Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Oct 1.
Published in final edited form as: Cultur Divers Ethnic Minor Psychol. 2021 Jul 22;28(4):493–502. doi: 10.1037/cdp0000485

The Ethnic Identity Scale: Affirmation, Really?

Alan Meca 1, Melinda Gonzales-Backen 2, Julie C Rodil 1, Isis Cowan 1, Sangeeta Sharma 1, Taylor N Webb 1, Timothy Hayes 3
PMCID: PMC8782914  NIHMSID: NIHMS1718169  PMID: 34291982

Abstract

Objectives:

The Ethnic Identity Scale (EIS) was developed to distinguish between process and content components of ethnic-racial identity. However, the affirmation subscale is composed entirely of negatively worded items, measuring negative feelings about one’s ethnic-racial group, rather than positive feelings as widely conceptualized. Addressing this gap, the current study examined the psychometric validity of a revised EIS with positively and negatively worded items to determine whether affirmation is best represented as a unidimensional construct, a bidimensional construct, or a combination of the two.

Methods:

The sample consisted of 280 college students (75.5% female; Mage = 20.95 years; SD = 1.98 years). The largest ethnic-racial group consisted of Black or African Americans (68.2%), followed by Asian/Asian Americans (12.1%), Hispanic/Latinos (9.6%), and other ethnic-racial groups (10%).

Results:

Confirmatory Factor Analysis provided evidence for both unidimensionality and multidimensionality. Indeed, although positively worded and negatively worded items of “affirmation” loaded onto a general factor representing affirmation, there was still a significant amount of variance captured by the negative ERI affect specific factor, indicating the presence of multidimensionality. Additionally, results indicated that negative ERI affect, over and above the general ERI affirmation factor, predicted psychosocial functioning.

Conclusion:

The current study expands our understanding of the multidimensionality of ethnic-racial identity, highlighting the need for examination of how we measure ERI affect at the very least, and possibly how we conceptualize it within the broader ERI literature.

Keywords: ethnic-racial identity, measurement, affirmation, emerging adults, well-being, depression

For ethnic-racial minority young adults, ethnic-racial identity (ERI) is a salient identity domain with implications for psychosocial and health outcomes (Rivas-Drake at el., 2014). Conceptualizations of ERI have varied across literature, but youths’ feelings about their ethnic-racial group (e.g., racial pride, private regard, affirmation) has been of primary interest. Indeed, positive feelings about one’s ethnic-racial group has been consistently found to be associated with positive psychosocial adjustment (e.g., self-esteem and well-being) and academic outcomes (e.g., positive academic self-efficacy and engagement) among ethnic-racial minority youth (for a review, see Rivas-Drake et al., 2014).

In this paper we point out that the measurement of ERI affirmation in a widely used measure of ERI, the Ethnic Identity Scale (EIS; Umaña-Taylor et al., 2004), may not measure what the field has called affirmation. Rather, because the subscale is negatively worded, it measures negative affect or negative feelings about one’s ethnic group, reverse coded such that higher scores suggest a lack of overtly negative feelings, rather than implying positive feelings about one’s ethnic group as is widely conceptualized in the literature. Indeed, a response of 1 (“Does not describe me at all”) for the item “I dislike my ethnicity” would not imply one likes their ethnicity, but simply that they do not dislike it. The lack of a positively worded measure of affirmation makes it difficult to determine whether affirmation represents a unidimensional construct, whereby the absence of negative feelings towards one’s ethnic-racial group membership implies the presence of positive feelings, a purely bidimensional construct consisting of positive and negative affect, or a combination of the two. Addressing this gap, the current study sought to examine the functionality of a revised affirmation subscale of the EIS with both positively and negatively worded items.

Conceptualizing Ethnic-Racial Identity Development

Ethnic-racial identity is a multidimensional psychological construct that reflects individuals’ beliefs and attitudes about their ethnicity and race as well as the process by which these beliefs and attitudes develop over time (Umaña-Taylor et al., 2014). Although some scholars note the distinctions between race and ethnicity and call for examination of associated constructs separately (e.g., Cokley, 2007), others have called for the use of the term ERI due to conceptual and experiential overlap between ethnicity and race (Umaña-Taylor et al., 2014). Further, Umaña-Taylor and colleagues note that the use of this combined term may be more consistent with youths’ lived experiences in that ethnic aspects may not be easily disentangled from the racial aspects of their identities. As such, based on the recommendations of Umaña-Taylor and colleagues, and its increased use in the literature, we will use ethnic-racial identity (ERI) throughout this paper. The multidimensional nature of ERI is founded on two largely disparate perspectives: a developmental process perspective and a content perspective (Yip et al., 2014).

The developmental process perspective largely draws on the ego identity perspective proposed in Phinney’s (1989) ethnic identity model. Phinney’s (1989) conceptualized ethnic/racial identity resolution as existing along a continuum of exploration, commitment, and affirmation. Drawing on Erikson’s (1950) and Marcia’s (1966) work, exploration refers to the process by which one considers what their ethnic-racial membership means to them; whereas commitment refers to an understanding of what ones’ ethnic-racial group membership means to them (Phinney, 1989). Drawing on social identity theory (Tajfel, 1981), affirmation refers to whether an individual feels positively about their ethnic group membership. Underlying this model however was the presumption that resolution necessitates a commitment to and the establishment of positive feelings towards ones’ ethnic/racial group. Addressing this assumption, Umaña-Taylor et al. (2004) disentangled affirmation from commitment and proposed a revised ERI model composed of exploration, resolution (i.e., clarity about the meaning of one’s ethnic-racial group membership), and affirmation.

In contrast, the content perspective has largely drawn on social identity theory (Tajfel, 1981), the multidimensional model of Black identity (Sellers et al., 1998), and Nigrescence (Cross & Vandiver, 2001; Worrell et al., 2001). Broadly, the content perspective emphasizes the salience, significance, meaning of one’s ethnicity/race, both broadly and situationally (Sellers et al., 1998; Wang et al., 2017). According to social identity theory, in an effort to protect one’s sense of positive self-concept, individuals are motivated to develop a positive view of marginalized in-groups. An additional tenet of social identity theory states that identities associated with marginalized groups will be more salient. Seller’s multidimensional model of Black identity (Sellers et al., 1998) applied this notion to ERI specifically through a multidimensional model of ERI content focused on the significance (i.e., salience, centrality, and regard) and the meaning of one’s ethnicity/race (i.e., ideology). Sellers et al. (1998) conceptualized regard as consisting of both public and private such that individuals form a perception of how society views their race that is distinct from their internalized racial pride. Finally, in association with involvement in the Black Power Movement in the late 1960s, Cross (1971) introduced his Nigrescence model that outlined the stages of Black consciousness development that range from pre-encounter to internalization–commitment (for a complete review of this multidimensionality, see Vandiver, 2001). Cross’s model highlights a progression through recognizing the social stratification of race to embracing culture and forming racial pride and emphasizes that it is the avoidance of explicitly negative forms of identity expression that is important for psychosocial functioning (Vandiver et al., 2002).

