Journal of Youth and Adolescence

, Volume 38, Issue 4, pp 544–559

Racial Identity and Academic Achievement in the Neighborhood Context: A Multilevel Analysis

Authors

  • Christy M. Byrd
    • Combined Program in Education & PsychologyUniversity of Michigan
    • Combined Program in Education & PsychologyUniversity of Michigan
Empirical research

DOI: 10.1007/s10964-008-9381-9

Cite this article as:
Byrd, C.M. & Chavous, T.M. J Youth Adolescence (2009) 38: 544. doi:10.1007/s10964-008-9381-9

Abstract

Increasingly, researchers have found relationships between a strong, positive sense of racial identity and academic achievement among African American youth. Less attention, however, has been given to the roles and functions of racial identity among youth experiencing different social and economic contexts. Using hierarchical linear modeling, the authors examined the relationship of racial identity to academic outcomes, taking into account neighborhood-level factors. The sample consisted of 564 African American eighth-graders (56% male). The authors found that neighborhood characteristics and racial identity related positively to academic outcomes, but that some relationships were different across neighborhood types. For instance, in neighborhoods low in economic opportunity, high pride was associated with a higher GPA, but in more advantaged neighborhoods, high pride was associated with a lower GPA. The authors discuss the need to take youth’s contexts into account in order to understand how racial identity is active in the lives of African American youth.

Keywords

Racial identitySocial disorganizationAcademic achievementHierarchical linear modelingEarly adolescents

Introduction

Adolescence is widely proposed as a time of intense identity development. Youth are concerned with establishing personal and collective identities, but not all identities are equal—some identities have greater implications for how youth are treated and how they respond to the world. For many African American adolescents, race becomes increasingly salient, thus making the process of racial identity development during this period very important. Researchers have demonstrated that racial identity, or the meanings youth attribute to their racial group membership, can be promotive of academic success (Chavous et al. 2003) and even serve protective roles for those experiencing risk factors in their environments, such as personal discrimination (Caldwell et al. 2004; Wong et al. 2003). Thus, this line of work suggests that a strong sense of racial group connectedness and high levels of group pride may serve as personal psychological resources to many youth.

We propose that in understanding the roles and functions of racial identity in relation to youth psychosocial adjustment, it is also necessary to consider the contexts in which adolescents’ identities develop. Ecological perspectives focused on ethnic minority development, such as Spencer’s Phenomenological Variant of Ecological Systems Theory (PVEST) (e.g., Spencer et al. 2001) suggest that youth’s social and psychological adaptation is influenced by the balance of risk and protective factors in their proximal contexts. While a strong, positive sense of racial identity may serve as a protective factor for many youth, it is less clear whether and how racial identity serves as such a resource in contexts that vary substantially in levels of risk and opportunities available to youth. This paper focuses on one specific context, the neighborhoods in which youth live, and the economic and social resources accessible to them.

In considering neighborhood context, we draw from social disorganization theory, which describes how different neighborhood characteristics are related to youth adaptation and adjustment outcomes. Using this perspective, we posit that (1) neighborhood characteristics structure youth’s contexts and, subsequently, can directly influence the meanings they attach to their racial group membership as well as the motivational orientations they develop around schooling, and (2) neighborhood opportunities and constraints influence how youth actualize a particular set of racial identity beliefs. In the current study, we sought to examine whether neighborhood characteristics are associated with youth’s racial identity attitudes. Also, we explored whether neighborhood-level factors exacerbate or attenuate the effects of racial identity on school engagement outcomes in early adolescence, even when controlling for youth’s family characteristics. This approach will help shed light on multilevel influences on African American youth development and interactions in the world.

Neighborhoods as Developmental Contexts

Neighborhoods offer both risks and resources to youth, some structural, others social. Ecological models of development posit that the balance of contextual risks and resources along with personal resources influences how youth adapt to their environments (Spencer et al. 2001). Social disorganization theory specifies neighborhood risks and resources as community features that prevent or enable residents in maintaining order and meeting common goals (Leventhal and Brooks-Gunn 2000). In other words, the quality and use of resident social networks, social capital, and other resources play important roles in how youth understand their chances for success and develop necessary values and skills. These resources can act beyond or even compensate for family characteristics. Structural causes of social disorganization include: poverty, which creates a lack of necessary resources; residential instability, which disrupts the formation of social networks; and racial/ethnic heterogeneity, which prevents residents from coming to consensus on values and goals (Sampson and Groves 1989). In this study, we examine social disorganization processes by focusing on three dimensions: collective socialization, economic opportunity, and institutional resources.

Collective Socialization

Collective socialization is the process through which adults in a neighborhood transmit their norms, values, and skills to adolescents (Ainsworth 2002). Communities that exemplify positive collective socialization are full of adults who have high levels of education and prestigious jobs, and who can therefore pass on skills and values that promote academic success. A neighborhood that exerts negative collective socialization, however, may present adolescents with role models who are indifferent to achievement and who may lead youth to take on similar values (Ainsworth 2002).

Economic Opportunity

A characteristic closely related to collective socialization is economic opportunity, which refers to actual and perceived financial success of a community. That is, whether jobs are available to community members and whether adolescents perceive them as so influence adolescents’ goals and aspirations (Leventhal and Brooks-Gunn 2000). Clearly, job opportunities determine the types of role models available, as well, and if youth do not see the chance to better themselves through education, they may not place much value in school. For instance, Mickelson (1990) demonstrated that, though African American youth may express support for educational achievement in the abstract, their concrete values, which are based on personal experiences, are more predictive of their academic behaviors. Neighborhoods should play a large role in shaping those concrete values by informing youth of the opportunities available.

Institutional Resources

In addition to socialized values and opportunities, the institutional resources within a neighborhood can influence youth’s personal identity development and adaptation. Institutional resources refer to the quality and availability of organizations and services that improve residents’ lives. Organizations like medical clinics, libraries, and YMCAs provide for adolescents’ basic needs as well as give them opportunities for pro-social structured activities and skill-building (Ainsworth 2002; Leventhal and Brooks-Gunn 2000). For example, a community center can serve as a place for youth to gather after school rather than engaging in delinquent behavior. Neighborhoods with more institutional resources are also more likely to provide opportunities for positive interactions between adolescents and adult figures.

