Introduction

In ethno-racially diverse societies, the processes through which individuals explore and come to understand their ethnic-racial group memberships are key developmental competencies that confer psychological, academic, and interpersonal benefits (Umaña-Taylor & Rivas-Drake, 2021). Ethnic-Racial identity (ERI) development also facilitates resilience for youth from ethnic-racial minoritized backgrounds by boosting efficacy and competence for co** with ethnic-racial stereotypes and discrimination that are pervasive in the context of systemic racism in the U.S. (Neblett et al., 2012). Adolescence and the transition to young adulthood are salient developmental periods for ERI due to cognitive maturation and multiple changes in social and cultural contexts as many youth navigate changes in school, work, family, and peer settings during this time (Umaña-Taylor et al., 2014). Currently, one in four U.S. K-12 public school students are Latino,Footnote 1 and the college enrollment rate for Latino youth increased from 22% in 2000 to 36% in 2018 (Hussar et al., 2020). Though there have been significant advances in theory and research examining the developmental progression of Latino youth’s ERI across the adolescent years (Kiang et al., 2010) and ERI among college students (Syed et al., 2007), limited work has examined change and continuity across the developmental and sociocultural transition from high school to college in support of a developmental lifespan model (Williams et al., 2020). In this study, Latino youth transitioning from high school to college reported on multiple ERI components (exploration, resolution, public regard) across 3 years. Longitudinal multilevel growth modeling was used to examine adolescents’ average ERI change patterns and explore social factors (e.g., Latino heritage culture orientation, gender) theorized to account for individual variability in ERI development across this transition.

Advancing a Lifespan Model of Ethnic-Racial Identity

ERI is a multidimensional construct informed by one’s ethnic heritage (e.g., traditions, values, language, etc.) and racialized experiences as part of a socially constructed group within a specific sociohistorical context (Umaña-Taylor et al., 2014). ERI includes the process through which individuals develop beliefs and attitudes about their ethnic-racial group membership over time through exploration and resolution (Umaña-Taylor et al., 2004), as well as the content or feelings attached to this aspect of identity, such as how much individuals feel that others positively view their ethnic-racial group (i.e., public regard; Sellers et al., 1998). Though much theoretical and empirical work has focused on adolescence, a lifespan approach argues for the significance of broadening the conceptualization of identity formation across transition periods into young adulthood to identify the course of ERI developmental changes and renegotiation over time as contexts change (Williams et al., 2020). Consistent with cultural-ecological models (García-Coll et al., 1996), the lifespan model theorizes that ERI unfolds across development as a function of identity-relevant experiences that prompt individuals to consider and evaluate their ERI, such as family and peer ethnic-racial socialization and ethnic-racial-based discrimination (Williams et al., 2020). Given that ERI, by definition, follows a developmental progression and is context-dependent (Umaña-Taylor et al., 2014), transition periods characterized by multiple changes in sociocultural settings of development and novel identity-relevant experiences (such as the transition from high school to college) offer unique opportunities to examine ERI over time. In addition to the developmental and cultural significance of ERI during adolescence and young adulthood, the longitudinal examination of ERI is important because growth in these developmental competencies has predicted positive outcomes over time, such as higher academic motivation, higher self-esteem, and lower levels of depressive symptoms (Kiang et al., 2013; Rogers-Sirin & Gupta, 2012; Umaña-Taylor et al., 2018).

In 2018, 36% of U.S. Latino youth between the ages of 18 and 24 were enrolled in college, and of Latino students who completed high school, 63% immediately enrolled in college the following fall (Hussar et al., 2020). Like earlier school transitions, the high school to college transition is characterized by contextual changes that prompt identity-relevant experiences (Williams et al., 2020), such as exposure to more diverse peers, discrimination experiences, and involvement in academic and extracurricular activities (Huynh & Fuligni, 2012). Many college-enrolled youth experience additional changes in proximal developmental settings during this time, including living, studying, and working in new spaces. On one hand, the college transition may afford new opportunities for youth to engage in issues of diversity and identity (e.g., ethnic studies curricula; student affinity groups) alongside same- and other-group peers. On the other hand, many college campuses remain majority White settings with a documented legacy of (re)producing White upper/middle class norms and expectations that marginalize the cultural histories, values, and identities of youth of color (Stephens et al., 2012; Yosso et al., 2009). Thus, for developmental and contextual reasons, the transition to college has been theorized to influence ERI developmental pathways into adulthood (Williams et al., 2020). Though prior studies have examined ERI among Latino adolescents (e.g., Umaña-Taylor et al., 2009a) and among Latino college students (e.g., Syed et al., 2007), extant research has tended to reinforce a distinction between these two populations without longitudinal attention to the transition period. Furthermore, studies of Latino college students have rarely examined separate components of the multidimensional construct of ERI over time.

Ethnic-Racial Identity Exploration and Resolution

As an extension of Erikson’s (1968) notions of identity formation, developmental theory indicates that the process of youth exploring their ERI (e.g., thinking/reflecting, learning, speaking with others, attending events and activities) and gaining clarity or resolution about what their ERI means to them are related and separate identity process components that contribute to a more positive overall sense of self, increase academic motivation and engagement, and boost abilities to cope with ethnic-racial stereoty** and discrimination (Umaña-Taylor et al., 2014). Aligning with this developmental theory, longitudinal studies have found that ERI exploration and resolution increase across adolescence (e.g., French et al., 2006), though these average developmental patterns vary with individual and social factors. In a study of Latino adolescents in the U.S. Midwest, ERI exploration and resolution increased in a linear pattern across 4 years of high school for Latina adolescents but not for Latino male adolescents (Umaña-Taylor et al., 2009a). In another study of ethno-racially diverse adolescents in southern California, Latin American adolescents’ ERI exploration increased across 4 years of high school to a greater extent for youth who gained more same ethnic-racial group friends and whose ERI centrality (i.e., importance of ethnic-racial group to one’s self-concept) increased over this period (Kiang et al., 2010). A lifespan approach to ERI suggests that youth continue to explore and reconsider the clarity with which they see their identity as they transition between contexts and negotiate role changes across developmental transitions as part of a lifelong journey (Williams et al., 2020), such as the transition to college and the beginning of the transition to adulthood.

