Dimensions of Academic Contingencies Among African American College Students

Abstract

Drawing from existing literature, the authors conceptualized a two-dimensional framework of African American students’ academic contingencies of self-worth. The results from exploratory and confirmatory factor analyses with a sample of African American college freshmen (N = 330) supported this prediction. Self-Worth Dependent academic contingencies were characterized by students’ self-worth being more wholly reliant on academic performance. Self-Worth Enhancing academic contingencies represented linkages between self-worth and academics where self-worth is reinforced by positive achievement but not necessarily diminished by negative feedback in the educational setting. Correlational and multiple regression analyses were used to test relationships between academic contingencies of self-worth dimensions, race-related factors (identity and discrimination experiences), academic outcomes, and psychological outcomes. Findings showed that Self-Worth Dependent academic contingencies predicted less positive beliefs about Blacks and worse psychological outcomes. Self-Worth Enhancing academic contingencies predicted more positive personal beliefs about Blacks and feeling more connected to Blacks. Implications for African American student motivation, achievement, and adjustment are discussed.

Keywords

contingencies of self-worth African American college students academic achievement academic motivation

Over and above the stressors that all students face, racial prejudice, stereotyping, and discrimination pose situational threats that can adversely influence African American students’ educational outcomes (Chavous et al., 2003; Sellers & Shelton, 2003; Smalls, White, Chavous, & Sellers, 2007). The prevalence of racial biases in education has led some scholars to suggest that many African American students perceive incompatibility between their personal identities and education, leading them to disconnect their personal assessments of worth from the educational domain. It is further posited that disconnecting one’s self-worth from scholastic achievement compromises academic motivation and leads to negative educational outcomes over time (e.g., Crocker & Major, 1989; Steele & Aronson, 1995). At the same time, however, there is empirical evidence indicating that students who strongly connect their self-worth to academics are at academic and psychological risk (e.g., Crocker & Luhtanen, 2003; Crocker, Luhtanen, Cooper, & Bouvrette, 2003; Crocker, Quinn, Karpinski, & Chase, 2003; Crocker, Sommers, & Luhtanen, 2002; Niiya, Crocker, & Bartmess, 2004). From these two perspectives, African American students face risks if their self-worth is linked too weakly or too strongly to education.

In response to this seeming paradox, researchers have asserted the need to assess variation among African American students as well as social and cultural factors unique to African Americans in studying connections between their self-worth and academics (Cokley, 2003; Cokley, Komarraju, King, Cunningham, & Muhammad, 2003; Graham, 1994). Such approaches can reveal complexity in attitudes and behaviors that comparative research tends to obscure (Chavous et al., 2003; Cokley, 2003; Graham, 1994; Sellers, Rowley, Chavous, Shelton, & Smith, 1997; Smalls et al., 2007), making within-group investigations particularly amenable to informing conflicting theories. Accordingly, the current study integrated research on African American students’ race-related beliefs and experiences with psychological academic motivation and performance frameworks to assess within-group diversity in how African American students connect their self-worth to educational achievement. In doing so, this study speaks to inconsistencies in extant literature on the supposed advantage or disadvantage of African American students connecting their self-worth to the academic domain and can inform practical efforts to bolster African Americans’ academic motivation, achievement, and adjustment.

Academic Contingencies of Self-Worth and the “Paradox” of African American Students

The current research used Crocker and Wolfe’s (2001) contingencies of self-worth theory as a primary conceptual framework for examining connections between students’ self-worth and academic achievement. Drawing from achievement motivation scholarship such as Covington’s (1984) self-worth theory, as well as from frameworks focused on the implications of self-esteem instability (e.g., Harter & Whitesell, 1999; Kernis, Cornell, Sun, Berry, & Harlow, 1993), Crocker and colleagues argued that individuals evaluate their self-worth in various domains organized along a continuum from internal to external. Internal contingencies of self-worth, such as virtue, are associated with stable self-perceptions and more positive well-being. Internal contingencies are thought to be beneficial because self-worth is based on sources that are controllable and presumably stable (e.g., Kernis et al., 1993). In contrast, external contingencies of self-worth, which are relatively inconsistent and uncontrollable, pose psychosocial threats. According to this model, basing one’s self-worth on academics, an external domain, exposes students to risks, including heightened ego threat, unstable self-esteem, negative affect, and lower levels of psychological functioning (Crocker et al., 2002; Crocker & Luhtanen, 2003; Crocker, Luhtanen, et al., 2003; Crocker, Quinn, et al., 2003; Crocker & Wolfe, 2001; Niiya et al., 2004).

Although contingencies of self-worth theory suggests that there are few benefits to strongly linking one’s self-worth to academic outcomes (Crocker & Luhtanen, 2003), related academic identification scholarship posits that African American students underperform relative to other groups of students because they do not link their self-worth to academic achievement strongly enough (Crocker & Major, 1989; Crocker & Wolfe, 2001; Steele & Aronson, 1995). Academic identification perspectives argue that experiencing racially threatening academic contexts lead many African American students, especially those whose self-worth is strongly contingent on educational performance to academically disidentify as a means to maintaining a positive sense of self-worth (Crocker & Major, 1989; Crocker & Wolfe, 2001; Osborne & Walker, 2006; Steele & Aronson, 1995). It is also suggested that over time, disconnecting the self from academics compromises scholarly performance (Steele & Aronson, 1995).

Taken together, academic contingencies of self-worth research and academic disidentification scholarship pose conflicting positions on the utility of students connecting their self-worth to the educational domain. Complicating matters further are equivocal empirical findings within and across academic contingencies and academic identification research literatures. Some research, for example, has shown racial differences in the correlation between self-esteem and grade performance, one common measure of academic identification (e.g., Osborne & Walker, 2006). In contrast, academic contingencies of self-worth research indicates that there are no significant group differences in Black, White, and Asian American college students’ reported academic contingencies of self-worth (Crocker, Luhtanen, et al., 2003; Crocker & Wolfe, 2001; Zeigler-Hill, 2007; see also Cokley, 2003; Graham, 1994). As such, theoretically and empirically, academic contingencies of self-worth and academic disidentification frameworks inconsistently forecast how racial stigma may shape African American students’ connections between their self-worth and academics.

