The Everyday Implications of Ethnic-Racial Identity Processes: Exploring Variability in Ethnic-Racial Identity Salience Across Situations

Abstract

Given the social and developmental relevance of ethnicity-race during adolescence, it is important to understand the meaning of ethnic-racial identity in adolescents’ everyday lives. The current study considered how individual differences in ethnic-racial identity exploration (i.e., the extent to which individuals have explored their ethnicity-race), and commitment (i.e., the extent which they have a clear sense of what it means to them) influenced variability versus stability in the awareness of ethnicity-race in a given situation (i.e., salience), and how this variability is related to mood in that situation. Within an ethnic/racially diverse sample of 395 adolescents (M_age = 15; 63 % female; 12 % Black, 26 % Latino, 34 % Asian, 23 % White), results indicated that ethnic-racial identity exploration was unrelated to variability in salience, while commitment promoted stability in salience across situations. Further, among adolescents who were generally very aware of their ethnicity-race, increases in situational salience were related to decreased negative and anxious mood. Among adolescents who were generally not aware of their ethnicity-race, increases in situational salience were related to increased positive and decreased negative mood. Implications for understanding the developmental and everyday experiences of ethnic-racial identity are discussed.

Introduction

As the ethnic and racial diversity of the United States continues to grow, ethnic-racial identity is increasingly recognized as a normative component of youth development (Lee Williams et al. 2012). Ethnic-racial identity is a multi-faceted construct that captures many different aspects of how individuals make personal sense of their own ethnicity-race (Umaña-Taylor et al. 2014). Recent conceptual advances propose that is important to understand how the degree to which adolescents have explored and made a commitment to this aspect of their identity is related to everyday significance of ethnicity-race, and what this means for general psychological functioning (Umaña-Taylor et al. 2014). Empirical work has made inroads in this area, demonstrating that there are mean differences in awareness of ethnic-racial identity based on where youth are in the process of developing their ethnic-racial identity (Yip 2014). Further, this research has indicated that this awareness of ethnic-racial identity is unrelated to subsequent mood (Yip 2014). However, there are many questions that remain unanswered. Is the degree to which individuals fluctuate in their awareness of ethnic-racial identity from moment to moment based on developmental processes of ethnic-racial identity? Do fluctuations in awareness matter for adolescents’ mood? In the current study, we extend previous research by examining (a) the relationship between ethnic-racial identity processes of exploration and commitment and salience variability, and (b) how salience variability is related to adolescents’ mood in a given situation.

Developmental Differences in the Variability of Salience

How individuals make personal meaning of their ethnicity-race is complex and includes many different considerations (Umaña-Taylor et al. 2014). One theoretical perspective, the multidimensional model of racial identity, establishes many different components that capture the meaning and significance that individuals draw from their ethnic-racial group membership (Sellers et al. 1998). Drawing heavily on social identity theory in particular (Tajfel and Turner 1986), this is referred to as a content approach (Umaña-Taylor et al. 2014). The current study focuses on just one of those content components: salience. Salience is defined as the relevance and significance of ethnicity-race in a specific situation. Thus, salience is expected to vary from moment to moment for the same person as a function of the interaction between the person and his/her immediate context (Sellers et al. 1998). Research has confirmed that it changes from one day to the next and one situation to the next (e.g., Yip 2005, 2009; Yip and Fuligni 2002); indeed, research estimates that 40 % of the variability in salience can be attributed to within-person experiences and the immediate context, such as intragroup interactions, that make ethnic-racial identity more or less salient (Yip 2005). Given this conceptualization, salience has most often been examined from a within-person approach; that is, by comparing the experience of salience for the same individual to him/herself across time. However, research also indicates that the remaining 60 % of variability is accounted for by individual differences, such as stable characteristics of the individual (Yip 2005). Therefore, it is also important to explore individual (e.g., between-person) contributors to ethnic-racial identity salience.

Yip (2014) adopted such a between-person perspective by examining whether ethnic-racial identity processes influenced average levels of situational salience over time. Consideration of ethnic-racial identity processes reflects another theoretical perspective that is informed by a psychosocial theory of identity development (Erikson 1968; Marcia 1966) and Phinney’s (1989) subsequent adaptation to ethnic-racial identity. From this theoretical perspective, exploration and commitment are ethnic-racial identity processes. Exploration captures the degree to which an individual has sought out ethnic-racial group information, and commitment captures the degree to which an individual has come to a sense of clarity about the personal meaning of it. Drawing on this perspective, Yip (2014) examined associations between relative levels of exploration and commitment and average levels of salience across situations. Results indicated that African American, Latino, Asian, White, and Native American adolescents exhibiting high exploration and low commitment (i.e., a moratorium status) or high exploration and high commitment (i.e., an achieved status) reported higher average ethnic-racial identity salience across situations compared to adolescents exhibiting low exploration and commitment (i.e., a diffused status). This research offered insight into between-person differences in levels of ethnic-racial identity salience depending upon ethnic-racial identity processes.

What remains unknown is whether exploration and commitment influences the degree to which salience varies from situation to situation. This is a significant shortcoming in developmental knowledge, as the variability of situational salience is also an important developmental indicator of ethnic-racial identity; the more stable an identity is, the more likely it is to be central to an individual’s overall sense of self (Sellers et al. 1998). Therefore, to a certain extent, exploring the variability of salience provides information about the development of more stable ethnic-racial identity. Further, the degree to which ethnic-racial identity is consistently salient is an important component of whether an individual will engage in the developmental processes of ethnic-racial identity exploration and commitment (Umaña-Taylor 2004). Consideration of variability requires a within-person perspective to examine consistency versus inconsistency of an individual’s salience from one situation to the next. Therefore, in order to address the interrelation between the ethnic-racial identity processes of exploration and commitment and the content of ethnic-racial identity salience variability, between- and within-person perspectives need to be captured simultaneously.

