Racial Discrimination Experiences and African American Youth Adjustment: The Role of Parenting Profiles Based on Racial Socialization and Involved-Vigilant Parenting

Abstract

Objectives

The purpose of this study was to test whether parenting profiles based on racial socialization and involved-vigilant parenting would compensate for or moderate associations between racial discrimination experiences and academic outcomes and psychological well-being among African American adolescents.

Methods

Participants were 1363 African American adolescents (mean age = 14.19; 52.3% female) from three Midwestern suburban school districts. Latent profile analysis was used to examine whether there were distinct combinations of parenting. The relationships among racial discrimination experiences, parenting profiles, and adjustment were examined using structural equation modeling.

Results

Three distinct parenting profiles were found: Moderate Positive (n=767; moderately high involved-vigilant parenting and racial barrier, racial pride, behavioral, and egalitarian messages, and low negative messages), Unengaged (n=351; low racial socialization messages and moderately low involved-vigilant parenting), and High Negative parenting (n=242; high negative messages, moderate other racial socialization messages, and moderately low involved-vigilant parenting). Racial discrimination experiences were negatively associated with youth adjustment. Moderate Positive parenting was related to the best academic outcomes and Unengaged parenting was associated with more positive academic outcomes than High Negative parenting. Moderate Positive parenting was associated with better psychological well-being than Unengaged or High Negative parenting although the benefits were greater for adolescents with fewer racial discrimination experiences.

Conclusions

Distinct patterns of racial socialization messages and involved-vigilant parenting contribute to differences in African American youth adjustment.

Many adolescents experience declines in academic engagement and psychological well-being (Anderman, Maehr, & Midgley, 1999). One area of potential vulnerability for African American adolescents’ development of competencies is their exposure to racial discrimination. Racial discrimination experiences have been linked to lower academic expectations, achievement motivation, and performance among African American adolescents (Burchinal, Roberts, Zeisel, & Rowley, 2008; Neblett, Philip, Cogburn, & Sellers, 2006; Wong, Eccles, & Sameroff, 2003). These experiences have also been linked to increases in depressive symptoms and problem behaviors (Clark, Coleman, & Novak, 2004; Simons, Murry, McLoyd, Lin, Cutrona, & Conger, 2002), as well as decreases in self-esteem and psychological well-being (Fisher, Wallace, & Fenton, 2000; Sellers, Copeland-Linder, Martin, & Lewis, 2006; Wong et al., 2003). Yet, not all African American adolescents who experience racial discrimination display maladjustment.

This study is informed by the Integrative Model of Minority Children's Competencies (Garcia Coll et al., 1996). The Integrative Model suggests that ethnic minority children develop in unique contexts due to interpersonal and institutional racism and discrimination. While discrimination can have negative effects on the development of minority children, there are adaptive qualities that can protect ethnic minority youth and contribute to the positive development of competencies such as psychological well-being, academic motivation, and prosocial behavior. At the family level, both culturally-specific parenting, including racial socialization, and general parenting practices are related to minority child development (Garcia Coll et al., 1996; Elmore & Gaylord Harden, 2013).

Children’s receptiveness to parenting practices with specific goals, such as racial socialization, may be influenced by the general parenting environment (Darling and Steinberg, 1993). Adolescents may respond more positively to racial socialization messages if they are in a responsive parenting context (Smalls, 2009). Very few studies have examined how different combinations of culturally-specific and general parenting explain variation in outcomes among African American adolescents who experience racial discrimination despite the reality that adolescents do not receive each parenting practice in isolation. The goals of this study were to 1) use a person-centered approach to explore profiles of parenting that include general and culturally-specific parenting practices; 2) examine the relationship between racial discrimination experiences and African American adolescents’ developmental competencies; and 3) to test whether profiles of parenting were directly associated with African American adolescents’ academic outcomes and psychological well-being or moderated the relationship between racial discrimination experiences and these developmental competencies.

Racial Socialization and Adolescent Outcomes

One factor that has been used to help explain variation in the effects of racial discrimination among African American adolescents is racial socialization. Racial socialization includes implicit and explicit messages about “the meaning and significance of race and racial stratification, intergroup and intragroup relations, and personal and group identity” (Lesane-Brown, 2006, p. 403). Boykin and Toms’ (1985) conceptual framework of socialization, along with more recent studies building off their framework (Bentley-Edwards & Stevenson, 2016; Coard & Sellers, 2005; Hughes et al., 2006; Lesane-Brown, 2006), provides a rationale for the content of racial socialization messages examined in this study. Boykin and Toms (1985) suggest that Africans Americans experience a triple quandary in that they must navigate mainstream, cultural, and minority experiences. These three areas of experience influence the racial socialization messages present in families. Mainstream experiences are oriented toward the dominant culture and may contribute to either silence about race or egalitarian messages that emphasize equality and harmony across race (Bowman & Howard, 1985; Boykin & Toms, 1985; Coard, Wallace, Stevenson, & Brotman, 2004; Hughes et al., 2006; Lesane-Brown, 2006). Egalitarian messages have been found to be related to fewer problem behaviors and higher psychological well-being among African American adolescents (Neblett, White, Ford, Philip, Nguyên, & Sellers, 2008).

Cultural experiences are related to group traditions, values, and practices (Boykin & Toms, 1985). Racial pride messages address these experiences by emphasizing the positive aspects of one's racial/ethnic group and are the most common type of racial socialization message (Coard et al., 2004; Lesane-Brown, 2006). Higher frequencies of racial pride messages have been positively related to academic motivation, academic achievement, and self-concept (Bowman & Howard, 1985; Hughes et al., 2006) and found to weaken negative associations between racial discrimination and grades (Wang & Huguley, 2012). Scholars have also indicated that negative messages that reinforce stereotypes about one’s racial group are understudied in the racial socialization literature (Bentley-Edwards & Stevenson, 2016; Coard & Sellers, 2005). Negative messages have been associated with depressive symptoms, problem behaviors, and lower psychological well-being among African American adolescents (Neblett et al., 2008). In addition, cultural messages include implicit behavioral messages transmitted through engaging in activities or having culturally relevant items related to one's ethnicity or racial group in the household (Caughy, O’Campo, Randolph, & Nickerson 2002; Coard et al., 2004; Lesane-Brown, 2006). Behavioral messages have been associated with fewer depressive symptoms and higher psychological well-being among African American adolescents (Neblett et al., 2008).