Although these frameworks differ substantially in the origin and focus, they all emphasize the importance of affect, or the positive and negative feelings individuals hold about their ethnic-racial group membership. Consistent with content models of ERI that emphasize the implications of forming a positive sense of one’s ethnicity and race for well-being, research has indicated affect towards ones ERI is associated with a variety of indicators of well-being and psychosocial adjustment (Rivas-Drake et al., 2014). For instance, in a study of Mexican-origin adolescents, higher levels of affirmation (as measured by reverse coding negative-affect items on the EIS) were associated with fewer depressive symptoms which in turn predicted better academic outcomes (Piña-Watson et al., 2018). Additionally, research suggests that ethnic-racial affect can reduce the negative effect of discrimination on depressive symptoms. Specifically, Kogan et al. (2014) found that racial self-concept (i.e., centrality and private regard, or positive affect) mediated the longitudinal association between discrimination and depressive symptoms among African Americans transitioning from adolescence to emerging adulthood. Given the implications of ERI affect for psychosocial outcomes and its role as a protective factor against racial stress, it is essential to critically examine current measures and conceptualizations of ERI affect.

Measures of Ethnic-Racial Identity Development

Drawing on Phinney’s ethnic identity model (1989), the Multigroup Ethnic Identity Measure (MEIM; Phinney, 1992) was devised to capture the degree to which individuals have explored and committed to their ethnic identity and established positive feelings toward their group membership (Phinney & Ong, 2007). Despite its original intent, a number of psychometric evaluations of the MEIM have indicated a two-factor structure consisting of exploration and identity belonging (i.e., commitment and affirmation) (Phinney & Ong, 2007). As such, the MEIM has failed to distinguish between process and content and is incongruent with Phinney’s (1989) original model (Umaña-Taylor et al., 2004). Addressing this limitation, the Ethnic Identity Scale (EIS; Umaña-Taylor et al., 2004) was created to assess the developmental processes of exploration and resolution as well as the content of identity (i.e., affirmation). Studies examining the psychometric properties have supported the 3-factor structure of the EIS (White et al., 2011; Umaña-Taylor et al., 2004). The fact that the EIS is able distinguish between developmental and content components of ERI is a strength of the EIS, as components of ethnic identity are differentially associated with adolescent adjustment (Rivas-Drake et al., 2014).

Although the contribution of the EIS should not be understated, an important methodological limitation of the EIS is the utilization of solely negatively worded items intended to capture affirmation. Umaña-Taylor et al. (2004) initially included positively worded items for affirmation, however, these items did not load onto a meaningful factor. As such, within the final version of the EIS, even when reverse coded, the subscale assesses negative ERI affect or the level of negative feelings about ethnic group membership. Because this is in direct contrast to how ERI affirmation is typically discussed, which is positive affect regarding one’s ethnic-racial group membership, the affirmation subscale of the EIS lacks face validity. It is also important to note that nearly 50% of Umaña-Taylor et al.’s (2004) participants were non-Hispanic White. Given that ERI tends to be particularly salient among ethnic/racial minorities (Syed & Mitchell, 2013), it is possible that positive and negative items failed to diverge among non-Hispanic Whites. In contrast, given the complex history in with race and ethnicity in the United States (García Coll et al., 1996), positive and negative affect might represent distinct constructs for ethnic/racial minorities.

The private regard subscale of the Multidimensional Measure of Black Identity (MMBI; Sellers et al., 1998) provides a promising example of including positively and negatively worded items. Given the establishment of psychometric properties of the MMBI (Scottham et al., 2008), it is worth examining the positively worded affirmation items in an ethnic-racial minority sample. Further, scholars often consider private regard and affirmation similar constructs, but this may be inaccurate given positively and negatively worded items. Exploring the dimensionality of positive and negatively worded ERI affect items would lead to a more consolidated literature. Indeed, in the absence of positively worded items, it is impossible to determine whether ERI affect represents a unidimensional construct (i.e., presence or absence of affirmation), a bidimensional construct consisting of independent, yet related positive (i.e., affirmation) and negative affect dimensions, or a combination of the two. Given that studies on self-esteem have identified a substantive distinction between positive and negative self-esteem (e.g., Alessandri et al., 2015), it stands to reason that positive and negative feelings towards ones’ ethnic-racial group may similarly represent distinct dimensions. Indeed, the expanded model of Nigrescence (Cross & Vandiver, 2001) emphasized a conceptual distinction between Pre-Encounter self-hatred and pro-Black attitudes. Additionally, Vandiver et al. (2002) found moderate correlations (−.3 to −.4) between self-hatred and private regard (i.e., positive feelings an individual has about being Black), which is conceptually similar to affirmation as typically conceptualized (Umaña-Taylor et al., 2014). In sum, there is a need to examine the role item wording plays in the affirmation scale of the EIS.

The Current Study

To address the above methodological and theoretical gaps in the literature, the current study sought to evaluate two potentially competing conceptualizations of affirmation. On one end, affirmation may represent a unidimensional construct, ranging from negative to positive affect. On the other end, affirmation may best be represented as a bidimensional construct, with ERI positive and ERI negative affect representing two distinct dimensions of ERI affect. Towards this end, the current study utilized confirmatory factor analysis (CFA) to explore whether ERI affect could be best represented as a one-factor model, a two-factor model, or a bifactor model. Whereas a one-factor model proposes a unidimensional conceptualization of ERI affect, a two-factor model would represent a bidimensional conceptualization. In contrast, the bifactor model separates the common variance across items into a single target construct, while specifying specifies two (or more) orthogonal “group” factors representing the remaining variance (Gibbons & Hedeker, 1992). In doing so, a bifactor model allows for the capacity of specifying a single important target construct (i.e., ERI affirmation), while either retaining the idea of multidimensionality or extracting “nuisance” variance associated with item wording that may potentially interfere with the measurement of ERI affirmation (Reise et al., 2010). Building on these competing models, the current study also sought to determine the differential utility of positively and negatively worded affirmation items as they relate to psychological well-being and symptoms of depression and anxiety, outcomes which have been extensively studied in relation to ethnic identity (Rivas-Drake et al., 2014).

