For many African American youth, adolescence is the developmental period in which the process of negotiating racial identity comes to the fore. Developmental changes, such as increased autonomy, physical maturation, and initial experiences with dating, set the stage for a wide range of situations and experiences that increase the prominence of race and invite these youth to consider the personal significance and meaning of race (Fisher et al. 2000; Sellers et al. 2006; Tatum 1997). At the same time these changes are occurring, cognitive gains, such as growth in abstract thinking, cognitive processing, social perspective-taking, and the ability to integrate one’s own experiences and the experiences of others (Cooper et al. 2008), heighten awareness of race and make it possible for African American youth to actively grapple with the significance and meaning of being African American. This process of negotiating racial identity is a complex and multi-faceted one with implications for developmental outcomes such as academic motivation (Chavous et al. 2008; Harper and Tuckman 2006; Smalls et al. 2007) and psychological well-being (Sellers et al. 2006; Simons et al. 2002; Wong et al. 2003).

The African American family has been one of the most important and oft-cited social contexts to play a formative role in how African American adolescents make sense of the significance and meaning of race (Bowman and Howard 1985; Boykin and Toms 1985; Hughes et al. 2006). Within the African American family, scholars have asserted that racial socialization, often conceptualized as parents’ implicit and explicit messages about race, help African American youth to develop a positive racial identity, particularly in the face of racial bias and adversity (Demo and Hughes 1990; Sanders Thompson 1994; Stevenson 1995). Unfortunately, a number of methodological challenges such as the conceptualization and treatment of racial socialization in statistical analyses and the preponderance of cross-sectional studies in the extant literature, make this premise a difficult one to evaluate. In the present study, we use two waves of data to examine the relationship between racial socialization and racial identity with a specific emphasis on patterns of African American racial socialization practices as they relate to the personal significance and meaning that African American adolescents ascribe to race.

Racial Socialization

Racial socialization has been defined as “the transmission of parents’ world views about race and ethnicity to children by way of subtle, overt, deliberate and unintended mechanisms” (Hughes 2003, p. 15). As noted above, these world views have often been viewed as a class of adaptive and protective messages and practices used to promote positive racial identity and prepare ethnic minority children to combat racism (Bowman and Howard 1985; Boykin and Toms 1985; Stevenson 1994). In light of a number of recent studies showing evidence that particular messages about race, as well as socialization behaviors, compensate for and protect against the harmful effects of African American racial discrimination experiences (e.g., Harris-Britt et al. 2007; Neblett et al. 2006, 2008), further research is necessary to learn more about the nature of racial socialization processes and the mechanisms that underlie the relationship between racial socialization and positive youth development. Adolescents’ racial identity attitudes may be an important mediating factor in these processes. As a result, it is necessary to further investigate the link between racial socialization and racial identity.

African American parents engage the racial socialization process in a variety of ways (Hughes et al. 2006; Spencer 1983; Thornton et al. 1990). Some parents consider the discussion of race a central component to raising their children (Bowman and Howard 1985; Hughes and Chen 1997), others discuss race-related issues only when their children bring them up, and still others convey messages to their children that completely de-emphasize the importance of race. When they do discuss race, African American parents also vary in the content of the messages they send their children regarding what it means to be African American. Although a detailed review of racial socialization typologies is beyond the scope of the current study, some of the primary themes to emerge in the extant literature include racial pride messages, racial barrier messages, egalitarian messages, self-worth messages, negative messages, and socialization behaviors (Bowman and Howard 1985; Hughes et al. 2006). Racial pride messages emphasize African American unity, teachings about heritage, and instilling positive feelings toward the racial group. Racial barrier messages emphasize an awareness of racial inequalities and strategies for coping with racial adversity. Egalitarian messages emphasize interracial equality and coexistence. Self-worth messages promote feelings of individual worth within the broader context of the child’s race. Negative messages, on the other hand, reinforce negative societal stereotypes about African Americans. Finally, recent studies have begun to include socialization behaviors––race-related activities and behaviors (e.g., buying African American literature and art) that convey implicit messages about race to children. These nonverbal messages work in concert with specific verbal messages to suggest to the child the importance and meaning of being African American (Coard and Sellers 2005; Hughes et al. 2006; Neblett et al. 2008).

Racial Identity

African American racial identity has been conceptualized in a number of ways. In the last ten years or so, however, the Multidimensional Model of Racial Identity (MMRI; Sellers et al. 1998) has provided an instrumental conceptual frame for advancing what we know about African American racial identity. According to the MMRI, racial identity is that part of individuals’ self-concepts that is related to their membership within a race. In the model, racial identity is comprised of both the significance individuals place on race in defining themselves and their interpretations of what it means to be Black.