Implications of Neighborhoods to Achievement Processes

Neighborhood attributes may be uniquely influential in the development of adolescents relative to younger children, as adolescents have more autonomy in their movement and interactions in academic and social contexts outside of the family household. Thus, they have more opportunities for exposure to individuals and settings relevant to the social disorganization dimensions—peers and non-caregiver adults, employment opportunities, and community and leadership roles. In fact, neighborhood-level risk factors have been associated with adolescents’ academic outcomes like school dropout (Crowder and South 2003; Ensminger et al. 1996), educational attainment (Boyle et al. 2007; Garner and Raudenbush 1991), grade achievement (Gonzalez et al. 1996), and educational values (Ceballo et al. 2004), even when accounting statistically for individuals’ family level factors. For instance, one neighborhood characteristic viewed as relevant for positive collective socialization—the presence of more positive role models—predicts more youth-reported time spent on homework, as well as higher standardized test scores and educational attainment in nationally representative samples (Ainsworth 2002; Brooks-Gunn et al. 1993). In inner-city African American samples, positive role models were positively related to educational values for girls (Ceballo et al. 2004) and negatively related to dropping out for boys (Ensminger et al. 1996).

Fewer social disorganization studies have focused on features other than collective socialization to explain adolescents’ academic behavior. One reason for this is that measures of economic conditions have not been distinguished from measures of positive or negative role models—both can be measured by the percentage of college graduates, the poverty rate, or the unemployment rate, for instance (Leventhal and Brooks-Gunn 2000). Another reason is that median income often is highly correlated with occupational data, leading some to discard it (Ensminger et al. 1996). Furthermore, indicators of low income, in particular, are not as strong predictors as indicators of affluence (Ainsworth 2002; Leventhal and Brooks-Gunn 2000). Rather than focus only on income, our conceptualization of economic opportunity also includes perceptions of opportunity, which should better capture the mediation between structure and achievement as suggested by Roosa and colleagues (Roosa et al. 2003). Finally, studies of institutional resources appear to have been limited to effects of school environment or employment for adolescents (Leventhal and Brooks-Gunn 2000). The current study fills an important gap in its consideration of multiple features of social disorganization and their effects on a range of academic outcomes.

Neighborhood Context and Adolescent Racial Identity

The development of a positive sense of personal identity is a primary developmental task of adolescence. Racial identity is a multidimensional construct that describes an individual’s sense of group identity based on membership in a particular racial group, and it includes both the importance and the meanings the individual attaches to her/his racial group membership (Sellers et al. 1998).1 Recent studies provide a growing body of evidence that as an individual-level attribute, racial identity can serve as a compensatory factor in relation to psychosocial adjustment among adolescents (Chavous et al. 2008; Sellers et al. 2003), in that youth with a stronger sense of connectedness to their racial group and feelings of group pride generally show more positive psychosocial and academic outcomes relative to those with less endorsement of these attitudes. For instance, Altschul et al. (2006) found that African American adolescents with strong connections to their group and who believed achievement was a characteristic of African Americans had higher grades, spent more time on homework, and missed fewer days of school. Conversely, high identification with Eurocentric racial attitudes has been related to lower grade achievement (Spencer et al. 2001) and to lower academic persistence and more school problem behaviors for early adolescents (Smalls et al. 2007).

Racial Identity Development in Context

Neighborhoods should play a role in how racial identity is developed and enacted in adolescents’ lives by influencing their beliefs and attitudes about race. Specifically, neighborhood social disorganization could be related to racial identity, just as it is related to behavior. For instance, community adults, especially African American adults, pass on their racial attitudes just as they pass on other values. Racial socialization is the process through which parents transmit values and beliefs about race to their children (Hughes et al. 2006), but community members also participate in collective racial socialization that may enhance or counter socialization from family. Lesane-Brown and colleagues, for example, note that a majority of adolescents and college students in their study reported receiving race socialization messages from other adults like teachers and neighbors (Lesane-Brown et al. 2005). Collective socialization may differ in communities of varying compositions; that is, youth in neighborhoods with few African Americans may have access to different role models and receive dissimilar messages compared to youth from predominantly African American neighborhoods.

Other neighborhood characteristics could influence youth’s beliefs about the meaning of being Black, as well. Youth in neighborhoods with low economic opportunity may make race-related attributions; that is, they could reason that limited opportunities exist because of discrimination against African Americans, or because of the abilities, skills, or work ethic of African Americans. Or, they could attribute constrained opportunity to unfair barriers for their group that can only be overcome with extra effort and persistence. Both types of beliefs have implications for how youth might evaluate and identify with their race. Similarly, institutional resources in neighborhoods may provide youth with opportunities to interact with other African Americans in positive contexts and to learn about Black history and culture in ways that influence positive constructions of their racial identity.

A few studies have investigated the link between racial identity and neighborhood context, and their findings are equivocal. Bennett (2007) found that perceptions of neighborhood risk, as measured by the Urban Hassles Scale, were negatively related to a positive ethnic group connection among African American urban adolescents, while Corneille and Belgrave (2007) found no direct relationship between racial identity and perceiving neighborhood risk. We are interested in whether, beyond youth-reported perceptions, actual levels of neighborhood risks and resources relate to adolescents’ racial identity beliefs.

Neighborhoods as Moderators of Racial Identity Effects

Scholars suggest that structural factors can influence the meanings youth attach to their racial group membership and the associations individuals make between their group and the academic domain. In fact, conceptual frameworks developed to explain African American achievement motivation processes implicate both the role of racial identity and contextual factors. However, the role of contextual factors has not been examined explicitly in empirical investigations. For instance, the aforementioned findings regarding the link between racial identity and academic achievement among Black youth are consistent with the racial identity-as-promotive perspective (Smalls et al. 2007), which asserts that because African Americans have been historically denied opportunities for educational and occupational mobility due to race, many Black communities place a high premium on academics as a means of attaining success and societal equality for the group. Thus, a strong connection with Black identity can be coupled with a sense of pride in the collective struggle of African Americans in ways that enhance achievement motivation, particularly in the face of group barriers.

The racial identity-as-promotive perspective raises the possibility that those who have been excluded from mainstream opportunities (those with access to fewer resources and opportunities) may be more likely to endorse the value of education as a means to mobility and thus show higher academic engagement. However, this possibility has not been examined empirically. In addition, research has demonstrated that racial identity may serve in a protective capacity, buffering youth from the deleterious effects of interpersonal risk factors, such as racial discrimination, on psychosocial and academic outcomes (e.g., Caldwell et al. 2004; Chavous et al. 2008; Wong et al. 2003; Sellers et al. 2003, 2006). For example, Caldwell et al. (2004) found that among young adult males reporting high levels of racial discrimination experiences, those with a higher identification had fewer violent behaviors than those who identified less with being African American. Additionally, Wong et al. (2003) found that reports of racial discrimination correlated negatively with valuing school and achievement, but that the negative effects were reduced for those with a positive racial group connection. These results suggest that having a strong and positive racial identity allows African American youth to persevere after experiencing racial discrimination by insulating them against negative messages and providing them with an alternative source of affirmation. However, within such studies, there has been little examination of whether the impact of racial identity on achievement varies among youth experiencing different levels of risk.