Only a handful of longitudinal studies have examined Latino youth’s ERI developmental trajectories across the college years, mostly indicating support for increases in exploration during this time. Ethier and Deaux (1994) interviewed Latino students at highly selective private universities on three occasions across the first year of college, finding that students who initially identified to a greater extent with their ethnic heritage group became involved in cultural activities that increased ethnic identification, whereas students who initially identified to a lesser extent in terms of ethnic heritage perceived more threat in the college environment and decreased ethnic identification over time. In a study of Latino students attending public universities in California, ERI exploration and commitment (i.e., positive feelings and sense of belonging to one’s group) did not increase, but cluster analysis revealed that more students moved from moratorium (i.e., higher exploration and lower commitment) to achieved (i.e., higher exploration and higher commitment) ERI status rather than the reverse (Syed et al., 2007). In an ethno-racially diverse sample of students (23% Latino) attending a public university in California, ERI exploration and commitment increased across 3 years of college, with exploration showing a curvilinear pattern of less initial growth followed by greater increases in the third year (Syed & Azmitia, 2009). In contrast, a study of ethno-racially diverse students of color (7% Latino) at a public university in the Midwest found the reverse; ERI exploration showed greater initial growth followed by less of an increase across 2 years of college (Zhou et al., 2019). Tsai and Fuligni (2012) conducted one of the few studies to assess adolescents’ ERI across the transition from high school to college in a sample of Asian, European, and Latin American students, finding that exploration decreased across 2 years to a greater extent for students attending 2-year compared to 4-year institutions; ERI belonging/commitment did not change. Building on this literature, ERI exploration and resolution were assessed in the current study as distinct developmental processes on five occasions from the final year of high school across 3 years in a sample of Latino students who were recruited from 92 high schools and planning to enroll in a 4-year public university in the U.S. Southwest, reflecting a sample and context that have not yet been a focus of longitudinal ERI studies.

Ethnic-Racial Identity Public Regard

Grounded in social identity theory (Tajfel & Turner, 1986) and the multidimensional model of racial identity (Sellers et al., 1998), the ERI content dimension of public regard refers to how much individuals feel that others positively view their own ethnic-racial group. Consistent with the notion of positive group distinctiveness, in which individuals are motivated to compare their social groups positively with other groups to maintain a positive self-concept (Tajfel & Turner, 1986), studies have shown that adolescents with higher ERI public regard report fewer mental health problems and higher academic adjustment (see Rivas-Drake et al., 2014, for review). ERI public regard is shaped by socio-contextual factors. For example, youth who experience more ethnic-racial-based discrimination may develop less positive feelings about how others view their group as they try to reconcile their identities and the realities of the unfair treatment they experience. In studies of ethno-racially diverse adolescents (including subgroup analyses for Latino youth), ethnic-racial discrimination from adults at school and from peers has been negatively associated with ERI public regard in middle school (Rivas-Drake et al., 2009) and high school (Douglass & Umaña-Taylor, 2017).

As Latino youth transition to college, many experience contextual shifts in their academic settings, including the structure, size, and ethnic-racial compositions of their classrooms. In a cross-sectional study, Latino students attending a highly selective private university in the U.S. Northeast who felt their teachers/professors held a more positive view of their ethnic-racial group also reported lower depressive symptoms and higher self-esteem (Rivas-Drake, 2011a). Furthermore, ERI public regard from teachers/professors mediated the negative association between parents’ ethnic-racial socialization practice of preparing their children for bias and adolescents’ psychosocial adjustment, indicating that Latino youth well-being is shaped by the extent to which youth internalize messages they receive about their ethnic-racial group membership from family and college professors. There has been limited longitudinal research examining developmental trajectories of ERI public regard. In the sample described above (Tsai & Fuligni, 2012), students’ perceived devaluation by society (i.e., reverse coding a measure of ERI public regard) increased from the final year of high school across 4 years (Huynh & Fuligni, 2012), suggesting this transition period may include changes in ERI public regard. Extending this prior research, the current study of Latino adolescents transitioning to college included longitudinal examination of ERI public regard on five occasions across 3 years.

Variability in Ethnic-Racial Identity Development by Heritage Culture Orientation and Gender

Latino youth ERI formation across the transition to college is situated in the context of family cultural traditions, values, and practices, including the enculturation processes of maintaining ties to the family’s heritage culture and the acculturation processes of adopting mainstream U.S. cultural expectations and ways of being (Gonzales et al., 2015). Immigrant youth and youth with immigrant parents and/or grandparents develop their ERI in the U.S. context with proximal connections to their family heritage cultural traditions and values, and these aspects of enculturation serve as cultural strengths supporting ERI exploration and resolution (García Coll et al., 1996; Umaña-Taylor et al., 2009b). In line with this model, aspects of orientation to Latino heritage culture such as Spanish language use and family ethnic-racial socialization have emerged as key influencers of ERI developmental processes in qualitative research with Latino college students (Torres, 2003) and have been positively associated with ERI exploration and resolution in longitudinal studies of Latino adolescents (Umaña-Taylor et al., 2009b) and college students (Syed et al., 2007). Following this research, and with attention to resilience models for immigrant-origin youth (Suárez-Orozco et al., 2018), the current study examined variability in ERI levels at the end of high school and trajectories across the college transition by Latino heritage culture orientation, while controlling for immigrant generation status.

Adolescents’ ERI development occurs in concert with other social identity dimensions, including gender (García Coll et al., 1996). As originally defined by Crenshaw (1989) and applied to psychological theory (e.g., Juan et al., 2016), intersectionality is a framework that centers the identities and experiences of women of color who face marginalization in society on the basis of their ethnic-racial background and gender, rendering identity studies incomplete without attention to intersecting systems of power (i.e., racism, sexism) that together shape multiple social identity domains. Recent guidelines for multicultural psychology research from the American Psychological Association encourage researchers to consider how developmental and contextual antecedents of identity can be acknowledged and addressed from an intersectionality perspective (Clauss-Ehlers et al., 2019), such as attending to ethnicity-race and gender. For all youth, and specifically for Latino youth from families with immigrant backgrounds, gender is an organizing feature of adolescent development that gains salience as youth navigate more contexts outside the home, negotiate intergenerational cultural differences within the family, and gain greater awareness of collective identities, including ethnicity-race and gender (Schroeder & Bámaca-Colbert, 2019).

