Virtual Environments, Online Racial Discrimination, and Adjustment among a Diverse, School-Based Sample of Adolescents

Abstract

Given the recent rise in online hate activity and the increased amount of time adolescents spend with media, more research is needed on their experiences with racial discrimination in virtual environments. This cross-sectional study examines the association between amount of time spent online, traditional and online racial discrimination and adolescent adjustment, including depressive symptoms, anxiety and externalizing behaviors. The study also explores the role that social identities, including race and gender, play in these associations. Online surveys were administered to 627 sixth through twelfth graders in K-8, middle and high schools. Multiple regression results revealed that discrimination online was associated with all three outcome variables. Additionally, a significant interaction between online discrimination by time online was found for externalizing behaviors indicating that increased time online and higher levels of online discrimination are associated with more problem behavior. This study highlights the need for clinicians, educational professionals and researchers to attend to race-related experiences online as well as in traditional environments.

INTRODUCTION

Racial discrimination is a common stressor and a growing threat to adolescent health and well-being. More specifically, within their lifetime, up to 94% of African American, Latino, and Asian youth have experienced traditional or face-to-face discrimination that was associated with their racial and ethnic background (Benner & Kim, 2009; Dotterer, McHale, & Crouter, 2009; Flanagan, Syvertsen, Gill, Gallay, & Cum-sille, 2009; Harris-Britt, Valrie, Kurtz-Costes, & Rowley, 2007; Huynh & Fuligni, 2010; Martin et al., 2011; Medvedeva, 2010; Neblett et al., 2008; Pachter, Szalacha, Bernstein, & Coll, 2010). Much of the research in the area of traditional racial discrimination focuses on the perceived frequency of these experiences within the classroom, including unfair treatment due to race (Chavous, Rivas-Drake, Smalls, Griffin, & Cogburn, 2008), where respondents may be treated with less respect or harassed because of their race or ethnicity (Rivas-Drake, Hughes, & Way, 2009; Shin, D’Antonio, Son, Kim, & Park, 2011). To extend this body of research, some scholars have explored disparities and experiences of tracking, unfair discipline, perceptions of lower levels of intelligence, or receiving less academic praise and reinforcement than their white counterparts (Benner & Kim, 2009; Dotterer et al., 2009; Cogburn, Chavous, & Griffin, 2011).

Race in Virtual Environments

While the foundation of literature for traditional discrimination is grounded in decades of empirical investigation, little is known about adolescents’ racial discrimination experiences in virtual environments. We define online racial discrimination as denigrating or excluding individuals or groups on the basis of race through the use of symbols, voice, video, images, text, and graphic representations. These experiences may resemble traditional discrimination and include being disrespected or being called race-related names (Gaylord-Harden & Cunningham, 2009; Roberts, Gibbons, Gerrard, Weng, Murry, Simons, & Lorenz, 2012; Umaña-Taylor, Wong, Gonzales, & Dumka, 2012). Online forms of racial discrimination occur in social networking sites, chat rooms, discussion boards, through text messaging, web pages, online videos, music, and online games. For example, a black female student from the prestigious Stuyvesant High School in New York was sent a video on Facebook in which white fellow students performed a five minute and forty second rap calling her “ni**er” and “ignorant,” and threatening her with sexual violence. Images and text also construct racial minorities as inferior, unintelligent, as criminals and, in many cases, animals. They also mock African American skin color and body types, cultural practices, and history (Tynes, Umaña-Taylor, Rose, Lin, & Anderson, 2012).

In the mid-1990s, the Internet was lauded for its potential to usher in a color-blind society. As the medium proliferated, scholars argued it could eliminate racial cues from communication and lead to a more egalitarian electronic global village, where there would be no race, gender or infirmities (Ess, 2001; Negroponte, 1995). Though visual signifiers of race may have been removed in early virtual environments, research on adults shows that across a range of online communication settings (in internet relay chat, for example; Glaser et al., 2002; Kang, 2000; Nakamura, 2002; Kendall, 1998), race takes on a linguistic form. Once made visible through text, it has been found to be central to the culture of computer-mediated environments. Further, many of the social norms and ills that exist offline are often reproduced in adult online communities (Burkhalter, 1999). This is increasingly evident as images, videos, and graphic representations of the body become more prevalent online.