Exploration and commitment were initially proposed as complementary components of a progressive stage model of identity development (Marcia 1966). However, research has found that individuals do not necessarily move through identity statuses in a prescribed sequence (e.g., Seaton et al. 2012). In turn, much research examines exploration and commitment in tandem to describe relative levels of exploration and commitment in terms of “statuses” (e.g., Douglass and Yip 2015; Syed et al. 2007; Seaton et al. 2006, 2012). Indeed, recent evidence suggests that identity development is neither uniform nor unidirectional (e.g., Meeus et al. 1999). For example, individuals can fluctuate between high and low commitment (Waterman 1999; Syed et al. 2007), or between high and low exploration (Seaton et al. 2012). Even so, exploration and commitment are generally expected to increase with age (Yip et al. 2006), as individuals gain the social and cognitive abilities to engage in these processes (Selman 1980; Quintana 1994), along with the developmental imperative to form a secure sense of self (Erikson 1968). In turn, it has been proposed that adolescents with higher exploration and commitment fare better with respect to psychological adjustment (Erikson 1968; Phinney 1990). The evidence is clear that both exploration and commitment are important considerations in ethnic-racial identity development. Therefore, in examining the variability of salience, the current study considers exploration and commitment as distinct, though interrelated, processes that may inform variability of salience. This departure from the approach of Yip (2014), who examined relative levels of exploration and commitment through a status approach, allows consideration of the potentially unique influence of each component for salience variability.

Variability (and stability) has long been recognized by foundational psychological theories as important features of the psychological experience (e.g., Rogers 1961, 1963). Conceptually, variability captures both between- and within-person perspectives; there can be individual differences in the degree to which people vary (i.e., some can vary more than others), and that degree of variability is defined by the amount that a person deviates from his/her own experiences over time. For ethnic-racial identity salience, variability would mean that adolescents are at times very aware of their ethnic-racial identity, and at other times not at all aware of their ethnic-racial identity. For example, qualitative work highlights such experiences of stability and variability in adolescents’ own words (Douglass et al. 2015 for study details). When asked by an interviewer “When you’re walking around and you’re going to classes and you’re hanging out with friends, are you aware of your race?” a 17-year old African American female responded:

Yeah I am, I don’t know…like it depends because sometimes when I’m in the cafeteria… I’m with certain groups right? …when I’m with this one certain group that’s basically, like all African Americans or Black…like when they get kind of mad, it’s soo…yeah I’m thinking about it then. And if I’m walking and I have a lot of [Hispanic] friends …I’m like ‘do you know I’m the only Black person here with y’all?’ and then I’m always just, whenever I’m with my friends I think about it.

Following up, the interviewer then asked “Do you think about it when you’re in your classes like maybe when you’re talking with your teachers?” and the adolescent responded “No not really. That’s not something I really focus on when I’m in class.” These responses suggest that between those situations with friends in the cafeteria and in the classroom talking to a teacher, this adolescent’s salience is highly variable. On the other hand, stability would mean that adolescents consistently experience a given level of awareness of their ethnic-racial identity, be it high or low. An example of high ethnic-racial identity salience stability is provided by an 18-year old Dominican, Haitian and Native American female who responded:

I mean…I think about it a lot if I’m walking through the hallways ‘cause honestly like this school is like, there’s more like Caucasian people and Chinese people so it’s like, I always look and I’m like ‘wow, like I’m really in this school’ because the school I came from before there was more Hispanics and Black people so it’s like, it was like a change for me, and I think about that all the time.

This response suggests that this adolescent experiences significant stability in her salience across situations. Together, these recollections recount not only how ethnic-racial identity salience varies from one context to the next for the same adolescent, but also how, across adolescents, some report consistent ethnic-racial identity salience whereas others report more variability in salience.

Although variability in psychological experiences has long been of interest (e.g., Fleeson 2004; Syed and Azmitia 2008), it was not until recently that analytical advances offered a measurement of variability that matches its conceptualization (Hoffman 2007). The approach allows for variability to be captured from within- and between-person perspectives simultaneously by modeling the heterogeneity of residual variances (i.e., how much an individual varies around their own experiences) as well as whether some individuals are more likely to be variable than others. Subsequently, individual factors can be considered to determine whether they explain such individual differences in variability. In the current study, this approach is applied to examine ethnic-racial identity processes of exploration and commitment as they relate to within-person fluctuations in salience across situations. Therefore, we propose hypotheses about the relative variability in salience based on exploration and commitment.

Salience is largely responsive to contextual cues (e.g., Yip 2005; Yip and Fuligni 2002) in ways that individuals cannot explicitly control. However, scholars have argued that identity processes may provide adolescents with representations of the self that subsequently influence even implicit experiences (Pittman et al. 2011). For example, previous research has shown that individual differences in characteristics such as self-esteem and depression influence variability in affect across time (Kuppens etal. 2007). Therefore, the processes of ethnic-racial identity development may subsequently influence the degree of variability in salience in predictable ways. Consistent with theory, we hypothesized that both exploration and commitment are developmental assets that would promote stability in salience. First, having personally explored the meaning of ethnic-racial identity provides youth with a working knowledge of how and when ethnicity-race is meaningful to them (Phinney 1989); adolescents who have this working knowledge will likely use it as a reference when faced with situations that may (or may not) raise salience. Having exploration experiences to serve as a stable reference would likely lead to greater stability in salience across situations. In contrast, adolescents who have not explored their ethnic-racial identity would not be equipped with such a working knowledge, and therefore may be more reliant on contextual cues to raise salience. Given that contextual cues are likely to be very unreliable across situations, this would lead to greater variability in salience. Second, having made a personal commitment to the meaning of one’s ethnic-racial identity provides youth with greater certainty about their identity (Umaña-Taylor et al. 2014). In any given situation, adolescents will likely draw on this commitment to a greater extent than they rely on external cues to inform their salience, leading to greater stability of salience across situations. In contrast, adolescents who do not have the internal resources that a strong sense of commitment provides may be more reliant on contextual cues, again leading to greater variability in salience.

Threshold Effects in the Variability of Salience

If there are in fact individual differences in salience variability based on ethnic-racial identity status, does this variability have psychological consequences? Yip (2014) found that absolute levels of ethnic-racial identity salience were positively associated with positive feelings about ethnic-racial identity (i.e., private regard) across situations. In turn, the benefits of salience were expected to generalize to overall positive feelings. However, there were no significant associations between levels of situational salience and positive mood, leading to the conclusion that there are not general psychological benefits to experiencing salience in a situation (Yip 2014). However, the level of salience in any given situation can be decomposed into two components; the experience of salience at that moment (i.e., within-person) and the level of salience that an individual generally experiences (i.e., between-person; Zeiders et al. 2015). Because ignoring the distinction between within- and between-person components can suppress the unique effects of each (Hoffman 2007), the current study reconsiders whether there are positive effects of salience on mood by distinguishing between between- and within-person effects.