Academic Contingencies and Racial Socialization, Identity, and Discrimination

Importantly, the academic contingencies of self-worth and academic identification theories both consider how racial stigma may act as a meaningful influence on African American students’ academic motivation, achievement, and adjustment. Neither theory, however, delineates how race-related factors other than racial stigma may influence the ways African American students connect their self-worth to academics. To consider a wider array of race-related influences would necessitate a conceptual framework that highlights a more comprehensive sociohistorical analysis of African Americans students’ relationship to racial bias generally, and to educational attainment more specifically. In the present study, we speak to the importance of representing a more inclusive set of race-related influences by considering how racial socialization, identity, and discrimination experiences may shape academic contingencies of self-worth among African American students. In this way, our approach speaks to the dearth of research on self-processes and education among African American students. The present study also shifts the prevalent focus in studies of Black student experiences from a deficit model explaining underachievement relative to Whites and Asians, to a strengths-based framework that has the potential to reveal pathways by which many African American students perform and adjust well in college.

African Americans’ distinctive sociocultural status in the United States places group members in a precarious position with respect to educational attainment. On one hand, many African American parents and community members emphasize educational attainment as a primary means to individual and collective mobility (Cokley, 2003; Galleta & Cross, 2007; Smalls et al., 2007). African American parents and community members often additionally convey messages about the potential for racial bias and exclusion in society and in the academic domain more specifically (e.g., Bowman & Howard, 1985; Hughes et al., 2006; Neblett, Phillip, Cogburn, & Sellers, 2006). These two sets of messages—that academics are important and that the educational environment may be biased—likely influence African American students’ academic contingencies of self-worth. Socialization related to the importance of educational attainment may promote a type of academic contingency where students’ self-worth is wholly tied to academics (Self-Worth Dependent academic contingencies). Racial socialization messages about potential racial biases in society and in educational contexts may additionally lead African American students to maintain a set of more tentative connections between their self-worth and education. These more tentative linkages could allow students to simultaneously reap the benefits of academic achievement as well as protect the self when racial stigma is salient or when academic outcomes are less optimal (Self-Worth Enhancing academic contingencies).

The proposed two-dimensional academic contingencies framework is also supported by research on African American racial identity and racial discrimination experiences. For instance, research has found that African American youth (Cokley, 2003; Galleta & Cross, 2007; Graham, 1994), particularly those with stronger African American identification and racial pride (e.g., Chavous et al., 2003), often place a high personal value on education. It is likely then that students who strongly identify with being African American and have higher awareness and pride around their racial heritage would be more likely to connect their self-worth to academic achievement. Students’ identity beliefs and discriminatory experiences related to racial bias, however, may lead them to additionally develop our proposed more cautious Self-Worth Enhancing academic contingency. We propose that Self-Worth Enhancing academic contingencies may allow students’ self-evaluations to benefit from positive achievement outcomes while remaining less vulnerable to negative evaluative outcomes and experiences.

Rationale for Self-Worth Dependent and Enhancing Academic Contingencies Dimensions

In addition to being conceptually supported by research on African American racial socialization, identity, and discrimination, the two predicted academic contingencies dimensions are supported indirectly by previous contingencies of self-worth research. Specifically, Crocker, Luhtanen, et al. (2003) confirmed a seven-factor structure of the Contingencies of Self-Worth Scale in research that first validated the measure. Of importance to the current study are the researchers’ results showing that the internal reliability of the academic contingencies subscale was lower among Black students than among White and Asian American students (Crocker, Luhtanen, et al., 2003). The difference in internal reliabilities provides some indication that the scale items were answered differently for Black students relative to other groups in the sample. The authors’ exploratory and confirmatory factor analyses, however, sought to confirm distinctions across items assessing seven contingencies of self-worth domains. Their focus was not on examining variation in the factor structures of items within particular domains. Thus, in the present study we extend findings from Crocker and colleagues’ prior research by examining whether African American students conceptualize academic contingencies of self-worth in ways that are consistent with our Self-Worth Dependent and Self-Worth Enhancing representations.

The proposed two dimensions are also corroborated by our face validity assessment of the items comprising the existing academic contingencies of Self-Worth Scale. One set of items appears to assess the extent to which students’ self-worth is dependent on achievement outcomes where self-worth increases in the context of positive performance and decreases in the context of negative performance (e.g., “When I do poorly on an exam or paper, my self-esteem suffers.”). A second set of items seems to measure whether students’ self-worth increases in response to positive achievement outcomes only (e.g., “My self-esteem gets a boost when I get a good grade on exam or paper.”). As such, superficially, the existing academic contingencies of self-worth items appear to reflect the two proposed dimensions. In sum, three rationales—support from scholarship on African Americans’ race-related beliefs and experiences, evidence from previous contingencies of self-worth research, and a face validity assessment of an existing academic contingencies of self-worth measure—substantiate the two predicted academic contingencies of self-worth dimensions.

Moreover, if our proposition is supported and African American college students conceptualize academic contingencies of self-worth along two distinct dimensions, the dimensions should be differentially associated with students’ academic and psychological outcomes as well as with their racial identity beliefs and racial discrimination experiences. For instance, according to Crocker and Wolfe (2001), strong, holistic connections between one’s self-worth and education are academically and psychologically maladaptive. Consequently, higher Self-Worth Dependent academic contingencies, which most closely align with Crocker and colleagues’ conception of academic contingencies of self-worth, should be associated with negative psychological and performance outcomes. Additionally, because Self-Worth Dependent academic contingencies may be related to students’ belief that their educational triumphs are essential for the racial group’s success and for their own mobility (Smalls et al., 2007), we also expect the dimension to be positively associated with strong racial group connectedness (racial centrality).

Contrastingly, stronger Self-Worth Enhancing academic contingencies should be associated with relatively positive academic and psychological outcomes. This is because Self-Worth Enhancing academic contingencies signify a connection where self-worth is boosted by positive feedback in academic settings, but not necessarily reduced by negative feedback. We suspect that such linkages may be driven by concurrent messages that educational attainment is important, yet racial barriers may need to be overcome in pursuing scholarly excellence. Awareness of barriers may lead students to connect to education enough to reap the benefits from positive academic outcomes, but not so holistically that students’ self-worth suffers in the face of negative academic feedback or racial experiences. Similar to the Self-Worth Dependent academic contingencies, this dimension will likely be associated with stronger connectedness to African Americans (racial centrality). Additionally, Self-Worth Enhancing academic contingencies may also relate to positive feelings about African Americans (private regard) and to more pessimistic views of societal regard of African Americans (public regard). Finally, previous research has shown that racial stigma negatively influences connections between African American students’ self-worth and their achievement (e.g., Steele & Aronson, 1995). Thus, we also explored whether students’ interpersonal racial discrimination experiences, one type of racial stigma experience, related to their Self-Worth Dependent and Self-Worth Enhancing academic contingencies.