Some longitudinal studies of Latino adolescents have shown that ERI exploration and resolution developmental trajectories vary by gender (Umaña-Taylor et al., 2009a), likely due to gendered differences in family ethnic-racial socialization emphasizing expectations for women to maintain cultural traditions (Umaña-Taylor & Guimond, 2010). For example, ERI exploration and resolution increased for Latina girls and did not increase for Latino boys across the high school years (Umaña-Taylor et al., 2009a), and family ethnic-racial socialization was associated with ERI to a greater extent for girls than for boys (Umaña-Taylor & Guimond, 2010). The authors suggested that earlier advances in ERI developmental processes among girls may be due to the salience of gender role expectations for carrying on cultural traditions and engaging in other aspects of cultural socialization, which are not uniform by gender in traditional Latino culture (Umaña-Taylor et al., 2009a). In contrast, other studies of ethno-racially diverse high school and college students have not found ERI developmental process differences by gender (Kiang et al., 2010; Syed & Azmitia, 2009; Zhou et al., 2019), which is consistent with notions from ERI developmental theory suggesting that exploration and resolution are universal developmental processes expected to progress over time for all youth (Umaña-Taylor et al., 2014). Prior research on public regard among Latino youth has resulted in mixed findings concerning gender, though longitudinal work is limited in this area. For example, in a longitudinal study with two measurement occasions, Huynh and Fuligni (2012) found that Latino and Asian American students’ societal devaluation (i.e., reverse scoring public regard) did not differ by gender across the college transition. In a cross-sectional study of Latino, Black, and White high school students, Douglass and Umaña-Taylor (2017) found that male youth reported higher ERI public regard than female youth. Given the paucity of research incorporating intersectionality frameworks into longitudinal studies of ERI (Rogers et al., 2020), examining potential gendered patterns of ERI public regard changes from high school into college is important based on consequential differences in the type of ethnic-racial stereotypes and discrimination experienced by male and female Latino youth (Umaña-Taylor & Guimond, 2010). These differences in lived experiences may inform gendered pathways of ERI public regard across the developmental and contextual changes of the college transition that have not been possible to examine with the study designs and methods of extant research.

Current Study

The current study examined Latino adolescents’ ERI trajectories from their final year in high school to 3 years later. This study helped to fill a gap in the literature because extant studies have examined ERI development separately across adolescence and across years of college, respectively, without focusing on change and continuity across the transition period from high school to college. This is important because the adaptation to college is a time characterized by multiple social and cultural contextual shifts that engender significant identity-relevant experiences for continued exploration, formation, and redefinition. The goals of this study were to examine whether and how ERI changed across this transition and to explore variability in ERI trajectories by Latino heritage culture orientation and gender. It was expected that ERI exploration would increase across this period, based on increasing developmental capacities for identity growth and context changes that heighten the salience of ethnic-racial heritage exploration during this contextual transition. Average increases in resolution were not expected across the transition to college. It was theorized that ERI public regard from academic instructors would change across the college transition. ERI public regard may increase due to new academic and social opportunities for engaging with issues of diversity and identity that center Latino cultural heritage and issues of race, thereby boosting students’ internalization of positive views of Latinos in the academic domain. On the other hand, ERI public regard may decrease across this period due to exposure to and greater awareness of stereotypes and the realities of discrimination that impact Latino students in higher education, thereby leading to internalization of negative societal views of Latino students. Regarding individual variability in these change patterns, it was expected that youth with higher orientation to Latino heritage culture would report higher ERI exploration and resolution prior to college compared to youth with lower heritage culture orientation, due to elements of the enculturation process that maintain ethnic-racial heritage practices. ERI exploration and resolution change patterns were not expected to vary by gender. It was expected that change patterns in public regard may differ by gender based on an intersectionality framework for considering the ways in which adolescents’ identity development is embedded within and informed by multiple systems of oppression (i.e., racism, sexism). Due to limited available longitudinal studies on public regard across this transition period, no directional hypotheses were specified regarding gender variability in rates of change; public regard was expected to be higher for male relative to female youth prior to college.

Methods

Participants

As part of a longitudinal study examining Latino youth’s experiences across the transition to college, 209 participants were recruited during their senior year of high school prior to enrolling at a large southwestern U.S. university (Doane et al., 2018). Participants were recruited through orientation sessions, e-mail, text messages, phone calls, university and community partnerships, and word of mouth, during which English/Spanish bilingual staff were available to answer questions from eligible participants and caregivers. Eligibility to participate included being a senior in high school, being accepted to the university, having paid an initial financial deposit or having selected to defer payment, self-identifying as Latino/Hispanic, and living within 60 miles of the university during their senior year in high school. Out of 239 youth who consented to the study, 209 (87.4%) participated in study procedures during the first wave of data collection (5.9% were excluded based on eligibility criteria and 6.7% did not respond to scheduling requests after initial consent). Three participants withdrew from the study during the longitudinal follow-up assessments, resulting in 206 participants included in the current analyses.

In this analytic sample (N = 206; Mage = 18.10 years, SD = 0.41; 65.0% female), 84.5% of the sample identified as being of Mexican descent; 8.7% as South or Central American; 5.3% Cuban; 2.4% Puerto Rican; and 2.4% other Latino(a)/Hispanic descent. With respect to immigration generation, 10.2% of participants were first-generation immigrants (i.e., born outside the U.S.), 62.0% second generation (i.e., at least one parent born outside the U.S.), and 28.2% third generation or above (i.e., both parents were born in the U.S.). The majority of participants (68.8%) were first-generation college students (i.e., neither parent completed a bachelor’s degree). See Doane et al. (2018), for additional demographic information. Participants were recruited from 92 high schools from the surrounding metropolitan area. Based on university institutional data identifying participants’ high schools and publicly available data from the National Center on Education Statistics [NCES] (2016), high schools represented in the sample ranged from 6 to 96% Latino student enrollment (M = 53%; SD = 26%; see Park et al., 2021; Sladek et al., 2020). At the time of participants’ college enrollment (Fall 2017), the focal institution included 49.6% White (non-Hispanic/Latino) undergraduate students and 22.8% Hispanic/Latino undergraduate students. Data were collected from the present sample through 2020, when the university undergraduate population was 46.6% White and 26.2% Hispanic/Latino (Arizona State University, 2021). As of 2022, the focal university was designated a Hispanic-serving institution (HSI), a federal designation for higher education institutions with at least 25% Hispanic enrollment for undergraduate full-time students that confers opportunities for grants to expand educational opportunities for Hispanic students (NCES, 2018). Throughout the data collection period, the institution’s focus on serving Hispanic/Latino students is reflected in a number of cultural, financial, and pre-professional enrichment opportunities available to students on campus (Arizona State University, 2022).

Procedure

The university Institutional Review Board approved all procedures. All participants provided consent/assent and a parent/guardian provided consent for participants younger than 18. Study consent forms and communication with research staff were offered in English and/or Spanish, based on family preference. All surveys were completed in English by youth participants. Participants completed surveys online either in their home or in a university lab based on their preference and were adequately financially compensated for their participation at each wave as approved by the institutional review board. Participants completed the first wave of survey responses prior to their college attendance, December 2016 – July 2017 (T1). Questionnaires were then administered online each semester thereafter, including: September – December 2017 (T2; M = 173 days since T1, SD = 46), January – July 2018 (T3; M = 141 days since T2, SD = 33), and September – December 2018 (T4; M = 224 days since T3, SD = 38). Although ERI measures were not collected in questionnaires administered January – July 2019 (T5; M = 144 days since T4, SD = 31), and September – December 2019 (T6; M = 173 days since T5, SD = 39), the measures were included in questionnaires administered January – July 2020 (T7; M = 156 days since T6, SD = 28). Over half (56%, n = 108) of T7 questionnaires were completed prior to transitioning to online learning (March 12, 2020) due to the COVID-19 pandemic. Sixteen (8%) T7 questionnaires were completed after the murder of George Floyd, an event with documented implications for ethnic-racial socialization that impacted many young adults’ perceptions of systemic forces that may influence ERI (Howard et al., 2022). Independent samples t-tests indicated that none of the study’s outcome variables significantly differed before or after these two events (ps > .05).