To address the question of whether similar findings to those found with adults would be found with adolescent populations, Tynes and colleagues (2004) conducted one of the earliest studies of race-related discourse in monitored versus unmonitored teen chat. They found that 37 out of 38 half-hour transcripts had at least one racial or ethnic utterance and that, in the absence of monitors, participants were more likely to engage in negative racial and ethnic discourse. Other studies on adolescents suggest that they often encounter stereotypic and racist images, including text and video about people of color, but lack the digital and media literacy skills to determine who produces this material and whether it is legitimate or from a credible source (Daniels, 2009). Moreover, hate groups may assign ambiguous titles to websites, so that they are easily mistaken for more innocuous sites. Adolescents can unknowingly stumble upon these sites while searching for information online. While doing a report on weather for example, individuals may find the Stormfront. org for kids website. Its home page reads: “Need to do a school report of M.L. King? Visit this website for all your needs: www.martinlutherking.org.” The viewer is then redirected to the site and can click on a number of links, including one that reads “the truth about Martin Luther King.” Those who go there encounter a myriad of disparaging articles, speeches, and “facts” about this historical figure. Since children and adolescents are generally less able than adults to differentiate between truth and fiction online, hate messages masquerading as facts may be interpreted as truth.

Despite the election of the nation’s first African American president and reports of a post- racial America, there has been an increase in online hate activity (Simon Wiesenthal Center 2009). The numbers of extremist and hate sites rose exponentially from 6,000 to 10,000 from 2006 to 2009 and to 15,000 by 2011 (Simon Wiesenthal Center, 2012). Additionally Storm-front.org, arguably the Internet’s first and oldest hate site, crashed on the night of the election because it witnessed unprecedented numbers of visitors. Moreover, the Knights of the Ku Klux Klan site has seen its numbers of unique daily visitors almost triple, from 15,000 to 40,000 (Chen, 2009). Not only does a fringe element of society engage is discriminatory practices online, but Daniels (2012) outlines how researchers are beginning to understand race and racism across a range of mainstream virtual environments. Though theorizing about race online is in its infancy, scholars argue that individuals engage in “two faced racism” (Picca & Feagin, 2007), where tolerance is performed in multicultural spaces and explicit racism in private white only spaces (Daniels, 2012, Steinfeldt et al., 2010). A blurring of public and private (Daniels & Hughey, 2012), along with a perceived lack of monitoring (Tynes et al., 2004), may contribute to virtual environments where adolescents frequently receive negative messages about their racial group.

TRADITIONAL VS. ONLINE RACIAL DISCRIMINATION AND ADOLESCENT ADJUSTMENT

The increase in hate activity coupled with the fact that 95% of adolescents of ages 12 through 17 have access to the internet (Lenhart, Madden, Smith, Purcell, Zickuhr, & Raine, 2011) point to a pressing need to understand online racial discrimination. As growing numbers of youth gain access to the internet, youth of color may be more susceptible to discriminatory experiences because of their race or ethnicity (Kahn, Spencer & Glaser, 2013). Subrahmanyam and Smahel (2011) describe the behavioral coherence between an individual’s offline identity and the activities they engage in online. This may increase the number of online contexts and the nature of online interaction that may increase the likely of being victimized. For example, Tynes, Giang, Williams, and Thompson (2008) found that 29% of African American and 42% of multiracial/other high school students reported experiencing racial discrimination in online environments (via text message, social networking sites, online games, and other internet-based sites). To extend these findings to Asian American youth, Shin and colleagues (2011) assessed the nature and frequency of bullying among Korean American high school students, and determined that approximately 25% of the respondents experienced online victimization. Among the identified victims, 29% reported their victimization was the result of their point of origin (i.e., home country), while 23% were victimized due to their skin color.