Method

Participants

A total of 330 self-identified African American freshmen (244 females, 86 males) were recruited from three, large, public, predominately White institutions (PWIs) in the United States. The sample included 141 students from a university in a Midwestern suburb, 80 students from an institution in a Midwestern city, and 109 participants attended a university in the Southeastern region of the country. The participants ranged in age from 16 to 21 years (M = 18 years, SD = 0.37). A measure of mother’s highest educational attainment assessed students’ social class (see Starfield, Riley, Witt, & Robertson, 2001). Approximately, 95% of mothers graduated from high school and 48% of mothers attained a college degree or more. For all analyses, we used a subsample of students who had complete data on all study variables (n = 310, 6% missing data, 228 females, 82 males). The excluded participants did not differ from included respondents in terms of gender, χ²(1, N = 330) = 0.46, p = .52, or mother’s education, t(324) = 0.70, p = .49. Students who were missing data were equally represented at each of the institutions, χ²(2, N = 330) = 1.47, p = .48.

Procedure

The data were collected as part of a larger investigation of college adjustment among African American college freshmen transitioning to PWIs. Participants were recruited by telephone and e-mail via a list of African American students from each institution’s registrar’s office. The data in this study represent the initial wave of investigation, which occurred in the first 2 weeks of the freshman year. The survey was administered in research labs by African American researchers via paper-and-pencil methods in small groups (3-10 students). On average, the survey administration took 35 to 50 minutes. At the end of the administration, participants were compensated with $15.

Measures

Academic contingencies of self-worth

The academic contingencies of Self-Worth Scale is a subscale of the Contingencies of Self-Worth measure (Crocker & Wolfe, 2001), containing seven items (see Table 1) on a response scale from 1 (strongly disagree) to 7 (strongly agree). Scale items assess the degree to which individuals connect their self-worth to academic performance (e.g., “Doing well in school gives me a sense of self-respect,” “When I do poorly on an exam or paper, my self-esteem suffers.”). Higher scores indicate stronger connections between students’ self-worth and achievement. The internal reliability for the current sample (α = .68) was consistent with alphas obtained for Black students in the study that validated the scale (Crocker, Luhtanen, et al., 2003).

Table 1.

Factor Loadings for Principal Components Analysis (PCA) With Direct Oblim Rotation of Academic Contingencies Scale

Item	Enhancing	Dependent
My self-esteem gets a boost when I get a good grade on exam or paper.	0.68	−0.07
When I do poorly on an exam or paper, my self-esteem suffers.	0.05	0.48
Whether or not I am a good student is unrelated to my overall opinion of myself.	0.13	0.89
My opinion of myself is not tied to how well I do in school.	0.20	0.85
I feel better about myself when I know I’m doing well academically.	0.82	0.08
Doing well in school gives me a sense of self-respect.	0.86	0.25
It is important to me to score well on standardized tests of academic ability such as the SAT or ACT.	0.58	0.37

Note. Factor loadings >.32 are in boldface. Enhancing = Self-Worth Enhancing dimension of academic contingencies of self-worth; Dependent = Dependent dimension of academic contingencies of self-worth.

Psychological outcomes

Three psychological outcomes were assessed—self-esteem, depressive symptoms and perceived stress. Self-esteem was assessed with the Rosenberg Self-Esteem Scale (Rosenberg, 1979). The measure consists of 10 items (e.g., “I take a positive attitude toward myself.”), which use a 4-point Likert-type scale ranging from 1 (disagree strongly) to 4 (agree strongly). In the current sample, the measure displayed strong internal reliability (α = .82). Higher scores indicate more positive self-esteem. Students’ depressive symptoms were measured with the Center for Epidemiological Studies Depressive Symptoms Scale (CES-D), a 20-item self-report measure designed for use in the general population and validated with multi-racial samples (Radloff, 1977). Participants were asked to report the frequency of 20 events (e.g., “I feel hopeful about the future.”) over the past week. Responses ranged from 1 (rarely, none of the time, less than 1 day) to 4 (most of the time, 5-7 days). Higher scores reflect more severe depressive symptoms (α = .88). Perceived stress was assessed using the Perceived Stress Scale (PSS), which has been validated with college samples (Cohen, Kamarck, & Mermelstein, 1983). Participants reported the frequency of 14 experiences (e.g., “felt that you were unable to control the important things in your life?”) in the past month (α = .86) on a response scale from 0 (never) to 4 (often).

College academic preparation

Students’ self-reported college entry standardized test scores (either the Scholastic Aptitude Test [SAT] or the American College Testing [ACT] exam) were used as a measure of academic preparation on entering college. We converted the scales to be on the same metric (using conversion guides from the College Board; see http://www.collegeboard.org; SAT-ACT concordance tables available at http://professionals.collegeboard.com/data-reports-research/sat/sat-act). We felt that this variable provided a consistent measure of college academic preparation, given the vast variation in participants’ high school grading systems and metrics (see Greer & Chwalisz, 2007).

Racial identity

Students’ racial identity beliefs were assessed with two dimensions of the Multidimensional Inventory of Black Identity–Short (MIBI-S; Martin, Wout, Nguyen, Sellers & Gonzalez, 2005; Sellers et al., 1997), racial centrality and racial regard. Racial centrality assessed the importance of being Black to students’ overall identity (e.g., “I have a strong sense of belonging to Black people.”) with four question items. The private regard subscale, composed of eight items, assessed racial group pride (e.g., “I feel good about Black people.”) while the public regard subscale, measured with five items, captured students’ affective beliefs about society’s views of Blacks (e.g., “Overall, Blacks are considered good by others.”). Participants responded on a scale from 1 (strongly disagree) to 7 (strongly agree). Higher scores indicate stronger race centrality and more positive regard. The scales’ internal consistencies in the present study, centrality (α = .81), private regard (α = .76), and public regard (α = .75), are similar to those found in previous research that validated the subscales (e.g., Sellers et al., 1997).

Racial discrimination

The Racism and Life Experiences Scale (RaLES; Harrell, 1997) assessed racial discrimination experiences. Participants reported how often in the past year they experienced 18 incidences of unfair treatment (e.g., “How often in the past year have your ideas or opinions been minimized, ignored, or devalued?”). Responses ranged from 0 (never) to 5 (once a week or more). Higher scores on the RaLES reflect more frequent racial discrimination experiences. The measure was internally reliability in the current sample (α = .90), consistent with previous research showing the scale to be both valid and reliable among African American college students (e.g., Sellers & Shelton, 2003; Smalls et al., 2007).