Of the 206 participants in the analytic sample, retention was strong at T2 (89.3%), T3 (94.7%), T4 (82.5%), T5 (86.4%), T6 (82.0%), and T7 (93.2%). Attrition across time was not significantly associated with participant’s gender, ps > .220, Latino heritage culture orientation, ps > .217, immigration generation, ps > .460, parents’ education level, ps > .110, or T1 ERI measures, ps > .070.Footnote 2 Across the study period, most participants reported that they remained enrolled as students in the focal university (T2: 94.7%; T3: 91.7%; T4: 83.0%; T7: 75.2%) or enrolled in another college or university (T2: 3.4%; T3: 4.9%; T4: 7.3%; T7: 9.3%), as classified based on self-report and cross-verified with institutional records. ERI measures did not significantly differ by college enrollment status within each wave, ps > .150.

Measures

Ethnic-Racial identity (ERI)

Participants completed the Ethnic Identity Scale - Brief (EIS-B; Douglass & Umaña-Taylor, 2015) to assess ERI exploration (3 items; e.g., “I have attended events that have helped me learn more about my ethnicity”) and resolution (3 items; e.g., “I am clear about what my ethnicity means to me”). The EIS-B has demonstrated reliability and validity in samples of Latino adolescents and university students (Douglass & Umaña-Taylor, 2015). Likert-scale responses range from 1 (does not describe me at all) to 4 (describes me very well). Means were calculated for each subscale, with higher scores indicating higher ERI exploration and resolution. Cronbach’s alphas for exploration and resolution were high across all waves (αs > .83; Table 1).

Table 1 Bivariate correlations and descriptive statistics
Full size table

ERI public regard was assessed with the Public Regard scale (Rivas-Drake, 2011b), based on Sellers et al.’s (1998) conceptualization of public regard as one’s perception of the extent to which society views one’s ethnic-racial group positively or negatively. As figures of authority in classroom settings are institutional agents through which ERI public regard is communicated (Rivas-Drake, 2011b), participants reported on their ERI public regard based on how their teachers (T1) and professors (T2, T3, T4, T7) viewed their ethnic-racial group. Participants responded to three items (e.g., “Some [teachers/professors] don’t expect my ethnic group to do well in life”) on a Likert scale of 1 (strongly disagree) to 5 (strongly agree). All items were reverse scored, and the mean was calculated across the items, such that higher values indicated higher ERI public regard. A version of this measure that referred just to “teachers” has been used reliably with Latino adolescents (Rivas-Drake, 2011b), and the modification in this study to include “professors” is consistent with the academic setting of participants in the current sample after T1. Cronbach’s alphas were high across all waves (αs > .92; Table 1). Analyses supporting the longitudinal measurement invariance of this modified version of the measure are presented in the Results section.

Latino heritage culture orientation

At T1, participants completed a subscale of the Acculturation Rating Scale for Mexican Americans – II (ARSMA-II; Cuellar et al., 1995), which was modified to be inclusive of all Latinos/Hispanics (e.g., “I associate with Mexicans and/or Mexican Americans” was reworded to “I associate with Latinos/Hispanics”). Instructions included acknowledging that Latino/Hispanic culture includes many different groups, such as Mexicans or Mexican Americans, Puerto Ricans, and others; if participants belonged to a specific Latino/Hispanic group, they were asked to think of this group when they answered the questions. Participants reported the extent to which they identify and are familiar with Latino/Hispanic people and culture, as well as how frequently they engage in behaviors and cultural practices associated with Latino heritage culture using a fully-anchored scale from 1 (not at all) to 5 (extremely often or almost always); e.g., “My friends, while I was growing up, were Latino or Hispanic.” “My family cooks foods from a Latino/Hispanic country (e.g., Mexican food, Cuban food.”). The mean of 16 items was calculated, and higher scores on Latino Orientation reflect a greater degree of enculturation and alignment with Latino heritage culture behaviors and practices. Cronbach’s alpha was high (α = .91)

Sociodemographic variables

Participants self-reported demographic characteristics at T1, including gender (0 = female, 65.0% of the sample, and 1 = male, 34.5% of the sample). Covariates included immigration generation (based on parents’ and grandparents’ nativity information, ranging from 0 = youth, parents and grandparents born outside of U.S. to 7 = youth, parents, and all grandparents born in U.S; Umaña-Taylor et al., 2009b) and parent education level (mean of mother and father education level, 1 = less than high school to 10 = Doctorate or other advanced degree).

Analytic Strategy

First, descriptive statistics were reviewed and all data were screened for multivariate outliers. Multigroup confirmatory factor analysis (CFA) models were fit to test longitudinal measurement invariance of ERI measures across 5 waves to ensure that measurement occasions were comparable. Following standard practice (Widaman et al., 2010), models were fit for each ERI measure (e.g., exploration, resolution, public regard) with sequentially stricter constraints to test increasing levels of invariance, starting with a multigroup CFA model with all parameters freely estimated across time to test the equivalence of factor structures over time (i.e., configural invariance). Support for configural invariance was evaluated based on the consideration of multiple SEM model fit indices, with acceptable fit indicated by standard guidelines: CFI > .95, RMSEA < .08, and SRMR < .08 (Hu & Bentler, 1999). Next, factor loadings (i.e., metric invariance) and item intercepts (i.e., scalar invariance, also known as strong invariance) were constrained to be equal across time. Support for metric and scalar invariance was evaluated based on recommendations to consider change in CFI between nested models < 0.01 indicating invariance (i.e., longitudinal measurement equivalence; Putnick & Bornstein, 2016).