Drawing from offline studies, African American adolescents who perceive higher levels of discrimination tend to exhibit lower levels of self-esteem (Seaton & Yip, 2009), exhibit more conduct problems (Brody et al. 2006), and report higher levels of depressive symptoms (Cogburn et al., 2011; Gaylord-Harden & Cunningham, 2009; Neblett et al. 2008). Perceived discrimination was also associated with increased depressive symptoms for Latino youth (Umaña-Taylor & Updegraff, 2007; Lorenzo-Blanco et al. 2011). Consequently, researchers have noted that discrimination is related to decreased psychological well-being for adolescents (Sellers et.al., 2006; Umaña-Taylor & Updegraff, 2007) and declines in grades, academic curiosity, and persistence (Alfaro, Umaña-Taylor, Gonzales-Backen, Bámaca, & Zeiders, 2009; Neblett, Philip, Cogburn, & Sellers, 2006; Wong et al., 2003). In a similar study of early adolescents, stress and psychological characteristics related to race were associated with behavioral and emotional adjustment, particularly for African Americans (DuBois, Burk- Braxton, Swenson, Tevendale, & Hardesty, 2002). Although it is conceivable that these experiences extend to online contexts, the association between these outcomes and online discrimination has been the focus of few empirical investigations.

From the limited number of existing studies, we know that experiences with online racial discrimination, independent of other types of discrimination and stress, are adversely related to adolescent mental health. For example, in the first study to explore online experiences with an ethnically diverse sample, racial discrimination was positively associated with depressive symptoms and anxiety, even after adjusting for offline racial discrimination and adolescent stress (Tynes et al., 2008). As noted by the authors, one limitation of the study was the creation of an “other” racial category, which included Latino, Asian, and multiracial youth. While this approach was necessary due to a limited sample size, it was not possible to assess the level of online discrimination for Latinos or Asians, nor the extent to which these experiences were related to their mental health.

In addition to the aforementioned methodological concerns, and the dearth of literature on online discrimination, a recent report using data from the Kaiser Family Foundation demonstrated that African American, Latino, and Asian youth spend approximately 4 and ½ more hours with media per day than their white counterparts (Rideout, Lauricella, & Wartella, 2011). Interestingly, even after controlling for parent education, and family structure, race was still a significant predictor of media exposure, where youth of color used forms of media, including mobile devices, an average of one hour more per day than their white counterparts.

The extant research on whether this extended media use is associated with mental health and behavioral outcomes is equivocal. For example, research has demonstrated that increased internet use is associated with depressive symptoms and loneliness (Belanger, Akre, Berchtold, & Michaud, 2011; Kraut et al., 1998; Lam & Peng, 2010; Park, 2009), yet some studies have also shown no association between time online, well-being (Gross 2004), or depressive symptoms (Sanders, Field, Diego, & Kaplan, 2000). Therefore, this study contributes to this debate by examining whether time online is associated with depressive symptoms, anxiety, and problem behavior among a sample of African American, Asian, and Latino adolescents, which are the specific subgroups of students noted in previous literature to have extended media exposure. Further, this study examines whether time online moderates the relationship between discrimination and the outcome mental health and behavioral variables.

In the current study it was hypothesized that higher levels of traditional or in school discrimination will be associated with higher levels of depressive symptoms, anxiety, and externalizing behaviors for all participants. Further, when accounting for traditional discrimination, it was hypothesized that online racial discrimination will also be a significant predictor for each outcome variable, consistent with foundational research by Tynes and colleagues (2008). Since past research is equivocal regarding the relationship between mental health and time online, interactions between time online and discrimination will also be tested. It was expected that experiences of discrimination will be moderated by time spent online indicating that time spent online for youth experiencing high levels of stress due to discrimination will be significant predictors for poor mental health and externalizing behaviors.