Methods

Participants

The data for this study came from a cross-sectional study on cultural identity development among ethnic-racial minority college students. The sample consisted of 280 participants (75.5% female; Mage = 20.95 years; SD = 1.98 years, range 18–26 years). The largest ethnic-racial group consisted of Black or African Americans (68.2%), followed by Asian/Asian Americans (12.1%), Hispanic/Latinos (9.6%); the remaining 10% consisted of Black Haitian/Caribbean Islander or other. In terms of nativity, 9.3% of the sample were born outside of the U.S., 27.8% of the sample were born in the U.S. but had at least one parent born outside of the U.S., and 62.8% were, along with their parents, U.S. born. Finally, the sample included 39.6% freshmen, 17.9% sophomores, 19.6% juniors, and 22.9% seniors.

Procedures

The participants for this study were recruited from a participant pool at a large, ethnically diverse public university in the South-East during the Spring 2018, Fall 2018, and Spring 2019 semesters. Participants completed an online anonymous survey at their convenience. Only participant ID numbers were included in our data files, and these numbers were not linked with any identifying information. In exchange for their participation, participants received credit toward a university research requirement. The study was approved by the Institutional Review Board at the participating university.

Measure

Ethnic Identity Affect.

Ethnic identity affect was measured utilizing the original negatively worded 6-items (sample item: “I am not happy with my ethnicity”) affirmation subscale of the Ethnic Identity Scale (EIS; Umaña-Taylor et al., 2004). For the current study, we also include 6 positively worded versions (sample item: “I am happy with my ethnicity”) of the original negatively worded items. The positively worded items of affirmation were revised to be as close to the original negatively worded items to ensure that any association between these two variables was not due to differences in the content of the items. All items were rated on a 4-point Likert scale ranging from 1 (does not describe me at all) to 4 (describes me very well).

Psychological Well-being.

The Scales of Psychological Well-Being (Ryff & Keyes, 1995) was used to assess psychological well-being. The SPWB consists of 18 items that measure aspects of wellbeing including autonomy, environmental mastery, personal growth, positive relations with others, purpose in life, and self-acceptance (sample item: “I have confidence in my opinions, even if they are contrary to the general consensus”). Items are rated on a 6-point Likert scale ranging from 1 (strongly disagree) to 6 (strongly agree) and averaged to form a composite score (α = .83).

Depressive symptoms.

Depressive symptoms were assessed with the Center for Epidemiologic Studies Depression Scale (CES-D; Radloff, 1977). The CES-D consists of 20 items that measure symptoms associated with depression that one may have experienced in the past week (sample item: “I was bothered by things that usually don’t bother me”). Items are rated on a 4-point Likert scale, scores ranged from 1 (strongly disagree) to 5 (strongly agree) and averaged to form a composite score (α = .87).

Anxiety.

Symptoms of anxiety were measured with the Beck Anxiety Inventory (BAI; Beck et al., 1988) adapted for use with non-clinical populations by removing items that referred to clinical levels of symptoms and adding items referring to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) criteria for generalized anxiety disorder such as excessive worrying and “butterflies” in one’s stomach (American Psychiatric Association, 2000). This modified version of the BAI consists of 18-items to assess degree to which participants experience various symptoms of anxiety (sample item: “I have been worrying a lot this week”). Items are rated on a 5-point Likert scale, scores ranged from 1 (strongly disagree) to 5 (strongly agree) and averaged to form a composite score (α = .95).

Analytic Steps

The analytic process proceeded in two steps and was conducted in Mplus v8.0 (Muthén & Muthén, 2017). Given the 4-point item scale coupled by the presence of non-normality, items were treated as categorical and thus all models were estimated with a WLSMV estimation method. Missing data was minimal (0% to 10%) and handled with full-information maximum likelihood estimation (FIML; Collins et al., 2001). Model fit was evaluated using the comparative fit index (CFI), the root mean square error of approximation (RMSEA), and the standardized root mean square residual (SRMR). According to values suggested by Little (2013), good fit is represented as CFI ≥ .95, RMSEA ≤ .06, and SRMR ≤ .06 and adequate fit is represented as CFI = .90-.95, RMSEA = .06-.08, and SRMR = .06-.08. We did not use the χ2 value to gauge model fit because it tests a null hypothesis of perfect fit, which is rarely plausible with large samples or complex models (Davey & Savla, 2010).

First, in order to establish whether positively and negatively worded items of affirmation represented a unitary construct or separate dimensions, we utilized confirmatory factor analysis (CFA) to compare a 1-factor solution and a 2-factor solution (i.e., positive and negative ERI affect). For the purposes of model identification, the variance of each factor was set to 1.0. To compare these models, we relied on the ΔCFI (>.010) and ΔRMSEA (>.010) criteria to determine significant change in model fit (Little, 2013). In interpreting factors, a cutoff of ≥ .40 was used to determine salient loadings. Additionally, we estimated a bifactor solution that posited a primary affirmation factor with two lower order specific factors that account for potential shared variance due to similar item wording (Reise et al., 2007). Within a bifactor model, the method factors (i.e., positive and negative) are restricted to be uncorrelated with each other and uncorrelated with the general affirmation factor.

Because the 2-factor model and the bifactor model do not represent nested models, we relied on general model fit and a variety of indicators that estimate the degree to which each factor within the bifactor model contributes to the reliable variance. Specifically, we utilized three model-based indicators; the percent of uncontaminated correlations (PUC), percent of explained common variance (ECV), and the coefficient omega hierarchical (ωH) values (Reise et al. 2010). PUC represents the overall percentage of covariance that reflects variance from the general dimension. ECV represents the proportion of all common variance explained by a specific factor. For the specific method effects, ECV represents the strength of a specific factor relative to all explained variance only of the items loading onto that specific factor (Stucky & Edelen, 2015). Finally, ωH reflects the percentage of systematic variance attributable to individual differences each factor. For the specific factor, ωH reflects the proportion of reliable systematic variance of a subscale score after partitioning out variability attributed to the general factor.