The MMRI proposes four dimensions of racial identity in African Americans: the salience of identity (for a detailed review of salience, see Sellers et al. 1998); the centrality of the identity; the regard in which the person holds the group associated with the identity; and the ideology associated with the identity. The first two dimensions address the significance of race in the individual’s self-definition while the second two dimensions address the qualitative meaning that the individual ascribes to being Black. The centrality dimension of racial identity refers specifically to the extent to which race is a core part of an individual’s self-concept. Racial regard refers to individuals’ affective and evaluative judgments of their race and consists of both a private and a public component. Private regard refers to the extent to which individuals feel positively or negatively towards African Americans and their membership in that group. Public regard, on the other hand, refers to the extent to which individuals feel that others view African Americans positively or negatively. Finally, racial ideology is a composite of individuals’ beliefs, opinions, and attitudes with respect to how Blacks should act. This dimension represents the person’s philosophy about the ways in which African Americans should live and interact with society. Four ideologies are proposed: (1) a nationalist philosophy, characterized by a viewpoint that emphasizes the uniqueness of being Black; (2) an oppressed minority philosophy, characterized by a viewpoint that emphasizes the similarities between African Americans and other oppressed groups; (3) an assimilationist philosophy, characterized by a viewpoint that emphasizes the similarities between African Americans and the rest of American society; and (4) a humanist philosophy, characterized by a viewpoint that emphasizes the commonalities amongst all humans. In contrast to unidimensional approaches to racial identity (e.g., treating racial identity as a composite sum score), these four dimensions allow for a more complex and nuanced view of the ways in which African American youth define themselves with respect to race.

Parental Racial Socialization and Racial Identity

Whether implicitly or explicitly, there is evidence to suggest that parental racial socialization is associated with a number of race-relevant attitudes, beliefs, and awareness (Branch and Newcombe 1986; Demo and Hughes 1990; Hughes and Johnson 2001). For example, children whose parents emphasize their group’s culture, history, and heritage in socializing their children have been found to report more knowledge about their group (Knight et al. 1993), more favorable in-group attitudes (e.g., Knight et al. 1993; Marshall 1995; Stevenson 1995), and more positive self-concepts (Ou and McAdoo 1993). Branch and Newcombe (1986) reported that parents who taught their children about the positive aspects of their race had children with high “racial awareness, knowledge and preference” (p. 36). In a study of Black adults, Demo and Hughes (1990) found a relationship between racial identity and parental socialization such that adults who reported receiving race-related socialization messages that emphasized racial pride and getting along with Whites from their parents, were more likely to have strong feelings of closeness to other Blacks and to hold stronger support for Black separatism. Most recently, Hughes and Johnson (2001) reported that African American children who received messages about prejudice and discrimination were more likely to engage in self-exploration around the significance and meaning of race.

There are a few published studies that have specifically examined the association between racial socialization and racial identity. In one study, Stevenson (1995) found that African American adolescents’ beliefs regarding what racial socialization messages should be conveyed by African American parents were related to the adolescents’ scores on the Racial Identity Attitudes Scale (RIAS). In another study of 11-year-old African American youth, Wills et al. (2007) found that perceived caregiver racial socialization was related to youth-reported ethnic pride. Although neither study directly measured the frequency of racial socialization messages received, the findings suggest a possible link between messages about race and adolescent racial identity. Despite this link, a handful of studies report evidence to the contrary (e.g., Parham and Williams 1993). Phinney and Chavira (1995), for example, found that although African American parents reported providing the most extensive racial socialization, racial socialization was unrelated to their children’s ethnic identity. In a study of 225 African American adults, Sanders Thompson (1994) found that parental racial socialization was related to “awareness, knowledge, and acceptance of the cultural and social traditions of African Americans,” but unrelated to three other racial identity parameters examined in the study.

A number of theoretical and methodological limitations may account, in part, for the discrepancies in findings across studies of racial socialization and racial identity. One particular limitation relates to the various ways in which racial identity has been conceptualized and operationalized across studies. With respect to racial identity, researchers have used unidimensional approaches to conceptualize and measure racial identity. However, a number of researchers have argued that racial identity is a multifaceted construct that should be conceptualized and operationalized as such (e.g., Ashmore et al. 2004; Sellers et al. 1998). Further confounding the interpretation of findings in some studies is the use of composite scores to operationalize racial identity (e.g., Phinney and Chavira 1995). Quantifying racial identity with a sum score fails to capture the multiple dimensions of identity. Even when researchers have taken a multidimensional approach, researchers have emphasized different aspects of identity (e.g., a relative emphasis on attitudes, opinions, and beliefs versus feelings or actual behavior) and rarely agree on which dimensions truly reflect racial identity.

If the problems with racial identity are not sufficient alone to raise concerns, the conceptualization and measurement of racial socialization comes with its own set of problems and methodological challenges. For one, racial socialization (a process thought to begin in childhood) has typically been assessed in adults or parents, leaving little room for the adolescent’s perspective of the racial socialization process. Second, it has not always been possible to measure actual racial messages that parents convey to their children about race. Earlier studies conducted by Stevenson (1995), for example, measured beliefs about racial socialization and not racial socialization messages themselves. More significant yet, is the failure to account for the synergetic nature of the racial socialization process. Typically, studies have utilized analytic approaches that focus on the relationship between individual racial socialization messages (e.g., racial pride) and specific outcomes after controlling for the influence of other socialization messages that may be present. Some have even collapsed different racial socialization messages into one racial socialization sum score making it difficult to examine how the specific content of messages is related to racial identity and how different kinds of messages work together to influence identity. Both approaches limit the extent to which conclusions can be drawn regarding the cumulative impact of multiple messages on youth racial identity.

A third, substantive consideration in the current research literature on the relationship between racial socialization and racial identity is the preponderance of cross-sectional studies. To date, we are unaware of any studies examining the relationship between racial socialization and racial identity using prospective data. Such a limitation makes it difficult to draw strong inferential conclusions about the relationship between the two, and even more specifically, the direction of that relationship. In addition to the general ineffectiveness of cross-sectional designs in establishing the direction of effects, many studies that have examined the relationship between racial socialization and racial identity have been conducted with African American adults recalling their childhood experiences, thus requiring them to remember back to what their parents told them many years before. In effect, the majority of available studies employ a retrospective mechanism of investigation for a question that would be best suited to a prospective approach.