METHOD

Participants

Participants for the present study were selected from the first wave of a three wave longitudinal study that investigates the risk and protective factors associated with online experiences of students from twelve Midwestern US schools in grades 6 – 12. The current sample included 627 adolescents (45.3% male, 54.5%% female), with ages ranging from 10 – 18 (M = 14.42, SD = 1.98). The racial makeup of the sample, according to self-report, consisted of 53.7% Black or African American (n = 337), 37.2% Hispanic or Latino (n = 233), and 9.1% Asian or Asian American (n = 57). Overall, the average participation rate across the twelve schools was 49.8%.

PROCEDURE

Research assistants recruited students from classrooms, which were selected by administrators based on access to laptops or computers labs. Classes selected were primarily technology, English, and homeroom. Parental consent forms and fliers were distributed to approximately 150 students per school with copies available in English and Spanish. During distribution, research assistants gave a brief 10-minute presentation to the selected classes to describe the purpose of the study. On a prearranged date, researchers returned to each school to administer surveys via web link to all students who returned the affirmative parental consent forms.

Online surveys were sent to email addresses that were provided by participants, which were accessed during the allotted classroom time. In the event that participants did not have a valid email address prior to survey administration, temporary email addresses were established for survey access. In a small number of cases, surveys were accessed via a web link. Once access was granted, survey administration occurred over one to two consecutive class periods. Research assistants were present to inform students of confidentiality, explain terms, and troubleshoot any technical difficulties. Prior to beginning the survey, the research team explained that the respondents would be asked about their online experiences and their feelings about themselves. They were informed of confidentiality and told that they had the right to stop at any point in the survey. Following the completion or termination of the survey, all students were provided with resources such as local counseling services and internet safety websites to report online predators. As an incentive, students received $15 Amazon.com gift certificates for their participation for year one, and participating schools were provided with a small stipend. Recruitment and consent procedures were reviewed by the Institutional Review Board of the principal investigator’s institution.

MEASURES

• Depressive Symptoms: A12-item version of the Center for Epidemiologic Studies Depression Scale (CESD-12) was used to measure adolescents’ depressive symptoms (Radloff, 1977; Roberts & Sobhan, 1992). Sample items include “I felt I was just as good as other people” and “I had crying spells”. Response options ranged from “0=Rarely or none of the time” to “3=Most or all of the time.” Cronbach’s alpha in the current study was .72;

• Anxiety: Adolescents’ anxiety was assessed using four items from the tension subscale of the Profile of Mood States-Adolescents (Terry, Lane, Lane, & Keohane, 1999). Participants were asked to describe the extent that they felt each of the following: “panicky,” “anxious,” “worried,” and “nervous.” Responses ranged from “0=Not at all” to “4=Extremely.” Cronbach’s alpha in the current study was .79;

• Externalizing Behaviors: The rule breaking subscale of the Youth Self Report of the Child Behavior Checklist (Achenbach, 1991) was used to measure externalizing behavior. It is a widely used measure of competencies, emotional and behavior problems in youth between the ages of 11–18. Sample items include “I break rules at home, school, or elsewhere” and “I smoke, chew, or sniff tobacco”. Responses ranged from “0=Not true” to “2=Very true or Often true”. Cronbach’s alpha in the current study was .80;

• Time online: Time online was assessed using a single item indicator that assessed the average number of hours per day an individual spent online (i.e., How many hours are you online on a usual day when you use the internet?). Response options included “1 = 0 hours,” “2 = 1 hour or less,” “3 = 1 to 2 hours,” “4 = 2 to 3 hours,” “5 = 3 to 4 hours,” “6 = 4 to 5 hours,” “7 = 5 to 6 hours,” “8 = 6 to 7 hours,” “9 = 7 to 8 hours,” “10 = 8 to 9 hours,” “11 = 9 to 10 hours,” and “12 = more than 10 hours”;

• Online Racial Discrimination: Adolescents’ online discrimination based on race and ethnicity was assessed using two subscales of the Online Victimization Scale (Tynes, Rose, & Williams, 2010). Individual Online Racial Discrimination refers to derogatory text, images, and symbols that directly target an individual because of his or her race or ethnicity. This subscale was comprised of 4 items (e.g., “People have said mean or rude things about me because of my race or ethnic group online”), with responses options ranging from “0=Never” to “5=Everyday.” Cronbach’s alpha in the current study was .74;