Examination of between-person effects allows us to consider whether an adolescent who reports generally high salience has better mood outcomes on average than an adolescent who reports generally low salience. In contrast, examination of within-person effects allows us to consider whether there are differences in mood outcomes for an individual when he/she is experiencing higher than average salience (i.e., positive variability) compared to when that same person is experiencing lower than average salience (i.e., negative variability). Methodologically, such an approach parses repeated measures into distinct between-person and within-person components; with respect to the current study, the between-person component represents an individual’s average level salience, calculated as the mean of all situational measures of salience. The within-person component represents the degree to which an individual varies in salience across situations with respect to their own average, calculated as the difference of each measurement of salience from the individual’s average.

One advantage of considering within- and between-person perspectives simultaneously is the ability to subsequently explore threshold effects (Abela and Hankin 2009). Drawing heavily on a dynamic systems approach to development, threshold effects consider whether within-person experiences vary at different levels of between-person experiences (van Geert 1994). Importantly, threshold effects take into account how varying aspects of developmental processes interact over time to influence psychological experiences, such as mood. For example, applied to the current study, threshold effects allow consideration of whether variability in salience is related to mood outcomes differently for people who have high versus low average levels of salience. Statistically, threshold effects can be captured through the interaction of between- and within-person effects (Abela and Hankin 2009); in this case, whether the relationship between variability in ethnic-racial identity salience and mood differs depending on adolescents overall levels of salience can be considered. Therefore, it allows us to consider for whom and how situational salience may become meaningful for adolescents (Zeiders et al. 2015). This approach may reveal important qualifications for the previously observed null effects of salience on mood.

To further expand on Yip (2014), we also include a broader conceptualization of psychological adjustment in the current study by examining negative and anxious mood in addition to positive mood as outcomes of interest. With respect to these mood outcomes, salience is generally expected to be beneficial. Therefore, from a between-person perspective, we expect that adolescents whose ethnic-racial identity is generally very salient will exhibit better mood outcomes (i.e., lower negative and anxious mood, higher positive mood) than their peers who generally experience low salience. From a within-person perspective, experiencing positive variability (i.e., higher than average salience) is expected to be related to better mood outcomes. With respect to threshold effects, we expect that among adolescents whose ethnic-racial identity is generally very salient, situational variability in salience will be unrelated to mood outcomes. This lack of association is expected because adolescents who are generally always aware of their ethnic-racial identity, and therefore have better mood overall, may not be as sensitive to these changes (i.e., a ceiling effect). However, among adolescents whose ethnic-racial identity is generally not salient, it is expected that situational variability in salience would be beneficial for mood outcomes, as these youth will benefit from atypical experiences of having their ethnic-racial identity relevant.

Finally, it is possible that in addition to being related to variability in salience, ethnic-racial identity processes of exploration and commitment influence the threshold effect of salience on mood. Broadly, ethnic-racial identity is conceptualized as a lens through which individuals experience the world around them (Sellers et al. 1998). Further, it is believed that having both explored and committed to ethnic-racial identity allows individuals to feel a coherent sense of self, while not engaging in these processes leaves individuals in a state of crisis and confusion (Phinney 1989). Therefore, it is likely that the degree to which people feel positively in light of being aware of their ethnic-racial identity is in part dependent on exploration and commitment; being aware of an identity that one feels that they understand is likely a more positive experience than being aware of an identity that one is confused about. Therefore, we expected that among adolescents who felt generally low salience, the relationship between positive variability and mood would be stronger amongst individuals who had explored and/or made a commitment to their ethnic-racial identity. That is, we expected that both exploration and commitment would amplify the threshold effect of salience on mood. However, among adolescents who felt generally high salience, we expected that the ceiling effect would persist and that exploration and commitment would not influence the relationship between positive variability and mood.

The Present Study

The current study considered the relationship between ethnic-racial identity processes and variability in salience, and the general psychological implications of this variability. The first aim of the study was to examine whether ethnic-racial identity status would predict variability in salience from one situation to the next. We expected that there would be significant differences in variability based on ethnic-racial identity exploration and commitment, such that exploration and commitment would both be related to greater stability (i.e., less variability) of salience across situations. The second aim of the study was to examine threshold effects in salience. We expected that among adolescents whose ethnic-racial identity is generally very salient, salience variability would be unrelated to mood outcomes. Among adolescents whose ethnic-racial identity is generally not salient, we expected that salience variability would be linked to beneficial mood outcomes (i.e., lower negative and anxious moods, higher positive mood), and that these associations would persist across situations. Finally, we expected that exploration and commitment would further moderate these relationships, such that exploration and commitment would amplify the strength of threshold effects.

Method

Participants

The participants used in Yip (2014) were also used in the current study. Participants were 395 ethnic-racially diverse adolescents (M = 15.22, SD = .67; 63 % female) who self-identified as Black or African American (12 %), Hispanic or Latino (26 %), Asian or Asian American (34 %), White (23 %), Native American or American Indian (1 %), or Other (i.e., did not specify, or identified as multi-ethnic; 4 %). Among adolescents not born in the United States (17.5 %), age of immigration ranged from 6 months to 14 years, and the largest proportion emigrated from China (6 %). The most common response to questions about parental education level was that participants did not know their parents’ highest level of education (29 %); the second most common response was that participants’ parents completed high school (13 %).

Procedure

Data on the racial composition of all New York City public high schools were obtained from the Department of Education and used to select five similar-sized and academically-comparable schools that varied in the ethnic/racial composition of the student body. Specifically, schools were selected to represent a predominantly Asian school, a predominantly White school, a predominantly Hispanic school and two racially heterogeneous schools. Due to the diversity of the NYC area, “predominantly” is defined by a single group representing at least 40 % of the school’s student population. In the heterogeneous schools, no group represented more than 40 % of the school’s population. Parental consent and youth assent letters were sent home to all 9th graders in the fall of 2008 and 2009, drawing in two cohorts (Cohort 1 n = 238, Cohort 2 n = 157). All students in each school were welcome to participate. Participation began in 9th grade, where the current data are drawn from.