Minority experiences encompass awareness and development of coping skills to deal with discrimination. Racial barrier messages address this realm of experience and are messages about potential challenges due to racial or ethnic group membership and how to cope with those obstacles (Bowman & Howard; Hughes et al., 2006; Lesane-Brown, 2006). While some studies have found racial barrier messages to be positively related to academic motivation (Neblett et al., 2006), other studies have found a curvilinear relationship where moderate racial barrier messages buffer the effects of racial discrimination but higher and lower frequencies of racial barrier messages were related to negative outcomes (Harris-Britt, Valrie, Kurtz-Costes, & Rowley, 2007). Still others have found no significant relationship between racial barrier messages and adolescent outcomes (Neblett et al., 2009).

These different racial socialization messages can have competing implications and the balance of the various types of racial socialization messages can vary in families (Boykin & Toms, 1985; Coard & Sellers, 2005). Parents give different types of messages based on their goals and perceptions of the world and their children’s characteristics and needs. The balance of messages used can influence adolescents’ understanding of race and thus, their responses to racial discrimination.

Several studies have examined racial socialization profiles. White-Johnson, Ford, and Sellers (2010) for example found three profiles: unengaged (below average messages except for negative messages), multifaceted (above average messages except for negative messages), and low race salience (emphasis on mainstream messages). They did not examine the relationship between profiles and adolescent outcomes. In one study, Neblett and colleagues (2008) found four racial socialization profiles: High Positive, Moderate Positive, Moderate Negative, and Low Frequency (Neblett et al., 2008). Moderate Negative and Low Frequency were associated with lower psychological adjustment and High Positive parenting buffered the negative effects of racial discrimination on stress. In another study examining boys’ academic outcomes, they also found 4 profiles except instead of Moderate Positive, they found a Self-Worth profile which emphasized mainstream messages over race-related messages (Neblett, Chavous, Nguyen, & Sellers, 2009). Adolescents in the Self-Worth and Positive Socialization profiles had more academic persistence and higher grades but the profiles did not buffer the negative effects of discrimination. These studies suggest that families vary in their composition of racial socialization messages but do not explore general parenting which could provide a more holistic view of adolescents’ family context.

Involved-Vigilant Parenting

Involved-vigilant parenting is a parenting process that has been identified by researchers from the Family and Community Health Study (FACHS) as an important predictor of African American youth competence and resilience (Brody et al., 2001; Brody & Ge, 2001; Brody, Murry, Kim, & Brown, 2002; Murry, Berkel, Simons, Simons, & Gibbons, 2014; Simons et al., 2002). It includes inductive reasoning, mutual problem-solving, and monitoring in order to assess aspects of both parents’ responsiveness and behavioral control (Murry et al., 2014). Inductive reasoning occurs when parents provide explanations for the rules and encourage children to think about the consequences of their actions (Le, Ceballo, Chao, Hill, Murry, & Pinderhughes, 2008). Mutual problem-solving involves parents and adolescents discussing problems and developing solutions together (Miller et al., 2010). Both inductive reasoning and mutual problem solving may increase adolescents' understanding and internalization of parents' values and rules, as well as increase their ability to apply reasoning and make healthy decisions in new settings and circumstances (Brody, Murry, Kim, & Brown, 2002; Le et al., 2008; Miller et al. 2005, 2010). Monitoring consists of parents' knowledge of their children's actions, whereabouts, and peer relationships (Brody et al. 2002; Berkel et al. 2009) and can protect adolescents from harm by reducing opportunities for adolescents to get into trouble or inappropriate situations (Ceballo, Ramirez, Hearn, & Maltese, 2003).

Among African American youth involved-vigilant parenting has been found to be related to higher social skills (Hurd et al., 2013), psychological well-being (Hurd et al., 2013), and self-regulation skills (Brody & Ge, 2001; Murry et al., 2014), as well as fewer depressive symptoms (Simons et al., 2002). Involved-vigilant parenting has also been found to weaken the positive associations between racial discrimination and conduct problems and depressive symptoms among African American adolescents (Brody, Chen et al., 2006). Parents who engage in more involved-vigilant parenting have also been found to give more cultural and preparation for bias racial socialization messages (Murry et al., 2014). Additionally, in a study that included involved-vigilant parenting and racial socialization as indicators of a latent factor of parenting (Brody, Murry et al., 2006), parenting was related to fewer high risk behaviors among youth. This study used racial socialization as a unidimensional indicator and did not account for families varying in their proportions of different types of racial socialization practices. Overall, parenting characterized by responsiveness and appropriate vigilance may work in tandem with racial socialization to promote developmental competencies and offset the noxious effects of experiences with racial discrimination on youth outcomes.

Influence of Racial Socialization Differs by Parenting Context

Recent studies have suggested that different combinations of general and culturally-specific parenting are associated with variance in adolescent outcomes. In one study of African American adolescents, racial barrier messages were positively associated with academic persistence and emotional school engagement when paired with highly democratically-involved parenting but negatively associated with academic engagement when paired with lower levels of democratically-involved parenting (Smalls, 2009). Elmore and Gaylord-Harden (2013) found mixed results. High racial pride and supportive parenting predicted low internalizing problems and high externalizing problems while moderate racial pride and high supportive parenting were related to fewer externalizing problems. Cooper and McLoyd (2011) unexpectedly found that racial barrier messages were related to higher self-esteem and fewer depressive symptoms for girls who perceived less positive relations with their mothers. These studies provide evidence that the relationship between certain types of racial socialization and youth outcomes is influenced by general parenting but in seemingly different ways based on the message and outcome examined. However, these studies only examined racial pride or racial barrier messages and did not reflect the multifaceted racial socialization messages parents provide their youth.

The Potential Moderating Role of Adolescent Gender

Gender was examined as a moderator of parenting profiles and racial discrimination in this study due to potential gender differences in African American adolescents’ racial discrimination, socialization, and academic experiences. African American boys reported more racial discrimination experiences than girls in a nationally representative sample (Seaton et al., 2008). African American parents also have higher academic expectations of girls and engage in more monitoring, rule enforcement, and communication with girls than boys (Mandara et al., 2010; Varner & Mandara, 2013; Wood, Kaplan, & McLoyd, 2008). In addition, Cooper and McLoyd (2011) found that the interaction between mother-adolescent-relationship and racial barrier socialization on African American adolescents’ self-esteem and depressive symptoms differed by gender. Academically, African American girls outperform African American boys in grades, test scores, and graduation rates (Saunders, Davis, Williams, & Williams, 2004). Due to these gender differences across primary study variables, we examined if the relationships among parenting profiles, racial discrimination, and outcomes differed by gender.