• Offline Racial Discrimination: Adolescents’ perceived discrimination in offline settings was assessed using an abbreviated version of The Perceived Discrimination by Adults/Peers Scale (Way, 1997). Items capture experiences in two of three offline contexts: Other Students in School (i.e., Peers; 5 items; e.g. “How often do you feel that other students in school treat you with less respect because of your race or ethnicity”) and Adults in School (5 items, e.g. “How often do adults in school treat you like you’re NOT smart because of your race or ethnicity?”). Responses were scored on a 5-point Likert scale ranging from “0=Never” to “4=All the time.” A mean score was calculated for each context, with higher scores indicating greater perceived discrimination in each context. Cronbach’s alpha for peer and adult discrimination was .90 and .92 respectively.

DATA ANALYSIS

Data were analyzed to explore the differences in online and offline racial experiences and the impact of time online and online racial discrimination on depressive symptoms, anxiety, and externalizing behaviors for the three racial groups. Hierarchical multiple regression was used to examine the predictive nature of (1) time online, and (2) online and (3) offline peer and adult discrimination on (a) depressive symptoms, (b) anxiety, and (c) externalizing behaviors. Of particular interest was time online and the interactions of time online by online and offline racial experiences.

For each model, depression, anxiety, and externalizing behaviors served as separate dependent variables, with gender, grade, race, and time online entered in step 1; traditional peer and adult discrimination in step 2; online racial discrimination in step 3; and interactions for time online by online racial discrimination, time online by traditional adult discrimination, and time online by traditional peer discrimination in step 4. Variables were entered in this order to examine whether each block of variables is predictive over and above the variables entered earlier in the model by examining whether the change in the F-Statistic is significant at each step. If the change in the F-statistic was not significant, no further steps were added to the model. All predictor variables, with the exception of gender and race, were centered prior to creation of the interaction terms and analysis (see Table 1 for descriptive statistics prior to centering).

Table 1.

Means and standard deviations for scales by racial group and total sample

Item/Scale (Scale Values)	Asian	African American	Latino	Total
Depressive Symptoms (0–3)	.79 (.42)	.86 (.44)	.84 (.44)	.85 (.44)
Anxiety (0–4)	.70 (.77)	.70 (.85)	.67 (.78)	.69 (.82)
Externalizing Behaviors (0–2)	.15 (.16)	.29 (.26)	.28 (.26)	.27 (.26)
Time Online (1–12)	4.32 (2.20)	4.86 (2.51)	3.92 (1.77)	4.46 (2.27)
Online Racial Discrimination (0–4)	.25 (.37)	.34 (.65)	.34 (.58)	.33 (.61)
Racial Discrimination by Peers at School (0–4)	.54 (.68)	.60 (.88)	.44 (.76)	.53 (.82)
Racial Discrimination by Adults at School (0–4)	.21 (.46)	.49 (.76)	.29 (.67)	.39 (.71)

RESULTS

• Depressive Symptoms: To explore the predictive nature of time online, traditional (adult and peer) racial discrimination, and online racial discrimination on depressive symptoms, a regression model for the entire group was constructed. Based on the significant changes in the F statistic (F(1.536) = 15.91 p < .001), the final model included the main effects, but did not include the interaction terms (i.e., Step 4), and accounted for approximately 17% of the variance in depressive symptoms (R² = .165). Significant main effects included traditional peer discrimination (β =.16, p < .01), traditional adult discrimination (β=.12, p < .05), and online racial discrimination (β =.18, p < .001), while accounting for gender (Male = 0, β =.16, p < .01). These findings suggest that higher levels of discrimination, both offline and online, predicted higher levels of depressive symptoms for the sample population (See Table 2);

• Anxiety: To explore the predictive nature of time online, traditional (adult and peer) racial discrimination, and online racial discrimination on anxiety, a regression model for the entire group was constructed. Based on the significant changes in the F statistic (F(3.526) = 6.13 p < .001), the final model included the main effects and interaction terms, and accounted for approximately 14% of the variance in anxiety (R² = .144). While the final model included both main effects and interaction terms, the only significant main effect was online racial discrimination (β =.13, p < .01), while accounting for gender (Male = 0, β =.17, p < .001). These findings suggest that higher levels of online discrimination predicted higher levels of anxiety for the sample population (See Table 2).