Demographic background

Background indicators included participants’ gender (female, male), academic institution (Midwestern suburban, Midwestern urban, Southeastern suburban), and self-reported mother’s educational attainment, which was reported on a scale from 1 (junior high or less) to 8 (PhD/MD).

Results

Data Analytic Strategy

This study investigated whether African American college students connected their self-worth to academics in ways that are consistent with our conceptualization of Self-Worth Enhancing and Self-Worth Dependent academic contingencies. The analyses involved three phases. First, according to established scale validation procedures (e.g., Thompson, 2004), the sample was randomly divided in half. With the first half of the sample, principal components analysis (PCA) was conducted to explore the factor structure for academic contingencies of self-worth items. Second, confirmatory factor analysis (CFA) with the second random half of the sample was used to validate the best-fitting model from the exploratory analysis. Third, correlation and ordinary least squares multiple regression analyses were performed to test whether academic contingencies of self-worth dimensions were differentially associated with students’ race-related beliefs and experiences, college academic preparation, and psychological outcomes.

Exploring Two Dimensions of Academic Contingencies for African American College Students

The first step was to explore the factor structure of the academic contingencies of Self-Worth Scale for the students in our sample. Approximately 50% of the sample (n = 142) was randomly selected and submitted to a PCA. The emergent components were allowed to correlate by using a direct oblim rotation (Fabrigar, Wegener, MacCallum, & Strahan, 1999), and the Kaiser Criterion (all factors with eigenvalues >1.0) was used to decide the data’s most appropriate factor structure (Kaiser, 1960). Two components emerged with eigenvalues greater than 1.0 (accounting for 36% and 23% of the variance, respectively). Items were assigned to components to the degree that they more heavily loaded on to a dimension and met the conventional .40 minimum loading (e.g., Hair, Anderson, Tatham, & Black, 1998). Using this criterion, four items loaded onto the first component and three items loaded onto the second component (see Table 1).

The content of each component was analyzed to determine whether the items loading onto each dimension represented our theoretical representation of Self-Worth Enhancing and Self-Worth Dependent academic contingencies. The first component contained the items: “My self-esteem gets a boost when I get a good grade on an exam or paper,” “I feel better about myself when I know I’m doing well academically,” “Doing well in school gives me a sense of self-respect,” and “It is important to me to score very well on standardized tests of academic ability such as the SAT or ACT.” This set of items is consistent with our representation of Self-Worth Enhancing academic contingencies. The items reflect a connection between students’ self-worth and education such that performance “boosts” self-worth and provides students with a feeling of “self-respect,” yet the language among these items does not indicate that negative academic outcomes reduce students’ self-worth.

Three items loaded onto a second component: “Whether or not I am a good student is unrelated to my overall opinion of myself (reverse-scored),” “When I do poorly on an exam or paper, my self-esteem suffers,” and “My opinion of myself is not tied to how well I do in school (reverse-scored).” These items closely correspond to our representation of Self-Worth Dependent academic contingencies, whereby students’ “overall opinions” of themselves are defined by their academic performance. In the case of our depiction of Self-Worth Dependent academic contingencies, and consistent with prevailing contingencies of self-worth theory, such heavy reliance on scholarly achievements as a source of self-worth likely makes students vulnerable to self-worth instability, leading to compromised mental health and poorer academic performance (Crocker & Wolfe, 2001; Harter & Whitesell, 1999; Kernis et al., 1993).

Cross-Validation of the Two-Dimensional Model

CFA, which allows researchers to specify a scale’s anticipated structure a priori, was performed to supplement the exploratory analyses. Using MPlus software (Muthén & Muthén, 1998-2007), two alternative models were estimated with the second random half of the sample (n = 170). First, we tested a model where all items load onto one factor as proposed by existing research. The second model corresponded to our suggested two-dimensional structure, which was preliminarily supported by the PCA. For each model, multiple model fit indices were evaluated, as each index has certain strengths and weaknesses. First, the likelihood ratio chi-square was observed, where a nonsignificant chi-square statistic indicates good fit to the data. The likelihood ratio chi-square is sensitive to sample size and to the size of the correlations in the sample (Kline, 2005). As such, three additional fit indices were evaluated. The root mean square error of approximation (RMSEA) is a measure of “misfit” that does not approximate a central chi-square distribution, is robust against the influence of sample size, and takes into account model complexity (Kline, 2005). An RMSEA of .08 is considered acceptable, and good fit is indicated by RMSEA of less than .05 (Browne & Cudeck, 1993). Finally, we observed the comparative fit index (CFI) and the standardized root mean square residual (SRMR), where CFI values more than .90 and SRMR values less than .10 represent good fit (e.g., Kline, 2005). Results for a one-dimensional model indicated poor model fit, ²(14, N = 170) = 116.16, p = .00, CFI = .69, SRMR = .10, RMSEA = .21 (see Figure 1). In contrast, results for a two-dimensional model indicated better fit to the data, χ²(13, N = 170) = 28.47, p = .01, CFI = .95, SRMR = .06, RMSEA = .08 (see Figure 2). A chi-square difference test comparing the one-dimensional and two-dimensional models showed a significant reduction in chi-square values, χ²(1) = 87.69, p =. 00, further supporting the two-dimensional model as the better solution (see Widaman, 1985).

Figure 1.

Confirmatory factor analysis assuming that academic contingencies fall along one dimension

Figure 2.

Confirmatory factor analysis assuming that academic contingencies fall along two dimensions

Academic Contingencies and Race-Related, Academic, and Psychological Outcomes

The final set of analyses consisted of correlational and multiple regression analyses, which were conducted to investigate whether Self-Worth Enhancing (α = .70) and Self-Worth Dependent (α = .65) academic contingencies were uniquely associated with students’ race-related beliefs and experiences, academic attributes, and psychological outcomes.