Current Study

This study builds upon previous work exploring the influence of parenting on African American youth outcomes in its use of latent profile analysis to holistically consider both general and culturally-specific parenting, examination of developmental competencies, and inclusion of a sample of African American youth from diverse racial and socioeconomic school contexts. The conceptual model is shown in Figure 1. We aimed to test the following hypotheses: 1) distinct parenting profiles based on involved-vigilant parenting and racial socialization will emerge, 2) racial discrimination experiences will be negatively associated with academic behaviors and psychological well-being, and 3) there will be significant moderating effects of the parenting profiles on the relationships between discrimination and youth outcomes. In addition, gender differences in these associations were explored due to past results indicating gender differences in racial discrimination experiences, socialization, and outcomes among African American youth (Stevenson, McNeil, Herrero-Taylor, & Davis, 2005; Thomas & Stevenson, 2009).

Figure 1.

Conceptual model.

Method

Participants

Data were from a multi-wave study of 1590 adolescents recruited from three school districts in a Midwestern metropolitan area. The current study used Wave 1 survey data of a sample of 1363 adolescents (6^th to 12^th graders) who self-identified as Black or African-American. Adolescents ranged in age from 11 to 19 years old (M = 14.19, SD = 1.68) and were 52.3% female. The percentage of adolescents in middle school (6^th to 8^th grade) was 46.1% and in high school (9^th to 12^th grade) was 53.9%. In the sample, 40.6% of the participants' parents were married, 35.3% were single, 18.4% were divorced, 4.3% were separated, and 1.4% were widowed. For parental educational level, most of the participants' parents were high school graduates (22.6%) or had some postsecondary education (22.0%), 19.5% were college graduates, 19.3% earned a Masters' degree, 9.3% did not graduate high school, 4.7% had some graduate school experience, and 2.6% earned a doctorate or professional degree.

Procedure

Adolescents who identified as Black, African American, or multiracial with an African American parent were recruited from three school districts in a Midwestern metropolitan area from the 2010–2011 to 2013–2014 school years. Wave 1 data consists of each participants’ first survey completion during that time period. In each school district, we included one middle school and one high school. The school districts were selected to include districts that varied in their socioeconomic and racial/ethnic composition to survey the experiences of adolescents in diverse school contexts. Participants in the first school district (63% of sample) attended schools that over the time of data collection were predominately African American (95%) and where most students (74%) were eligible for free or reduced-price lunch. Participants in the second school district (24.9% of sample) attended a racially diverse middle school (44% European American, 38% African American, 18% in other racial/ethnic categories). The high school in the second district was predominately European American (60% European American, 26% African American, 14% Other) and 22% of the student body was economically disadvantaged. The schools in the third school district (12.3% of sample) were both predominately European American (Middle School: 63% European American, 20% African American, 17% Other; High School: 64% European American, 21% African American, 15% Other). Student eligibility for free or reduced-price lunch was 24% of the high school student body and 30% of the middle school student body.

Students were recruited through informational packets that were mailed to families and fliers distributed during parent meetings and lunch periods. Only students whose parents completed informed consent forms were eligible to participate. Each adolescent provided informed assent before participating. The survey included questions about family life, discrimination experiences, academic motivation, and social and emotional functioning. Each adolescent who completed the approximately 45 minute survey was given a $20 Visa gift card. The data were collected in compliance with [masked for review] Institutional Review Board.

Measures

Racial discrimination experiences

Adolescents’ racial discrimination experiences were measured by 18 items from the Daily Life Experience subscale of the Racism and Life Experience scales, which has shown good validity and reliability with African American families (Harrell, 1994). These items assessed the frequency of microaggressions related to race (e.g., “Being treated rudely or disrespectfully”). Adolescents rated how often each experience occurred in the past year on a scale from 0 (never) to 5 (once a week or more). Higher mean scores indicated more experiences of being the target of discrimination (α = .97).

Racial socialization messages

Adolescents answered 26 items related to the messages their caregivers gave them about race from the Racial Socialization Questionnaire-teen (Lesane-Brown, Scottham, Nguyen, & Sellers, 2006). This measure was theoretically derived based on racial socialization messages previously identified in the literature and has shown good reliability and validity with African American families, for example, it has been used to predict African American adolescents’ academic and psychological functioning (Neblett et al., 2008; 2009; Smalls, 2009). Five types of racial socialization messages were assessed. Negative messages reinforced harmful stereotypes about African Americans (5 items; α = .80; e.g., "Told you Blacks are not as smart as people of other races."). Racial barrier messages discussed potential race-related challenges adolescents may face and how to cope with them (4 items; α = .73; e.g., "Told you that some people may dislike you because of the color of your skin."). Behavioral messages were implicit and included cultural items in the home and attending cultural events (5 items; α = .72; e.g., "Gone with you to Black cultural events (i.e. plays, movies, concerts, museums)"). Egalitarian messages emphasized equality and harmony among races (4 items; α = .69; e.g., "Told you that Blacks and Whites should try to understand each other so they can get along"). Racial pride messages emphasized positive aspects of one’s group membership (4 items; α = .66; e.g., "Told you never to be ashamed of your Black features"). For all items, adolescents reported how often they received these messages from parents in the past year, on a scale from 0 (never) to 2 (more than twice). Mean scores for each of the five subscales were used as five indicators in the latent profile analyses.

Involved-vigilant parenting

Involved-vigilant parenting is characterized by a combination of monitoring, inductive reasoning, and mutual problem-solving (Murry et al., 2014). The current study used a 13 item version of the involved-vigilant parenting scale used in the Family and Community Health Study (FACHS) that has shown good validity and reliability with African American families (Brody et al., 2001; Kim, Brody, & Murry, 2003). Items were adapted from the Iowa Youth and Families Project (Conger & Elder, 1994) and focus groups were completed to assess the items’ cultural relevance with African American families when they were first used in the FACHS (Murry et al., 2014). Subscales included monitoring (5 items; α = .73; e.g., "In the course of a day, how often does your caregiver know where you are?"), inductive reasoning (4 items; α = .59; e.g., "How often does your caregiver give reasons to you for his/her decisions?"), and problem solving (4 items; α = .75; e.g., "When you and your caregiver have a problem, how often can the two of you figure out how to deal with it?"). Participants answered these items using a 4-point scale from 1 (never) to 4 (always). The mean score for all items was used in all analyses (α = .83 for the whole scale).