In a multilevel modeling framework, two-level growth models were used to address longitudinal measurement occasions (Level 1) nested within individuals (Level 2; analytic N = 946 observations for 206 persons). First, data were visually plotted to examine average change patterns and variability in change patterns over time for each outcome (i.e., ERI exploration, resolution, and public regard). Next, unconditional multilevel models with random intercepts and no predictors were fit for each outcome. Growth parameters were then introduced to the models, starting with a linear slope term, to test whether there was an average ERI change pattern over time. Based on variability in participants’ survey completion dates within each study wave, and the advantages of modeling developmental trajectories with time treated as a continuous variable (Grimm et al., 2016), ERI linear growth trajectories were modeled at Level 1 by including a wave- and person-specific time variable indicating the number of months participants completed each survey since the first date of T1 survey completion (e.g., 4.5 = four and a half months since the start of the study). Thus, differences between participants in the length of time between survey assessments across time were accounted for in the modeling approach and this centering strategy results in the model intercept representing average levels of the outcome at participant-specific T1 survey date. Along with visual inspection of the data and evaluating statistical significance of random effects estimates, likelihood ratio chi-square difference (i.e., nested model) tests were used to test the significance of growth parameters in sequential fashion (e.g., linear, quadratic, cubic). After the most appropriate growth model was determined for each outcome, random slope terms were introduced to account for between-person differences in growth trajectories (Snijders & Bosker, 2011). Finally, person-level factors (e.g., gender, Latino heritage culture orientation) theorized to be related to differences in ERI prior to college and change in ERI over time were entered as Level 2 predictors of intercepts (i.e., T1 ERI, prior to college transition) and slopes (i.e., ERI trajectories over time), controlling for immigration generation and parent education level. All models were fit in Mplus version 8 (Muthén & Muthén, 1998-2017) using full information maximum likelihood estimation with robust standard errors.

Results

See Table 1 for descriptive statistics and bivariate correlations. Significant positive correlations across time indicated that ERI exploration (rs: .41 – .57, ps < .05), ERI resolution (rs: .56 – .68, ps < .05), and ERI public regard (rs: .35 – .42, ps < .05) were moderately stable. Latino heritage culture orientation was positively correlated with ERI exploration (rs: .34 – .49, ps < .05) and resolution (rs: .30 – .45, ps < .05) from T1 to T7 and negatively correlated with ERI public regard (rs: −.26 – −.17, ps < .05) from T1 to T7, with the exception of T2 (r = −.07, p > .05). Having more U.S.-born family members was correlated with lower ERI exploration (rs: −.25 – −.17, ps < .05) and ERI resolution (rs: −.23 – −.16, ps < .05). Compared to female youth, male youth reported lower ERI exploration from T3 to T7 (rs: −.24 – −.17, ps < .05) and higher ERI public regard at T1, r = .21, p < .05. Multigroup CFA models demonstrated support for scalar invariance (i.e., equivalence of factor structures, factor loadings, and item intercepts) across time, supporting longitudinal measurement equivalence for all ERI measures (Table 2).

Table 2 Model comparisons for longitudinal measurement invariance of ethnic-racial identity scales
Full size table

Ethnic-Racial Identity Exploration

Regarding the average change pattern for ERI exploration, a linear growth model fit the data significantly better than an unconditional model that posits no changes, χ2(1) = 6.170, p = .013. As adding a quadratic growth term did not significantly improve fit relative to the linear model, χ2(1) = 1.942, p = .163, model building proceeded with a linear growth model. On average, ERI exploration positively increased over time, and there were significant individual differences in this rate of increase (i.e., random slope; Table 3; Fig. 1a). T1 Latino heritage culture orientation was positively associated with ERI exploration at T1 prior to the college transition (i.e., model intercept; Table 3). Gender was not significantly associated with ERI exploration prior to college. Regarding between-person variability in linear slopes, increases in ERI exploration across time did not vary by gender, but there was a marginally significant trend indicating that Latino heritage culture orientation accounted for variability in the rate of change in exploration, p = .087. This trend suggested that adolescents with higher orientation to Latino culture prior to college did not increase in exploration over time to the same extent as adolescents with lower orientation to Latino culture. See Fig. 1a for visual plots of ERI exploration over time.

Table 3 Ethnic-Racial identity trajectories from multilevel growth modeling
Full size table
Fig. 1
figure 1

Variability in Ethnic-Racial Identity Development Across College Transition. Note. N = 206. Figure 1a = ERI exploration. Figure 1b = ERI resolution. Figure 1c = ERI public regard. T1 = pre-college. T7 = approximately 3 years later. Estimated marginal means plotted by wave for the sample average (black line with square markers) and by factors predicting significant differences in ethnic-racial identity (gray lines). Error bars are standard errors. Latino heritage culture orientation plotted +1 SD (“High”; 18% of sample) and −1 SD (“Low”; 18% of sample) from mean. 65.0% of participants identified as female; 34.5% of participants identified as male

Full size image

Ethnic-Racial Identity Resolution

Regarding the average change pattern for ERI resolution, a linear growth model did not fit the data significantly better than an unconditional model that posits no changes, χ2(1) = 0.162, p = .687. Visual inspection of the data indicated there was likely a ceiling effect, such that ERI resolution was high at T1 prior to the college transition and remained relatively high over time (see Fig. 1b). Despite no average change, adding a random slope term indicated there were significant individual differences in rates of change (Table 3); thus, model building proceeded with a linear growth model. Similar to ERI exploration, T1 Latino heritage culture orientation was significantly positively associated with ERI resolution at T1 prior to the college transition (Table 3). Gender was not significantly associated with ERI resolution prior to college. Regarding between-person variability in linear slopes, T1 Latino heritage culture orientation significantly predicted variability in the rate of change in resolution, such that adolescents with higher orientation to Latino culture prior to college did not increase in resolution over time to the same extent as adolescents with lower orientation to Latino culture. There was also a marginally significant trend indicating that gender also accounted for some of this slope variability, p = .087, suggesting that male youth did not increase in resolution over time to the same extent as female youth. See Fig. 1b for visual plots of ERI resolution over time.

Ethnic-Racial Identity Public Regard

Turning to the average change pattern for ERI public regard, though a linear growth model did not fit the data significantly better than an unconditional model that posits no changes, χ2(1) = 0.160, p = .689, visual inspection of the data suggested potential non-linear change patterns. A quadratic growth model fit significantly better than a linear model, χ2(1) = 9.014, p = .003, and a cubic growth model fit significantly better than the quadratic model, χ2(1) = 3.892, p = .049. Thus, model building proceeded with a cubic growth model. On average, ERI public regard positively increased over time (i.e., significant linear slope), this linear rate of increase slowed over time (i.e., significant quadratic), and this rate of deceleration in the positive change pattern also changed over time (i.e., significant cubic; see Table 3; Fig. 1c). Adding random effects to account for between-person variability in these growth parameters was a marginal improvement compared to a random intercept only model where rates of change were considered equal across individuals, χ2(9) = 14.844, p = .095. Gender was significantly associated with ERI public regard at T1 prior to the college transition (i.e., model intercept), such that male youth reported higher ERI public regard than female youth (Table 3). A marginally significant trend indicated that the positive linear rate of increase in ERI public regard was lower for male youth than female youth. A post-hoc comparison focusing on the change from T1 to T2 indicated that ERI public regard increased across the college transition to a greater extent for female youth, t(117) = 4.689, p < .001, Cohen’s d = 1.139, compared to male youth, t(57) = 1.648, p = .105, Cohen’s d = 0.983. T1 Latino heritage culture orientation was not significantly associated with ERI public regard (intercept or linear slope). Regarding between-person variability in quadratic and cubic growth functions, changes in the linear increases in ERI public regard across time did not vary by gender or Latino orientation. See Fig. 1c for visual plots of ERI public regard over time.