One of the more well-known frameworks for explaining African American achievement and underachievement is Fordham’s and Ogbu’s (1986) cultural-ecological framework. This framework asserts that African Americans’ experiences with constrained opportunity structures resulted in their developing beliefs systems that position Black identity in opposition to mainstream avenues of mobility, such as educational achievement. Thus, an emphasis on one’s Black identity would relate to decreased academic motivation and engagement. There is little evidence of this “acting White” phenomenon, or the idea that Black youth as a whole view doing well in school as a characteristic of Whites and incompatible with their Black identity (Smalls et al. 2007). In fact, a major critique of this perspective is that while it may useful in our understanding of processes through which some youth may disengage with schooling, it ignores variability among Blacks (O’Connor and Horvat 2006). Neighborhood-level factors may be one source of variation. Mickelson’s (1990) research highlights the importance of considering adolescents’ experience of context-level opportunities and constraints in relation to their achievement motivation attitudes. Thus, it may be that Fordham’s and Ogbu’s perspective better explains ways racial identity may develop and function in areas lacking opportunity rather than a normative process characteristic of all Black youth. For instance, it is possible that youth experiencing greater barriers in their communities are more likely to develop negative linkages between their racial group identity and the academic domain. However, this often is assumed (and is prevalent in popular discourse on Black youth) but not investigated empirically.

Could a stronger connection to one’s racial identity be a promotive or protective factor in relation to neighborhood-level risk factors, or does it exacerbate the effects of community risk? A few studies provide information that addresses this question. For instance, Corneille and Belgrave (2007) studied a sample of urban girls and found that as neighborhood risk increased, girls who endorsed more Pro-Black attitudes had less intention to use drugs than girls who did not endorse Pro-Black attitudes. Additionally, Miller and MacIntosh (1999) found that only when youth were weakly identified with being African American did general neighborhood risk relate to GPA. These studies suggest that when community-level risks are high, having a positive identification is even more adaptive than when risks are low. When considering family SES rather than neighborhood context, Chavous et al. (2002) found that for African American college students from less affluent and educated families, higher racial connectedness (centrality) related to high academic competence self-perceptions, while there was no significant relation for those from more affluent families. Among a sample of Latino college students, Ong et al. (2006) found that ethnic identity had a stronger relationship to GPA for youth from lower SES backgrounds than for those from higher SES backgrounds, suggesting a differential impact by background, which provides some support for the idea that racial identity may serve as a compensatory factor among youth coming from higher risk environments. The current study seeks to expand previous research by considering neighborhood-level risk factors and resources in diverse neighborhoods among adolescents.

Research Design and Hypotheses

The goal of this study is to determine how neighborhood context-level factors are associated with adolescent racial identity and the relationships between racial identity and academic achievement. To do so, we used hierarchical linear modeling (HLM). HLM allows us to take into account the fact that youth are nested within neighborhoods and to model individual and neighborhood effects at the same time. Analyzing multilevel data at a single level of analysis introduces bias and can lead to underestimated group effects. The technique is often used in neighborhood research (e.g., Garner and Raudenbush 1991) but rarely in racial identity studies (for an example, see Simons et al. 2002). Nevertheless, it is the most fitting technique to appropriately address our research question.

We examined six different outcomes as our academic indicators: Grade point average, school utility values, time spent on homework, number of absences, skipping class, and being sent to the principal’s office. Much of the social disorganization literature has focused on problem behaviors, while much of the racial identity literature has considered academic achievement, so there is a need to consider both types of outcomes. Additionally, by assessing both, we not only capture the end result of youth’s efforts in school (grade achievement), we also capture their behavior in and engagement with school. These behaviors should have consequences for grades but may have different relationships to our variables of interest. Our research questions are:
  1. 1.

    How do neighborhood characteristics relate to racial identity?

     
  2. 2.

    How do neighborhood characteristics and racial identity relate to academic engagement outcomes?

     
  3. 3.

    How do neighborhood characteristics moderate the relationships of racial identity with academic engagement outcomes?

     

We expected neighborhood characteristics to relate to racial identity, such that more positive neighborhood features, especially positive role models, would predict greater group attachment, importance, and pride. Since we used hierarchical models, our second hypothesis was that sufficient variation would exist between neighborhood types in our outcomes and that neighborhood characteristics would predict academic engagement. Specifically, neighborhoods high in social disorganization indicators would have more negative outcomes. Based on previous studies linking racial identity and achievement, we expected racial identity to be positively related to academic outcomes. In examining whether the relationship between racial identity and outcomes varied by neighborhood type, we were in a position to test the competing hypotheses suggested by the racial identity-as-promotive perspective and cultural-ecological perspective. Specifically, we considered whether higher levels of neighborhood disorganization would evidence a stronger relationship between racial identity and academic engagement in neighborhoods with more risk and fewer resources than more organized neighborhoods. Alternatively, we considered that racial identity may be promotive and have an even stronger relationship to outcomes in more organized neighborhoods.

Method

Participants

The participants are 564 African American adolescents (56% male), who are part of the Maryland Adolescent Development in Context Study (MADICS), a longitudinal study being conducted in Prince George’s County, Maryland. In 1993, when the data were collected, the county was predominantly African American and included a wide range of neighborhood types and income levels (more information can be found at http://www.rcgd.isr.umich.edu/pgc/). We included only self-reported African American youth from Wave 3 of the study, when participants had just completed eighth grade, and only youth who had complete neighborhood data (We chose to focus on Wave 3 because of our desire to study early adolescents. We also wanted the Census data to be as relevant as possible. Wave 4 was completed at the end of eighth grade, in 1996). The 42 youth excluded did not significantly differ from the youth who were included on any study variables. The average age of the participants was 12.27 (SD = 0.57), and their parents’ income ranged from less than $5,000 to more than $75,000 (median between $40,000 and $45,000). About 92% of the youth’s primary caregivers had graduated from high school; the median education level was about 1 year of postsecondary education. For the study, youth and their primary caregivers each completed 50 min face-to-face interviews and 50 min self-administered questionnaires.

Measures

Neighborhood-Level Variables

Neighborhoods are typically conceptualized as Census tracts, which contain between 3,000 and 8,000 individuals (Leventhal and Brooks-Gunn 2000). The Census Bureau creates the boundaries to reflect physical and social markers that residents recognize. In the current study, we used a combination of 1990 Census tract data and parent reports to characterize the neighborhood-level factors in order to capture both structural and perceptual features.