Adolescent psychological well-being

Adolescents’ psychological well-being was assessed using a 12-item modified version of the widely used Ryff Scales of Psychological Well-being, which has shown good validity and modest reliability across various populations (Ryff, 1989; Ryff & Keyes, 1995; Springer & Hauser, 2006). For the current study, three subscales were used: personal growth (4 items; α = .66; e.g., "I am the kind of person who likes to give new things a try."), purpose in life (4 items; α = .64; e.g., "I have a sense of direction and purpose in life."), and self-acceptance (4 items; α = .66; e.g., "I like most aspects of my personality."). Adolescents indicated the extent to which they agreed or disagreed with each statement on a scale from 1 (strongly disagree) to 5 (strongly agree). Some items were reverse coded so that higher scores on all items indicated higher psychological well-being. Mean scores of each dimension were used as three indicators of a latent variable for psychological well-being. There were significant correlations between the three indicators, rs = .46 to .59, p < .001.

Adolescent academic outcomes

Adolescents completed 3 sets of items designed to assess their academic persistence, education aspiration, and perceived likelihood of achieving education aspiration. Academic persistence was assessed through four items adapted from the Scale for Academic Engagement (Wellborn, 1991) that has been used in a number of studies with African American youth (Butler-Barnes, Chavous, Hurd, & Varner, 2013; Neblett et al., 2006; Smalls, White, Chavous, & Sellers, 2007). A sample item is, "If I can't get a problem right the first time, I just keep trying." Adolescents reported on a scale from 1 (not true at all) to 4 (very true). High mean scores represent higher levels of academic persistence (α = .68). For educational aspiration, adolescents answered one item modified from the Maryland Adolescent Development in Context Study (MADICS; Eccles, 1997; i.e., “How far do you want to go in school?”) on a scale from 1 (some high school) to 6 (graduate degree). For perceived likelihood of achieving educational aspiration, adolescents answered one item modified from the MADICS (Eccles, 1997; i.e., “Sometimes people do not get what they want. How likely is it that you will go as far as you want in school?”) on a scale from 1 (not at all likely) to 5 (almost certain). Adolescents’ scores on academic persistence, education aspiration, and perceived likelihood of achieving education aspiration were used as three indicators of a latent variable, academic outcomes. There were significant correlations between the three indicators, rs = .23 to .30, p < .001.

Adolescent gender

A dummy variable was created for adolescent gender (0=male, 1=female).

Demographic variables

We measured a set of demographic variables including age, parental education, and school district. For parental education, adolescents reported the highest educational level of their primary caregivers on an 8-point scale ranging from 1 (Junior high school or less) to 8 (Ph.D./M.D./J.D.). For school district, we created two dummy variables racially-diverse school (0 = African American-dominant school, 1 = racially-diverse school) and European American-dominant school (0 = African American-dominant school, 1 = European American-dominant school) with African American-dominant school as the reference group.

Results

Descriptive Statistics

We first conducted descriptive and correlational analyses for all study variables (see Table 1). Among the independent variables (racial socialization domains, involved-vigilant parenting, discrimination, and adolescent gender), racial socialization domains were positively associated with each other (rs = .20 to .61, p < .001) except that negative messages were not significantly associated with egalitarian messages and racial pride. Involved-vigilant parenting was negatively associated with negative messages (r = −.19, p < .001) but positively correlated with other domains of racial socialization (rs = .10 to .31, p < .001). Racial discrimination experiences were positively associated with all domains of racial socialization (rs = .11 to .38, p < .001) but negatively correlated with involved-vigilant parenting (r = −.16, p < .001). Being female was associated with fewer racial discrimination experiences (r = −.13, p < .001), parental negative messages (r = −.17, p < .001), and racial barrier messages (r = −.07, p < .01), but more racial pride messages (r = .06, p < .05) and involved-vigilant parenting (r =.11, p < .001).

Table 1.

Descriptive statistics and zero-order correlations among all study variables

Variable	1	2	3	4	5	6	7	8	9	M	SD	N
1. Psychological well-being1	--									0.00	0.44	1362
2. Academic outcomes1	.86***	--								0.00	0.36	1362
3. Racial discrimination	−.21***	−.29***	--							1.20	1.18	1333
4. Negative messages	−.38***	−.42***	.38***	--						0.30	0.44	1358
5. Racial barriers	.06*	.02	.31***	.22***	--					1.00	0.59	1357
6. Behavioral messages	.12***	.09***	.23***	.20***	.45***	--				0.98	0.53	1357
7. Egalitarian messages	.27***	.25***	.11***	.03	.43***	.48***	--			1.18	0.54	1358
8. Racial Pride	.31***	.28***	.14***	−.02	.49***	.56***	.61***	--		1.33	0.51	1358
9. Involved parenting	.41***	.44***	−.16***	−.19***	.10***	.20***	.31***	.30***	--	2.81	0.55	1358
10. Female	.14***	.16***	−.13***	−.17***	−.07**	−.01	.03	.06*	.11***	--	--	1358
11. Age	−.20***	−.25***	.08**	.24***	.10***	.00	−.08**	−.10***	−.10***	14.19	1.68	1356
12. Parental education	.05	.08**	.08**	−.08*	.08*	.11***	.11***	.14***	.05	4.54	1.71	1095
13. European-American-dominant school2	−.02	−.02	.12***	−.08**	.05	−.05	.00	.02	−.06*	--	--	1361
14. Racially-diverse school2	.13***	.15***	.04	−.15***	.13***	.13***	.14***	.16***	.08**	--	--	1361

Note.

¹latent factor scores,

²dummy variables of school district, with the African-American-dominant school as reference group.

^*p < .05,

^**p < .01,

^***p < .001.

For the associations between independent variables and adolescent outcomes, adolescent psychological well-being and academic outcomes were correlated negatively with racial discrimination experiences and negative messages (rs = −.42 to −.21, p < .001), but correlated positively with behavioral, egalitarian, and racial pride messages, involved-vigilant parenting, and being female (rs = .09 to .44, p < .001). In addition, psychological well-being was positively associated with racial barrier messages (r = .06, p < .05).

