup discrimination and self-esteem. Some earlier studies indicated an apparent protective function of ethnic-racial identity when faced with outgroup discrimination (Cobb et al., 2019;Romero & Roberts, 2003;Umaña-Taylor & Updegraff, 2007). However, other studies found ethnic identity was not a buffer against discrimination (Huynh & Fuligni, 2010). Similar to our study, much of the prior work on discrimination included measures using wording that was designed to apply to broad ethnic groups. On the one hand, there are similarities in the oppressions that Black and Latinx people face in US society, which may make this approach appropriate. For example, Rosenbloom and Way (2004) found that Black and Latinx youth reported similar experiences with discrimination during police interactions, while shopping, and from teachers. On the other hand, there are some important differences Torres et al., 2010). For example, stereotypes about Black people are more likely to be based on attributes such as criminality or attitudes, while stereotypes about Latinx people are more likely to include assumptions about immigration status or occupations (Ghavami & Peplau, 2013). It could be that our measure of discrimination was too broad to ascertain the ways ethnic-racial identity can protect against specific discriminatory experiences that are unique to each group.

Ingroup Marginalization
We hypothesized that ingroup marginalization would be related to lower self-esteem and that either felt typicality or ingroup ties would augment this negative relation. This was found with felt typicality but not with ingroup ties. Ingroup marginalization had a particularly strong negative effect on esteem for those with high ethnic-racial typicality. This result is consistent with social identity theory (e.g., Turner et al., 1987) and cognitive consistency or balanced identity theoretical approaches (e.g., Gawronski & Strack, 2012), which posit that increased identification with a group is related to higher self-esteem. It stands to reason that those who strongly identify with their group would feel particularly upset when marginalized from those within their valued ingroup than would those with weaker ingroup identification. To our knowledge, this is the first study to test and document this potential effect.
In addition, ingroup ties and felt typicality may possibly mitigate the relation between ingroup marginalization and self-esteem. Once these two identity variables were added to the regression model, ingroup marginalization was no longer a significant predictor of self-esteem. Even though it did not appear to moderate the relation of ingroup marginalization to self-esteem (via an interaction effect), ingroup ties may have buffered the possible impact of marginalization on esteem. In this regard, individuals who had strong ties to ingroup members may have been less susceptible to those who might marginalize them within their ingroup. Indeed, previous research found that feeling pride with one's ethnicracial identity and a sense of connection to others within one's group appeared to provide a barrier to internalizing biased messages from ingroup peers (Harris-Britt et al., 2007;McCoy & Major, 2003;Yip, 2018).
A question to consider in future research is whether the ingroup members with whom individuals have ties and feel similar are different than the ingroup members from whom they experience marginalization. That is, having a supportive set of friends and family from one's ingroup may help ameliorate the negative effects of feeling marginalized from a different set of ingroup members. Whereas support from one's ethnic-racial ingroup is also known to buffer the effects of outgroup discrimination (e.g., Leach et al., 2008;Umaña-Taylor & Updegraff, 2007), perhaps ingroup support can more easily counter the negative effects of ingroup marginalization.

Correlations Among Study Variables
Exploratory analyses revealed noteworthy relations among study variables. The bivariate correlations indicated a positive correlation between age and reported experiences with outgroup discrimination. This finding supports prior work suggesting perceptions of ethnic-racial discrimination and awareness of difference in treatment may increase over time due to life transitions such as schooling, entering the workforce, and moving residences (Gee et al., 2012;Syed & Azmitia, 2008). Our undergraduate sample were students enrolled at a university with a high percentage of nontraditional students who may have experienced many of these transitions. Some characteristics of nontraditional students include those who either start college after the age of 25 and work full-time while enrolled (Cho, 2019;Choy, 2002). These experiences and life transitions may be related to increased occasions to witness forms of discrimination across a range of diverse contexts-which may heighten awareness of prejudice.

3
Outgroup discrimination and ingroup marginalization were each related to lower ethnic-racial ingroup ties. Additionally, ingroup marginalization was related to lower ethnic-racial felt typicality. Experiencing outgroup discrimination or ingroup marginalization based on a stable group membership may lead some individuals to internalize negative societal messages and hold fewer positive feelings about their own group (Speight, 2007). Future longitudinal or experimental investigation could possibly elucidate the causal nature of this relationship (e.g., Gonzales-Backen et al., 2018).
Finally, ethnic-racial felt typicality and ingroup ties were each positively related to esteem. The findings are consistent with prior research suggesting that group identity has a positive impact on well-being among ethnic-racial minority groups (e.g., Phinney & Alipuria, 1990).