When PUC and the ECV value for the general factor is greater than .70, common variance can be regarded as unidimensional (Rodriguez et al., 2015). When PUC is less than .80, ECV value for the general factor greater than .70 coupled by an ωH value for the general factor greater than .70 suggest the presence of multidimensionality - although not sufficient to rule out unidimensionality (Reise et al. 2013). Beyond these mode-based indicators, the adequacy of a bifactor model can also be determined in relation to its parameter estimates. Towards this end, we relied on standardized factor loadings and the item explained common variance (IECV). As recommended by Reise et al. (2010), when items load more strongly onto the general factor than on their respective method factor, there is support for a unidimensional scoring scheme. The IECV values provide the extent to which the variance of a specific item is accounted by the general dimension alone, with values greater than .80 suggesting a unidimensional item.

In the second step, building on the championed model, we utilized a structural equation modeling (SEM) to establish predictive and incremental validity by exploring the relationship between latent ERI affective subscales and observed measures of adaptive and maladaptive psychosocial functioning (i.e., psychological well-being, symptoms of depression and anxiety). To do this, we retained the measurement model from step 1 and added psychological well-being and symptoms of depression and anxiety as observed outcomes.

Results

Step 1 –Factor Structure of the Revised Ethnic Identity Scale

As indicated in Table 1, the 2-factor model1 was not only associated with good fit [χ2 (53) = 117.30, p<.001; CFI = .990; RMSEA = .066; SRMR = .038] but was a significant improvement upon the 1-factor model [ΔCFI = .067; ΔRMSEA = .114]. Latent bivariate factor correlation between positively and negatively worded items (henceforth referred to as positive and negative ERI Affect, respectively) correlating at −.634. Additionally, we calculated McDonald’s omega (McDonald, 2013), an indicator of factor reliability, for each of the subscales. Interpreted like Cronbach’s alphas, McDonald’s omega was .94 and .96 for positive and negative ERI affect, respectively.

Table 1.

Overview of Model fit for Confirmatory Factor Models (CFA)

Model χ2 (df) CFI ΔCFI RMSEA (90% CI) ΔRMSEA TLI SRMR
1-Factor Model 542.13 (54)* .923 .180 (.166 to .194) .905 .121
2-Factor Model 117.30 (53)* .990 .067 .066 (.050 to .082) .114 .987 .038
BiFactor Model1 100.03 (43)* .991 -- .069 (.051 to .087) -- .986 .034
Open in a new tab

Note.

*

p < .050.

1

The loading of POS #6 on the specific methods factor was constrained to 0 in order to eliminate the presence of a potential linear dependency.

Next, we estimated a bifactor model with a general affirmation factor and two-method effects (i.e., positive and negative effects). However, this bifactor model produced a linear dependency associated with item 6 (“I feel positive about my ethnicity”). Given that this item did not significantly load onto the specific method factor, we constrained the factor loading and estimated the model. Model fit for this bifactor model was comparable to that of the 2-factor model [χ2 (43) = 100.030, p<.001; CFI = .991; RMSEA = .069; SRMR = .034]. The PUC was at .54 and ECV of the general factor was .66, suggesting that 54% of the covariance reflected variance from the general dimension and that the general affirmation factor only accounted for 66% variance in affirmation items.

That being said, as indicated in Table 1, standardized factor loadings for the negatively worded items were nearly all higher on their respective specific effect than on the general factor. Indeed, as indicated by IECV values (.371 to .484), the negatively worded method effect accounted for the majority of the variance (59%) in these items, indicating that negative ERI affect may represent a unique construct over and above ERI Affirmation. In contrast, standardized factor loadings for the positively worded items were not only all higher on the general factor than the respective specific factor, but loadings on the specific factor were either non-significant or below the typical meaningful thresholds (e.g., |.40|, Brown, 2015). Unsurprisingly, the majority of the variance within the positively worded items were captured by the general factor (.853–1.00), with the positive ERI affect specific factor accounting for only 9.6% of the variance among the positively worded items.

Taken together and given that the ωH value for the general factor was .780, based on criteria suggested by Reise and colleagues (2013), there was sufficient multidimensionality to completely rule out unidimensionality. At the same time, although the positive ERI affect specific factor largely represented residual “nuisance” variance, there was still a significant amount of variance captured by the negative ERI affect specific factor. Although the 2-factor model was more consistent with our hypothesis that positive and negative affect represent distinct yet related components of ones’ ERI, given that the bifactor model makes no a priori assumption on the relative footing of either the general factor or the specific factor and allows both the general factor and the specific factors to compete in explaining item variance (Reise et al., 2010), we championed the bifactor model.

Step 2 – Differential Effects of Ethnic-Racial Affirmation

Next, to assist in determining whether the variance extracted by the specific factors represented meaningful constructs, we examine the association between the latent variables with psychological well-being and symptoms of depression and anxiety. Bivariate correlations between latent ERI affect and psychosocial functioning were estimated in Mplus. As indicated in Table 3, the general ERI affirmation factor was negatively associated with symptoms of depression (r = −.129) and anxiety (r = −.179) and positively associated with psychological well-being (r = .493). However, the negative ERI affect specific factor was positively, and more strongly, associated with symptoms of depression (r = .418) and anxiety (r = .379), and negatively associated with psychological well-being (r = −.231). Finally, and surprisingly given the low factor loadings and reliability, the positive ERI affect specific factor was significantly and positively associated with symptoms of depression (r = .277) and anxiety (r = .252).

Table 3.

Bivariate Correlations Between Latent ERI Affirmation and Observed Outcomes

Variable M (SD) 2 3 4 5 6
1. General Factor -- 000 (.000) .000 (.000) −.129 (.058)* −.179 (.068)* .493 (.056)***
2. Negative ERI Affect -- -- .000 (.000) .418 (.065)*** .379 (.075)*** −.231 (.067)***
3. Positive ERI Affect -- -- .277 (.089)** .252 (.096)* −.029 (.097)
4. Symptoms of Depression 1.354 (.093) -- .760 (.023)*** −.388 (.044)***
5. Symptoms of Anxiety 2.480 (.814) -- −.399 (.055)***
6. Psychological Well-being 4.436 (.221) --
Open in a new tab

Note. Bivariate correlations between latent ERI affirmation and observed outcomes were estimated in Mplus utilizing Tech 4. For the purposes of model identification, the mean and variance of each factor is set to 0 and 1.0, respectively. Standard errors are reported in parentheses.