Parenting Profiles

Latent profile analyses

The first goal of the current study was to examine whether there were groups of families with different combinations of parenting (i.e., parental racial socialization domains and involved-vigilant parenting). To do so, we used latent profile analysis with Mplus 7.31 (Muthén & Muthén, 1998 – 2015). Mplus uses the full information maximum likelihood (FIML) estimation method to handle missing data, which enables full usage of all available data in the model. A series of models were specified (i.e., 1-, 2-, 3-, 4-, 5-, 6-class). Several model fit indices were used to compare models with varying numbers of classes in addition to examining whether the classes appeared substantively and conceptually meaningful and qualitatively unique from other classes in the model (Nylund, Asparouhov, & Muthén, 2007). Specifically, Bayesian information criteria (BIC), sample-size adjusted Bayesian information criteria (ABIC), entropy, and the Lo–Mendell–Rubin likelihood ratio test (LMR) were used. Smaller values on the BIC and ABIC are indicative of a better fitting model and a significant p value on the LMR indicates that a model with k classes had better fit to the data than a model with k-1 classes.

Results of latent profile analyses

Model fit indices of latent profile analyses are presented in Table 2. Based on the model fit indices and the identification of conceptually meaningful and interpretable classes, the 3-class solution was the optimal solution. Specifically, BIC and ABIC tended to level off after the 3-class solution which had a p-value of LMR smaller than .001. Although the 4-class solution had smaller BIC and ABIC compared to the 3-class solution, one group of the 4-class solution had a small sample (n = 53, 4%) and two profiles had similar combinations of indicators.

Table 2.

Model Fit Indices Examining Latent Profiles of Parenting and Class Proportions and Counts

	1 class	2 classes	3 classes	4 classes	5 classes	6 classes
Log likelihood	−11558.673	−10840.442	−10446.727	−10222.273	−10039.480	−9971.772
Number of parameters	12	19	26	33	40	47
BIC	23203.938	21817.988	21081.068	20682.673	20367.599	20282.694
ABIC	23165.819	21757.633	20998.477	20577.845	20240.536	20133.395
Entropy	n/a	0.779	0.880	0.907	0.839	0.844
p-value of LMR	n/a	0.000	0.000	0.000	0.002	0.135
Proportions	100	34–66	26-56-18	24-17-56-04	16-34-16-30-04	14-7-33-30-13-3
Counts	1361	462–899	346-768-247	322-229-757-53	221-458-223-406-53	193-97-446-409-178-38

Note. BIC=Bayesian information criteria, ABIC=sample-size adjusted Bayesian information criteria, LMR=Lo–Mendell–Rubin likelihood ratio test. N=1361.

Figure 2 presents a graphic summary of the three parenting profiles using standardized means of each indicators. The largest class was labeled Moderate Positive (n = 767, 56%). This class was characterized by scores near the sample mean on all indicators (within the range of −0.5 SD to 0.5 SD) and higher scores on positive indicators (egalitarian messages, racial pride messages, and involved-vigilant parenting) relative to the sample mean. The second largest class was labeled Unengaged (n = 351, 26%). This class was characterized by low scores on all six indicators relative to the sample mean. The third class, High Negative (n = 242, 18%), had high scores on negative messages (1.8 SD above the sample mean) and scores near the sample mean on the other five indicators.

Figure 2.

Summary of Parenting Profiles (Standardized Means)

Parenting Group Differences among Study Variables

We used analysis of variance (for continuous variables) and chi-square tests (for categorical variables) in SPSS 22 to assess whether there were significant differences on study variables across the three parenting groups. Results are presented in Table 3. For the six parenting indicators, the Unengaged group had significantly lower scores on all parenting indicators except negative messages and involved-vigilant parenting, compared to the other two groups. The Unengaged group were not significantly different from the Moderate Positive group on negative messages and were not significantly different from the High Negative group on involved-vigilant parenting. Compared to the Moderate Positive group, the High Negative group had significantly higher scores on negative messages, and significantly lower scores on egalitarian messages, racial pride, and vigilant-involved parenting. For racial discrimination and demographic variables, the High Negative group had the highest frequency of racial discrimination experiences, followed by the Moderate Positive group. The High Negative group consisted of older adolescents and a lower percentage of female adolescents than the other groups. The Unengaged group had older adolescents than the Moderate Positive group. The Moderate Positive group had higher parental education levels and a higher percentage of students in the European American-dominant school district. For psychological and academic outcomes, the Moderate Positive group had the highest scores, followed by the Unengaged group, and the High Negative group reported the lowest scores.

Table 3.

Descriptive Statistics of Study Variables across Parenting Profiles

Variable	Unengaged	Moderate Positive	High Negative	F or Chi-square test
Negative messages	0.13 (0.21) a	0.11 (0.18) a	1.09 (0.33) b	F(2, 1355) = 1909.88***
Racial barriers	0.49 (0.44) a	1.17 (0.57) b	1.18 (0.38) b	F(2, 1354) = 227.27***
Behavioral messages	0.47 (0.41) a	1.15 (0.46) b	1.14 (0.41) b	F(2, 1354) = 311.06***
Egalitarian messages	0.60 (0.40) a	1.44 (0.42) b	1.18 (0.43) c	F(2, 1355) = 492.52***
Racial Pride	0.72 (0.40) a	1.62 (0.29) b	1.26 (0.41) c	F(2, 1355) = 821.67***
Involved parenting	2.58 (0.55) a	2.98 (0.48) b	2.59 (0.57) a	F(2, 1355) = 99.47***
Psychological well-being 1	−0.10 (0.44) a	0.16 (0.37) b	−0.34 (0.42) c	F(2, 1358) = 161.64***
Academic outcomes1	−0.07(0.35) a	0.13 (0.32) b	−0.31 (0.28) c	F(2, 1358) = 180.39***
Racial discrimination	0.81 (1.00) a	1.11 (1.08) b	2.05 (1.28) c	F(2, 1330) = 95.45***
Age	14.18 (1.72) a	13.97 (1.54) b	14.85 (1.87) c	F(2, 1351) = 26.33***
Parental education	4.24 (1.65) a	4.70 (1.70) b	4.40 (1.74) a	F(2, 1091) = 7.42***
Female (%)2	51%	58%	36%	χ2(2) =35.60***
School district:				χ2(4) =77.66***
Racial-diverse school (%)2	14%	13%	9%
European-American-dominant school (%)2	16%	33%	12%
African-American-dominant school (%)2	70%	54%	79%

Note.

¹latent factor scores.

²Percentage within each parenting group. For continuous variables, means are presented outside the parentheses and standardized deviations are presented in the parentheses. In a row, means with different superscripts were significantly different from each other.

^***p < .001. Degree of freedom for F test vary across variables due to missing data.