Finally, the relationship between racial socialization and racial identity must be understood within a developmental context. The range of ages across populations spanning from elementary school-aged children (e.g., Hughes and Johnson 2001) to adults (e.g., Sanders Thompson 1994) is one factor that may partially explain the diversity of findings across studies. Parental messages about race vary across ages and developmental periods (Hughes and Chen 1997) and the significance and meaning of race is likely fluid, even across the several years that span the developmental period of adolescence, as a function of different experiences with race. Gender and socioeconomic context are also important developmental influences to consider in the study of the relationship between racial socialization and racial identity. Bowman and Howard (1985) and others (e.g., Hughes and Chen 1997) have reported gender differences with respect to the kinds of messages that parents emphasize to their sons and daughters, while socioeconomic backgrounds may shape different ideas that parents have about race and pass onto their children (Neblett et al. 2008). These sociodemographic factors may shape adolescent racial identity directly or indirectly via racial socialization, and thus are important to consider in assessing the relationship between racial socialization and racial identity.

The Present Study

In the present study, we investigate the relationships among parent racial socialization practices and racial identity in African American middle and high-school students. As with prior work, we attempt to address several shortcomings in the extant literature by utilizing cluster analytic techniques to capture the synergetic nature of racial socialization within a longitudinal investigation. Two primary objectives comprise the study. First, we identify different patterns of adolescent-reported racial socialization across five different types of verbal messages as well as the nonverbal messages conveyed by racial socialization behaviors. This approach allows us to focus on the individual adolescent as opposed to the individual socialization variables as the unit of analysis. Consistent with prior theory and the one available study to use a cluster analytic approach with multiple racial socialization messages (Neblett et al. 2008), we predict at least two different socialization patterns which include a relative emphasis on racial pride, racial barrier, and self-worth messages, and a second pattern characterized by the relative low frequency of racial socialization messages. Our second aim is to examine whether patterns of racial socialization are associated with adolescent-reported racial identity at a later time point. We do so while taking into consideration the role of developmental context (as indexed by adolescent gender, developmental period, and socioeconomic background) in the interpretation of the findings. Consistent with the idea that racial socialization promotes racial identity (Peters 1985), we predict that racial socialization will be associated with subsequent racial identity. While we are unable to make specific predictions, we expect that patterns of racial socialization that emphasize racial pride, racial barrier, and self-worth messages will be associated with racial centrality and public and private regard, all of which have been identified as important compensatory and resilience factors in the context of African American adolescents’ experiences with racial discrimination.

Method

Participants

Data for the present study come from the first two waves (referred to as Time 1 and Time 2) of a longitudinal study of race, psychological adjustment, and academic achievement. Three hundred fifty-eight self-identified African American youth in grades 7 through 11 were recruited from all eleven middle and high schools in a medium-sized public school district in the Midwestern United States. The sample consisted of 144 males (40%) and 214 females (60%). Participants ranged in age from 11 to 17 years of age with a mean age of 14 at Wave 1. Approximately 65% of the sample were in the 7th (n = 119) and 8th grades (n = 113) during the first wave of the study. The remaining 35% of students were spread across the 9th (n = 66), 10th (n = 56), and 11th (n = 4) grades. A primary caregiver was identified for each child, and the primary caregiver’s highest level of education completed (as reported by the child) was used as an indicator of educational attainment. The highest level of educational attainment was less than a high school diploma for a small minority of the primary caregivers (7.5%), a high school diploma for 20% of the sample, 43.5% attended some college or received a college diploma, and 29.1% completed some graduate study or attained a master’s degree, Ph.D., J.D., or M.D.

The study was conducted in a city with a population of approximately 110,000. The public school district was comprised of roughly 9,613 students in middle and high school. The student population was comprised primarily of White students (57%) followed by African American students (20.2%). The racial composition of the student body varied significantly across the 11 schools in the study, with the percentage of African American students ranging from 7.2% to 64.9%. On average, 18% of the students in the district were eligible for free or reduced lunch. Whereas the median family income for African American parents in the city surpasses the national average (parent-reported median family income $30,000–39,000), there is significant variation across the sample, with parent-reported household incomes ranging from less than $10,000 to more than $130,000.

Procedure

The sample was recruited through each of the eleven middle and high schools in a Midwestern public school district. Parents were initially contacted through information provided by their child’s school district. The school district provided the study team with contact information for parents of children identified as African American, Biracial, or other. Letters endorsed by the school district were sent to parents to invite them and their child to participate in a longitudinal study exploring the diverse ways parents discuss race and race-related issues with their children, how these messages contribute to their children’s feelings about themselves, and how these feelings relate to their school performance. Follow-up phone calls were conducted to ensure that contact information was accurate and to address any concerns about participating in the study. Youth administrations were approximately an hour in length and were held after school and monitored by trained African American research assistants. Participants were compensated with $20 or $30 certificates to the local mall for their participation in the first and second waves of the study, respectively.