Descriptive statistics

All means, standard deviations, and bivariate correlations are reported in Table 2. Generally, students reported very high self-esteem (M = 3.53, SD = 0.46) and moderate depressive symptoms (M = 1.73, SD = 0.49) and perceived stress (M = 1.91, SD = 0.70). Students reported high racial centrality (M = 5.64, SD = 1.11) and private regard (M = 6.42, SD = 0.65) and moderate public regard (M = 3.62, SD = 0.94). On average, students reported few interpersonal racial discrimination events in the past year (M = 1.67, SD = 0.84), with male students (M = 1.91, SD = 0.86) reporting more frequent discrimination than their female counterparts (M = 1.59, SD = 0.81); t(308) = 2.95, p < .01. Multivariate analyses of variance revealed significant institutional differences in college academic preparation, F(2, 307) = 51.78, p < .001 and Self-Worth Enhancing academic contingencies, F(2, 307) = 3.00, p = .05. Post hoc Scheffe analyses indicated significant differences in college academic preparation among students at the urban Midwestern institution (M = 3.62, SD = 1.44), the suburban Midwestern institution (M = 5.23, SD = 1.65), and the suburban Southeastern institution (M = 5.89, SD = 1.27). Post hoc analyses did not reveal significant pairwise differences in Self-Worth Enhancing academic contingencies. Students reported high Self-Worth Enhancing academic contingencies scores (M = 5.87, SD = .92) and moderate Self-Worth Dependent academic contingencies scores (M = 4.41, SD = 1.45). Men and women varied in Self-Worth Dependent academic contingencies, t(308) = −1.99, p < .05, with males (M = 4.14, SD = 1.30) reporting lower scores than female students (M = 4.51, SD = 1.49).

Table 2.

Summary of Intercorrelations, Means, and Standard Deviations for Major Study Variables (n = 310)

Measure	1	2	3	4	5	6	7	8	9	10	11
1. College Prep	—	.32**	.09	.09	.07	−.04	−.09	−.06	.00	−.05	.08
2. Mother’s Ed		—	.03	.08	.06	−.07	−.10	−.05	.05	−.08	−.07
3. Enhancing			—	.30**	−.02	.05	.02	.08	.12	−.05	.07
4. Dependent				—	−.32**	.33**	.32**	.02	−.14*	−.21*	.12
5. Self-esteem					—	−.67**	−.59**	.33**	.26**	.07	−.13*
6. DepSym						—	.70**	−.17**	−.32**	.16*	.25**
7. PerStr							—	−.15*	−.25**	−.13*	.15*
8. Centrality								—	.49**	−.05	.18**
9. Private regard									—	.14*	.05
10. Public regard										—	−.13*
11. Discrimination											—
M	5.21	na	5.87	4.41	3.53	1.73	1.91	5.64	6.42	3.61	1.67
SD	1.63	na	0.92	1.45	0.46	0.49	0.70	1.10	0.65	0.94	0.84
Actual range	1-10	1-8	1-7	1-7	1-4	1-4	0-4	1-7	1-7	1-7	0-5
Observed range	1-10	1-8	1-7	1-7	1.60-4	1-3.40	.36-3.50	2-7	3.88-7	1-6.50	0-4.80

Note. Prep = preparation; Ed = educational attainment; Enhancing = Self-Worth Enhancing dimension of academic contingencies of self-worth; Dependent = Dependent dimension of academic contingencies of self-worth; DepSym = Depressive Symptoms; PerStr = perceived stress; na = not applicable.

p < .05. **p < .01.

Correlations

Bivariate correlations between study variables are presented in Table 2. Spearman’s rho was used to calculate the correlation between mother’s educational attainment and other study variables. Pearson’s product–moment correlations were used to assess the associations between all other study variables. Of note, the Self-Worth Dependent and Self-Worth Enhancing dimensions were positively correlated (r = .30, p < .01). Additionally, Self-Worth Dependent academic contingencies scores were positively associated with depressive symptoms (r = .33, p < .01) and perceived stress (r = .32, p < .01) and negatively associated with self-esteem (r = -.32, p < .01), private regard (r = −.14, p < .05), and public regard (r = −.21, p < .05). Self-Worth Enhancing academic contingencies were not associated with psychological, academic, or race-related variables. Neither dimension of academic contingencies of self-worth was associated with students’ college preparation.

Regression models

Ordinary least squares multiple regression models were used to estimate the effects of academic contingencies of self-worth dimensions on three sets of outcomes variables: racial identity dimensions (centrality, private regard, public regard), racial discrimination experiences, and psychological outcomes (self-esteem, depressive symptoms, perceived stress). In each model, gender, mother’s educational attainment, and institution (dummy coded with the urban, Midwestern institution as the reference category) were entered as controls based on previous literature (e.g., Starfield et al., 2001) and in light of the reported descriptive analyses. Control and predictor variables were simultaneously entered into the regression models.

First, we examined the relationship between academic contingencies of self-worth dimensions and race-related beliefs and experiences (see Table 3). The overall racial centrality model was not significant, F(6, 303) = 1.55, p = .16, f² = .01; however, stronger Self-Worth Enhancing academic contingencies scores were related to more racial centrality, β = .16, p < .01. The overall private regard model was statistically significant, F(6, 303) = 3.77, p = .001, f² = .05 and explained approximately 5% of the variance in private regard. Higher Self-Worth Enhancing academic contingencies scores predicted higher private regard (β = .20, p < .01), and higher Self-Worth Dependent academic contingencies scores predicted lower private regard (β = −.17, p < .01). The overall public regard model was not significant, F(6, 303) = 1.41, p = .21, f² = .01, yet Self-Worth Dependent academic contingencies scores related to more negative public regard (β = −.16, p < .01). Neither Self-Worth Enhancing nor Self-Worth Dependent scores predicted frequency of interpersonal racial discrimination experiences, F(6, 303) = 1.91, p = .08, f² = .02.

Table 3.

Multiple Regression Analyses Predicting Race-Related Beliefs and Experiences (n = 310)

	Racial centrality			Private regard			Public regard			Racial discrimination
Predictor	b	SE b	β	b	SE b	β	b	SE b	β	b	SE b	β
Gender	−.05	.15	−.02	−.13	.08	−.09	−.04	.12	−.02	−.34	.11	−.18**
Mother’s education	−.03	.04	−.04	−.03	.02	−.06	.01	.03	.01	.01	.03	.02
Suburban Midwestern institution	.16	.16	.07	.22	.09	.17*	−.11	.14	−.06	−.04	.12	−.03
Suburban Southeastern institution	.09	.17	.04	.17	.10	.12	−.09	.15	−.05	.07	.13	.04
Enhancing	.19	.07	.16**	.14	.04	.20**	.03	.06	.03	−.01	.05	−.01
Dependent	.02	.05	.03	−.08	.03	−.17**	−.10	.04	−.16**	.04	.03	.07

Note. Enhancing = Self-Worth Enhancing dimension of academic contingencies of self-worth; Dependent = Dependent dimension of academic contingencies of self-worth.

p < .05. **p < .01.