Participants were administered pre-surveys in groups of 10–30 students, and then given a cellular phone to complete five randomly prompted experience sampling reports per day for the next 7 days, totaling 35 surveys. Times for the random prompts each day were determined using a random number generator and pre-programmed into each phone using a calendar function that was hidden from participants. Because participants were prompted at unknown times, missing data were inevitable. On average, participants completed 17 (SD = 9, range from 1 to 34) surveys. Participants were compensated $50 after completing the study. Data from the Youth Experience Study have been published elsewhere (Douglass and Yip 2015; Douglass et al. 2014; Shelton et al. 2014; Yip 2014, 2015; Yip et al. 2013).

Measures

Background Characteristics

Participants reported their ethnic-racial identification and country of birth. Ethnic-racial identification was recoded to indicate whether participants were a member of an ethnic-racial minority group (1 = ethnic-racial minority, including Black/African American, Hispanic/Latino, Asian, and Native American/American Indian; 0 = White). Country of birth was recoded to represent nativity status (1 = non-U.S. born; 0 = U.S. born). See Table 1 for descriptives and correlations for all study variables.

Table 1.

Correlations among key study variables

	1	2	3	4	5	6	7	8
1. Minority status		−.07	.08	.14**	.13*	−.10*	−.18***	−.02
2. Nativity status			.08	.09	.01	−.04	−.05	.03
3. ERI exploration				.55***	.14**	.03	.11*	.16**
4. ERI commitment					.20***	−.08	−.02	.09
5. ERI salience						−.20***	−.20***	.14**
6. Negative mood							.77***	−.15**
7. Anxious mood								−.01
8. Positive mood
M (SD)	.77 (.42)	.83 (.38)	2.61 (.54)	2.97 (.44)	4.25 (1.67)	.72 (.70)	1.00 (.68)	2.04 (.63)

Correlations are computed using the within-person average of all situational measures. Minority status was coded as 1 = ethnic-racial minority, 0 = White. Nativity status was coded was 1 = non-U.S. born, 0 = U.S. born

ERI ethnic-racial identity

^*p < .05;

^**p < .01;

^***p < .001

Ethnic-Racial Identity Exploration and Commitment

The exploration and commitment subscales of the Multigroup Ethnic Identity Measure (MEIM; Phinney 1992) were used to assess exploration and commitment. The exploration subscale includes five items and assesses the extent to which an adolescent has actively examined the role of ethnicity in his or her identity. Participants responded to a scale ranging from 1 (strongly disagree) to 4 (strongly agree) where higher scores indicate greater exploration (M = 2.61, SD = .54, Cronbach’s α = .74). The commitment subscale includes seven items and taps into a sense of resolution around the role of ethnicity in one’s life. Once again, participants responded to a scale ranging from 1 (strongly disagree) to 4 (strongly agree) where higher scores indicate a greater sense of commitment (M = 2.97, SD = .44, Cronbach’s α = .87). Tests of measurement invariance conducted by Yip (2014) suggested that there were no differences in the MEIM structure between White and ethnic-racial minority participants.

Situation-Level Salience

Ethnic-racial identity salience, operationalized as awareness of one’s ethnic identity, was assessed at each prompt. As others have done (e.g., Yip 2005; Yip and Fuligni 2002), participants responded to a single item that asked “How aware are you of your race/ethnicity right now” on a scale ranging from 1 (not at all) to 7 (extremely) where higher scores indicate higher salience (given the repeated nature of the measure, M and SD were calculated as the average of within-person assessments; M_average = 4.25, SD_average = 1.67). Because the measure is a single item, typical internal consistency measures such as Cronbach’s α cannot be computed. Instead, average inter-item correlation (r̄) has been proposed as a comparable measure of internal consistency for single items assessed through repeated measures (Dennisen et al. 2008); for salience, r̄ = .76.

Situation-Level Mood

Positive, negative, and anxious mood were assessed at each prompt using items drawn from the Profile of Mood States (POMS; McNair et al. 1971) abbreviated for experience sampling studies (see Cranford et al. 2006). Four items were used for each of the positive mood (happy, calm, joyful, excited), negative mood (sad, hopeless, discouraged, blue), and anxious mood (anxious, nervous, on edge, unable to concentrate) subscales. Items were adapted by adding “right now” to each item stem (e.g., “How on edge do you feel right now?”). Participants responded on a 5-point scale ranging from 0 (not at all) to 4 (extremely); higher scores reflect higher levels of positive mood (M_average = 2.04, SD_average= .63, Cronbach’s α_average = .85), negative mood (M_average = .72, SD_average = .70, Cronbach’s α_average = .90) and anxious mood (M_average = 1.00, SD_average = .68, Cronbach’s a α_average = .90), respectively.

Aim 1 Analytic Plan: Developmental Differences in the Variability of Salience

The first aim of the study was to examine differences in within-person variability of salience based on ethnic-racial identity process. That is, we explored whether exploration and commitment were associated with how much adolescents’ salience fluctuated from one situation to the next. Analyses followed the steps for examining within-person variability detailed in Hoffman (2007) and were conducted in SAS 9.2. A series of preliminary analyses were conducted; first, two-level unconditional models were estimated to determine whether there was significant within-person variation in salience. Next, prompt order (i.e., 1–35) was examined as a predictor of level-1 salience to establish any method artifacts; this was necessary to consider variation across situations that was a function of participants’ repeatedly answering the same items, rather than the individual differences of primary interest. Finally, tests for heterogeneity of level-1 residual variances were conducted to examine whether there were significant individual differences in variation (Hoffman 2007). Examining whether some people vary more than others is a necessary prerequisite to examining whether individual differences predict variability. Following these preliminary analyses, the main analyses of interest were conducted in which level-1 residual variances of salience were modeled as a function of exploration and commitment. Level-1 residual variances represent variability across situations (Hoffman 2007). High variability represents fluctuations in salience, and low variability represents stability in salience. Based on previous conceptual and empirical work highlighting important differences in experiences of ethnic-racial identity based on ethnic-racial minority status and nativity status (e.g., Helms 1990; Phinney 1989; Umaña-Taylor et al. 2013; Yip et al. 2008), these demographic characteristics were included as covariates in the model.