A cross-sequential design was employed in which a new cohort of participants was recruited to participate in the study during each spring from 2002 through 2004. As a result, we collected data three times from our first cohort, twice from our second cohort, and once from our third cohort. The data were collected in the spring and summer across the three years and spacing between data collections points was relatively stable across and within cohorts. Students who participated in more than one wave of the study were recontacted at school and at home via phone for follow-up participation. The present study focuses on the first two cohorts and the first two waves of data. During the three years of the study, the households of 742 children were contacted to participate in the study. A total of 546 adolescents participated in one of the three cohorts for an overall participation rate of 74%. Of the 465 students in cohorts 1 and 2 who completed Wave 1, 358 students also completed wave 2 for a retention rate of 77%. Of these 358 students, 258 (72%) were in cohort 1 and 100 (28%) were in cohort 2. There were no appreciable differences in the retention rates across the two cohorts (83–76%). Adolescents who participated in both Waves 1 and 2 did not differ in gender composition, age, or parent educational attainment from those who dropped out after Wave 1.

Measures

Racial Socialization

The Racial Socialization Questionnaire-Teen (26 items; Lesane-Brown et al. 2006) was used to assess the frequency of six types of racial socialization practices. Reliabilities are reported for Time 1. Participants were asked to respond to each of items using a 3-point Likert-type rating scale (0 = never to 2 = more than twice) indicating the frequency with which their parents had engaged in various racial socialization practices. The Racial Pride subscale (4 items; α = .63) measures the extent to which primary caregivers encourage their child to take pride in their racial group, and the history, norms, customs, and traditions of that group (e.g., “You should be proud to be Black”). The Racial Barrier subscale (4 items; α = .69) measures the frequency of messages that prepare the child for racial adversity in the broader society (e.g., “Blacks have to work twice as hard as Whites to get ahead”). The Egalitarian subscale (4 items; α = .64) measures the extent to which primary caregivers communicated that people of all races are equal and should not be treated differently (e.g., “you should try to have friends from all different races”). The Self-Worth subscale (4 items, α = .74) measured the frequency with which the primary caregiver communicated that the child has value both as an individual and as a person of color (e.g., “you can be whatever you want to be”). The Negative subscale (5 items; α = .66) measured the extent to which caregivers conveyed messages that disparaged Black people (e.g., “Told you that learning about Black history is not that important”). The Racial Socialization Behaviors subscale (5 items; α = .73) assessed the frequency of socialization activities or behaviors related to Black culture (e.g., “Bought you books about Black people”).

Racial Identity

The Multidimensional Inventory of Black Identity––teen (Scottham et al. 2008) was used to assess racial identity at Time 1 and Time 2. All reliabilities are reported for Time 2. Participants were asked to respond to each of 21 items using a 5-point Likert-type rating scale (1 = really agree to 5 = really disagree) indicating the extent to which they agreed with each statement. The scale consists of seven dimensions, with three items comprising each of the subscale dimensions. Scores were recoded such that higher scores represented higher endorsement of the subscale items and scales. Centrality assessed the extent to which race is an important part of the individual’s overall personality (e.g., “I have a strong sense of belonging to other Black people”) (α = .56). Private regard captured positive feelings toward one’s racial group (e.g., “I feel good about Black people”) (α = .70). Public regard assessed the extent to which the adolescent believed that others view Blacks in a positive manner (e.g., “People from other races think that Blacks have made important contributions”) (α = .68). Assimilationist ideology assessed the view that Blacks should become more like Whites and emphasize a mainstream American identity over an African American one (e.g., “Blacks should act more like Whites to be successful in this society”) (α = .66). Humanist ideology measured the belief that people should be viewed in terms of their individual traits and similarities with all human beings instead of being viewed through social identities such as race (e.g., “Blacks should think of themselves as individuals, not as Blacks”) (α = .46). Minority ideology measured the perspective that there are similarities between Blacks and other oppressed ethnic minorities (e.g., “People of all minority groups should stick together and fight discrimination”) (α = .47). Nationalist ideology emphasized the uniqueness of Blacks’ experience as an oppressed group in America (e.g., “Blacks should support Black entertainment by going to Black movies and watching Black TV shows”) (α = .67). Due to problems with the internal reliability of the humanist and minority subscales, these subscales were dropped from subsequent analyses.

Demographic Variables

In addition to gender and primary caregiver’s highest level of educational attainment, we used the participant’s current grade to assign a value for adolescent developmental period. Developmental period was designated as ‘Middle School’ for students in Grades 7 and 8, and ‘High School’ for students in Grades 9–11.

Results

Data were analyzed for adolescents with complete data for both waves of the study (N = 358). We used Analysis of Variance (ANOVA) and Chi-Square analyses to determine whether participants who completed only one wave of the study differed from those who completed both waves of the study on any of the racial identity, racial socialization, or sociodemographic variables at Time 1. The analyses indicated no significant differences between the two groups on any of these study variables.

Preliminary Analyses

Preliminary analyses focused on descriptive statistics among racial socialization variables at Time 1 and racial identity variables at Time 2. Participants reported receiving racial barrier messages (M = 1.16, SD = .56), egalitarian messages (M = 1.35, SD = .52), and behavioral socialization (M = 1.17, SD = .53) from their parents approximately once or twice. They reported receiving slightly more racial pride and self-worth messages (M = 1.47, SD = .49; M = 1.71, SD = .44, respectively). Adolescents in the sample reported receiving few negative race messages (M = .17, SD = .31). Examination of the racial identity variables indicated high levels of agreement with items endorsing racial centrality (M = 3.82, SD = .83) and private regard (M = 4.58, SD = .63). The sample endorsed public regard (M = 3.20, SD = .92) and nationalist ideology (M = 3.36, SD = .89) at slightly lower levels. Assimilationist items (M = 1.73, SD = .80) garnered the lowest level of endorsement.