Limitations and Suggestions for Future Research
Some notable strengths of our study include its consideration of two separate dimensions of ethnic-racial identity and testing them as potential moderators of the relations of ingroup marginalization and outgroup discrimination to Black and Latinx undergraduates' self-esteem. Looking ahead, we highlight a few areas for future research that can address some of our study's limitations.
First, we recommend investigating the effects of ingroup marginalization and outgroup discrimination on esteem and identity among those of various gender identities, sexual orientations, ethnic-racial groups, social classes, and other social identities. For instance, those from multiracial backgrounds may straddle between ethnic-racial groups in phenotypic appearance and cultural practices (Giamo et al., 2012;Poston, 1990). As a result, they may be likely to experience greater scrutiny regarding their ethnic-racial identity and belonging (Rockquemore et al., 2009;Franco & O'Brien, 2018). In addition, ingroup marginalization may influence women and men in different ways, specifically concerning issues related to colorism and phenotypic bias (Alexander & Carter, 2022). Because colorism is often rooted in traditional stereotypes of femininity related to beauty and attractiveness, women may face harsher judgments about skin tone than men (Hill, 2002). Relatedly, skin tone is more likely to affect self-esteem for women than for men (Thompson & Keith, 2001). It would be beneficial for future research to investigate the ways in which the intersection of identities may influence experiences with ingroup marginalization and outgroup discrimination.
Second, we recommend considering if and how individuals' identification with an ethnic-racial or cultural outgroup might moderate the impact of discrimination on well-being. Individuals may at times navigate between cultures or identify more closely with experiences and customs of those outside of their ingroup (Cooper et al., 2015;Manzi & Benet-Martinez, 2022). As such, recent research has suggested that felt typicality should be investigated as a dual identity by comparing felt similarity to one's own group as well as an outside group (Martin et al., 2017;Nielson et al., 2020). Our measure of ethnic-racial typicality only evaluated individuals' felt typicality to their own ethnic-racial group. In future research, we suggest separately evaluating individuals' identification to their ethnic-racial group and to other groups (e.g., Josselson & Harway, 2012).
Third, because our findings are correlational, the direction of effect cannot be inferred. For example, it could be that if one's behaviors deviate from group norms, this will attract more ingroup marginalization and conformity pressure from peers. In contrast, ingroup marginalization may cause individuals to believe that they do not fit in with peers and are thus not typical of their group. In a similar way that experiencing outgroup discrimination encourages ethnicidentity exploration (Cross, 1985), experiencing discrimination from ingroup peers may be an impetus to reflect on one's identity and how they fit in with their group.
Finally, we suggest building on the measures and methods that we employed. As mentioned earlier, it would be important to use group-specific measurements of discrimination. For example, a measure of discrimination might include specific experiences that are common among Black and Latinx groups, such as being followed in stores, as well as those that are unique to each group. Also, when assessing experiences with discrimination, it may prove useful to distinguish between overt forms of racism and more subtle microaggressions, as these experiences can have varying effects on well-being (Solorzano et al., 2000;Sue et al., 2007;Yosso et al., 2009). We also expect that qualitative analyses could complement our questionnaire methods. For example, personal narratives could illuminate how specific forms of microaggressions (e.g., invalidation of identity, colorblindness, and assumptions regarding criminality and intelligence) and institutional discrimination uniquely affect identity and well-being (Sue et al., 2007).

Conclusions
The present research highlighted the possible roles of ethnicracial identity on the impact of discrimination on Black and Latinx undergraduates' well-being. Our study was distinctive in its consideration of both ingroup marginalization and outgroup discrimination as well as two dimensions of ethnic-racial identity. Specifically, we found that experiencing bias from one's own group has significant implications for self-esteem, especially among those who feel typical of their group. Accordingly, our findings suggest the importance of feeling ingroup support and belonging among Black and Latinx adults. In promoting positive adjustment among Black and Latinx groups, it is important to take note of the various forms of outgroup discrimination and ingroup marginalization that these groups may face.
Acknowledgements The research was supported by a Cota-Robles Fellowship at the University of California, Santa Cruz and a UHD Organized Research and Creative Activities Award to Antoinette Wilson and a grant from the UCSC Academic Senate Committee on Research to Campbell Leaper. The authors thank Kristin Anderson at the University of Houston-Downtown for her support and assistance in arranging data collection for the study. Timea Farkas is appreciated for her advice during the planning of the study. We also thank Brenda Gutierrez for her thoughtful feedback on an earlier draft of this manuscript. Finally, the authors are grateful to the participants.
Author Contributions AW was primarily responsible for conceiving the study. AW and CL equally contributed to writing the manuscript and conducting the data analyses.
Funding The research was supported by a Cota-Robles Fellowship at the University of California, Santa Cruz, a UHD Organized Research and Creative Activities Award, and a grant from the UCSC Academic Senate Committee on Research.
Data Availability Upon request, the first author will provide information about the study's materials. To ensure participants' privacy, prior IRB approval stipulated that access of data was limited to the researchers.

Conflict of interest
We have no known conflicts of interest to disclose.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.