Collective socialization was the average of the standardized percentage of college graduates and the standardized percentage of adults employed in managerial or professional occupations. We measured economic opportunity by standardizing and averaging median family income and a scale of the likelihood of future success. On this scale, parents reported the likelihood of adolescents in their neighborhood graduating from high school, completing college, and finding a well-paying job on a scale of 1 (very low) to 5 (very high). The average responses for both African American and White parent participants in the study sample were then aggregated to the tract level so that the indicator measured overall perceptions of neighborhood success. Though the study focuses on the African American participants, we used all of the parent responses to increase the reliability of the neighborhood reports. Furthermore, we combined these two indicators to capture not only current economic standing, as other studies do, but also to capture views of neighborhood potential.

We assessed institutional resources by two scales. The first asked about the availability of six community services, including scouting troops and after-school programs. Parents responded “yes,” “no,” or “I don’t know.” The second scale consisted of four items asking for agreement on whether services were available (two items), schools were good, and how difficult it was to get help for their children on a scale from 1 (strongly disagree) to 5 (strongly agree). Responses were also aggregated to the Census tract level.

Racial Identity

Three aspects of youth’s racial identity beliefs were assessed. The racial connection measure (Wong et al. 2003) was composed of four items asking whether youth felt they had a close community of friends because of their group, whether they felt their group had a rich cultural heritage and meaningful traditions, and whether people of their group were supportive of each other (α = 0.69). Participants answered on a 5-point scale ranging from 1 (not at all true of me) to 5 (very true of me), such that a higher score represented greater connection to the group. Racial importance, the relevance of understanding the African American experience, taps into the centrality of individuals’ racial group membership and was assessed by one item asking youth how important knowing their racial background was to them. Racial pride, or individuals’ affective feelings about their racial group was assessed by one item asking youth how proud they were of their racial background. Both questions were on a four point scale from 1 (not at all) to 4 (very). Although our one-item indicators were not ideal, they each tapped into dimensions of racial identity established as conceptually distinct and differentially predictive of psychosocial outcomes (e.g., Ashmore et al. 2004; Chavous et al. 2003; Sellers et al. 1998). Thus, the racial importance and racial pride items were the most appropriate available in our data for a multidimensional examination of racial identity.

Academic Outcomes

We examined academic performance, attitudes, and behaviors as indicators of youth’s academic engagement. To measure grade performance, youth’s eighth GPAs were obtained from school records. A perceived school utility/values scale (Wong et al. 2003) included items taken from Mickelson’s (1990) abstract values scale. Items assessed how much youth valued school and consisted of four items: how much youth felt they had to do well in school to be successful in life, whether they felt kids in their neighborhood benefited from school, whether they felt school was not so important (reversed) and whether they felt they learned more useful things from friends than school (reversed). Items were on a scale of 1 (strongly disagree) to 5 (strongly agree). The reliability was 0.68. We assessed school engagement behavior in several ways. Youth were asked to report the number of days they had been absent in the previous year. They were also asked the average amount of time they had spent on homework each night in hours and minutes, and reports were converted into minutes. Finally, they reported whether in the last 6 months they had ever skipped class without a valid excuse or been sent to the principal’s office (0 = no, 1 = yes).

Control Variables

As neighborhood resources are tied to patterns of racial stratification in neighborhood racial composition, we included the neighborhood percentage of Black households in the models tested. To account for family level socioeconomic status (SES) in our examination of neighborhood context effects, we included parents’ reported highest education levels and income before taxes in tested models. Finally, we included youth’s age and gender in our models, because Black girls tend to have higher achievement than do boys (e.g., Kaba 2005). We also controlled for age, as in our sample older students were most likely held back, and thus would be more likely to show lower achievement than on-time peers.

Results

Table 1 gives descriptive statistics for all of the study variables. The average GPA was moderately high, nearly 3.50 on a 5-point scale, as was the average academic importance. The youth on average spent about an hour and a half on homework each night and missed 10 days of school, though the variation in both was large. About half of the sample had skipped class or been sent to the principal’s office in the previous 6 months. The mean level of connection to racial group was slightly below the midpoint (M = 2.77, SD = 0.89). Importance and pride were both high, with most youth answering “very” to both questions (M = 3.59, SD = 0.71, and M = 3.85, SD = 0.43, respectively).
Table 1

Means, standard deviations, and bivariate correlations for study variables

N = 564

M

SD

1

2

3

4

5

6

7

8

1. Age

12.27

0.57

1

−0.10

0.13

−0.11

−0.14

−0.14

−0.14

−0.16

2. Gendera

 

56%

 

1

0.03

−0.03

−0.04

−0.04

−0.06

−0.03

3. Percent Black HH

64.31

20.50

  

1

−0.26

−0.29

−0.64

−0.67

−0.59

4. Caregiver education

13.67

2.30

   

1

0.50

0.31

0.36

0.32

5. Caregiver income

9.56

4.28

    

1

0.40

0.50

0.39

6. Positive role models

0.00

0.84

     

1

0.86

0.79

7. Economic opportunity

0.00

0.82

      

1

0.76

8. Institutional resources

0.00

0.64

       

1

9. Connection

2.77

0.89

        

10. Importance

3.59

0.71

        

11. Pride

3.85

0.43

        

12. GPA

3.47

0.84

        

13. School utility value

3.81

0.62

        

14. Time on homework

86.70

52.68

        

15. Absences

10.38

10.18

        

16. Sent to principalb

 

51%

        

17. Skipped classb

 

44%

        
 

9

10

11

12

13

14

15

16

17

1. Age

−0.14

−0.08

−0.08

−0.18

−0.15

−0.04

0.17

0.05

0.07

2. Gender

−0.06

0.07

0.01

0.29

0.18

0.03

−0.01

−0.15

0.03

3. Percent Black HH

−0.02

−0.05

−0.01

−0.15

−0.02

−0.08

0.19

0.11

0.06

4. Caregiver education

0.13

0.08

0.05

0.26

0.15

0.08

−0.25

−0.04

0.01

5. Caregiver income

0.11

0.09

0.10

0.24

0.10

0.12

−0.32

−0.13

−0.10

6. Positive role models

0.04

0.07

0.10

0.18

0.05

0.09

−0.23

−0.11

−0.17

7. Economic opportunity

0.08

0.10

0.05

0.18

0.01

0.12

−0.23

−0.05

−0.10

8. Institutional resources

0.07

0.05

0.05

0.22

0.08

0.09

−0.24

−0.04

−0.14

9. Connection

1

0.16

0.16

0.11

−0.04

0.09

−0.06

−0.01

0.07

10. Importance

 