Supplementary Analyses

Supplementary analyses were conducted to examine the extent to which findings were similar when restricting the sample (N = 206) to youth who identified as Mexican descent (N = 174) and controlling for first-generation college status (i.e., parents have not completed a bachelor’s degree). Possibly due to restriction of statistical power, significance levels for this smaller subsample (N = 174) were not identical to those of the primary results (N = 206), but all growth trajectories and associations were in the same direction. See Table S1 in the online supplementary materials for these results.

Discussion

U.S. youth are growing up in a society that is increasingly ethno-racially diverse and marked by tension and hostility regarding issues of ethnicity-race. This context renders identity formation around the meaning of one’s ethnic-racial background a normative developmental competency with significant implications for the health and adjustment of all youth (Umaña-Taylor & Rivas-Drake, 2021). Latino youth are a growing and diverse portion of the U.S. population who currently comprise 25% of K-12 public school students, and 63% of Latino high school graduates enroll in higher education immediately after high school (Hussar et al., 2020). In these educational systems Latino adolescents face substantial barriers and threats to their development as a function of institutionalized racism (Yosso et al., 2009), while building developmental competencies and forging affirmative communities for successfully navigating a hostile climate (Tsai & Fuligni, 2012). The current study advanced the developmental literature to longitudinally examine multiple components of ethnic-racial identity (ERI) among Latino adolescents across a 3-year period spanning their final year of high school and transitions to a 4-year public university. As hypothesized, prior to college, ERI exploration and resolution were higher for adolescents with a higher orientation to Latino heritage culture. Also consistent with hypotheses, ERI exploration increased in a linear fashion, on average, ERI resolution was consistently high and did not show uniform changes, and there was significant individual variability in both of these ERI developmental process change patterns. The extent to which adolescents believe that their teachers/professors view members of their ethnic-racial group positively (i.e., ERI public regard) increased from high school to the first college semester, a positive change that was more pronounced for female-identifying compared to male-identifying adolescents; these increases in ERI public regard were less prominent over time after the first college year, exemplifying a non-linear change pattern. These models controlled for immigration generation and parent education level. Discussion of these findings focuses on novel contributions of this longitudinal study to context-dependent, lifespan, and intersectional models of ERI development.

For youth from immigrant families, the enculturation process includes family engagement in sharing and maintaining aspects of ethnic-racial heritage, such as values, traditions, and language (Suárez-Orozco et al., 2018). Latino youth with a higher orientation to aspects of their heritage culture may be encouraged to explore their ethnic-racial background through hearing stories at home, attending cultural heritage activities and events as a family in community contexts, and through myriad other ways in which adolescents learn about their ERI from family explicitly (e.g., direct communication and conversations between elders and youth) and implicitly (e.g., music, food, and language at home). Given the positive links between family ethnic-racial socialization and multiple indices of positive adjustment, including ERI development (see Umaña-Taylor & Hill, 2020, for review), this points to a significant strength of Latino youth with a higher orientation to heritage cultural practices prior to the transition to college. The current study joins prior research in supporting this association between heritage culture orientation and ERI development among Latino adolescents (Umaña-Taylor et al., 2009b) and helps to clarify that this pattern emerges prior to the college transition. Research is needed to measure multiple aspects of family ethnic-racial socialization in relation to ERI development to further understand potential shifts in the family role across this transition period. It is important for this research to also attend to the broader sociocultural context in which these processes unfold, such as examining the impact of anti-immigrant policies and legislative discrimination (e.g., Aguirre, 2012) and policies for cultural diversity and inclusion at the university level that may promote and/or inhibit ERI development.

As has been suggested by scholars advocating for a lifespan approach (Williams et al., 2020), and in agreement with prior studies (e.g., Zhou et al., 2019), Latino adolescents’ process of exploring their ERI increased from the final year of high school through the subsequent 3 years. The effect size of this average increase over a 3-year period was approximately 0.20 SD units, which is larger than expected increases in high school-based intervention programs designed to promote adolescents’ ERI exploration, and thereby psychosocial and academic adjustment (e.g., Umaña-Taylor et al., 2018), suggesting the college transition may stimulate considerable opportunities for the exploration process. This exploration process likely gains prominence across this transition period due to students’ active engagement in information-seeking and learning about their ethnic-racial heritage, participating in cultural campus and community activities and events, and the many other identity-relevant experiences that coincide with transitioning to a new academic, social, and cultural context. There were individual differences in this average pattern, and these differences were not related to gender. Though the findings suggested that youth with higher familiarity with and engagement in heritage culture practices prior to college may not increase in ERI exploration over time to the same extent as youth with lower orientation to Latino culture, this finding did not reach conventional levels of statistical significance and should be interpreted with caution. It is possible that developmental increases in ERI exploration into young adulthood reflect a relatively universal process that expands long-held notions of identity search beyond the bounds of adolescence. As prior research has shown a greater decline in ERI exploration over time for students attending 2-year colleges (Tsai & Fuligni, 2012), it is also possible that the transition to a 4-year university, specifically, raises the salience of ethnicity-race for Latino students. For many youth, the greater degree of ethnic-racial diversity on residential college campuses relative to high schools may encourage the process of exploration through interactions with peers. For example, prior research on the transition from middle to high school has suggested that students moving from smaller schools in ethnic-racial enclave neighborhoods to larger and relatively more diverse schools may experience heightened salience of ethnicity-race in their everyday lives that prompts exploration (French et al., 2006). In the present sample, some youth may have experienced this change across the transition to college, but this was likely not the case for all youth in the sample, as there was variability in the ethnic-racial composition of high school contexts represented in the study. Increases in ERI exploration across the initial college years suggest it is vital for universities to commit to efforts that support this developmental process, such as providing curricular offerings in ethnic studies and critical race perspectives, taking steps to diversify faculty and staff, and adequately funding student affinity groups. As studies have shown in K-12 education, promoting ERI exploration has substantial promise for boosting students’ psychosocial and academic adjustment (e.g., Bonilla et al., 2021; Umaña-Taylor et al., 2018).