Table 2.

Summary of final step of regression models for depression, anxiety, and rule breaking

	Depression (Step 3)		Anxiety (Step 4)		Rule Breaking (Step 4)
	B (SE B) β		B (SE B) β		B (SE B) β
Final Model: Step 1
Gender (Female = 1)	.14 (.04)	.16**	.29 (.07)	.17***	−.01 (.02)	−.02
Grade	.00 (.01)	.01	.00 (.02)	.00	.02 (.01)	.11**
Ethnicity (Asian = 1)	−.01 (.06)	−.01	.12 (.12)	.04	−.10 (.04)	−.11**
Ethnicity (Latino = 1)	.02 (.04)	.03	.04 (.08)	.02	.02 (.02)	.04
Time Online	.01 (.01)	.07	.01 (.02)	.00	.02 (.01)	.13**
Step 2
Peer Discrimination	.08 (.03)	.16**	.09 (.06)	.09	.03 (.02)	.08
Adult Discrimination	.07 (.03)	.12*	.11 (.06)	.10	.05 (.02)	.13*
Step 3
Online Discrimination	.13 (.03)	.18***	.18 (.06)	.13**	.06 (.02)	.14**
Step 4
Time Online by Online Discrimination	---	---	−.03 (.03)	−.05	.03 (.01)	.17***
Time Online by Peer Discrimination	---	---	.07 (.03)	.18**	.00 (.01)	.03
Time Online by Adult Discrimination	---	---	.01 (.03)	.02	−.01 (.01)	−.04
	Δ F(1.536) = 15.91 p <.001, R2 = .165, R2 Δ = .03		Δ F(3.526) =6.13 p < .001, R2 = .144, R2 Δ = .03		ΔF(3.522) = 6.08 p < .001, R2 = .211, R2 Δ = .03

Note: African American is reference group for all ethnicity analyses.

^*p<0.05,

^**p<.01,

^***p<.001.

In addition to the significant main effects, the interaction of time online by peer discrimination emerged as a significant predictor (β =.18, p < .001) of anxiety. While it should be noted that the main effects of time online (β =.00, p > .05) and peer discrimination (β =.09, p >.05) were not independently significant, all main effects are retained in the final model, and based on the a priori decision rules of significant changes in the F statistic, the significant interaction was probed and graphed based on ± 1 standard deviation, and simple slopes were evaluated with time online serving as the moderating variable (see Figure 1). The simple slope for individuals who reported high time online (i.e., one standard deviation above the mean) was significant (t(623) = 3.40, p < .001) for anxiety. More specifically, students who spent more time online and experienced high levels of traditional peer discrimination (i.e., one standard deviation above the mean) reported higher levels of anxiety than students who spent less time online:

• Externalizing Behavior: To explore the predictive nature of time online, traditional (adult and peer) racial discrimination, and online racial discrimination on externalizing behaviors, a regression model for the entire group was constructed (see Figure 2). Based on the significant changes in the F statistic (F(3.522) = 6.08 p < .001), the final model included the main effects and interaction terms, and accounted for approximately 21% of the variance in externalizing behaviors (R² = .211). Significant main effects included time online (β = .13, p < .01), traditional adult discrimination (β = .13, p < .05), and online racial discrimination (β = .14, p < .01), while accounting for grade (β = .11, p < .01) and ethnicity (Asian = 1, β = −.11, p < .001). These findings suggest that more time online, higher levels of adult offline discrimination, and higher levels of online discrimination, predicted higher levels of externalizing behaviors for this sample population (See Table 2).