Table 4 summarizes regression analyses for self-esteem, depressive symptoms, and perceived stress outcomes. Racial identity dimensions and racial discrimination have been previously shown to predict psychological outcomes (e.g., Sellers & Shelton, 2003). As such, we controlled for these factors in the following models. The self-esteem model accounted for 19% of the variance and was significant, F(10, 299) = 8.01, p < .01, f² = .23. Self-Worth Dependent academic contingencies scores predicted lower self-esteem (β = −.32, p < .01). The depressive symptoms model was also significant, F(10, 299) = 10.78, p < .01, f² = .32. Stronger Dependent academic contingencies of self-worth scores predicted higher depressive symptomatology β = −.27, p < .01). Finally, the perceived stress model was significant, F(10, 299) = 6.61, p < .01, f² = .18. Stronger Dependent academic contingencies of self-worth scores related to more perceived stress (β = −.29, p < .01).

Table 4.

Multiple Regression Analyses Predicting Academic and Psychological Outcomes (n = 310)

	Self-esteem			Depressive symptoms			Perceived stress
Predictor	b	SE b	β	b	SE b	β	b	SE b	β
Gender	.01	.06	.01	.03	.06	.03	.1	.09	.06
Mother’s education	−.01	.02	−.02	.02	.02	.05	.02	.02	.04
Suburban Midwestern institution	−.01	.06	−.02	.04	.06	.04	.01	.1	.01
Suburban Southeastern institution	−.02	.07	−.02	.07	.07	.06	.1	.1	.07
Racial centrality	.12	.03	.30**	−.05	.03	−.11	−.04	.04	−.07
Private regard	.03	.04	.05	−.16	.05	−.22**	−.17	.07	−.15*
Public regard	.01	.03	.02	−.01	.03	−.02	−.04	.04	−.05
Discrimination experiences	−.08	.03	−.14**	.18	.03	.30**	.14	.05	.17**
Enhancing	.01	.03	.02	−.02	.03	−.05	−.03	.04	−.04
Dependent	−.1	.02	−.32**	.09	.02	.27**	.14	.03	.29**

Note. Enhancing = Self-Worth Enhancing dimension of academic contingencies of self-worth; Dependent = Dependent dimension of academic contingencies of self-worth.

p < .05. **p < .01.

Taken together, our findings reveal two positively correlated, yet distinct academic contingencies of self-worth dimensions among the African American college students in our sample. Neither dimension was associated with students’ reported academic preparation or interpersonal racial discrimination experiences. As expected, the dimensions were differentially related to students’ racial identity beliefs and to their psychological outcomes. Self-Worth Enhancing academic contingencies predicted more positive private regard (personal affective beliefs about Blacks) and stronger racial centrality (connectedness to Blacks). Self-Worth Dependent academic contingencies scores related to lower private regard, more negative public regard (beliefs about societal views of Blacks), lower self-esteem, more depressive symptoms, and more perceived stress. In sum, the findings partially supported our original predictions and offer clear directions for future inquiries.

Discussion

Between-group comparisons and a near-exclusive focus on racial disparities between African Americans and other groups of students have made race stigma perspectives quite prevalent in explaining African Americans’ academic experiences and outcomes. Research on racial socialization, identity, and discrimination however, show that for many African American students, the importance and value placed on being African American can be a source of pride and can also protect students from negative outcomes in the face of racial and academic barriers (Chavous et al., 2003; Neblett et al., 2006; Sellers et al., 1997; Small et al., 2007). The present study applied such findings from racial socialization, identity, and discrimination literatures to research on self-processes and academic motivation to propose and empirically support a two-dimensional model of academic contingencies of self-worth among African American college students. This objective was informed by our observation that distinct psychological theories are in conflict as to whether strong connections between students’ self-worth and education are advantageous or detrimental. Our research was also driven by recommendations from scholars advocating for the use of culturally informed, within-group assessments as a way to clarify the assumptions that drive studies of motivation and achievement among African American students (e.g., Cokley, 2003; Cokley et al., 2003; Graham, 1994). Such propositions led to our prediction that African American students maintain both Self-Worth Enhancing and Self-Worth Dependent academic contingencies.

Self-Worth Enhancing Academic Contingencies

Self-Worth Enhancing academic contingencies reflect the extent to which self-worth is boosted by educational achievement, but not necessarily diminished in the face of negative outcomes and experiences. We posited that rather than completely disconnecting the self from academics when faced with academic or racial difficulties, some African American students may opt to more cautiously connect their self-worth to the educational domain. Speculating further, we expected Self-Worth Enhancing academic contingencies to predict stronger connectedness to Blacks (racial centrality), more positive feelings about Blacks (private regard), and the belief that society holds more negative views of Blacks (public regard). Finally, we anticipated Self-Worth Enhancing academic contingencies to be associated with more positive college preparation and psychological outcomes on entering college.

The findings partially supported these predictions. As expected, stronger Self-Worth Enhancing academic contingencies related to more positive private regard. Self-Worth Enhancing academic contingencies did not predict stronger centrality or more negative public regard. In some ways however, our predictions positioned racial centrality and public regard as proxies for racial socialization messages (Hughes et al., 2006). Thus, an important future direction would be to investigate whether racial socialization messages are directly related to Self-Worth Enhancing academic contingencies, rather than associated with this dimension via racial identity beliefs.

The findings did not support our expectation that Self-Worth Enhancing academic contingencies would be associated with more positive academic and psychological outcomes, as scores were unrelated to students’ college academic preparation, self-esteem, depressive symptoms, and perceived stress. This may have been because we conceptualized Self-Worth Enhancing academic contingencies as being protective when students experience racial stigma and other negative circumstances in educational settings. It is thus possible that higher Self-Worth Enhancing academic contingencies are less directly predictive of academic achievement and adjustment and instead, buffer the negative experience of racial and academic adversities when they occur. Future research investigating whether Self-Worth Enhancing academic contingencies moderate the association between academic and racial difficulties and motivation, achievement, and adjustment could inform this hypothesis more directly.