Results

Aim 1 Results: Developmental Differences in the Variability of Salience

Preliminary Analyses

Unconditional models indicated that there was significant within-person variability in salience (p < .001); that is, adolescents reported significant variation in salience from one situation to the next. Specifically, 31 % of the variability in salience was accounted for by something about the specific situation. The remaining 69 % of variability was accounted for by something about the adolescent. Prompt order (i.e., 1–35) was a significant predictor for variability in situational salience (b = −0.01, p < .001), indicating that ratings of salience became slightly more consistent over the course of the study. Prompt order was subsequently included as a covariate in all subsequent analyses. Additionally, day type (weekday versus weekend) was also included as a covariate. Next, the test for heterogeneity of residual variances was significant (H(281) = 5526.98, p < .001), indicating that some adolescents were significantly more variable in their ratings of salience across situations than others.

Predicting Situational Variability of Salience

Next, we explored whether ethnic-racial identity exploration and commitment accounts for why some adolescents are significantly more variable in salience than others. First, significant results from the covariates indicated that adolescents born outside of the U.S. were more variable than adolescents born in the U.S., and adolescents were more stable on weekends than weekdays. For the main effects of interest, results indicated that exploration was unrelated to variability in salience, while commitment was negatively related to variability in salience (see Table 2). That is, as predicted, adolescents with high commitment reported more stable salience than their peers with low commitment.

Table 2.

Parameter estimates for residual variance of situational salience

Residual variance of situational salience
Parameter	Est. (SE)
Intercept	2.62 (.20)***
Prompt order	.00 (.00)
Day type	−.91 (.06)***
Minority status	−.05 (.04)
Nativity status	−.11 (.04)**
ERI exploration	.01 (.03)
ERI commitment	−.14 (.04)**

Parameter estimates (Est.) for the residual variance of situational salience are presented, along with the standard error (SE) of the estimates in parentheses. Day type was coded as 1 = weekend, 0 = weekday. Minority status was coded as 1 = ethnic-racial minority, 0 = White. Nativity status was coded was 1 = non-U.S. born, 0 = U.S. born

ERI ethnic-racial identity

^*p < .05;

^**p < .01;

^***p < .001

Aim 2 Analytic Plan: Threshold Effects in the Variability of Salience

The second aim of the study examines the psychological effects of salience by exploring whether between-person salience and within-person changes in salience interact to inform mood outcomes. Multilevel path analyses were conducted in Mplus 7.2 in which repeated measures at level 1 were nested within individuals at level 2. Analyses followed steps for examining threshold effects detailed in Zeiders et al. 2015. As a preliminary step, two-level unconditional models were estimated for negative, anxious, and positive mood as outcomes to examine whether there was significant within-person variation in mood across situations. Next, models were conducted to examine the threshold effects of interest. Between-person salience was calculated as each adolescent’s average salience across situations. Since this is a person-level construct, this average value was included at Level 2 of the HLM models. To estimate how much variability adolescents reported around their own means, we also estimated within-person salience. Within-person salience was computed using the repeated measures of salience, which were person-mean centered (i.e., each situational measure of salience was centered around the adolescent’s own mean across situations) to represent within-person variability above and below an adolescent’s average. Since within-person salience varies across situations, it was included at Level 1 of the HLM models (Zeiders et al. 2015). Main effects for within- and between-person salience were examined first to explore how adolescents’ means and their variability around their means were related to psychological outcomes (see Fig. 1a). Next, the cross-level interaction of between- and within-person salience was entered in the model to explore how mean levels and variability around that mean interacted to influence psychological outcomes (i.e., threshold effects; see Fig. 1b). Finally, two separate three-way interactions were examined: (1) the interaction of exploration, between-person salience, and within-person salience, and (2) the interaction of resolution, between-person salience, and within-person salience. To probe any significant interactions, simple slopes analyses were conducted at one standard deviation above and below the mean of the between-person effect (Aiken et al. 1991). As in earlier analyses, ethnic-racial minority status, nativity status, prompt order, and day type were included as covariates in all models.

Fig. 1.

a Model heuristic depicting the main effects of within-person and between-person salience predicting concurrent mood outcomes. Note: BP between-person, WP within-person, ERI ethnic-racial identity. Solid black lines indicate significant paths; dashed grey lines indicate non-significant paths. Covariances among endogenous and exogenous variables were modeled but are not depicted here. Covariates are depicted to the right of the outcomes.*p < .05. **p < .01.***p < .001. b Model heuristic depicting the interaction of within-person and between-person salience predicting concurrent mood outcomes. Note: BP between-person; WP within-person, ERI ethnic-racial identity. All paths depicted in a were estimated but are not depicted here for clarity of presentation. All covariates are depicted in grey. Path estimates are only shown for interaction effects. Solid lines depict significant paths. *p < .05. **p < .01. ***p < .001

Aim 2 Results: Threshold Effects in the Variability of Salience

Preliminary Analyses

Unconditional models indicated that there was significant within-person variability in all mood outcomes (ps < .001); that is, adolescents reported significant variation in all mood indicators from one situation to the next. For both negative and anxious moods, 54 % of variability was accounted for between-person and 46 % accounted for within-person, suggesting that a slight majority of the variability can be attributed to differences between adolescents rather than differences across situations. This pattern was reversed for positive mood, with 45 % of variability accounted for between-person and 55 % accounted for within-person.

Main Effects

There were a number of significant main effects for covariates, which are reviewed first. Prompt order was positively related to negative mood and negatively related to positive mood, indicating the presence of a method artifact such that that reports of negative mood increased slightly over the course of the study, and reports of positive mood decreased slightly. Ethnic-racial minority status was significantly related to anxious mood such that ethnic-racial minorities reported lower anxious mood on average than White adolescents over the course of the study. Additionally, exploration was positively related to anxious mood and positive mood, indicating that adolescents with higher exploration reported higher average anxious mood and positive mood over the course of the study than their peers with lower exploration. With respect to the main effects of interest, there was a negative main effect of between-person salience on negative and anxious moods, and a positive main effect on positive mood (see Fig. 1a). That is, as predicted, beneficial mood effects were observed for adolescents who had generally high salience to the extent that they felt less negative, less anxious, and more positive on average than their peers who had generally low salience. There was a main effect of within-person salience on negative mood and a main effect on positive mood, although in opposite directions. Specifically, in situations where adolescents reported salience that was higher than their own average, they reported less negative mood and more positive mood. In contrast, within-person salience was unrelated to anxious mood (see Fig. 1a).