*

p < .05

**

p <.010

***

p<.001

Next, to determine the incremental validity, we estimated an SEM model to examine the unique effect of the general ERI affirmation and the positive and negative ERI affect specific factors, with gender and age included as covariates. Towards that end, we estimated the SEM model utilizing the bifactor measure model. This model was also associated with good fit [χ2 (88) = 149.093, p = .001; CFI = .990; RMSEA = .050; SRMR = .034]. As indicated in Table 4, and consistent with the bivariate correlations, the general ERI affirmation factor was negatively associated with symptoms of depression (β = −.138, p = .012) and anxiety (β = −.198, p = .002) and positively associated with psychological well-being (β = .487, p <.001). At the same time, the negative ERI affect specific factor was none the less positively and more strongly associated with symptoms of depression (β = .426, p <.001) and anxiety (β = .392, p <.001) and negatively associated with psychological well-being (β = −.242, p < .001). Finally, and surprisingly, given that the positive ERI affect specific factor largely reflected minimal residual variance, the positive ERI affect specific factor was none the less positively associated with symptoms of depression (β = .236, p = .007) and anxiety (β = .249, p = .010).

Table 4.

Differential Standardized Effects of Ethnic-Racial Affect Utilizing the Bifactor Model

Outcome Predictor Estimate (S.E.) 95% CI p-value
Symptoms of Depression General Factor −.138 (.055) −.245 to −.031 .012
Positive Affect .236 (.088) .063 to .408 .007
Negative Affect .426 (.071) .286 to .566 <.001
Age .013 (.077) −.138 to .164 .865
Gender −.012 (.056) −.123 to .099 .829
Symptoms of Anxiety General Factor −.198 (.064) −.324 to −.073 .002
Positive Affect .249 (.097) .060 to .439 .010
Negative Affect .392 (.082) .231 to .552 <.001
Age .014 (.071) −.126 to .154 .844
Gender .046 (.064) −.079 to .170 .475
Psychological Well-Being General Factor .487 (.053) .384 to .591 <.001
Positive Affect .006 (.093) −.176 to .188 .945
Negative Affect −.242 (.064) −.366 to −.117 <.001
Age .064 (.064) −.060 to .189 .313
Gender .081 (.070) −.057 to .219 .251
Open in a new tab

To further probe these counterintuitive finding, given the lack of interpretability associated with this factor, we re-estimated the bifactor model after constraining the effects of positive ERI affect specific factor on psychosocial functioning to 0. The revised model was associated with good model fit [χ2 (91) = 155.263, p < .001; CFI = .990; RMSEA = .050; SRMR = .036] and there was no substantive drop in model fit [ΔCFI = .001; ΔRMSEA < .001]. In contrast, although a model with the effects of both negative and positive ERI affect constrained to 0 was still associated with good to mediocre model fit [χ2 (94) = 342.479, p < .001; CFI = .961; RMSEA = .097; SRMR = .366], there was a substantive drop in fit [ΔCFI = .030; ΔRMSEA = .047].

Discussion

The Ethnic Identity Scale (EIS; Umaña-Taylor et al., 2004) was specifically developed to distinguish between developmental and content components of ERI. However, the affirmation subscale is composed entirely of negatively worded items, measuring negative feelings about one’s ethnic-racial group, rather positive feelings about one’s ethnic-racial group. The exclusive focus on negatively worded items not only limits the interpretation of the associations between the EIS’ affirmation subscale and psychosocial functioning, but also makes it difficult to determine the dimensionality of ERI affect. Accordingly, the current study sought to determine whether affirmation represented a unidimensional construct, a bidimensional construct, or a combination of the two and determine the differential utility of positively and negatively worded ERI affect items. As a whole, the bifactor model provided evidence for both unidimensionality and multidimensionality. Indeed, although positively worded and negatively worded items of “affirmation” significantly loaded onto a general factor representing affirmation, there was still a significant amount of variance captured by the negative ERI affect specific factor, indicating the presence of multidimensionality.

The variance captured by the general factor across positively and negatively worded items reflects more closely what researchers have traditionally conceptualized as affirmation (Schwartz et al., 2014) or private regard (Sellers et al., 1998). Consistently, the overwhelming majority of the variance (85% to 100%) among positively worded items was captured by the general factor. In contrast, the residual variance captured by the negative ERI affect specific factor may represent feelings and sentiments akin to self-hatred, or the extent to which individuals dislike their ethnic-racial group, which is a core component of Nigrescence theory (Cross & Vandiver, 2001; Worrell et al., 2001). According to Cross (1971), self-hatred occurs during the early, pre-encounter stage, of racial identity development whereby individuals embrace anti-Black attitudes emphasized within the broader mainstream society resulting in low self-esteem and overall maladaptive functioning. The fact that negatively worded items loaded more strongly onto the negative ERI affect specific factor may indicate that the presence or absence of positive feelings towards ones’ ethnic/racial group is not equivocal to presence or absence of self-hatred directed at ones’ ethnic/racial group.

Although these may indicate a theoretical distinction between ERI positive and negative affect, it is worth noting that these findings may stem from the current Likert scale. As previously noted, endorsing the option “Does not describe me at all” for a positively worded ERI item such as “I am happy with my ethnicity” does not constitute endorsement of the antithesis. Taken together, this manuscript emphasizes the need for a reexamination within the ERI literature on how we measure, and possibly, conceptualize ERI affect. Indeed, a clear implication of the current findings at the very least, is that EIS, as currently established, has drawn on items that contained shared variance reflecting ERI affirmation and substantive shared variance reflecting either a distinct construct (e.g., negative ERI affect, self-hatred, etc.) or measurement error. If similar findings emerge even after moving to a Likert scale that ranges from endorsement to opposition (e.g., strongly disagree to strongly agree), it may be necessary for the ERI literature to move beyond isolated terms such as affirmation, self-hatred, and private regard and draw on a broader conceptualization focused on ones’ positive and negative affect towards ones’ ERI.

Providing greater support for the presence of multidimensionality, results also indicated important differential associations between the general affirmation ERI factor and the negative ERI affect specific factor across psychological well-being and symptoms of depression and anxiety. To begin with, and consistent with social identity theory (Tajfel, 1981), multidimensional model of Black identity (Sellers et al., 1998), Nigrescence (Cross & Vandiver, 2001; Worrell et al., 2001), and the broader literature as a whole (for review, see Rivas-Drake et al., 2014), the general ERI affirmation factor was negatively associated with symptoms of depression and anxiety and positively associated with psychological well-being. These findings provide further support for the role that youths’ feelings about their ethnic-racial group (e.g., racial pride, private regard, affirmation) plays in overall psychosocial functioning among ethnic/racial minority youth.