Self-Worth Dependent Academic Contingencies

Whereas Self-Worth Enhancing academic contingencies reflect more tentative connections between students’ self-worth and education, Self-Worth Dependent academic contingencies represent the extent to which students wholly based their self-worth on academics. This dimension aligned with Crocker and colleagues’ conceptualization of academic contingencies more broadly. Accordingly, we expected having stronger Self-Worth Dependent academic contingencies would relate to lower college preparation and more negative psychological outcomes. Similar to the findings for Self-Worth Enhancing academic contingencies, Self-Worth Dependent academic contingencies were not associated with college preparation. The findings may have differed however, if students’ high school grades were examined, which might prove to be a more explicit indicator of motivation and performance, relative to the standardized test scores used in this study. Consistent with expectations, stronger Self-Worth Dependent academic contingencies scores predicted lower self-esteem, higher depressive symptomatology, and more perceived stress.

The findings for Self-Worth Dependent academic contingencies are generally consistent with our expectations as well as with previous research by Crocker and colleagues. The results however, complicate previous academic identification research (e.g., Steele & Aronson, 1995), which maintains that strong connections between academics and self-worth are necessary for sustained academic motivation, engagement, and achievement. In the current study, stronger Self-Worth Dependent academic contingencies predicted psychological costs, which may actually compromise academic outcomes across students’ educational trajectories. Future longitudinal studies should assess the association between Self-Worth Dependent academic contingencies, academic performance, and psychological outcomes to determine whether Self-Worth Dependent academic contingencies indirectly diminish performance over time.

Finally, endorsing Self-Worth Dependent academic contingencies related to lower private and public regard, such that students who viewed African Americans more negatively and perceived more societal bias against African Americans based their self-worth on academics more strongly. Considered along with the negative associations found between Self-Worth Dependent academic contingencies and psychological outcomes, the results suggest that personally viewing one’s racial group less positively and in the context of racial stigma may place students at risk for developing connections between their self-worth and achievement that are psychologically taxing.

Considerations

This study provides initial support for a two-dimensional academic contingencies framework, yet in interpreting the present findings, a few considerations are worth noting. First, preliminary support for the two academic contingencies of self-worth dimensions was found using an existing measure, not a measure specifically designed to test nuances in how Black students connect their self-worth to education. Thus, although our results support our proposed framework, one must refrain from making definitive conclusions until our hypothesized model can be tested with measurement items purposely designed to reflect fine-grained distinctions between Self-Worth Enhancing and Self-Worth Dependent academic contingencies of self-worth. Second, the sample was predominately female and all respondents attended large, public, predominately White institutions. This limitation speaks to our inability to firmly generalize our findings to African American male students and to African American students who attend other types of institutions.

Future Directions

Consideration of this study’s limitations highlights clear steps for future research on academic contingencies of self-worth among African American students. Future psychometric studies, for example, should investigate the suggested dimensions with items specifically created for the purpose of highlighting complexity in the ways that African Americans students link their self-worth to the academic domain. Also, it will be important for future research to investigate the proposed dimensions with a more gender-balanced sample and with students attending different types of institutions (e.g., HBCUs, private PWIs, etc.). In the future, investigators may also want to extend the variable-centered approach that we used in the present study to conduct person-centered (e.g., cluster analytic) analyses that examine how variation in Self-Worth Enhancing and Self-Worth Dependent academic contingencies, within students, helps explain outcomes. Finally, a key conceptual distinction between Self-Worth Enhancing and Self-Worth Dependent academic contingencies pertains to how self-worth instability may shape responses to negative feedback. Our representation of Self-Worth Dependent academic contingencies suggests that students’ self-worth will fluctuate heavily in response to negative feedback, whereas our depiction of Self-Worth Enhancing academic contingencies implies that self-worth will be more robust to academic and racial obstacles. Future investigations ought to test this contention directly.

Contributions and Conclusions

Notwithstanding these considerations, this study makes important theoretical and practical contributions to the study of African American student motivation, achievement, and adjustment. Theoretically, our two-dimensional academic contingencies of self-worth framework is the novel by-product of synthesizing social psychological frameworks (e.g., academic contingencies of self-worth, academic identification) with research on African American racial socialization, identity, and discrimination. Whereas many previous studies have focused on the implications of weak or strong connections between students’ self-worth and academics, we suggested that different types of academic contingencies have unique implications for students’ educational outcomes. In many ways, the conception of academic contingencies types is analogous to recognizing that private and public regard are two dimensions of simultaneously held affective evaluations that students can concurrently vary in intrinsic and extrinsic motivation, and that individualism and collectivism signify two distinct, albeit related worldviews. Hence, our framework allowed us to conceptualize an approach for studying academic contingencies that recognizes how African American students may adapt to adversities in the educational context in ways other than completely disconnecting their self-worth from academics. Specifically, the Self-Worth Enhancing dimension of academic contingencies, which embodies a more tentative connection between students’ self-worth and academics, complements the traditionally conceived notion of academic contingencies as holistic linkages between the self and education. Practically, this study has the potential to inform programmatic efforts designed to improve African American student outcomes and to bolster students’ academic and psychological resilience—not by identifying deficits, but by reinforcing African American students’ sources of strength. If educational practitioners, for instance, can help African American students build support networks that reinforce Self-Worth Enhancing academic contingencies, students may be able to sustain academic identification and resist self-worth instability in the face of educational and racial struggles.

Historical and contemporary discrimination create a unique paradox for African American students—education is important in moving toward social and economic equality, but the educational system is not always equitable in its treatment of distinct racial group members. Our proposed academic contingencies dimensions highlight a novel approach for examining how African American students may respond to these conflicting messages. In doing so, the framework attempts to reconcile prevalent images of African American students as inherently at risk for academic underachievement with the fact that many African American students excel despite racial stigma. Although a preliminary step, we hope that future work taking such a strengths-based approach will build on the conceptual framework put forth in the current study.

Footnotes

The authors declared no conflicts of interest with respect to the research, authorship, and/or publication of this article.

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by National Institute of Mental Health Grant NIMH 5 P01 MH58565-03.

References

Bowman

P. J.

Howard

(1985). Race-related socialization, motivation, and academic achievement: A study of Black youths in three-generation families. Journal of the American Academy of Child & Adolescent Psychiatry, 24, 134-141.

Browne

M. W.

Cudeck

(1993). Alternative ways of assessing model fit. In Bollen

K. A.

Long

J. S.

(Eds.), Testing structural equation models (pp. 136-162). Newbury Park, CA: Sage.

Chavous

T. M.

Bernat

Schmeelke-Cone

Caldwell

Kohn-Wood

L. P.

Zimmerman

(2003). Racial identity and academic attainment among African American adolescents. Child Development, 74, 1076-1091.