Parenting Profiles Predicting Adolescent Outcomes

Structural equation modeling

Our second goal was to examine whether parenting profiles, racial discrimination, and adolescent gender independently or interactively predicted adolescent outcomes. For this goal, we conducted structural equation modeling (SEM) in Mplus 7.31 (Muthén & Muthén, 1998 – 2015). We first tested a measurement model with latent variables. Specifically, in a measurement model, we tested: a) whether personal growth, purpose in life, and self-acceptance indicated a latent variable psychological well-being, and b) whether academic persistence, educational aspiration, and perceived likelihood of achieving educational aspiration indicated a latent variable academic outcomes. The model fit for the measurement model was excellent, χ²(7) = 9.52, p = .22, CFI = .996, RMSEA = .03 [.000, .07], SRMR = .02. Factor loadings were all significant (p<.001), ranging from .67 to .80 for psychological well-being, and from .35 to .70 for academic outcomes.

Then, we tested structural models using parenting profiles, racial discrimination, and adolescent gender, as well as their interactions, to predict adolescent psychological well-being and academic outcomes simultaneously. We created dummy variables to represent parenting profiles. The coefficient estimation for each dummy variable indicated how each parenting profile was associated with adolescent outcomes relative to the reference parenting profile. The reference group was rotated to obtain all possible comparisons among parenting profiles. Thus, two structural models with different reference groups of parenting profiles were analyzed. In the first model, we included two dummy variables of parenting profiles with the Moderate Positive group as the reference group (coded as 0). In the second model, we included two dummy variables of parenting profiles with the Unengaged group as the reference group (coded as 0).

Interaction terms were produced by multiplying the predictors and moderators. We created two-way interactions between parenting profiles and racial discrimination, between parenting profiles and adolescent gender, and between racial discrimination and adolescent gender, as well as three-way interactions among parenting profiles, racial discrimination, and gender. For significant interactions, simple slope analyses were conducted and plotted. Adolescents’ age, parental education, and school district were included in the models as covariates. All continuous independent variables including covariates were standardized before modeling.

Results of structural equation modeling

Model fit for the two models were good and the same, χ²(67) = 171.83, p < .001, CFI = .95, RMSEA = .03 [.03, .04], SRMR = .02. Standardized path parameters are presented in Table 4. For both academic outcomes and psychological well-being, we found significant main effects of discrimination and parenting profiles. More racial discrimination experiences were related to lower academic outcomes and psychological well-being. The Moderate Positive group had higher academic outcomes and psychological well-being than the other two parenting groups. The Unengaged group had higher academic outcomes and psychological well-being than the High Negative group.

Table 4.

Standardized Parameters of Paths among Study Variables

Variable (RF: Moderate Positive)	Well-being			Academic Outcomes			Variable (RF: Unengaged)	Well-being			Academic Outcomes
	B	SE(B)	β	B	SE(B)	β		B	SE(B)	β	B	SE(B)	β
Discrimination (RD)	−.06	.03	−.13*	−.12	.03	−.26***	Discrimination (RD)	−.07	.04	−.14	−.16	.05	−.37***
Female (SX)	.00	.04	.00	.05	.04	.06	Female (SX)	.14	.06	.15*	.10	.06	.11
High Negative (NE)	−.64	.06	−.50***	−.47	.06	−.41***	High Negative (NE)	−.24	.06	−.18***	−.19	.07	−.17**
Unengaged (UE)	−.40	.05	−.36***	−.28	.05	−.28***	Moderate Positive (PO)	.40	.05	.40***	.28	.05	.31***
RD*NE	.12	.05	.14**	.01	.05	.02	RD*NE	.13	.06	.15*	.06	.06	.07
RD*UE	−.01	.05	−.01	−.05	.06	−.05	RD*PO	.01	.05	.01	.05	.06	.07
SX*NE	.20	.09	.10*	.03	.09	.02	SX*NE	.05	.10	.03	−.02	.11	−.01
SX*UE	.15	.07	.10*	.04	.08	.03	SX*PO	−.15	.07	−.14*	−.04	.08	−.05
RD*SX	−.03	.04	−.04	−.01	.04	−.02	RD*SX	−.03	.06	−.03	−.04	.07	−.05
RD*SX*NE	−.07	.07	−.04	.00	.08	.00	RD*SX*NE	−.07	.09	−.04	.03	.10	.02
RD*SX*UE	.00	.08	.00	−.02	.08	−.02	RD*SX*PO	.00	.08	.00	.02	.08	.03
Adolescent age	−.06	.01	−.11***	−.10	.02	−.23***	Adolescent age	−.06	.01	−.11***	−.10	.02	−.23***
Parental education	−.01	.02	−.02	.03	.02	.07	Parental education	−.01	.02	−.02	.03	.02	.07
European-American-dominant school	−.02	.04	−.02	−.03	.05	−.02	European-American-dominant school	−.02	.04	−.02	−.03	.05	−.02
Racially-diverse school	.03	.04	.03	.05	.04	.05	Racial-diverse school	.03	.04	.03	.05	.04	.05

Note. RF=Reference group of parenting profiles. N=1363.

^*p < .05,

^**p < .01.

^***p < .001.

Regarding interaction effects, we found two significant two-way interactions on psychological well-being. First, there were significant interaction effects between racial discrimination experiences and parenting profiles on adolescent psychological well-being. Thus, for simple slope analysis, we ran a multi-group SEM model with parenting profiles as the grouping variable to examine the association between perceived racial discrimination and psychological well-being in each parenting group. As shown in Figure 3, the association between perceived racial discrimination and psychological well-being was negative in the Moderate Positive group but was not significant in the Unengaged and High Negative parenting groups. Although adolescents in the Moderate Positive group experienced decreases in psychological well-being as racial discrimination experiences increased, they generally had higher psychological well-being than adolescents in the other two parenting groups regardless of the level of discrimination.

Figure 3.

Interaction between Racial Discrimination and Parenting profiles on Adolescent Psychological Well-being. The association between racial discrimination and psychological well-being was negative in the Moderate Positive group (significant) and in the Unengaged group (not significant), but was positive in the High Negative parenting group (not significant).

Second, there were significant interaction effects between parenting profiles and adolescent gender on psychological well-being. For simple slope analysis, we ran multi-group SEM models with gender as the grouping variable to examine how parenting profiles relate to psychological well-being for boys and girls separately. We found that the associations between parenting profiles and adolescent psychological well-being were stronger for boys than for girls. Specifically, for both male and female adolescents, there were significant differences in psychological well-being between each of the parenting profiles. Relative to the Moderate Positive parenting group, adolescents in the High Negative parenting group (β = −.49, p < .001 for males, β = −.34, p < .001 for females) and Unengaged group (β = −.34, p < .001 for males, β = −.25, p < .001 for females) had lower psychological well-being. In addition, adolescents in the High Negative parenting group had lower psychological well-being than the Unengaged group (β = −.16, p < .01 for males, β = −.15, p < .01 for females). However the differences between the Moderate Positive parenting group and the other two groups were larger for males than for females. Figure 4 presents the standardized means of psychological well-being across the six gender by parenting groups.