1

0.31

0.06

0.17

0.14

−0.08

0.01

−0.01

11. Pride

  

1

0.01

0.03

0.04

−0.02

0.01

0.03

12. GPA

   

1

0.36

0.20

−0.36

−0.30

−0.25

13. School utility value

    

1

0.17

−0.10

−0.23

−0.26

14. Time on homework

     

1

−0.10

−0.15

−0.13

15. Absences

      

1

0.09

0.08

16. Sent to principal

       

1

0.37

17. Skipped class

        

1

Note: Correlations do not reflect the nested structure of the data. Therefore, significance levels are not reported

aPercent male, male = 1, female = 2; b percent that answered “yes”

The means for the non-standardized neighborhood indicator variables indicate that the average percentage of college graduates and those in professional occupations was about 21 (SD = 11.20) and 28% (SD = 8.75), respectively. Median family income at the tract-level was about $47,000 (SD = $12,079). Likelihood of success was on average high, mean 3.51 out of 5 (SD = 0.47). Finally, the two institutional resources scales averaged 3.51 (SD = 0.87) and 3.84 (SD = 0.21). The youth on average lived in generally well-resourced neighborhoods.

Preliminary analyses included examining bivariate correlations; then we clustered the Census tracts to create suitable groups for analysis and conducted the HLM or regression analyses for the six academic outcomes. We also examined the relationships of the neighborhood characteristics to the racial identity variables.

Bivariate Correlations

Table 1 also shows the bivariate correlations between the study variables. The neighborhood characteristics were generally related to the outcomes in expected directions. The characteristics also had high correlations with each other, but we found little evidence of multicollinearity in preliminary regression analyses (i.e., none of the variance inflation factors were greater than 10; Kutner et al. 2005). For racial identity, connection and importance were related to some of the outcomes. The racial identity variables had small to moderate positive correlations with each other, which supports the multidimensional nature of racial identity, as it shows that the variables are indeed measuring disparate aspects of identity and not one unitary construct. Similarly, the academic outcomes were moderately correlated with each other. Our controls were significantly related to the outcomes and racial identity, so we included them in all models.

Neighborhood HLM Analyses

Participants lived in 123 Census tracts, with between 1 and 17 African American families in each tract (M = 4.59, SD = 3.99). Since about 60% of the tracts contained only one family, we clustered all of the tracts to create homogeneous neighborhood types following a procedure similar to Simons et al. (2004). Using Ward’s minimum-variance method, they clustered 259 block group areas into 46 clusters containing between 15 and 30 families based on five Census variables. Our clustering variables were median household income, percentage of female-headed households, percentage on public assistance, percentage of unemployed males, the percentage of families in poverty, and city distance. City distance was an ordinal variable created by MADICS staff indicating whether a tract was inner-city, suburban, or rural, and was included to ensure that tracts were geographically close. We specified 40 clusters for the K-means analysis, as this gave us more than ten participants per neighborhood. Decreasing the number of clusters to 30 did not greatly increase reliability of the estimates, and specifying 50 clusters tended to retain very small groups. On average, clusters consisted of three tracts. In the final 40 clusters, there were between 10 and 21 families in each cluster (M = 14).

All continuous variables were standardized before being entered into the HLM program (HLM for Windows 6.07), and models were estimated using full maximum likelihood estimation. First, we ran a fully unconditional model for each of the six outcomes to determine how much variance existed between communities and whether it was a significant amount. The intraclass correlation coefficients were between 2 and 7% (see Table 2), indicating that most of the variability lay within neighborhoods. Nevertheless, the amount of variance was significant and enough to justify further HLM analysis. Only time spent on homework and whether youth had been sent to the principal’s office did not vary significantly between neighborhoods. For those two outcomes, we conducted ordinary least squares (OLS) and logistic regression analyses.
Table 2

Final HLM models, by outcome

Outcome, ICC

GPA 4%

School utility value 2%

Number of absences 7%

Skipping classa,b

Fixed Effects

Coefficient

Std. error

Coefficient

Std. error

Coefficient

Std. error

Coefficient

Std. error

Average value

−0.01

0.04

0.00

0.05

0.00

0.04

−0.21*c

0.10

Level 2 control

    Percentage of Black households

        

Level 2 effects

    Positive role models

    

−0.10*

0.04

−0.69**

0.19

    Economic opportunity

  

−0.15+

0.08

  

0.39*

0.18

    Institutional resources

0.11*

0.05

0.13+

0.07

    

Level 1 controls

    Parent income

0.12*

0.05

  

−0.21***

0.05

  

    Parent education

0.15*

0.05

0.16***

0.03

−0.11**

0.04

  

    Genderd

0.58***

0.07

0.30***

0.07

    

    Age

−0.13**

0.05

−0.09*

0.04

0.13*

0.05

  

Level 1 effects

    Connection

0.08+

0.04

−0.09*

0.04

    

    Pride

−0.06

0.05

      

    Importance

  

0.15***

0.03

−0.05

0.07

  

Cross-level interactions

    Pride × Economic opportunity

−0.14**

0.04

      

    Importance × Economic opportunity

    

0.21*

0.09

  

    Importance × Institutional resources

    

−0.30**

0.09

  

Random Effects

Variance

df

Variance

df

Variance

df

Variance

df

Intercept

0.00

26

0.03*

37

0.01

37

0.04

37

Slope

0.01

26

  

0.09***

36

  

Level 1

0.75

 

0.83

 

0.71

   

aPopulation average model, b not calculated because outcome was dichotomous, c log-odds, d male = 1, female = 2, + p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001

Our method of analysis was to begin at the individual level (Level 1). We tested the controls (excluding neighborhood composition) and the racial identity variables individually and in subsets, proceeding both backward and forward. These variables were grand-mean centered. Those that were the strongest predictors were retained in the model. Next, we allowed each racial identity slope to vary in order to test whether the relationship between racial identity and the outcome was different in different neighborhood types. If the slope did vary, it was retained as a random effect and group-mean centered. In the second part of each analysis, we entered the neighborhood composition control variable and the social disorganization variables (Level 2), again testing each individually and in subsets. We tested for effects on both the outcomes and any significant slopes, that is, the relationships between racial identity and the outcomes. Only significant variables were retained, and all were grand-mean centered. Results are presented in Table 2. Table 3 gives the OLS and logistic regression results. Since the variables are standardized and grand-mean centered, the coefficients should be interpreted as the average effect for a standard deviation increase in the predictor, relative to the overall mean for the sample.
Table 3

OLS and logistic regression analyses, by outcome (all variables standardized)

 