At the same time, results also indicated that the negative ERI affect specific factor was negatively associated with psychological well-being and positively associated with both symptoms of depression and anxiety. In doing so, the unique effects of the negative ERI affect specific factor on psychosocial adjustment provides some evidence that the negative ERI affect specific factor may represent more than just nuisance variance. Indeed, consistent with Nigrescence theory (Cross & Vandiver, 2001), the negative ERI affect specific factor was more strongly associated with symptoms of depression and anxiety then the general ERI affirmation factor. As it relates to psychopathology, findings seem to indicate that it may not be positive feelings regarding one’s ethnicity/race that is particularly associated with symptoms of depression and anxiety, but instead, a lack of negative feelings. This notion has been previously suggested and supported in a study by Cross et al. (2012), which found a similar pattern when examining ethnic-racial salience and affect.

On the other hand, the general ERI affirmation factor was more strongly associated with psychological well-being than the negative ERI affect specific factor. This positive association between ERI affirmation and psychological well-being is highly consistent with social identity theory (Tajfel, 1981). Indeed, SIT postulates that an individuals’ well-being is informed by their capacity to establish a sense of group distinctiveness by favorably comparing their in-group against an out-group. In the context of marginalized groups however, because individuals are unable to change their group membership, individuals are tasked with adopting “social creativity” strategies that reframe their negative status more positively (Tajfel & Turner, 1986), As a result, individuals’ capacity to maintain a positive self-evaluation is largely contingent on their capacity to view their ethnic/racial group favorably. Consistently, the degree to which one has developed a sense of pride and belonging to ones’ ethnic/racial group is more likely to bolster youths’ well-being.

Limitations and Future Directions

The present results should be interpreted in light of several limitations. First, it is important to note that positively worded items were simply reworded and slightly modified versions of the original negatively worded affirmation items. Although this resulted in items that more closely paralleled the original items, thereby increasing the robustness of the findings, it may serve as a potential confound. Future research should utilize existing scales that tap into positive (i.e., Private Regard) and negative (negatively worded EIS subscale) affect. Second, and as previously noted, the presence of multidimensionality may, in part, be due to the utilization of a neutral anchor within the EIS scaling. Endorsing the option “Does not describe me at all” for a negatively worded ERI item does not constitute endorsement of the antithesis. Future studies are necessary to replicate the findings utilizing scale options that may more closely reflect a continuum of endorsement.

Third, whereas the sample was fairly ethnically and racially diverse, the small sample size made it impossible to examine whether there were differences across ethnic-racial minority groups. Additionally, the current sample was predominantly African American. Although research supports measurement equivalence of the EIS across ethnic-racial groups (Sladek et al., 2020), prior research has found differential effects between ERI components across ethnic-racial groups (Rivas-Drake et al., 2014). Future research is necessary to explore whether the dimensionality of ERI affirmation is similar across individuals from various ethnic-racial groups. Finally, it is important to contextualize the current findings within the context of the developmental period of emerging adulthood. Although emerging adulthood has been increasingly recognized as a key period for ERI development (Syed, 2010), it is important to note that ERI represents a development process (Umaña-Taylor et al., 2014). It is possible that the presence of multidimensionality that we observed may be developmentally specific to emerging adulthood. Future research is necessary to examine this distinction among adolescents.

The findings of this study have implications for future research and prevention and intervention efforts among ethnic-racial minority youth. First, a better understanding of the dimensionality of ERI affect allows for stronger conceptual and methodological approaches to the study of ERI. Further, increasing our understanding of how ERI affect is associated to psychosocial functioning may provide important avenues for the development of culturally relevant prevention and intervention and more precise evaluation of existing intervention efforts (e.g., Umaña-Taylor et al., 2018). Finally, there is a substantial body of literature that highlights the role of positive ethnic-racial affect with psychosocial functioning (Rivas-Drake et al., 2014). However, should the current findings be replicated, findings from the current study highlight a more nuanced association between dimensions of ERI affect and psychological functioning.

Conclusion

Despite these limitations, the current study has important theoretical and methodological contributions to the broader literature on ERI. Indeed, the findings from the current study further expand our understanding of the multidimensionality of ethnic-racial identity. Moreover, the findings from the current study highlight the need for examination of how we measure ERI affect at the very least, and possibly how we conceptualize it within the broader ERI literature. We hope that the present study will inspire a line of research in this direction.

Supplementary Material

Supplemental Material

Table 2.

Parameter Estimates for Positively Worded and Negatively Worded Affirmation Items

Item Wording Item/Factor 2-Factor Model Bifactor Model
General Method IECV
I am happy with my ethnicity. Pos01 .928 (.042) .900 (.025) .307 (.067) 0.896
If I could choose, I would not change my ethnicity. Pos02 .725 (.032) .707 (.045) .199 (.083) 0.927
I do not wish I was of a different ethnicity. Pos03 .737 (.022) .713 (.039) .239 (.092) 0.899
I like my ethnicity. Pos04 .852 (.039) .815 (.039) .338 (.081) 0.853
My feelings about my ethnicity are mostly positive. Pos05 .889 (.023) .925 (.027) −.226 (.176) 0.944
I feel positive about my ethnicity Pos06 .950 (.032) .946 (.030) .000 (.00) 1.000
My feelings about my ethnicity are mostly negative. Neg01R .817 (.017) .556 (.050) .574 (.050) 0.484
I feel negatively about my ethnicity. Neg02R .877 (.039) .596 (.056) .627 (.047) 0.475
I wish I were of a different ethnicity. Neg03R .943 (.036) .606 (.056) .718 (.042) 0.416
I am not happy with my ethnicity. Neg04R .848 (.027) .523 (.046) .681 (.046) 0.371
If I could choose, I would prefer to be of a different ethnicity. Neg05R .916 (.018) .564 (.059) .735 (.041) 0.371
I dislike my ethnicity. Neg06R .894 (.015) .550 (.069) .716 (.056) 0.371
Open in a new tab

Note. All loadings were significant and standard errors are reported in parentheses. IECV item explained common variance

Public Significance Statements.

A substantial body of research has found important relationships between positive feelings about one’s ethnic-racial group (e.g., racial pride, private regard, affirmation) and psychosocial and health outcomes for ethnic-racial minority youth. The current study sought to examine the functionality of a revised affirmation subscale of the Ethnic Identity Scale, a widely utilized measure of ethnic-racial identity, with both positively and negatively worded items. Results indicated negative ERI affect represents something distinct, yet related to ERI affirmation as whole, thereby highlighting the need for examination of how we measure ERI affect at the very least, and possibly how we conceptualize it within the broader ERI literature.