Figure 4.

Interaction between Adolescent Gender and Parenting Profiles on Adolescent Psychological Well-being. For both male and female adolescents, there were significant differences in psychological well-being across parenting profiles. However the differences between the Moderate Positive parenting group and the other two groups were larger for males than for females.

We did not find significant three-way interactions (parenting profile × discrimination × gender) for psychological well-being. We did not find any significant interactions for academic outcomes.

Discussion

Discrimination is an important social context when studying African American adolescents’ developmental competencies (Garcia Coll et al., 1996; Wong et al., 2003). It is important to identify factors that help explain variation in the association between discrimination and adolescents’ academic and psychological well-being (Brown, 2008). The major goal of the current study was to test whether parenting profiles based on both culturally-specific (racial socialization) and general parenting (involved-vigilant parenting) served as a compensating or moderating factor. In addition, we also examined adolescent gender as a potential moderator. The findings supported three profiles of parenting that consisted of varying levels of racial socialization and involved-vigilant parenting: Moderate Positive, Unengaged, and High Negative. The Moderate Positive parenting group demonstrated higher levels of psychological well-being and academic outcomes than the other two parenting groups, even though racial discrimination experiences were negatively related to psychological well-being and academic outcomes for this group. The association between parenting profiles and adolescent psychological well-being was stronger for boys than for girls.

Parenting Profiles

Three profiles of parenting emerged from the data: 1) Moderate Positive parenting, marked by slightly above average involved vigilant parenting, racial pride, egalitarian messages, behavioral messages, and racial barrier messages and slightly lower than average negative messages; 2) Unengaged parenting with lower than average levels of involved-vigilant parenting and all racial-socialization messages; and 3) High Negative parenting, characterized by high negative messages, slightly above average racial barrier and behavioral messages, average egalitarian messages, and lower levels of racial pride messages and involved-vigilant parenting. These profiles generally correspond to the Moderate Positive, Moderate Negative, and Low Frequency profiles found in previous research by Neblett and colleagues (2008). We did not find a High Positive group as in that study.

There were also some similarities with two out of three profiles found by White-Johnson et al. (2010). Their Unengaged profile (below mean for all but negative messages) is similar to our Unengaged profile except that negative messages were also below the mean in our study. The multifaceted group (above mean for all but negative messages) also corresponds with our Moderate Positive parenting group. However, we did not find a profile similar to the low race salience group that deemphasized racial messages while emphasizing more mainstream messages. Across all three studies, parenting profiles indicated at least 1 group that included a balance of cultural, minority, and mainstream experiences (e.g., Moderate Positive) and one group that could be considered mainstream in their use of silence about race (e.g., Unengaged; Boykin & Toms, 1985; Hughes et al., 2006). As this study included involved-vigilant parenting in the latent profile analysis, it is not surprising that the profiles did not completely replicate all profiles from previous studies. In addition to the inclusion of involved-vigilant parenting, differences in participants’ contexts could contribute to differences in the latent profiles. For example, our participants on average attended schools with larger African American student populations than the participants in the two previous studies. These differences in context could shape the parenting adolescents received, particularly related to racial socialization.

The results also underscored the utility of using latent profiles of parenting versus individual parenting practices for understanding the family contexts in which adolescents are embedded. For example, adolescents in the Moderate Positive and High Negative groups received similar levels of racial barrier and behavioral messages but had very different outcomes in psychological well-being perhaps due to differences in the levels of negative messages and involved vigilant parenting they received. In addition, the adolescents in each group seemed to have unique experiences beyond the parenting they received. Consistent with Neblett and colleagues’ (2008) findings, youth in the High Negative group reported more racial discrimination experiences than youth in the Moderate Positive and Unengaged groups. This difference suggests that High Negative parenting may be in response to youths’ discrimination experiences or that High Negative parenting contributes to greater awareness or perception of discriminatory events. For example, hearing more stereotypical messages about one’s own group from a parent may make an adolescent more stigma conscious. Individuals with high stigma consciousness have been found to perceive more discrimination (Pinel, 1999).

It is also possible that age and gender differences are contributing to the differences in socialization and racial discrimination experiences across parenting profiles. The High Negative group had more boys and older adolescents. Both boys and older adolescents reported more racial discrimination which is consistent with previous studies that have found African American boys to report more discrimination than African American girls (Matthews, Salomon, Kenyon & Zhou, 2005; Seaton, Caldwell, Sellers, & Jackson, 2008) and that have found African American adolescents to report more racial discrimination as they aged (Greene, Way, & Pahl, 2006). Boys receiving more negative messages is also consistent with a previous study that found boys received more messages containing stereotypes and strategies for coping with racism (Thomas and Speight, 1999). Older adolescents also tend to report having more conversations about race with their parents than younger adolescents (Hughes et al., 2006), which could increase their exposure to negative messages. Parents may think that boys and older adolescents can handle more negative messages than girls or younger adolescents. Our results, however, indicate that High Negative parenting is more strongly related to lower psychological well-being for boys than girls and that High Negative parenting is negatively associated with academic and psychological outcomes for adolescents across age.

Discrimination, Parenting Profiles, Adolescent Gender, and Adolescent Outcomes

We found significant main effects of discrimination and parenting profiles on adolescent psychological well-being and academic outcomes, as well as two-way interactions between discrimination and parenting profiles, and between parenting profiles and adolescent gender on psychological well-being. Unsurprisingly, racial discrimination experiences were found to predict lower academic outcomes and psychological well-being in this sample. These results correspond to previous studies that suggest that racial discrimination experiences undermine academic engagement and performance (Smalls et al., 2007; Wong et al., 2003) and are related to lower psychological functioning (Seaton & Yip, 2009; Sellers et al., 2006) among African American adolescents.

Parenting profiles were associated with academic outcomes. Adolescents in families with Moderate Positive parenting had better academic outcomes than adolescents with Unengaged or High Negative parenting. This finding suggests that adolescents have higher academic aspirations and engage in more consistent academic behaviors when they are receiving more positive messages about their racial/ethnic group and equality of people across race/ethnicity, combined with higher levels of involved-vigilant parenting. Adolescents who were in the Unengaged group may have done less well academically because they may not have built skills related to self-regulation due to not being involved in joint decision-making with their parents or being provided for reasons for parents’ rules (Brody & Ge, 2001; Le et al., 2008). In addition, they may not be equipped to deal with racialized interactions that occur in academic settings because they receive few messages from parents to prepare them for discrimination or to instill a sense of pride in their group (Lesane-Brown, 2006).