Time spent on homework

Sent to principal’s office

B

Std. error

B

Std. error

Intercept

−0.03

0.14

1.10**

0.33

Parent education

0.02

0.05

0.04

0.12

Parent income

0.05

0.06

−0.37**

0.13

Percent Black HH

−0.02

0.06

0.32*

0.14

Gender

0.02

0.09

−0.68**

0.21

Age

−0.04

0.05

0.07

0.11

Positive role models

−0.16+

0.09

−0.39+

0.22

Economic opportunity

0.18*

0.09

0.37+

0.22

Institutional resources

0.02

0.08

0.23

0.18

Connection

0.06

0.05

−0.03

0.11

Importance

0.13**

0.05

0.05

0.12

Pride

−0.02

0.04

0.03

0.11

Model R2

0.05

 

0.09

 

Model F2

2.44**

 

29.21**

 

+p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001

HLM Analyses

GPA

GPA is a standard achievement measure reflecting youth’s overall school success. It can be thought of as an interaction between youth’s school efforts and factors within the school beyond their control, such as teacher perceptions. We expected both neighborhood advantage and a positive racial identity to be related to higher GPAs. Table 2 shows that one hypothesis was confirmed: Youth in neighborhoods with greater institutional resources had higher grades (B = 0.11, p = 0.03). Additionally, connection was marginally significant (B = 0.08, p = 0.07).

Neighborhood context moderated the relationship between racial identity and GPA in a way that partially supported our hypothesis. Specifically, when economic opportunity was low, youth reporting greater pride had higher achievement, as expected. Yet when economic opportunity was high, youth reporting greater pride had lower achievement (B = −0.14, p = 0.004), when we had expected to find a weaker relationship. In fact, the relationship was stronger and in a negative direction. Figure 1 demonstrates this finding. As can be seen, youth with higher pride in advantaged neighborhoods (one standard deviation above the mean) had lower GPAs than youth with high pride in disadvantaged neighborhoods (one standard deviation below the mean). The relationship for average economic opportunity was not significant. The model explained all of the between-neighborhood variance, as indicated by a non-significant variance component, and 21% of the variance between youth. Economic opportunity explained all of the variance in the pride slope.
https://static-content.springer.com/image/art%3A10.1007%2Fs10964-008-9381-9/MediaObjects/10964_2008_9381_Fig1_HTML.gif
Fig. 1

Relationship between pride and standardized GPA as a function of economic opportunity. High and low opportunity are defined as one standard deviation above and below the mean

School Utility Value

Utility reflects youth’s perceptions of how useful school is for their future. As expected, racial importance predicted greater valuing of school (B = 0.15, p < 0.001). Having access to better institutional resources also trended toward greater valuing (B = 0.13, p = 0.05). Contrary to our expectations, however, youth with greater connection reported lower value (B = −0.09, p = 0.02), and youth in neighborhoods with greater economic opportunity reported marginally lower value (B = −0.15, p = 0.05), Even accounting for these variables and the controls, a significant amount of between-neighborhood variance remained unexplained. The model did explain 13% of the variance between youth.

Absences

Both absences and skipping class reflect youth’s engagement in school and willingness to be present. Some absences are unavoidable, but it is assumed that excessive absences indicate a larger problem. We expected greater neighborhood disorganization and weaker racial identity to be associated with more absences, and this was the case for positive role models (B = −0.10, p = 0.03). We again found moderation of the neighborhood context on the effects of racial identity. In the same manner as pride and GPA, greater economic opportunity predicted a positive relationship between racial importance and absences, meaning that for youth in more advantaged neighborhoods, seeing the group as more important was related to more absences, while in the neighborhoods one standard deviation below the mean in economic opportunity, importance was negatively related to absences (B = 0.21, p = 0.03; see Fig. 2). Institutional resources had an even stronger effect in the opposite direction, which we also did not expect. That is, for youth in neighborhoods with more resources, greater racial importance was related to fewer absences; the opposite was the case in less resourced neighborhoods (B = −0.30, p = 0.004; see Fig. 3). In both cases, the relationship at the mean was not significant. For both slopes, we had expected to find a stronger relationship in more disorganized neighborhoods than in more organized neighborhoods. Interestingly, these two variables explained only 28% of the differences in the racial importance slope across neighborhoods. The full model explained all of the variance between neighborhoods and 25% of the variance between youth.
https://static-content.springer.com/image/art%3A10.1007%2Fs10964-008-9381-9/MediaObjects/10964_2008_9381_Fig2_HTML.gif
Fig. 2

Relationship between importance and standardized number of absences as a function of economic opportunity. High and low opportunity are defined as one standard deviation above and below the mean

https://static-content.springer.com/image/art%3A10.1007%2Fs10964-008-9381-9/MediaObjects/10964_2008_9381_Fig3_HTML.gif
Fig. 3

Relationship between importance and standardized number of absences as a function of institutional resources. High and low resources are defined as one standard deviation above and below the mean

Skipping Class

The model for skipping class was different from the model for absences. None of the individual-level variables predicted this outcome, but again, having more positive role models was associated with lower chances of having skipped class, as anticipated (B = −0.69, p = 0.001). Unexpectedly, however, we found that greater economic opportunity increased the chance that youth had skipped class (B = 0.39, p = 0.04). We also expected racial identity to be positively associated with this outcome, but none of the dimensions were. The final model explained 77% of the variance between neighborhoods. We could not calculate the variance explained at the individual level because the outcome was dichotomous.

OLS Analysis

Homework

We viewed the time youth reported spending on their homework as another indicator of their school engagement. The final two outcomes did not vary between neighborhoods, meaning that an HLM analysis was not appropriate. We expected youth in more organized neighborhoods and with stronger racial identities to report doing more homework. Interestingly, positive role models trended toward less time on homework (B = −0.16, p = 0.08) and economic opportunity was significantly related to more time on homework (B = 0.18, p = 0.049). Youth reporting greater importance of their racial group also reported more homework time (B = 0.14, p = 0.006). The model explained 5% of the variance between youth (Table 3).

Logistic Regression Analysis

Disciplinary Problems

Whether youth had been sent to the principal’s office is an indicator of their misbehavior in school, and over half of our sample had experienced this event. Positive role models showed a trend toward a lower chance of this occurring (B = −0.39, p = 0.08) and economic opportunity trended toward a greater chance (B = 0.37, p = 0.09) (Table 3).

In summary, institutional resources and positive role models had anticipated effects on youth’s outcomes, while higher economic opportunity was associated with more undesirable behavior. Furthermore, racial identity did not have uniformly positive effects, either across outcomes or across neighborhoods. In some cases, there was no effect or even a negative one.