Figure 1.

Time online as a moderator of the association between peer discrimination and anxiety

Figure 2.

Time online as a moderator of the association between online racial discrimination and externalizing behaviors

In addition to the significant main effects, the interaction of time online by online racial discrimination emerged as a significant predictor (β =.17, p < .001) of externalizing behavior. The significant interaction was probed and graphed based on ± 1 standard deviation of time online, and simple slopes were evaluated with time online serving as the moderating variable (see Figure 1). The simple slope for individuals who reported high time online (i.e., one standard deviation above the mean) was significant (t(623) = 5.234, p < .001) for externalizing behaviors. More specifically, students who spent more time online and experienced high levels of online racial discrimination (i.e., one standard deviation above the mean), reported more externalizing behaviors than students who spent less time online.

DISCUSSION

The purpose of this study was to explore the association between racial discrimination, both online and offline, and adjustment among a diverse sample of adolescents. Given the inconsistencies in previous research, and the overall dearth of online discrimination literature, time spent in virtual environments was of direct interest in this association. Results demonstrated that the total sample, on average, spent between two to four hours online per day and that extended time online was associated with increased online racial discrimination and higher levels of externalizing behaviors, as well as peer discrimination and anxiety. Moreover, consistent with the traditional discrimination literature, the results of the present study demonstrated associations between race-related victimization online and depressive symptoms, anxiety, and externalizing behavior (Brody et al. 2006; Coker et al. 2009; Greene Way, & Pahl, 2006; Huynh & Fuligni 2010; Pachter et al. 2010; Sellers, Caldwell, Schmeelk-Cone, & Zimmerman, 2003).

These findings parallel extant literature that has found associations between discrimination and adjustment for youth of color (Coker et al., 2009; Fisher, Wallace, & Fenton, 2000). Coker and colleagues (2009), for example, found for African American and Latino fifth graders, discrimination was associated with depressive symptoms and conduct disorder (attention deficit hyperactivity disorder and oppositional defiant disorder were also found for Latinos). In addition, Russell et al. (2012) found that youth who experienced bias-based harassment due to race, sexual orientation, disability, or religion reported lower mental health status and increased substance use levels. Consequently, it appears that being victimized because of these social identities has a particularly detrimental impact on health.

Though the current study results are only partially consistent with research that shows an association between internet use and externalizing behaviors (Holtz & Appel, 2011; Tsitsika et al., 2014), there is evidence that suggests it is the activities in which adolescents engage that may contribute most to adjustment outcomes; not simply time spent on the internet (Bell, 2007). It should also be noted that for the current study, and contrary to previous research, time online was not an independent significant predictor of anxiety or depressive symptoms (Belanger et al., 2011; Kraut et al., 1998; Lam & Peng, 2010; Park, 2009). However, a significant interaction between time online and peer discrimination in school was found for anxiety. This finding suggests that those who experience both discrimination at school and spend an increased amount of time online report more anxiety. However, it is not clear if those who are discriminated against at school turned to online environments or if the increased amount of time online happened concurrently with the discriminatory experiences and increased levels of anxiety. Further longitudinal research is needed to better understand the relationship between internet use and face-to-face discrimination. In addition, more research is needed to better understand the overlap between online and offline discriminatory experiences over time.

Another finding of note is that in each of the models, online discrimination was associated with negative outcomes for the youth of color in the study. This adds to the small body of literature examining online discrimination. Consistent with earlier findings by Tynes and colleagues (2008), online discrimination is related to poor mental health outcomes. The current study also found that online discrimination is associated with externalizing behaviors. Further, an interaction between time online and online discrimination demonstrates that individuals who spend an increased amount of time online and experience high levels of online discrimination reported more externalizing behaviors than those who spend an increased amount of time online and experienced lower levels of online discrimination. Specifically this suggests that the experiences youth have in virtual environments and the time spent online are important factors in externalizing behaviors.