The associations among Time 1 racial socialization and Time 2 racial identity were examined using zero-order correlations. In general, adolescent reports of the six racial socialization subscales were positively related to one other. Reports of negative messages, however, were more modest, and inversely related to reports of racial pride (r = −.13, p < .01) and self-worth messages (r = −.22, p < .01). Adolescents’ reports of racial identity were also related to each other, with correlations ranging from slight to moderate (Table 1).

Table 1 Zero-order correlations among demographic, racial socialization, and racial identity (Time 2) outcome variables (N = 358)
Full size table

There were also significant bivariate relationships between the Time 1 racial socialization variables and the Time 2 racial identity variables. For example, higher levels of Time 1 racial barrier messages, racial pride messages, and socialization behaviors were associated with higher levels of Time 2 centrality and nationalist ideology. Higher levels of Time 1 racial barrier messages were associated with lower levels of Time 2 public regard. Time 1 negative messages were associated with higher levels of assimilationist ideology at Time 2.

Racial Socialization Clusters

Latent GOLD (Vermunt and Magidson 2005) was used to perform cluster analyses on the six subscales of the Racial Socialization Questionnaire-Teen at Time 1 in order to identify different profiles of racial socialization messages and behaviors that existed within the sample. Latent class analysis is a multivariate technique in which latent constructs are created from indicator variables and used to create clusters. Identifying the appropriate cluster solution using this method involves several steps. First, model fits and comparisons are assessed using the likelihood ratio chi-squared statistic (L2), an index of the association between the variables that remains unexplained. Lower values indicate a better fit to the data. Typically L2 is compared to the chi-squared distribution (producing a significance level that serves as a second diagnostic tool). However, since L2 is not well-approximated when the number of indicators or the number of categories of these indicators is large, the alternative bootstrap p-value is recommended (Langeheine et al. 1996). A non-significant bootstrap p-value (>.05) indicates that the model is a better fit to the data than the other models. Finally, the Bayesian Information Criterion (BIC) serves as an alternative indicator of overall goodness-of-fit. Lower BIC values indicate a better fit to the data.

Five latent class models (ranging from 1 to 5 clusters) were estimated using the subscales of the Racial Socialization Questionnaire-Teen. Summary statistics for these models are displayed in Table 2. Of the five models estimated, the three-cluster solution appeared to be the most appropriate fit to the data. The three-cluster model had the lowest BIC (5619.07) compared to all of the other models. In addition, this model had a substantial reduction in L2 (24.59%) over the baseline compared to the 2-class model, and a non-significant bootstrap p-value (.17). A comparison of the 3-class model to the 2-class model (using a conditional bootstrap method) was significant, (p < .01), confirming that the 3-class model provided a significantly better fit to the data.

Table 2 Model fit statistics for latent class analyses of racial socialization classes
Full size table

Local dependence for each variable pair was also examined. The Latent Gold program provides the bivariate residual (BVR) as a diagnostic statistic that assesses the bivariate relationships among indicators (Magidson and Vermunt 2004). A BVR that is larger than 3.84 suggests that the model falls somewhat short of explaining the association between a pair of indicators (Magidson and Vermunt 2004). By allowing dependence among the indicators (in this case, the racial socialization subscales), models with better fit to the data can be estimated. In the present study, the self-worth and negative messages pair had a substantially large BVR (4.99). Consequently, a three-cluster model with the direct effect between self-worth and negative messages was estimated. The direct effect accounted for the residual correlation between the two indicators and provided a more parsimonious model with a better fit. The modified model had a smaller BIC (5611.63), a larger reduction in L2 (25.30%), and acceptable BVRs. This model was adopted as the final cluster solution and used in the study analyses. Raw and standardized means of each racial socialization variable (Table 3) were used to label and describe the clusters. Standardized means of the indicator variables by cluster group are summarized graphically in Fig. 1.

Table 3 Raw means, standardized means, and (standard deviations) of racial socialization subscales by racial socialization cluster group (N = 358)
Full size table
Fig. 1
figure 1

Standardized mean frequencies of racial socialization subscales by racial socialization class

Full size image

High Positive

The first cluster was labeled ‘High Positive’ and comprised 36.6% of the sample (n = 131). This cluster was characterized by high means relative to the rest of the sample on racial pride, behavioral socialization, and self-worth messages. This group was also high on egalitarian and racial barrier messages. In contrast, scores for negative messages were below the sample mean.

Moderate Positive

The second cluster was labeled ‘Moderate Positive’ and comprised 40.8% of the sample (n = 146). This group was characterized by racial socialization subscale scores that were near the sample mean for most of the subscales. With respect to the socialization subscale raw means, the Moderate Positive cluster had relative high scores for egalitarian, racial pride, and self worth messages, moderate scores on the racial barrier and socialization behavior subscales, and low scores on the negative subscale.

Low Frequency

The ‘Low Frequency’ cluster comprised 22.6% of the sample (n = 81) and was characterized by low scores on most of the racial socialization variables. Participants’ scores were approximately 1 standard deviation below the sample mean on five of the six socialization subscales. A notable exception was negative messages, for which youth in this cluster reported a frequency of messages close to the sample mean.