Neighborhood Effects on Racial Identity

The purpose of the final analyses was to determine whether neighborhood characteristics predicted racial identity. The fully unconditional HLM models revealed that significant variation did not exist between neighborhood types, so we conducted traditional regression analyses with all three variables. The regressions controlled for parent education and income, percentage of Black households, and the other racial identity variables. They revealed no significant neighborhood effects on racial identity.

Discussion

The purpose of this study was to use a multilevel analysis to examine how racial identity and neighborhood effects individually and in combination related to the academic achievement of African American early adolescents. We analyzed data on youth’s reported racial identity and academic outcomes in the context of Census data and parent reports about the types of neighborhoods they lived in, guided by social disorganization theory. Our measures of community characteristics were different from many neighborhood effects studies, in that we used both perceptual and structural indicators, and our approach revealed important findings.

Neighborhoods

The results supported many of our hypotheses about neighborhood effects. We predicted that significant variation would exist between neighborhood types, and that was true for youth’s GPA, school utility value, absences, and skipping class. As in other studies (e.g., Cook et al. 2002; Leventhal and Brooks-Gunn 2000), neighborhoods accounted for small, yet important, amounts of variance in outcomes. More institutional resources related to higher GPAs, and more positive role models related to fewer reported absences and less skipping class. Greater economic opportunity predicted spending more time on homework but also more maladaptive engagement behavior. As we expected, more organized neighborhoods generally were associated with better outcomes.

Our positive role models findings are similar to other social disorganization studies (Ainsworth 2002; Crowder and South 2003; Brooks-Gunn et al. 1993). Unlike other studies, however, we also found positive effects for institutional resources but negative effects for economic opportunity. We did not expect to find that higher median incomes and parents’ perceptions that youth would be more successful would be associated with youth valuing school less than those in settings with lower levels of these characteristics. The more economically privileged youth may be able to afford to have lower school utility attitudes, however, since their contexts likely provide them with supports in other ways that may attenuate the effects of low school values. Alternatively, adolescents in lower-income areas may value school more because they see it as a way to better their position in life. It also appears that the school utility scale measures abstract values that are less connected to youth’s behavior (Mickelson 1990). The scale had only small (but significant) correlations to the other outcomes—the highest correlation was with GPA (r = 0.36). Thus valuing school may not translate directly to school achievement in the same ways for youth in more affluent and less affluent contexts. Finally, since school values were overall high in this sample, the difference in levels of these scores for those in more affluent and less affluent contexts may be less troubling than they seem.

Interestingly, positive role models were related to academic behaviors and not school utility value, while institutional resources and economic opportunity were marginally related to value, suggesting that adults with higher incomes and better educations may indeed be modeling behavior and values for youth. On the other hand, institutional resources, which included YMCAs and scouting troops, may help promote the importance of school for future success. Without looking at multiple indicators of academic engagement and neighborhood characteristics, we would not have observed these differences. Furthermore, we might not have observed the counter-intuitive results without including a perceptual variable of economic opportunity. Other studies using only indicators of poverty or income without perceptions of opportunity have not found significant results relative to collective socialization (Ainsworth 2002). Our results may point to the need for a re-examination of how neighborhood economic conditions relate to outcomes.

Racial Identity

We predicted that more connection, greater racial pride, and greater racial importance would be related to better academic outcomes, and these relationships would vary by neighborhood type. Also, we predicted a stronger relationship between racial identity and outcomes in more disadvantaged neighborhoods. Some of our hypotheses were supported. Youth reporting higher racial group importance tended to value school more and to spend more time on their homework, findings consistent with research pointing to the promotive functions of strong racial identity in relation to achievement (Altschul et al. 2006; Miller and MacIntosh 1999). However, the findings for racial group connection were more mixed, with stronger connection related to lower school utility value, but marginally related to higher GPA.

The findings highlight the importance of considering different dimensions of racial identity. The very small correlation between racial importance and connection suggests that youth’s reports of high racial importance and high racial connection tap into conceptually different belief systems. From a motivational perspective, the relevance of being African American to adolescents’ personal identities, as measured by racial importance, may relate to youth linking their personal success with being successful on behalf of their racial group. Thus, the value of school may not only be individual but also based in a desire for the advancement of African Americans (e.g., Smalls et al. 2007). These highly identified youth not only valued school more, they reported working harder in school. However, having a strong psychological attachment to members of one’s racial group and feeling supported by group members, as measured by connection, may influence motivational beliefs about education in different ways. Since the connection measure taps into youth’s personal connection to their group heritage and perceived support from more proximal members of their group, adolescents’ reports of higher group connection may reflect their observations and understandings of obstacles to achievement historically experienced by those in their community contexts. With this understanding, they may not endorse the value of education as much as those with lower group connection because they recognize the possibility that for people like them (i.e., African Americans), being successful in school will not necessarily pay off with success later in life (Mickelson 1990). However, the fact that racial connection also related to higher GPA shows that racial identity attitudes may influence abstract educational attitudes and performance outcomes in different ways. It is possible that adolescents with greater racial connection may feel pessimistic about the general payoff of school, yet still show personal engagement in school, which may imply a desire to overcome obstacles even while recognizing them. Future research can untangle these effects by measuring youth’s specific beliefs about African Americans and achievement. Another explanation is that highly connected youth may feel supported by others of their race, which enables them to achieve, while highly identified youth may not feel the same support or need it in order to achieve. Our findings suggest research examining different dimensions of racial identity in relation to achievement is needed.

Racial importance was not related to GPA, but it was related to engagement in the form of school absences, and this relationship differed across neighborhood types. The resulting relationships point to the complexity of neighborhood risk and the importance of particular resources in neighborhoods lacking in economic ones. Opportunities for structured activities may provide support for highly identified youth in ways that are absent for youth without those resources, even when the neighborhoods are similar in economic opportunity. Such neighborhood resources may play a particular role in shaping what it means to identify with being African American. We could not assess, however, how youth’s use of those resources shaped their identity, as we only had information on the presence and quality. Future research could explore the subtleties of these relationships by assessing how youth’s specific interactions with their neighborhood contexts shape their identities. For instance, a study could examine how youth’s participation in scouting troops or after-school programs is associated with achievement for adolescents of varying racial group identification.