Interaction Effects

Next, the interaction of between-person and within-person salience was added to the model. Results indicated that the interaction was significant for negative, anxious, and positive moods (see Fig. 1b). Simple slopes analyses indicated that among adolescents with high between-person salience (i.e., high average levels of salience), within-person salience was negatively associated with negative (b = −.07, SE = .02, p < .001) and anxious (b = −.06, SE = .01, P < .001) moods. In other words, adolescents who generally reported high salience felt less negative and less anxious in situations where their salience was above their own average; see Fig. 2). Within-person salience (i.e., changes in salience from one situation to the next) was not significantly associated with positive mood (b = −.01, SE = .01, p = .49). Among adolescents with low between-person salience (i.e., low average levels of salience), within-person salience was also negatively associated with negative mood (b = −.02, SE = .01, p = .03), but positively associated with positive mood (b = .03, SE = .01, p = .01). In other words, adolescents who generally reported low salience feel less negative and more positive in situations where their salience was higher than average (see Fig. 2). Although the simple slope effects for negative mood were significant for adolescents with both high and low average (i.e., between-person) salience, the interaction indicates that this effect was stronger for adolescents with high average salience. There was no association between within-person salience (i.e., changes in salience from one situation to the next) and anxious mood (b = −.01, SE = .01, p = .25.

Fig. 2.

Effect of variability in WP salience for high BP and low BP salience adolescents on concurrent mood. Note: WP within-person, BP between-person. 0 value for variability in WP salience represents situations in which adolescents’ salience is equal to their own average. For adolescents with high BP salience, 0 is equal to a 5 on a 0–6 scale; thus, they can vary between 5 below and 1 above their own average. For adolescents with low BP salience, 0 is equal to a 2 on a 0–7 scale; thus, they can vary 2 below and 4 above their own average. Significant slope are depicted in solid lines; non-significant slopes are depicted in dotted lines

Finally, two three-way interactions were tested to consider whether either exploration or commitment further moderated these threshold effects. Statistics reported in text are for the exploration and commitment models, respectively. Results indicated that there were no significant three-way interactions for negative mood (b = −.01, SE = .01, p = .51; b = .00, SE = .02, p = .95), anxious mood (b = −.01, SE = .01, p = .46; b = .00, SE = .02, p = .84), or positive mood (b = .00, SE = .01, p = .73; b = .01, SE = .02, p = .38) indicating that the way that the threshold effects of salience influenced mood were not significantly different based on exploration or commitment.

Discussion

The relevance of one’s ethnic-racial identity varies moment by moment, and such variation has critical implications for ethnic-racial identity development and adolescent well-being (Sellers et al. 1998; Yip 2005). While theories have long recognized the variability of salience and its possible developmental implications, little attention has been paid to this issue by empirical work. The present study fills this void by directly examining salience variability, and linking this variability to psychological outcomes. Specifically, this study is the first to show that between-person differences in ethnic-racial identity processes of exploration and commitment influence the degree to which salience varies across situations, and that this variability has important meaning for adolescents’ general psychological adjustment in a given situation. Overall, the present study provides nuanced insight into for whom and when experiences of ethnicity-race become psychologically relevant in adolescents’ everyday lives. Below, we discuss how the findings of the current study are both a novel advancement in and complementary to existing knowledge.

Developmental Differences in the Variability of Salience

The present study found that individual differences in ethnic-racial identity commitment were associated with the degree to which individuals experienced variability in salience from situation to situation. This is consistent with an agenetic sociocognitive view of the self that conceptualizes individuals as self-organizing and proactive, rather than simply subject to and shaped by their contexts (Bandura 2001). More specifically, it is also consistent with ethnic-racial identity theory which asserts that the extent to which salience varies from situation to situation is determined, in part, by person characteristics of stability (Sellers et al. 1998). The hypotheses that both exploration and commitment would be significantly related to stability in salience were only partially supported. Although commitment was related to variability in the expected direction, such that adolescents with high commitment exhibited greater stability of salience over time, exploration was unrelated to variability in salience. As individuals with more stable characteristics are expected to be less influenced by situational cues, this study suggests that ethnic-racial identity commitment, but not exploration, is one such personal characteristic of stability.

This apparently unique influence of commitment is consistent with previous research; Toomey and colleagues found that among a sample of Mexican-origin adolescent mothers, experiences of discrimination were positively related to later risky behaviors; however, adolescents who exhibited an achieved or foreclosed status were buffered from this effect (Toomey et al. 2013). They concluded that the high level of commitment, which marked both achieved and foreclosed statuses, was likely the operating protective factor. Among Latino adults, Torres and colleagues found that commitment, but not exploration, served as a buffer of the negative effects of discrimination on mental health (Torres et al. 2011). Additionally, Brittian et al. found that among Latino and Black college students, commitment (referred to as resolution) but not exploration was negatively related to depressive symptoms (Brittian et al. 2014). In this study, it appeared that the protective nature of commitment was conferred through associated positive feelings about one’s ethnic-racial identity. In the current study, the apparent relevance of commitment above exploration makes sense; commitment is conceptualized as an internal process of coming to a sense of clarity in self-concept with respect to ethnicity-race, while exploration is conceptualized as more of an outward facing process of involvement in behaviors and activities (Phinney 1989). Since coming to awareness of ethnic-racial identity in a given situation is also an internal process that draws upon the working self-concept (Sellers et al. 1998), it makes sense that it would be closely aligned with commitment. The current study contributes to the growing literature which suggests that commitment may be a particularly critical component of ethnic-racial identity.

These findings can also be considered in tandem with prior work that has documented differences in average levels of salience based on ethnic-racial identity statuses. Using this same sample, Yip (2014) found that youth with high exploration and commitment (i.e., an achieved status) demonstrated the highest average salience, and that youth with low exploration and commitment (i.e., a diffused status) demonstrated the lowest average salience (Yip 2014). Taken together, the current study and Yip (2014) suggest that high average salience and stability may not always go hand in hand; while youth in statuses characterized by high exploration demonstrated the highest salience on average (Yip 2014), the current study finds that exploration is unrelated to variability in salience. We suspect that while average salience may be more closely associated with exploration, variability of salience may be more closely tied to commitment. With respect to exploration, it may be that youth who are engaged in active exploration may be attending to matters of ethnicity/race in their everyday lives to a greater degree, or that high salience prompts adolescents to explore identities. With respect to commitment, it may be that having a strong sense of clarity allows individuals to experience salience in a more consistent way over time, or that consistently stable salience facilitates a sense of clarity. Future work using a variable-centered approach is needed to disentangle the directionality of these relationships over time.