As young people gain widespread access to the internet, they may face additional discrimination in virtual environments as demonstrated by the current study. Although this study adds to the existing body of literature regarding online discrimination and adolescent adjustment, some limitations should be noted. These limitations include the fact that this study is cross-sectional. Though significant relationships between discrimination and the psychological and behavioral outcomes did emerge, the temporal ordering of the variables in the associations of interest is unclear. Conceptually, we expected that discrimination would precede depressive symptoms, anxiety, and externalizing behaviors; however, the cross-sectional nature of our data did not enable testing that hypothesis directly. Additionally, it is unclear how long- term exposure to online discrimination may be related to mental health and behavioral outcomes. Future research should examine this relationship for adolescents of color.

CONCLUSION

This study is the first to explore associations between internet use, online racial discrimination, offline racial discrimination, mental health, and externalizing behavior among a diverse sample of students. Findings from the current study, along with previous literature, demonstrated that experiences online have unique contributions to mental health that are over and above offline experiences with discrimination (Tynes et al., 2008). Furthermore, these findings reinforce the necessity for mental health and educational professionals to assess and provide adolescents with coping strategies for online and traditional racial discrimination experiences. Overall, the findings from this study suggest that both time spent in virtual environments and the nature of adolescent interactions may impact mental health and behavioral outcomes, rather than increased time spent online alone.

Biographies

Brendesha M. Tynes is an associate professor of Education and Psychology at the USC Rossier School of Education. Her research focuses on the role of the Internet in child and adolescent development with special attention to academic performance, mental health and problem behavior. She is also interested in the design of online interventions that reduce cyber-bullying. She is the principal investigator of a longitudinal study of the risk and protective factors associated with online victimization as well as a study of the use of mobile devices to enhance STEM performance and social-emotional learning. Dr. Tynes earned a master’s in Learning Sciences from Northwestern University and a doctoral degree in Human Development and Psychology from UCLA.

Chad Rose holds a doctoral degree in Special Education from the University of Illinois at Urbana-Champaign. He is currently an assistant professor in the Department of Special Education at the University of Missouri. His research focuses on bullying and victimization among students with disabilities, bully prevention efforts within a multi-component framework, and the intersection of social/emotional learning and bully prevention for students with disabilities.

Sophia Hiss is a PhD student at the USC Rossier School of Education. Her research examines the social and peer influences of adolescents’ educational outcomes and psychological well-being. Specifically, she has studied the impact of peer victimization in online settings examining general victimization and racial discrimination. She is also interested in the ways that adolescents cope with victimization and the possible buffers that may mitigate the negative outcomes associated with victimization online.

Adriana J. Umaña-Taylor is Foundation Professor at Arizona State University in the T. Denny Sanford School of Social and Family Dynamics. She received her PhD in Human Development and Family Studies from the University of Missouri-Columbia. Dr. Umaña-Taylor’s research focuses on ethnic identity formation, familial socialization processes, culturally informed risk and protective factors, and psychosocial functioning among ethnic minority youth.

Kimberly Mitchell is a Research Associate Professor of Psychology at the Crimes against Children Research Center, located at the University of New Hampshire. Her areas of research focus on youth Internet victimization and how technology can be used as a tool for prevention and intervention. She has been studying Internet use among youth for over 12 years. Dr. Mitchell has directed or co-directed several national projects including the First and Second Youth Internet Safety Studies (YISS); the Survey of Internet Mental Health Issues; and the First, Second, and Third National Juvenile Online Victimization Studies. She was the Principal Investigator of the third YISS, a grant to investigate the commercial exploitation of children through the Internet; and a grant exploring the role of technology in youth harassment. She is the author or co-author of over 70 peer-reviewed papers in her field and has spoken at numerous national conferences.

David R. Williams holds a doctoral degree in Sociology from the University of Michigan. He is currently the Florence and Laura Norman Professor of Public Health at the Harvard School of Public Health, Professor of African and African American Studies and an Affiliate of the Sociology Department at Harvard University. His research has enhanced understanding of the complex ways in which race, racial discrimination, socioeconomic status and religious involvement can affect physical and mental health.