Second, we found buffering effects of racial importance and pride in neighborhoods low in economic opportunity, as in other studies (Corneille and Belgrave 2007; Miller and MacIntosh 1999; Ong et al. 2006). Youth whose race was more important/central to them missed fewer days in school. In a context where youth are less likely to succeed, adolescents may feel even more motivated on behalf of their race than youth in lower-risk economic contexts. Furthermore, we find the only positive effects of racial pride in these neighborhood types. Youth with higher group pride had higher overall achievement. That is, having a negative evaluation of their group may be an added risk to adolescents, in addition to the risks of being in a lower-income area. Therefore, high pride may serve as an essential resource to help counter contextual risks. These results are consistent with studies suggesting the utility of a strong and positive group identification among youth experiencing neighborhood economic risks (Chavous et al. 2003; O’Connor 1997) as well as the racial identity-as-promotive perspective (Smalls et al. 2007), asserting that recognition of constrained opportunities for group members can promote achievement motivation. We find little support for the popular cultural-ecological perspective (Fordham and Ogbu 1986), which posits that ecological restraint leads to Black youth defining a strong racial identity in opposition to academic achievement.

Interestingly, the picture is very different in the more economically resourced neighborhoods. Both Corneille and Belgrave’s (2007) and Miller and MacIntosh’s (1999) studies were conducted with urban samples, and none of the parents of the participants in Ong et al. (2006) study had completed college. Our analyses in an economically diverse sample revealed surprising results for youth in the more advantaged, suburban communities. Adolescents with higher racial importance reported more absences than less identified youth. Also, youth with higher pride had lower overall achievement.

One important aspect of these particular contexts is that youth in areas with more economic advantage also lived among fewer African American families—the correlation between percentage of Black households and economic opportunity was r = −0.67. Though living in a more predominantly White neighborhood may offer numerous resources for adolescents, it may pose some risks in terms of racial identity. For instance, African Americans may be less valued as a group in those neighborhoods; subsequently African American adolescents may have experiences that lead them to identify less strongly with or feel less proud of their group. The stress associated with feeling that race is salient in settings where youth are the minority may be amplified for early adolescents as they struggle to find their place among their peers and establish their identities. Without messages to support a positive racial identity, youth may adopt the less positive evaluation of those around them in order to succeed. Youth who remain highly identified or proud in such negative settings may feel stigmatized in ways that lead to school disengagement or avoidance. This finding challenges Fordham and Ogbu’s (1986) contention that Black youth’s academic disengagement occurs because they perceive little opportunity for mobility and come to feel that achievement is not congruent with being African American. Instead, we speculate based on our findings that youth’s disengagement may relate to feelings of discomfort or devaluing in a setting where their group is the minority, even in contexts where youth observe more opportunities for economic mobility. Xue et al. (2007) found protective benefits of living in predominantly Black neighborhoods on cigarette smoking and suggests positive group socialization effects as a mechanism. Similarly, in our study it may be that living among members of their own race has positive consequences for some aspects of academic identity development. An alternative explanation, given the cross-sectional study design, is that low achieving and disengaged youth in more advantaged (and more predominantly White) contexts may choose to identify with their group more strongly as an alternative source of affirmation.

Finally, we found no direct, significant associations between neighborhood characteristics and racial identity variables. The findings are counter to our predictions based on social disorganization theory about the deleterious impact of neighborhood disorganization on the meanings adolescents attach to their racial group identity. However, in a study of adults, Parham and Williams (1993) found no relationship between family level socioeconomic status and racial identity attitudes. Together, the findings suggest there may be more variation within neighborhood contexts than across contexts, but that racial identity functions differently across neighborhood settings.

Limitations and Considerations

This study has illuminated how neighborhood-level characteristics interact with racial identity to predict adolescent achievement outcomes. Though we highlighted the need to take neighborhoods into account, we could not explicate more specific linkages between how adolescents identify with their racial group and their engagement with and achievement in school. Furthermore, since this is a cross-sectional study, we cannot determine causality from our results. Other limitations include our use of single indicators of racial importance and pride. When these data were collected, established multidimensional measures of racial identity (e.g., Sellers et al. 1997), were not available. However, the indicators we used are conceptually similar to such measures. In fact, given the various significant findings using these measures, it may be the case that the effects we found may be underestimating the associations among racial identity, neighborhoods, and achievement, relative to what we would have found with current multi-item measures.

A consideration in interpreting our findings is that our neighborhood characteristics variables were based on parent data rather than youth perceptions. A strength of this approach is that we demonstrated the influences of neighborhood characteristics beyond youth’s own perceptions and family level factors, which was a main study goal. Linking neighborhood-level factors with youth’s own neighborhood perceptions and racial identity beliefs would provide insights into mechanisms more proximal to the youth, however.

Another consideration is that our sample is primarily middle-class. About 90% of the sample had incomes higher than half of the Black population in 2006 dollars (U.S. Census Bureau 2007). While these results may not represent all African American youth, they do speak to the experiences of a population that often has been overlooked in research (Pattillo 2005). Furthermore, our use of this sample allowed us to show the risks and opportunities associated with neighborhood levels of resources opportunity, even among families in relatively advantaged locations.

A strength of our study is that we used HLM and thus avoided common problems associated with analyzing data at one unit of analysis; unfortunately, we could not completely avoid aggregation bias in our measures of the neighborhood context, nor did we have access to more systematic measurements of neighborhood social characteristics (i.e., community surveys or observer ratings). Additionally, we could not examine gender differences because dividing the sample would not have left enough individuals per group to conduct the analyses. Nevertheless, our results showed what traditional regression analyses could not: that the meaning of racial identity is not the same in every neighborhood.

Conclusions

Neighborhood context matters in the meanings and functions of racial identity among African American early adolescents. The socioeconomic and social organizational features of neighborhoods are associated with differences in how racial identity relates to youth’s academic values, behaviors, and achievement. A sense of racial pride and importance can buffer some negative effects of living in more disorganized, less advantaged neighborhoods on school engagement outcomes, and there may be some negative effects associated with being highly racially identified in more advantaged neighborhood contexts. In the end, our results show that, in order to positively affect youth achievement, researchers and educators not only need to consider race differences in their approaches to understanding African American students, they also need to consider variation among youth in their beliefs about race and how they link these beliefs to achievement motivation. Our study demonstrates the utility of considering as sources of variation the levels and types of opportunities and constraints youth experience where they live.

Footnotes
1

Recently, various racial identity scholars have asserted that racial identity and ethnic identity are similar constructs that often are referred to interchangeably (see review by Quintana (2007), for a discussion). For African Americans in particular, it has been suggested that it is difficult to distinguish and disentangle beliefs systems that might be considered “racial” from those considered “cultural”, due to the historical and structural forces related to race that have at least in part circumscribed African Americans’ development of beliefs systems, traditions, and practices (e.g., Miller and MacIntosh 1999). We choose to use the term “racial identity” in the present paper because we were interested in youth’s beliefs about their group in relation to its function and status as a racial group in society.

 

Copyright information

© Springer Science+Business Media, LLC 2009