Threshold Effects in the Variability of Salience

In addition to establishing links between salience variability and ethnic-racial identity developmental processes, the present study also established a relationship between salience variability, as assessed by deviation from one’s average salience, and adolescents’ mood in a given situation. Previous experimental research has shown that identity salience influences how individuals process information; for example, in a laboratory study, Maitner and colleagues found that individuals who were primed to make their student identity salient were more analytical in their review of a university proposal than individuals who were primed to make their individuality or American identity salient (Maitner et al. 2010). The current study extends this knowledge to indicate that identity salience influences how adolescents emotionally perceive situations in their everyday lives.

Yip (2014) found that absolute levels of salience in a given situation were unrelated to positive mood in the sample as a whole. Based on these findings, one would conclude that being aware of one’s ethnic-racial identity does not have general psychological implications. However, it appears that these null results may have been the result of not distinguishing between within- and between-person components of salience. By examining situational variability in salience (i.e., within-person salience) simultaneously with one’s average salience (i.e., between-person salience), the current study found that salience does in fact have implications for situational mood. First, the results indicated that adolescents who generally report high salience across situations experience better mood overall. Second, the results indicated that, when they report salience that is higher than their own average, they experience better overall mood in that situation. These findings are consistent with the original theoretical position that the extent to which ethnic-racial identity is salient at a particular moment influences how the person construes the situation (Sellers et al. 1998). Importantly, it also suggests that specific instances of salience should be interpreted simultaneously with adolescents’ cumulative experiences of salience over time (i.e., their average experience). Developmentally, this highlights the importance of combining within-person experiences that occur at a given point in time and between-person experiences that accumulate over time (Hoffman and Stawski 2009) in capturing adolescents’ experiences of salience.

Further, although deviation from one’s average salience level was related to adolescents’ mood, this relationship took on different forms depending on one’s average salience level (i.e., the threshold effect). Although threshold effects have long been recognized in developmental models of stress (e.g., Abela and Hankin 2009; Zuckerman 1999), this is the first study to examine threshold effects in relation to identity, and it proved meaningful. In order to fully appreciate the psychological benefits of ethnic-racial identity in adolescents’ everyday lives, salience in a specific situation should be interpreted not only in conjunction with, but relative to, an adolescent’s own mean. The interrelation of these two components is perhaps most clearly depicted on the x-axis of Fig. 2; adolescents in the current study who have generally high levels of salience (i.e., 1 SD above the mean of the sample) are subject to a ceiling effect such that they can only vary above their own average to a very limited extent, whereas there is much greater possibility for variation below their own average. In contrast, adolescents who have generally low levels of salience (i.e., 1 SD below the mean of the sample) are subject to a floor effect, such that they can only vary below their own average to a very limited extent, whereas there is much greater possibility for variation above their own average.

These differences in the possible range of variability in salience between individuals appear to matter when it comes to mood. Although we hypothesized that a threshold effect would emerge, the exact nature of this effect was surprising. We expected that adolescents who generally experienced high salience would be unaffected by variability, but results indicated that negative variability (i.e., experiencing salience below their own mean) was related to poorer mood outcomes among these adolescents. Specifically, adolescents who experienced generally high salience reported more negative and anxious mood in situations where they were less aware of their ethnic-racial identity than average. For adolescents who are used to feeling that their ethnic-racial identity is relevant, a lack of relevance appears to be a negative experience for them. In addition, results indicated that among adolescents who experienced generally low salience, positive variability (i.e., experiencing salience above their own mean) was related to better mood outcomes. Specifically, adolescents who experienced generally low salience reported more positive and less negative mood in situations where they were more aware of their ethnic-racial identity than average.

Personality theories (e.g., Rogers 1961, 1963) have long conceptualized variability through “authenticity;” feeling authentic is believed to be a centering experience that allows individuals to feel an organized, coherent sense of self. What has long been contested is what leads to authenticity; experimental research has shown that variability in personality traits across lab-manipulated situations is negatively linked to well-being (e.g., Sheldon et al. 1997). This supports a trait consistency perspective which proposes that deviation from one’s traits is psychologically taxing to the extent that it decreases authenticity (Fleeson and Wilt 2010). The findings for adolescents who generally experienced high salience are consistent with this trait consistency perspective, as being less aware of ethnic-racial identity than usual was associated with worse mood outcomes. It may be that when adolescents are used to being aware of their ethnic-racial identity, being unaware feels inauthentic, leading to worse mood.

However, research utilizing experience sampling measures has indicated that a flexible enactment of personality traits (e.g., acting in a manner indicative of one personality trait at one point, and another personality trait at a different point) is actually related to greater authenticity (Fleeson and Wilt 2010). This supports a state-content significance perspective which proposes that ways of acting feel authentic because of their content and consequence, regardless of an individual’s traits. This state-content significant perspective is useful in interpreting the findings for adolescents who generally experienced low salience; when adolescents who are generally unaware of their ethnic-racial identity experience higher than normal levels of salience, it is beneficial for their mood. It is likely that among these individuals, an increase salience is a result of some aspect of the situation that triggers it. In such cases, being genuinely reactive to the situation also helps individuals feel authentic, again leading to better mood outcomes.

Overall, these findings point to salience as a positive psychological experience, although it is important to note that there were some inconsistencies in the form of mood benefits between adolescents with high and low average levels of salience. First, although adolescents with high average salience experienced changes in negative mood outcomes (i.e., negative and anxious mood) in relation to salience variability, they did not experience changes in positive mood. While possibly surprising, this distinction is not unprecedented; the absence of negative feelings is well recognized as distinct from the presence of positive feelings (Cacioppo and Berntson 1999), and decades of learning research suggest that the presence of positive feelings is generally a stronger affective experience than the absence of negative feelings. In the current study, it is likely that adolescents with high levels of average salience have a different baseline when it comes to salience variability and positive versus negative feelings; that is, while feeling a level of salience that is closer to their own (generally high) average may help decrease negative feelings, it may not be a strong enough experience to incite positive feelings. Second, although adolescents with low average salience experienced beneficial changes in negative and positive mood in relation to salience variability, they did not experience changes in anxious mood. Taken in tandem with the individual difference findings, it appears that youth who are generally unaware of their ethnic-racial identity are more anxious on average than their peers who are generally aware of their ethnic-racial identity, and they do not become less anxious or worried in situations where they became more aware of their ethnic-racial identity. It may be that, in order for salience to confer benefits for anxiety, adolescents must already be used to experiencing such awareness on a regular basis.