Acknowledgement

Julie Rodil was bright and caring third year doctoral student in the Health Psychology program at Old Dominion University (ODU). In the last 3 years, Julie has made significant contributions to the literature, breaking new ground in her work on helping establish the United States Identity Scale, examining identity shifting, and in her Master’s Thesis, which turned its attention of developmental and cultural processes of the oppressor as it relates to understanding race, racism, and importance of diversity. Beyond her academic contribution, Julie was a mentor to the undergraduate and graduate students within the TARDIS lab, a friend, a sister, a daughter, and so much more. The world has lost a bright young scholar and an amazing and caring person.

Research reported in this publication was supported by the National Institute of General Medical Sciences of the National Institutes of Health under Award Number T34GM118259. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Matching funds from Old Dominion University are also acknowledged.

Footnotes

1

Path models with standardized estimates for the 2-factor and bifactor models can be found within the Supplemental Materials.

References

  1. Alessandri G, Vecchione M, Eisenberg N, & Laguna M (2015). On the factor structure of the Rosenberg (1965) General Self-Esteem Scale. Psychological Assessment, 27(2), 621–635. doi: 10.1037/pas0000073 [DOI] [PubMed] [Google Scholar]
  2. American Psychiatric Association (2000). Diagnostic and statistical manual of mental disorders (4th ed.). American Psychiatric Association. [Google Scholar]
  3. Beck AT, Epstein N, Brown G, & Steer RA (1988). An inventory for measuring clinical anxiety: Psychometric properties. Journal of Consulting and Clinical Psychology, 56(6), 893–897. 10.1037/0022-006X.56.6.893 [DOI] [PubMed] [Google Scholar]
  4. Cokley K (2007). Critical issues in the measurement of ethnic and racial identity: A referendum on the state of the field. Journal of Counseling Psychology, 54(3), 224–234. 10.1037/0022-0167.54.3.224 [DOI] [Google Scholar]
  5. Collins LM, Schafer JL, & Kam C-M. (2001). A comparison of inclusive and restrictive strategies in modern missing data procedures. Psychological Methods, 6(4), 330–351. 10.1037/1082-989X.6.4.330. [DOI] [PubMed] [Google Scholar]
  6. Cross WE (1971). The Negro-to-Black conversion experience: Toward a psychology of Black liberation. Black World, 20(9), 13–27 [Google Scholar]
  7. Cross WE, Grant BO, & Ventuneac A (2012). Black identity and well-being: Untangling race and ethnicity. In Sullivan JM & Esmail AM (Eds.), African American identity: Racial and cultural dimensions of the Black experience. (pp. 125–146). Lexington Books. [Google Scholar]
  8. Cross WE Jr., & Vandiver BJ (2001). Nigrescence theory and measurement: Introducing the Cross Racial Identity Scale (CRIS). In Ponterotto JG, Casas JM, Suzuki LA, & Alexander CM (Eds.), Handbook of multicultural counseling (pp. 371–393). Sage Publications, Inc. [Google Scholar]
  9. Davey A, & Savla J (2010). Statistical power analysis with missing data: A structural equation modeling approach. Routledge/Taylor & Francis Group. [Google Scholar]
  10. Erikson EH (1950). Childhood and Society. W. W. Norton & Company [Google Scholar]
  11. García Coll C, Lamberty G, Jenkins R, McAdoo HP, Crnic K, Wasik BH, & Vázquez García H (1996). An integrative model for the study of developmental competencies in minority children. Child development, 67(5), 1891–1914. 10.2307/1131600 [DOI] [PubMed] [Google Scholar]
  12. Gibbons RD & Hedeker DR (1992). Full-information item bi-factor analysis. Psychometrika, 57, 423–436. 10.1007/BF02295430 [DOI] [Google Scholar]
  13. Kogan SM, Yu T, Allen KA, & Brody GH (2014). Racial microstressors, racial self-concept, and depressive symptoms among male African Americans during the transition to adulthood. Journal of Youth and Adolescence, 44(4), 898–909. 10.1007/s10964-014-0199-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Little TD (2013). Longitudinal structural equation modeling. New York, NY: Guilford Press. [Google Scholar]
  15. Marcia JE (1966). Development and validation of ego-identity status. Journal of Personality and Social Psychology, 3(5), 551–558. 10.1037/h0023281 [DOI] [PubMed] [Google Scholar]
  16. McDonald RP (2013). Test theory: A unified treatment. Psychology Press. [Google Scholar]
  17. Muthén LK and Muthén BO (1998–2017). Mplus User’s Guide. Eighth Edition. Los Angeles, CA: Muthén & Muthén [Google Scholar]
  18. Phinney JS (1989). Stages of ethnic identity development in minority group adolescents. The Journal of Early Adolescence, 9(1–2), 34–49. 10.1177/0272431689091004 [DOI] [Google Scholar]
  19. Phinney JS (1992). The multigroup ethnic identity measure: A new scale for use with diverse groups. Journal of Adolescent Research, 7(2), 156–176. 10.1177/074355489272003 [DOI] [Google Scholar]
  20. Phinney JS, & Ong AD (2007). Conceptualization and measurement of ethnic identity: Current status and future directions. Journal of Counseling Psychology, 54(3), 271–281. 10.1037/0022-0167.54.3.271 [DOI] [Google Scholar]
  21. Piña-Watson B, Martinez AJ, Cruz LN, Llamas JD, Lopéz BG (2018). Ethnic identity affirmation as a strength for Mexican descent academic outcomes: Psychological functioning and academic attitudes as mediators. Psychology of Schools, 55(10), 1155–1170. 10.1002/pits.22188 [DOI] [Google Scholar]
  22. Radloff LS (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1(3), 385–401. 10.1177/014662167700100306 [DOI] [Google Scholar]
  23. Reise SP, Bonifay WE, & Haviland MG (2013). Scoring and modeling psychological measures in the presence of multidimensionality. Journal of Personality Assessment, 95(2), 129–140. 10.1080/002223891.2 [DOI] [PubMed] [Google Scholar]
  24. Reise SP, Moore TM, & Haviland MG (2010). Bifactor models and rotations: Exploring the extent to which multidimensional data yield univocal scale scores. Journal of Personality Assessment, 92(6), 544–49. 10.1080/00223891.2010.496477 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Reise SP, Morizot J, & Hays RD (2007). The role of the bifactor model in resolving dimensionality issues in health outcomes measures. Quality of Life Research: An International Journal of Quality of Life Aspects of Treatment, Care & Rehabilitation, 16(Suppl1), 19–31. 10.1007/s11136-007-9183-7 [DOI] [PubMed] [Google Scholar]