Average increases in Latino adolescents’ ERI exploration across the 3-year period from the final year of high school through the transition to college were contrasted by no average change pattern in ERI resolution across this same period, though significant individual differences in both patterns highlight diverse pathways of ERI developmental processes over time. ERI resolution was relatively high on average in the final year of high school (M = 3.05, on a 1.00–4.00 scale), which is consistent with average levels reported in other studies of Latino adolescents in high school (range of Ms: 3.27–3.39; Douglass & Umaña-Taylor, 2015; Umaña-Taylor et al., 2009a). Thus, on average, the maintenance of Latino adolescents’ high levels of resolution across the college years might reflect a promotive ceiling effect, whereby many youth establish clarity and understanding about the meaning of their ERI during adolescence and prior to college. This finding is similar to prior results indicating no changes in ERI commitment (i.e., positive feelings and sense of belonging to one’s group) across the college transition (Syed et al., 2007; Tsai & Fuligni, 2012), though this is a different construct assessed with a different measure. ERI resolution is a developmental competency consistently linked with positive indicators of mental health and academic adjustment (Rivas-Drake et al., 2014), suggesting that many Latino adolescents transitioning to college bring with them a significant asset that may confer both promotive and protective stress-buffering effects in multiple adjustment domains.

Despite this average pattern, individual differences in ERI resolution growth trajectories indicated that there was variability among Latino adolescents during this transition, and a portion of this variability was accounted for by Latino heritage culture orientation prior to college and (based on a marginal trend) gender. Specifically, youth with higher familiarity with and engagement in heritage culture practices prior to college did not increase in ERI resolution over time to the same extent as youth with lower orientation to Latino culture, which might suggest that a promotive ceiling effect of high resolution levels for Latino adolescents is most pronounced for those with stronger alignment to the enculturation process. Identifying and being familiar with aspects of Latino culture, as well as practices like using Spanish and participating in traditions and customs with family, may provide youth a foundation that crystallizes the clarity with which they view their ERI across the college years, even as they also continue to explore aspects of their ethnic-racial heritage afforded by this developmental period (Williams et al., 2020). On the other hand, Latino adolescents with less connection to heritage culture practices prior to college may benefit from gains in ERI resolution through exposure to novel exploration opportunities afforded in the college context, such as meeting new peers of the same or of a different ethnic-racial group or college curriculum supporting the consideration of issues of culture, ethnicity, and race (Torres, 2003). Though the findings suggested that female youth may have more gains in resolution over time relative to male youth, this finding did not reach conventional levels of statistical significance and should be interpreted with caution pending further research.

As posited by developmental ERI theory, individuals navigate a journey of identity development that involves renegotiation of meaning-making about the role of ethnicity-race in their lives, ultimately coming to an understanding of their ERI in different ways at different points across the lifespan (Umaña-Taylor et al., 2014; Williams et al., 2020). Together with positive increases in exploration, the variability in the pattern of resolution in this study suggests both continuity and change regarding ERI development across the college transition that exemplifies the multidimensionality and dynamic nature of advancing experience and growth in coming to see the role of ethnicity-race in one’s identity. Specifically, the findings highlight the diversity among Latino youth in developmental trajectories of ERI process components as a function of heritage culture orientation prior to the college transition. From the perspective of identity developmental status (Erikson, 1968), one explanation for the current findings could be that the most common pathway for Latino youth ERI during this time includes (for some) identity achievement, characterized by a coherent and consolidated identity following exploration, and (for others) identity foreclosure, characterized by an established identity not yet corresponding with dedicated exploration. Yet, this remains speculation, as the current study did not examine exploration in relation to resolution. Examining the dynamics of the process components of exploration and resolution in relation to one another, along with other ERI components, may help to clarify what it means for exploration to advance over time while resolution remains stable during the identity-salient years of the college transition. Future research using alternative methods is needed for this endeavor, such as person-centered approaches that account for configurations of ERI developmental status (i.e., achieved, foreclosed, etc.; Syed et al., 2007) and ERI process and content together (Wantchekon & Umaña-Taylor, 2021).

Latino adolescents’ ERI public regard, specifically in relation to academic instructors, generally showed positive changes across the college transition, with variability by gender. As youth develop the belief that others hold more positive views about their ethnic-racial group, they tend to report better mental health and self-esteem (Rivas-Drake, 2011a), suggesting that overall this is a positive pattern observed across the college transition. This study was unique in measuring ERI public regard from teachers in high school and from professors in college over time, with support for the longitudinal equivalence of this measure across the transition. It is likely that differences Latino adolescents experience in their interactions with high school teachers and their college professors contributes to this change pattern. Prior research examining ERI public regard across this transition is extremely limited. In contrast to the current finding, one prior study of ethno-racially diverse adolescents found evidence that perceived devaluation from society (i.e., a measure of ERI public regard reverse coded) increased across the college years with no gender differences (Huynh & Fuligni, 2012). As this prior study was conducted almost a decade before the current study, followed students from three high schools to many different colleges/universities, and did not specifically measure this construct with respect to academic instructors, it is possible that the contrasting results reflect generational changes, college context variation, and/or academic domain-specific processes that should be explored further with study designs capable of testing these possibilities. For example, the focal university is situated within a county that is 31% Latino (U.S. Census Bureau, 2017), a feature of the local context that may shape ERI development in ways that differ from contexts characterized by more or less Latino representation.

Furthermore, grounded in an intersectionality perspective to consider ERI through the lens of gender, female-identifying youth increased more (relative to male-identifying youth) in the belief that their academic instructors positively view their ethnic-racial group. This change was evident in the first college semester. Part of the reason for this observed pattern could be that Latina adolescents received more messages from their high school teachers that either implicitly or explicitly conveyed views that perpetuate sexism and racism, reinforcing traditional gender role expectations and stereotypes for the academic pursuits and possibilities of women of color. For example, prior studies have shown that academic sexism experienced by Latina high school students is linked with less perceived competence in mathematics and science (Brown & Leaper, 2010), and lack of Latina role models among high school educators and counselors contributes to Latina students’ academic concerns and hesitation about career possibilities (Sparks et al., 2021). It is also possible that upon arrival to a university campus, Latina adolescents saw more empowering examples and built supportive connections with university faculty that shaped more positive self-definitions attached to social identities like ERI. For example, in a prior study, Latina students participating in a yearlong STEM fellowship program reported that learning from Latina academic role models, as well as those from other social identity groups who shared experiences of marginalization and stereoty** themselves, contributed to their confidence in pursuing STEM careers and positively defining their identities with respect to intersections of ethnicity-race and gender (Sparks et al., 2021). As national data indicate that more Latino college students are female than male (a trend represented in the present sample; Hussar et al., 2020), this raises important questions about supporting all students’ identity development across this transition in ways that attend to dynamics of ethnicity-race, gender, and their intersections. As prior studies have shown (e.g., Torres, 2003), qualitative research may be uniquely positioned to gain further insights into Latino students’ awareness of systems of power and oppression (e.g., racism, sexism) and how they relate to identity-building processes. An intersectionality lens is important for this work moving forward to account for youth’s gendered and racialized experiences in society as influences on development, such as incorporating more nuanced measures of gender identity constructs into ERI studies.