Importantly, we found no evidence that the threshold effect of salience on mood differed between individuals based on either ethnic-racial identity exploration or commitment. Therefore, commitment appears to meaningfully influence the degree to which salience varies across situations, but it does not alter how this variability is linked to subsequent mood outcomes. That is, individuals with higher commitment are less likely to exhibit variability in salience; however, experiences of variability in salience are related to mood in the same way for all adolescents, regardless of commitment. Further, exploration appears to have no role in these relationships. One possible reason for these null findings is that average levels of salience over time and the developmental processes of exploration and commitment are tapping into similar processes; scholars have long argued that salience is a precursor for engagement in ethnic-racial identity processes (e.g., Cross 1971; Umaña-Taylor 2004). Without being consistently aware of any given social identity, one is not likely to explore or reflect on its personal meaning. Therefore, average levels of salience across situations may reflect general levels of engagement in processes of ethnic-racial identity, which may be accounting for the variance that exploration and commitment would otherwise capture. Empirically, the interrelation and directionality between long-term development in ethnic-racial identity and everyday, situational experiences of ethnic-racial identity are not well understood. Alternatively, it may also be that salience is a mechanism by which ethnic-racial identity commitment influences the daily psychological adjustment of adolescents. In this case, the salience mechanism appears to have two equally important components: the average level of salience experienced over time and the variability in salience in a given situation. Notably, we did not directly examine a mediational model due to statistical limitations; the modeling of variability in salience can use heterogeneity of residual variances as a dependent variable, but not an independent variable. Since a mediating variable is simultaneously an independent and dependent variable (Baron and Kenny 1986), it was impossible in the current case to examine such a comprehensive model. However, the findings in this study are consistent with the conceptual underpinnings of mediation.

The current study also offers important practical insight into everyday experiences of ethnic-racial identity during adolescence. Conceptually, issues of ethnicity-race are presumed to become salient when “[one’s] group is devalued and subject to negative stereotyping in the larger society (Spencer 1999).” In line with a colorblind perspective, one may then conclude that it is adaptive to deny or diminish the salience of ethnic-racial identity (Neville et al. 2000). However, research on colorblindness has largely dismissed the utility of this approach (Plaut et al. 2009; Richeson and Nussbaum 2004). By underscoring the psychological benefits of acknowledging and accepting the salience of ethnic-racial identity in everyday interactions with adolescents, this study adds to this accumulation of literature suggesting that ethnic-racial identity can be a positive component youth development (Lee Williams et al. 2012).Further, the current study supports recent intervention efforts recognizing the importance of salience in promoting developmental processes of ethnic-racial identity (Umaña-Taylor and Douglass., in press). For parents, practitioners, and teacher alike, this study points to the utility of a multicultural, pluralistic perspective.

Limitations and Future Directions

Although this study makes several contributions to the existing literature on ethnic-racial identity, some limitations and future directions should be noted. First, the present study used data from adolescents’ self-reports, and thus the observed relationships may be inflated by shared variation. Future work using multiple informants is needed to replicate the current findings. Second, while the intensive repeated measures of salience were aligned with the conceptual emphasis on salience variability, we were unable to disentangle the directionality of the studied associations given the cross-sectional nature of the repeated measures (i.e., collected over 1 week). Long-term longitudinal data are needed to better understand whether the extent to which adolescents perceive varying salience is determined by their identity development or the varying salience from situation to situation begets identity development. Additionally, the results indicated that there are significant differences in the variability of salience based on ethnic-racial minority status and nativity status. Unfortunately, we did not have the statistical power to examine whether more nuanced distinctions within these categories, such as specific ethnic-racial identification or age since immigration, also played a role. Future work with larger samples that are well-represented with respect to these demographic distinctions is needed.

Conclusion

The present study represents one of the first steps in examining variability of adolescents’ ethnic-racial identity experiences. Colloquially known as a period of storm and stress, psychologists, educators, and parents have long understood that adolescents experience significant mood swings; indeed, research shows that adolescents exhibit greater variability in mood than adults do (Larson et al. 1980). This research indicates that adolescents are also contending with variability in awareness of their ethnic-racial identity.

We found that clarity about the personal meaning of ethnic-racial identity (i.e., commitment) promoted stability in salience across situations, while having personally explored the meaning of ethnicity-race (i.e., exploration) was unrelated to variability in salience. The prominence of identity processes during adolescence are recognized by both theory and research, but this study is an important step towards understanding how those processes of exploration and commitment translate into experiences of ethnicity-race on an everyday basis (Umaña-Taylor et al. 2014). Developmentally, these findings show that, in any given moment, adolescents’ experiences of ethnicity-race are a product of both their experiences over time and their experiences in that moment. Taken together, the current study highlights the complexities at play in how developmental processes of coming to a personal sense of meaning and clarity are related to variability in the content of adolescents’ ethnic-racial identity, and how such variability influences everyday psychological adjustment.

Biographies

Sara Douglass is a Postdoctoral Research Fellow in the T. Denny Sanford School of Social and Family Dynamics at Arizona State University. She received her doctorate in Applied Developmental Psychology from Fordham University. Her research examines adolescent ethnic/racial identity development, adolescent peer interactions as they relate to race and ethnicity, and the influence of diversity on everyday experiences for minority adolescents.

Yijie Wang is a Postdoctoral Research Fellow in the Department of Psychology at Fordham University. She received her doctorate in Human Development and Family Sciences at the University of Texas at Austin. Her research investigates how adolescent adjustment is influenced by cultural and racial/ethnic contexts across developmental settings (e.g., family, peer, school).

Tiffany Yip is a Professor in the Department of Psychology at Fordham University. She received her doctorate in Community Psychology with concentrations in Developmental and Quantitative Methods from New York University. Her research interests include ethnic identity development among minority youth, young adults, and adults, direct and indirect association between ethnic identity and psychological adjustment, and the impact of ethnic-specific and general stressors for well-being outcomes.