  26. Rivas-Drake D, Seaton EK, Markstrom C, Quintana S, Syed M, Lee RM, Schwartz SJ, Umaña-Taylor AJ, French S, Yip T, & Ethnic and Racial Identity in the 21st Century Study Group (2014). Ethnic and racial identity in adolescence: implications for psychosocial, academic, and health outcomes. Child development, 85(1), 40–57. 10.1111/cdev.12200 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Rodriguez A, Reise SP, & Haviland M (2015). Evaluating bifactor models: Calculating and interpreting statistical indices. Psychological Methods, 21(2), 137–150. 10.1037/met0000045 [DOI] [PubMed] [Google Scholar]
  28. Ryff CD, & Keyes CLM (1995). The structure of psychological well-being revisited. Journal of Personality and Social Psychology, 69(4), 719–727. 10.1037/0022-3514.69.4.719 [DOI] [PubMed] [Google Scholar]
  29. Schwartz SJ, Syed M, Yip T, Knight GP, Umaña-Taylor AJ, Rivas-Drake D, & Lee RM (2014). Methodological issues in ethnic and racial identity research with ethnic minority populations: theoretical precision, measurement issues, and research designs. Child development, 85(1), 58–76. 10.1111/cdev.12201 [DOI] [PubMed] [Google Scholar]
  30. Scottham KM, Sellers RM, Nguyên HX (2008). A measure of racial identity in African American adolescents: The development of the Multidimensional Inventory of Black Identity -- Teen. Cultural Diversity & Ethnic Minority Psychology, 14(4), 297–306. 10.1037/1099-9809.14.4.297 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Sellers RM, Smith MA, Shelton JN, Rowley SAJ, & Chavous TM (1998). Multidimensional model of racial identity: A reconceptualization of African American racial identity. Personality and Social Psychology Review, 2(1), 18–39. 10.1207/s15327957pspr0201_2 [DOI] [PubMed] [Google Scholar]
  32. Sladek MR, Umaña-Taylor AJ, McDermott ER, Rivas-Drake D, & Martinez-Fuentes S (2020). Testing invariance of ethnic-racial discrimination and identity measures for adolescents across ethnic-racial groups and context. Psychology Assessment. 32(6), 590–526. 10.1037/pas0000805 [DOI] [PubMed] [Google Scholar]
  33. Stucky BD, & Edelen MO (2015). Using hierarchical IRT models to create unidimensional measures from multidimensional data. In Reise SP & Revicki DA (Eds.), Multivariate applications series. Handbook of item response theory modeling: Applications to typical performance assessment (p. 183–206). Routledge/Taylor & Francis Group. [Google Scholar]
  34. Syed M (2010). Developing an integrated self: Academic and ethnic identities among ethnically diverse college students. Developmental Psychology, 46(6), 1590–1604. 10.1037/a0020738 [DOI] [PubMed] [Google Scholar]
  35. Syed M, & Mitchell LL (2013). Race, ethnicity, and emerging adulthood: Retrospect and prospects. Emerging Adulthood, 1(2), 83–95. 10.1177/2167696813480503 [DOI] [Google Scholar]
  36. Tajfel H (1981). Human groups and social categories: Studies in social psychology. Cambridge University Press. [Google Scholar]
  37. Tajfel H and Turner JC (1986) The Social Identity Theory of Intergroup Behavior. In Worchel S and Austin WG (Eds.), Psychology of Intergroup Relation. Hall Publishers. [Google Scholar]
  38. Umaña-Taylor AJ, Kornienko O, Bayless SD, & Updegraff KA (2018). A universal intervention program increases ethnic-racial identity exploration and resolution to predict adolescent psychosocial functioning one year later. Journal of Youth and Adolescence, 47, 1–15. 10.1007/s10964-017-0766-5 [DOI] [PubMed] [Google Scholar]
  39. Umaña-Taylor AJ, Quintana SM, Lee RM, Cross WE Jr., Rivas-Drake D, Schwartz SJ, Syed M, Yip T, Seaton E, & Ethnic and Racial Identity in the 21st Century Study Group (2014). Ethnic and racial identity during adolescence and into young adulthood: An integrated conceptualization. Child Development, 85(1), 21–39. 10.1111/cdev.12196 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Umaña-Taylor AJ, Yazedjian A, & Bámaca-Gómez M (2004). Developing the ethnic identity scale using Eriksonian and social identity perspectives. Identity: An International Journal of Theory and Research, 4, 9–38. 10.1207/S1532706XID0401_2 [DOI] [Google Scholar]
  41. Vandiver BJ (2001). Psychological nigrescence revisited: Introduction and overview. Journal of Multicultural Counseling and Development, 29(3), 165–173. 10.1002/j.2161-1912.2001.tb00515.x [DOI] [Google Scholar]
  42. Vandiver BJ, Cross WE Jr., Worrell FC, & Fhagen-Smith PE (2002). Validating the Cross Racial Identity Scale. Journal of Counseling Psychology, 49(1), 71–85. 10.1037/0022-0167.49.1.71 [DOI] [Google Scholar]
  43. Wang Y, Douglass S, & Yip T (2017). Longitudinal relations between ethnic/racial identity process and content: Exploration, commitment, and salience among diverse adolescents. Developmental Psychology, 53(11), 2154–2169. 10.1037/dev0000388 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. White RMB, Umaña-Taylor AJ, Knight GP, & Zeiders KH (2011). Language measurement equivalence of the Ethnic Identity Scale with Mexican American early adolescents. The Journal of Early Adolescence, 31(6), 817–852. 10.1177/0272431610376246 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Worrell FC, Cross WE Jr., & Vandiver BJ (2001). Nigrescence theory: Current status and challenges for the future. Journal of Multicultural Counseling and Development, 29(3), 201–213. 10.1002/j.2161-1912.2001.tb00517.x [DOI] [Google Scholar]
  46. Yip T, Douglass S, & Sellers RM (2014). Ethnic and racial identity. In Leong FTL, Comas-Díaz L, Nagayama Hall GC, McLoyd VC, & Trimble JE (Eds.), APA handbook of multicultural psychology, Vol. 1. Theory and research (pp. 179–205). Washington, DC, US: American Psychological Association. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Material
NIHMS1718169-supplement-Supplemental_Material.pdf (352.5KB, pdf)

RESOURCES