Cluster Group Differences in Demographic Variables

Analyses were conducted to assess whether cluster groups differed by developmental period, gender, or primary caregiver’s educational attainment. Cluster group membership was related to developmental period, χ2 (2, N = 358) = 6.88, p < .05. Inspection of crosstabulation cell frequencies suggested that middle school students were more likely to be in the Moderate Positive cluster group, whereas high school students were more likely to be in the High Positive group. Cluster group membership was also related to adolescent gender, χ2 (2, N = 358) = 12.06, p < .01. Based on the distribution of crosstabulation cell frequencies, it appeared that girls were more likely to be in the Moderate and High Positive groups, whereas boys were more likely to be in the Low Frequency cluster group. Finally, the significance test for parental educational attainment approached significance [F(2, 331) = 2.56, p = .08]. Post-hoc analyses revealed that parents of adolescents in the Low Frequency cluster (M = 4.47) attained less formal education (p < .05) than parents of adolescents in the High Positive cluster (M = 5.00).

Relationship Between Racial Socialization and Subsequent Racial Identity

To investigate the role of racial socialization on subsequent racial identity, we estimated a series of General Linear Model (GLM) Analyses of Covariance (ANCOVA) with five of the seven racial identity subscales at Time 2 as dependent variables. As previously noted, oppressed minority and humanist ideology were not included in the analyses due to problems with the internal consistency of the scales. Developmental period, gender, and parent’s highest level of education completed were included as covariates. The main effect term for the cluster group membership variables represented our measure of the relationship between racial socialization at Time 1 and racial identity at Time 2. We interpreted a significant main effect for the cluster group membership variable as indicative of the fact that certain patterns of racial socialization may influence subsequent racial identity. We utilized findings from our post-hoc tests of these significant main effects analyses to identify what patterns of racial socialization were associated with racial identity at Time 2. Analyses were adjusted for multiple comparisons using a Bonferroni correction.

The GLM ANOVAs found significant cluster effects for three of the racial identity variables: centrality, and assimilationist and nationalist ideologies. With respect to racial centrality, the GLM model explained 4% of the variance. The main effect term for cluster group membership was significant [F(2, 340) = 4.35, p < .05: partial eta squared = .03] indicating that certain patterns of racial socialization are associated with racial centrality. The highest level of racial centrality at Time 2 was found in the High Positive cluster (M = 3.96). Individuals in the Low Frequency (M = 3.61) cluster reported significantly lower levels of centrality at Time 2 than the High Positive cluster. The group mean for the Moderate Positive group (M = 3.81) was not significantly different from the mean for the High Positive cluster group. In addition, parental educational attainment at Time 1 was significantly related to racial centrality at Time 2 such that adolescents whose parents had higher levels of educational attainment reported that race was less central to their identity at Time 2 (b = −.06; p < .05; partial eta squared = .02).

The overall GLM model explained 5% of the variance in assimilationist ideology. There was a significant main effect for cluster membership in assimilationist ideology [F(2, 340) = 3.98, p < .05; partial eta squared = .02]. Post-hoc analyses indicate that adolescents in the High Positive cluster group (M = 1.60) reported significantly lower levels of assimilationist ideology at Time 2 than those in the Moderate Positive cluster (M = 1.82) and Low Frequency clusters (M = 1.84). Gender was also related to assimilationist ideology (b = .21, p < .05; partial eta squared = .02) such that boys were more likely to feel that Blacks should emphasize a mainstream American identity over an African American one.

Finally, the GLM accounted for 7% of the variance in nationalist ideology. The analysis yielded a significant main effect for cluster group membership on nationalist ideology, [F(2, 340) = 4.41, p < .05: partial eta squared = .03]. Individuals in the High Positive cluster group reported significantly higher levels of nationalist ideology at Time 2 (M = 3.58) than the Low Frequency cluster groups (M = 3.22). Developmental period was associated with nationalist ideology (b = −.37, p < .01; partial eta squared = .04) such that high school students were more likely to endorse views emphasizing the uniqueness of being African American.

Discussion

Racial Socialization Profiles

Our first objective was to identify different patterns of adolescent-reported racial socialization across six different types of verbal and nonverbal messages. In general, African American adolescents reported receiving a wide range of messages from their parents regarding race, as well as engaging in various socialization activities with their parents. We identified three patterns of racial socialization experiences. Consistent with a prior study in which we investigated patterns of racial socialization experiences and psychological adjustment in an African American adolescent sample (Neblett et al. 2008), our two largest clusters (High and Moderate Positive) were comprised of adolescents who reported a parental emphasis on racial pride and self-worth messages. Racial barrier and egalitarian messages, as well as socialization behaviors were also significant components of racial socialization in these groups, particularly for the High Positive cluster group. These results corroborate preliminary evidence that African American parents are likely to combine different socialization messages and activities when socializing their children about race (Phinney and Chavira 1995). Furthermore, parents of African American adolescents provide socialization messages and engage in socialization behaviors with specific racial content or overtones (e.g., racial pride, racial barrier messages, and socialization behaviors) within a self-affirming context (e.g., self-worth messages).

Although the difference between the High and Moderate Positive cluster groups is largely a distinction with respect to the relative frequency of different messages, it is interesting to note the greatest differences in the relative emphasis of messages and activities between the two groups, particularly as indicated by the standardized means. Specifically, the High Positive Cluster group consists of a higher proportion of socialization behaviors (+.84 SD), racial barrier messages (+.83 SD), and racial pride messages (+.68 SD) relative to the corresponding standardized mean frequencies for socialization behaviors (−.30 SD), racial barrier (−.19 SD), and racial pride (−.04 SD) messages in the Moderate Positive group. Although the messages and activities operate in concert, these differences may provide some important clues as to which messages are particularly influential in shaping African American adolescent racial identity. We return to this idea when we further examine the relationship between racial socialization and racial identity below.