Cohen

Kamarck

Mermelstein

(1983). A global measure of perceived stress. Journal of Health and Social Behavior, 24, 385-396.

Cokley

K. O.

(2003). What do we know about motivation of African American students? Challenging the “anti-intellectual” myth. Harvard Educational Review, 73, 524-558.

Cokley

K. O.

Komarraju

King

Cunningham

Muhammad

(2003). Ethnic differences in the measurement of academic self-concept in a sample of African American and European American College Students. Educational and Psychological Measurement, 63, 707-722.

Covington

M. V.

(1984). The self-worth theory of achievement motivation: Findings and implications. Elementary School Learning, 85, 5-20.

Crocker

Luhtanen

R. K.

(2003). Level of self-esteem and contingencies of self-worth: Unique effects on academic, social, and financial problems in college students. Personality and Social Psychology Bulletin, 29, 701-712.

Crocker

Luhtanen

R. K.

Cooper

M. L.

Bouvrette

S. A.

(2003). Contingencies of self-worth in college students: Measurement and theory. Journal of Personality and Social Psychology, 85, 894-908.

10.

Crocker

Major

(1989). Social stigma and self-esteem: The self-protective properties of stigma. Psychological Review, 96, 608-630.

11.

Crocker

Quinn

D. M.

Karpinski

Chase

S. K.

(2003b). When grades determine self-worth: Consequences of contingent self-worth for male and female engineering and psychology majors. Journal of Personality and Social Psychology, 85, 507-516.

12.

Crocker

Sommers

S. R.

Luhtanen

R. K.

(2002). Hopes dashed and dreams fulfilled: Contingencies of self-worth and admissions to graduate school. Personality and Social Psychology Bulletin, 28, 1275-1286.

13.

Crocker

Wolfe

C. T.

(2001). Contingencies of self-worth. Psychological Review, 108, 593-623.

14.

Fabrigar

L. R.

Wegener

D. T.

MacCallum

R. C.

Strahan

E. J.

(1999). Evaluating the use of exploratory factor analysis in psychological research. Psychological Methods, 4, 272-299.

15.

Galleta

Cross

W. E.

(2007). Past as present, present as past: Historicizing Black education and interrogating integration. In Fuligni

A. J.

(Ed.), Contesting stereotypes and creating identities: Social categories, social identities, and educational participation (pp. 15-41). New York, NY: Russell Sage Foundation.

16.

Graham

(1994). Motivation in African Americans. Review of Educational Research, 64, 55-117.

17.

Greer

T. M.

Chwalisz

(2007). Minority-related stressors and coping responses among African American college students. Journal of College Student Development, 48, 388-404.

18.

Hair

J. F.

Anderson

R. E.

Tatham

R. L.

Black

W. C.

(1998). Multivariate data analysis with readings (5th ed.). Englewood Cliffs, NJ: Prentice Hall.

19.

Harrell

(1997). The Racism and Life Experience Scales (RaLES). Unpublished manuscript.

20.

Harter

Whitesell

N. R.

(1999). Beyond the debate: Why some adolescents report stable self-worth over time and situation, whereas others report changes in self-worth. Journal of Personality, 71, 1027-1058.

21.

Hughes

Rodriguez

Smith

E. P.

Johnson

D. J.

Stevenson

H. C.

Spicer

(2006). Parents’ ethnic-racial socialization practices: A review of research and directions for future study. Developmental Psychology, 42, 747-770.

22.

Kaiser

H. F.

(1960). The application of electronic computers to factor analysis. Educational and Psychological Measurement, 20, 141-151.

23.

Kernis

M. H.

Cornell

D. P.

Sun

Berry

Harlow

(1993). There’s more to self-esteem than whether it is high or low: The importance of stability of self-esteem. Journal of Personality and Social Psychology, 65, 1190-1204.

24.

Kline

R. B.

(2005). Principles and practice of structural equation modeling. New York, NY: Guilford.

25.

Martin

P. P.

Wout

Nguyen

Sellers

R. M.

Gonzalez

(2005). Investigating the psychometric properties of the Multidimensional Inventory of Black Identity in two samples: The development of the MIBI–S. Unpublished manuscript.

26.

Muthén

L. K.

Muthén

B. O.

(1998-2007). Mplus user’s guide (5th ed.). Los Angeles, CA: Muthén & Muthén.

27.

Neblett

Philip

Cogburn

C. D.

Sellers

R. M.

(2006). African American adolescents’ discrimination experiences and academic achievement: Racial socialization as a cultural compensatory and protective factor. Journal of Black Psychology, 32, 199-218.

28.

Niiya

Crocker

Bartmess

(2004). Learning orientations buffer contingent self-esteem from failure. Psychological Science, 15, 801-805.

29.

Osborne

J. W.

Walker

(2006). Stereotype threat, identification with academics, and withdrawal from school: Why the most successful students of colour might be the most likely to withdraw. Educational Psychology, 26, 563-577.

30.

Radloff

L. S.

(1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychosocial Measurement, 1, 385-401.

31.

Rosenberg

(1979). Conceiving the self. New York: Basic Books.

32.

Sellers

R. M.

Rowley

S. J.

Chavous

T. M.

Shelton

J. N.

Smith

(1997). Multidimensional inventory of Black identity: Preliminary investigation of reliability and construct validity. Journal of Personality and Social Psychology, 73, 805-815.

33.

Sellers

R. M.

Shelton

J. N.

(2003). The role of racial identity in perceived racial discrimination. Journal of Personality and Social Psychology, 84, 1079-1092.

34.

Smalls

White

Chavous

T. M.

Sellers

R. M.

(2007). Racial ideological beliefs and racial discrimination experiences as predictors of academic engagement among African American adolescents. Journal of Black Psychology, 33, 299-330.

35.

Starfield

Riley

A. W.

Witt

W. P.

Robertson

(2001). Social class gradients in health during adolescence. Journal of Epidemiological Community Health, 56, 354-361.

36.

Steele

Aronson

(1995). Stereotype threat and the intellectual test performance of African Americans. Journal of Personality and Social Psychology, 69, 797-811.

37.

Thompson

(2004). Exploratory and confirmatory factor analysis: Understanding concept and applications. Washington, DC: American Psychological Association.

38.

Widaman

K. F.

(1985). Hierarchically nested covariance structure models for multitrait-multimethod data. Applied Psychological Measurement, 9, 1-26.

39.

Zeigler-Hill

(2007). Contingent self-esteem and race: Implications for the Black self-esteem advantage. Journal of Black Psychology, 33, 51-74.