What are the implications of these findings for practice and policy, and what concrete steps can institutions take to support ERI development for Latino adolescents? From a strengths-based and identity-affirming perspective, educational institutions (both high schools and universities) should consider ways in which they can recognize and learn from Latino students’ identity-related assets to reorganize priorities for policies, curricular offerings, and student support services that engender a greater sense of belonging and boost academic adjustment. In alignment with the contributions of the current findings regarding typical ERI developmental patterns, there are several examples of evidence-based curricular offerings in K-12 education that promote adolescent ERI development and positive adjustment, which could be expanded upon to support youth during the transition to college. First, Umaña-Taylor and colleagues (2018) developed a universal health promotion intervention program, the Identity Project, which includes a manualized 8-session curriculum and training resources for educators to provide adolescents the time and space during the school day with which to explore their ethnic-racial backgrounds and share in and discuss this process with their peers, teachers, family, and community members. A randomized controlled trial of the program in 9th grade showed that increases in adolescents’ ERI exploration as a function of the program (relative to an attention control group) produced increases in ERI resolution, and thereby higher identity cohesion, self-esteem, academic performance, and lower depressive symptoms at 1-year follow up (Umaña-Taylor et al. 2018). To date, there is no formal college-level version of the Identity Project, though efforts are underway to train college students and graduates as facilitators of the program (Juang et al., 2020), suggesting considerable opportunities for future program development, translation, and evaluation in this regard. Second, Bonilla and colleagues (2021) conducted an analysis of the effects of ethnic studies curricula in U.S. K-12 public education, a category of courses that emphasize culturally-relevant and critically-engaged content such as social justice and stereoty**. Using a rigorous regression discontinuity design to analyze school district-level data, results showed that assignment to ethnic studies in 9th grade increased high school graduate rates, attendance, and probability of later college enrollment (Bonilla et al., 2021). Thus, concrete steps that can be put into practice and solidified with policy decisions by university administrators include introducing evidence-based curricula designed to promote adolescents’ ERI exploration and ethnic studies course offerings with adequate support for training and evaluation. Given the current findings showing typical increases in ERI exploration and public regard across the high school to college transition, such decisions would be in concert with recognizing and supporting the lived experiences of Latino adolescents’ ongoing journeys of ERI development as they enroll in higher education. The significant overrepresentation of White educators and faculty in the U.S. (Hussar et al., 2020) means that these efforts will require members of the privileged majority to learn about and reckon with issues of ethnic-racial inequity and how they inform youth development.

Limitations of this study include the inability of this study design to clarify whether changes in ERI were a result of the college transition, of developmental changes, or the interplay of both of these mechanisms of change. Furthermore, data were not collected regarding the specific programs, activities, or curricula in which youth participated during the study, limiting the extent to which this study can evaluate whether the findings may be due to these hypothesized aspects of the college transition. In addition, further research is needed that includes youth across a wider array of educational and career pathways from adolescence into young adulthood. The generalizability of these findings is unknown pending replication in other samples, regions, and school settings. Adolescents were recruited from one metropolitan area through their enrollment in one 4-year university, a design similar to most extant research on ERI across the college years (e.g., Syed et al., 2007; Zhou et al., 2019), with the exception that this study utilized a within-group design to focus on variability among Latino adolescents and included baseline survey measures prior to the college transition. Thus, future research would benefit from alternative designs that allow for examining questions related to university context influences on ERI development, such as following longitudinal samples of high school students over time beyond age 18 and including similar ERI measures across multiple college sites. Also important is the consideration of ethnic-racial composition of high school contexts, such as whether Latino students comprise a numerical minority or majority in their school context and the extent to which this composition changes across the college transition (see Park et al., 2021). Finally, as with all longitudinal research, the patterns observed in this study are subject to the number of measurement occasions and the time intervals between assessments. For example, changes were based on assessing youth approximately every 6 months over a period of 3 years. Additional information may be gained by assessing ERI in context on a more frequent basis (e.g., daily diary methods) or, on the other hand, across a wider developmental window (e.g., longitudinal follow up into adulthood). Additional research with sophisticated methods is needed to detect the complexity and fluidity with which ERI development changes across time and across contexts.

There are a number of future directions for this research based on the current findings. First, future work should consider incorporating measures of discrimination into the study of ERI, following transactional models that posit dynamic interactions between discrimination and identity in sha** the experiences and adjustment of youth of color (Hughes et al., 2016) and prior research examining these questions from a risk and resilience perspective (see Umaña-Taylor & Rivas-Drake, 2021, for review). Prior results from the current sample have identified associations of ethnic-racial discrimination experiences with internalizing symptoms and physiological stress processes across the college transition (Castro et al., 2022; Sladek et al., 2021), and the joint consideration of Latino adolescents’ experiences of ethnic-racial discrimination and ERI may afford unique insights into development and health patterns across time. Second, an important next step includes examining how changes in Latino youth’s ERI relate to other key developmental and health outcomes (e.g., academic adjustment, mental health) across the college transition, following prior research that has established the positive benefits of ERI development across the high school years (e.g., Rogers-Sirin & Gupta, 2012). Finally, as ERI is relevant to all youth, future research should also consider these developmental questions in ethno-racially diverse samples.

Conclusion

The study of ethnic-racial identity (ERI) and its developmental implications are needed to better understand and promote pathways that facilitate co** and resilience for youth from ethnic-racial minoritized groups who face pervasive ethnic-racial-based stressors. This study of U.S. Latino adolescents transitioning to college filled a critical gap in the research literature on ERI development that has tended to reinforce a distinction between contexts before and after high school, thus hel** to advance a lifespan model of ERI development that attends to contextual shifts and transition points with key significance for identity formation. Furthermore, this study addressed other limitations of prior studies by including analyses of multiple ERI dimensions, including exploration, resolution, and public regard, and provided evidence for longitudinal invariance of these measures over time. Change patterns observed in this sample of college-attending Latino youth indicated both consistency and variability in ERI developmental patterns within this diverse group, suggesting that efforts to support youth social identities regarding ethnicity-race will require attention to intersectionality and more than one-size-fits-all approaches. Results from multilevel growth modeling of longitudinal survey data collected before and after the college transition across 3 years showed that ERI exploration increased over time, ERI resolution was high with no change, and public regard from academic instructors increased in a non-linear fashion differentiated by gender. The current study also replicated the results of prior studies showing that Latino heritage culture orientation was positively associated with ERI exploration and resolution prior to college, and extended prior research to show that heritage culture orientation also accounted for part of the variability in resolution change patterns over time. Given the nuance with which adolescents continue to explore and define how they view themselves with respect to their ethnic-racial group memberships beyond the high school years, it is imperative that policymakers (e.g., university administrators, legislators) and practitioners (e.g., high school educators, university faculty) invest in supporting positive ERI development. Investments in supporting healthy ERI development have implications for student mental health and academic success, as well as broader societal consequences to ensure that future generations are equipped to consider, discuss, and address ethnic-racial inequities that constrain opportunities and social progress.