Consistent with prior studies that have found that a significant minority of African American parents engage in little to no explicit racial socialization (e.g., Hughes and DuMont 1993; Neblett et al. 2008; Parham and Williams 1993), the pattern of racial socialization observed in the Low Frequency cluster group provides additional empirical evidence that not all parents emphasize messages about race or engage in activities that convey messages about race. Whereas standardized scores for various kinds of racial socialization were approximately one standard deviation below the mean (with the exception of negative messages), and in some cases more than one standard deviation, the data reveal that adolescents in this cluster group still report receiving all socialization messages (with the exception of negative messages) and engaging in socialization activities on occasion (i.e., once or twice). As we have previously suggested, this pattern of findings may be indicative of an approach to racial socialization in which parents make a conscious decision not to focus on race with their children as an explicit racial socialization strategy (Hughes and Chen 1999).

Racial Socialization and Racial Identity

The second research objective was to investigate whether patterns of racial socialization are associated with adolescent-reported racial identity at a later time point. In the present study, bivariate and multivariate analyses supported the notion that racial socialization plays some role in shaping African American adolescent racial identity. At the multivariate level, High Positive and Low Frequency patterns of racial socialization were specifically associated with adolescent racial centrality, assimilationist ideology, and nationalist ideology. These results are consistent with prior studies that have reported associations between racial socialization and racial identity (e.g., Bennett 2006; Branch and Newcombe 1986; Wills et al. 2007), but which have failed to examine the relationship beyond a single time point.

Racial socialization scholars have suggested that parental messages concerning the meaning of being African American, as well as actions that convey meaning about race, provide important information that African American adolescents evaluate and incorporate into their racial identities (Demo and Hughes 1990; Thornton et al. 1990). In the present study, we found that adolescents in the High Positive cluster at Time 1 felt that race was more central to their self-concept at Time 2, were less likely to emphasize similarities between African Americans and all Americans, and more likely to emphasize attitudes highlighting the uniqueness of being African American. Adolescents in the Low Frequency cluster, on the other hand, tended to feel race was less a part of their core identity and interpreted being African American as similar, and not distinct from being American. This latter finding can be likened to an earlier finding by Spencer (1983) that the absence of parental engagement around race was associated with Eurocentric racial attitudes. In light of the patterns of racial socialization described above, the present results suggest that the differences in the relative emphasis on socialization behaviors, racial barrier and racial pride messages, may play a role in both the significance and meaning that African American youth attribute to being African American a year later.

African American racial socialization messages, activities, and experiences may provide an organizing construct or schema for how African American adolescents come to think about the personal significance and meaning of race. Socialization practices that highlight the importance of the African American experience (e.g., racial pride and racial barrier messages, socialization behaviors, etc.) may convey the message that race is significant and lead African American adolescents to integrate being African American as a core part of their identity (racial centrality). Furthermore, these messages convey the uniqueness and cultural distinctiveness of the African American experience (nationalist ideology) as opposed to the similarities between African Americans and all Americans (assimilationist ideology). Youth whose parents do not focus on race on the other hand, may not see race as important and come to understand what it means to be African American as something entirely different. It is plausible that for youth whose parents do not mention or engage race in any meaningful way, the message that race is significant is not conveyed or incorporated into the self concept, and being African American is no different than being American.

The relationship between racial socialization and racial identity in the present study is particularly exciting in light of several recent studies that suggest racial identity is an important resilience factor in how African American adolescents negotiate experiences with racial discrimination (Neblett et al. 2004; Sellers et al. 2006; Smalls et al. 2007; Wong et al. 2003). Neblett et al. (2004), for example, found that individuals with the highest level of centrality were immune from the damaging impact of racial hassles on subsequent levels of anxiety, stress, and depressive symptoms. In a series of studies examining the protective role of racial identity, Sellers and colleagues (Sellers and Shelton 2003; Sellers et al. 2003, 2006) found that racial centrality and nationalist ideology buffered the deleterious effects of racial discrimination on the psychological impact of racism as well as several indices of psychological health. These findings, taken in conjunction with the relationships between racial socialization and racial centrality and nationalist ideology found in the present study, suggest that family socialization with regard to race may play a formative role in shaping some of the dimensions of racial identity that are known to protect youth against the harmful influence of racism. If this is the case, these findings highlight the need for youth interventions and policy efforts that promote the African American family as a social institution that provides opportunities to celebrate being African American and learn about the reality of being African American in this society within the context of a supportive and self-affirming environment.

Although we were most interested in the relationship between patterns of racial socialization and subsequent racial identity, several interesting findings emerged with respect to sociodemographic group differences between the racial socialization cluster groups. These differences highlight the importance of considering the study findings within a developmental context. For example, middle school students appeared to be more likely to be in the Moderate Positive cluster group, while high school students were more likely to be in the High Positive group. This finding is consistent with age differences in various types of racial socialization messages that have been reported by Hughes and Chen (1997) and others (e.g., Demo and Hughes 1990) and suggests that parents may emphasize different racial socialization practices as a function of their child’s developmental period or stage. Consistent with the argument for increasing cognitive sophistication with respect to race with increasing age (e.g., Aboud 2008), differences in hypothetico-deductive thinking across adolescence might also lead older adolescents who can reason and think more abstractly to interpret messages they receive from their parents differently than younger adolescents.