Adolescents who received High Negative parenting performed less well academically relative to adolescents who were in the Unengaged group. Adolescents who were in the High Negative group received higher levels of each type of parenting except involved-vigilant parenting relative to their counterparts in the Unengaged group. Adolescents who receive relatively high negative messages, even in the context of moderate levels of other types of racial socialization messages including racial pride, may have a lower sense of self and internalize negative stereotypes about African Americans that devalue their academic ability (Constantine & Blackmon, 2002). These stereotypic messages, even when paired with some positive racial socialization messages, is worse for academic behaviors than receiving low levels of racial socialization messages. In the case of Unengaged messages, adolescents may be left unprepared for racial discrimination (Lesane-Brown, 2006), but may not have their sense of academic ability or group identity undermined to the extent that receiving higher than average levels of negative messages undermine academic aspirations and persistence.

There were also main effects of parenting profiles on psychological well-being. Adolescents with Moderate Positive parenting had better psychological well-being relative to adolescents with High Negative and Unengaged parenting. Adolescents in the Unengaged group also had higher psychological well-being group than adolescents in the High Negative group. Parenting profiles also interacted with racial discrimination in its association with psychological well-being but not academic outcomes. This pattern of results corresponds with previous studies that found parenting to buffer the negative results of racial discrimination experiences on psychological well-being (Neblett et al., 2008) but only found compensatory effects of parenting on boys’ academic adjustment (Neblett et al., 2009). Adolescents who experienced Moderate Positive parenting but high levels of racial discrimination had lower psychological well-being than adolescents who experienced Moderate Positive parenting and low levels of psychological well-being. This finding is consistent with some parenting studies that find that protective family factors have some reduction in benefit in high risk conditions (Cleveland, Feinberg, & Greenberg, 2010). There were not significant differences in adolescents’ psychological well-being within the Low Frequency and High Negative parenting groups based on their racial discrimination experiences but their overall level of psychological well-being was low. Still, Moderate Positive parenting was associated with better psychological well-being than Unengaged and High negative parenting, regardless of level of racial discrimination.

Overall, the results suggest that adolescents benefit psychologically when they receive positive racial messages in a context of supportive parental involvement, even when they experience racial discrimination. These results align with some previous studies (Smalls, 2009; Elmore & Gaylord-Harden, 2013) that found that adolescents had better academic and behavioral outcomes when they had moderate to high levels of racial socialization messages in the context of supportive, involved parenting (Elmore & Gaylord-Harden, 2013). The results contrast with Cooper and McLoyd (2011) who found more depressive symptoms and lower self-esteem when both supportive parenting and racial socialization was high. More studies must be conducted to understand this conflicting pattern of results.

There also was a significant interaction between gender and parenting on psychological well-being, but not on academic outcomes. Moderate Positive parenting was related to better psychological well-being than Unengaged or High Negative parenting for both boys and girls. However, there were greater differences in psychological well-being between boys receiving Moderate Positive parenting and boys receiving other types of parenting than for girls. Unengaged and High Negative parenting had stronger negative associations with boys’ psychological well-being relative to girls, although these types of parenting were also negative for girls. It is possible that African American boys are more sensitive to negative parenting due to their greater likelihood of experiencing stressors such as racial discrimination (Swanson, Cunningham, & Spencer, 2003; Seaton et al., 2008) and negative peer relations (Carter, 2006). In addition, a previous study has found parents of African American boys give incongruent messages about race and gender with more communal racial messages and more individualistic gender messages (Howard, Rose, & Barbarin, 2013). Boys receiving High Negative or Unengaged parenting may also receive more messages about competition and toughness than girls, which could contribute to differences in psychological outcomes.

Limitations

The results of this study must be considered in light of several limitations. First, we were unable to examine the relationship between the sources of racial discrimination and outcomes based on our measure of racial discrimination experiences. Benner and Graham (2013) found that only school-based discrimination was significantly related to academic achievement. It is possible that adolescents respond differently to parenting based on the source of racial discrimination. For example, parenting might interact with school-based discrimination but not societal discrimination on academic outcomes. Future studies using multidimensional measures of racial discrimination should consider interactions between source of discrimination and parenting on adolescent outcomes. Second, the results could be influenced by common method variance as reports of parenting, adolescents’ racial discrimination experiences, and outcomes were all reported by the adolescents. Future studies using multiple reporters and modes of data collection are necessary to further understand the links among parenting, discrimination, and adolescent outcomes. Finally, the associations found among parenting, psychological well-being, and academic outcomes are correlational. It is possible that parents' involved-vigilant parenting and racial socialization practices are in response to adolescents' previous psychological well-being and academic behaviors and attitudes. In addition, adolescents’ psychological well-being could influence their perceptions of racial discrimination. However, longitudinal studies examining psychological well-being and racial discrimination have found evidence in support of causal effects of racial discrimination on mental health over time rather than the alternative (Greene, Way, & Pahl, 2006; Hurd, Varner, Caldwell, & Zimmerman, 2014).

Implications

While there were some limitations, there are several implications of this study for research and for African American families. First, while most studies of African American adolescents focus on negative outcomes, this study examined their developmental competencies. Examining factors that promote or protect youths’ competencies can provide a more well-rounded understanding of the experiences of African American adolescents and be informative for development of interventions. Second, studying parenting profiles that take into account both cultural-specific and general parenting practices can contribute to a more holistic view of parenting in African American families and provide a more comprehensive understanding of African American youths’ experiences in families. Third, many parents of African American youth have concerns about how to prepare their children for discrimination and about the frequency of messages about race they should give. These findings suggest that parents who give moderate levels of messages about race and potential discrimination in an environment characterized by involvement, monitoring, and mutual problem solving will likely aid their adolescents in persisting in academic tasks and developing psychological well-being relative to those who give few or stereotypically negative racial messages in a context of less parental involvement. Furthermore, stereotypic racial messages, even when paired with some positive racial messages, seem to be more harmful than few messages, particularly in the context of low parental involvement. Finally, parenting is related to academic and psychological outcomes for African American boys and girls but low engagement and stereotypic messages may be particularly harmful for African American boys’ psychological well-being.