Gender is a second contextual factor to consider in understanding the relationship between parental racial socialization and adolescent racial identity. In the current study, girls were more likely to be in the Moderate and High Positive groups, while boys were more likely to be in the Low Frequency cluster group, arguably lending support to the old adage that “mothers raise their daughters and love their sons.” At the very least, it seems plausible that mothers are providing messages to their daughters in an attempt to prepare them for the world in ways that they are not doing for their sons. Sanders Thompson (1994) reported a similar finding whereby females reported receiving more extensive messages concerning race and race-related matters from the immediate family than male children. She suggested that the frequency and intensity of interactions and contacts around issues of race might vary as a function of gender roles in the African American community and society.

Finally, socioeconomic status must be considered as a critical contextual factor in assessing the relationship between racial socialization and racial identity. We found evidence of a statistical trend such that parents of adolescents in the Low Frequency cluster attained less formal education than parents of adolescents in the High Positive cluster. This pattern of results suggests that parents with different socioeconomic backgrounds may have different ideas about race and ethnicity (Hughes et al. 2006). Perhaps parents with higher levels of education convey a wider range of messages about race as a function of their own experiences with racial discrimination (Williams 1999). Gecas (1979) suggests that the social class of the family of origin plays an important role because it structures opportunities and resources for children, the types of schools that they attend, friends, and values and attitudes to which they are exposed. Adolescents who come from more affluent families may have more frequent interracial interactions that prompt their parents to discuss issues of race more readily (Neblett et al. 2008). In light of the relationships between racial socialization and racial identity in the present study, differential effects of these and other sociodemographic contextual factors (e.g., racial milieu of school and neighborhood contexts) may play a significant role in directing the focus of parental racial socialization efforts, and indirectly influence African American adolescent racial identity.

Study Limitations

The present study addresses a number of previously identified shortcomings in the extant literature; however several limitations will need to be addressed in future research. First, the internal reliability estimate of the racial centrality scale in the present study was just below the typical standard (i.e., .60) used in research. The low reliability estimate may, in part, be a function of the small number of scale items; however, future work will need to re-evaluate the internal consistency of this and other problematic subscales in the target population. Second, future studies will need to adopt multiple-informant designs in the study of adolescent racial socialization processes. While consideration of the adolescents’ perspective is certainly a strength of the current design, the use of parent- and child- accounts of racial socialization together might provide important clues about the extent to which adolescents and their parents experience racial socialization in similar (or dissimilar) ways, and add further depth and complexity to our understanding of these processes in shaping youth racial identity. For example, multiple-informant studies may shed light not only on how parental messages shape adolescent racial identity but also how adolescent experiences shape parental racial socialization and how this interplay between parent and child influences identity over time. A third consideration is the generalizability of the findings beyond the current sample. The differences in cluster group membership across several sociodemographic characteristics suggests that the patterns of socialization observed in this study may not generalize to other samples of African Americans. As noted above, differences in sociodemographic contexts might influence the types of socialization practices in which parents engage and so, future studies will need to examine the role of racial socialization in adolescent racial identity across youth who inhabit diverse developmental contexts.

Future Research

In addition to future work addressing methodological limitations of the current study, we propose that future research examine racial identity as a possible mediator of the relationship between racial socialization and psychological adjustment. Given that racial socialization appears to have some role in subsequent racial identity, it would be worthwhile to conduct prospective studies that examine racial socialization, racial identity, and developmental outcomes over time. As such, we reiterate the need for studies to formally test possible mediating relationships among racial socialization, racial identity, and developmental outcomes (Neblett et al. 2008). Also, we note that the present study has focused exclusively on parental racial socialization, but further research is needed that focuses on other sources of racial socialization. Sanders Thompson (1994) found that racial socialization by “other adult family members” was more strongly related to racial identity than parental racial socialization and future studies may be wise to consider not only other adult family members as socialization agents, but also adolescent peers. As youth become more autonomous, parents’ messages may play less of a role for older adolescents than younger adolescents. Given the developmental salience of peers during adolescence, adolescent peer relations may play a more significant role than has been previously recognized in how youth come to ascribe significance and meaning to being African American. Such investigations may prove useful in explaining a greater proportion of the variance in racial identity. Finally, we note the need for future work to consider multiple ecological contexts in understanding the relationship between racial socialization and racial identity over time. Future studies will need to move beyond statistical controls to truly begin to appreciate the interplay between individual factors such as age and gender and larger systemic influences such as socioeconomic, neighborhood, and other contexts in moderating the influence of the family context on the development of African American adolescent racial identity over time.

Conclusion

In the present study, we adopted a person-centered approach to the study of racial socialization to offer sorely needed empirical data regarding the content of parental racial socialization and its role in the development of African American racial identity. Our results indicate that the messages that build self-respect and pride about being African American, as well as the activities in which African American adolescents engage around African American culture, are both important factors in the significance and meaning that African American adolescents attribute to being African American. In light of recent evidence that racial identity attitudes are important factors in understanding the resilience of African American adolescents to experiences with racial discrimination, future work should continue to illuminate the ways in which African American parents and other socialization agents transmit messages to their children about the meaning of race and effectively prepare their children to successfully negotiate the range of experiences that African American adolescents encounter in the broader American society.