Social Media Use and Externalizing Behaviors During Early Adolescence

Abstract

A primary goal of this study was to examine adolescents’ use of the most commonly used social media apps in the United States. An additional goal was to examine the relationship between social media use and adolescent externalizing behaviors. The moderating influences of gender and race/ethnicity also were assessed. The sample included 1,072 adolescents (M age = 12.74 years, SD = 0.67, 51% girls; 64% White). Participating adolescents completed a survey in school. Most of the adolescents did not use Facebook or Twitter. However, the majority of adolescents used Instagram and Snapchat, and girls were more likely to use these social media apps than were boys. Adolescents who regularly used social media reported more delinquent behavior and poorer behavioral conduct in comparison with nonusers. In addition, boys who used Facebook regularly and Hispanic adolescents who used Instagram regularly were at an elevated risk for delinquent behaviors and poorer behavioral conduct.

Keywords

adolescence social media technology delinquency externalizing behaviors

Adolescents today are immersed in technology. Social media applications (“apps”), defined as web-based applications that allow users to generate and share content with others (Kaplan & Haenlein, 2010), are especially popular with young people. Social media apps possess a wide range of purposes and usage options, ranging from text-focused apps (e.g., Twitter) and image-centric apps focused on sharing videos and photos (e.g., Instagram, Snapchat) to apps that encompass a wide variety of usage capabilities (e.g., Facebook; Associated Press-NORC, 2017). In the United States, virtually all adolescents (93%–97%) use social media apps (Barry et al., 2017; Madden et al., 2013; Pew Research Center, 2018). Moreover, many adolescents (24%) report that they are online “almost constantly” (Lenhart, Duggan, et al., 2015). Contemporary youth spend approximately 7 to 9 hours a day engaging with technology, primarily via social media apps (Common Sense Media, 2015, 2019). On average, adolescents use three different social media apps, with some using up to eight different apps daily (Barry et al., 2017). Among adolescents, the most popular apps, in order, are Instagram, Snapchat, Facebook, and Twitter (Associated Press-NORC, 2017). Notably, much of the research conducted on social media use has focused on high school and college-age youth; younger adolescents have been overlooked. However, the specific social media apps used, time spent using social media, and the relationship between social media use and psychological adjustment may differ by developmental stage.

According to relational developmental systems models of human development (Bronfenbrenner, 2005; Bronfenbrenner & Morris, 2006; Lerner et al., 2014, 2015), human behavior results from the interaction between the individual and the individual’s contexts. Given that social media has become a central context during adolescence, it is critical to understand how social media use may influence behavior during this critical period of development. A growing number of studies have focused on the benefits and risks of youth social media use. Social media use may benefit youth by enhancing communication, providing access to novel information, and contributing to identity development (O’Keeffe & Clarke-Pearson, 2011; Shapiro & Margolin, 2014). However, social media use also has been associated with increased depressive symptoms (Ilakkuvan et al., 2019; Vannucci & Ohannessian, 2019), anxiety symptoms (Vannucci et al., 2017; Vannucci & Ohannessian, 2019), and alcohol and drug use (Ohannessian et al., 2017). In addition, social media use may provide a means for adolescent externalizing behaviors, including poor behavioral conduct and problem behaviors that are directed outward toward the external environment such as aggression, bullying, and disobeying rules (Carpenter, 2012; Elsaesser et al., 2017; Liu, 2004; Patton et al., 2014). For example, cyberbullying is widespread with close to 75% of school-age youth experiencing cyberbullying at least once a year (Dredge et al., 2014; Elsaesser et al., 2017).

The anonymous nature of social media may be attractive for youth lacking in social skills and social competence, making it easier for such youth to act aggressive toward others. Moreover, some research and theory (e.g., social compensation theory; Campbell et al., 2006) has suggested that youth with preexisting conduct problems tend to have poor social skills and are attracted to social media and use it more frequently because it allows for anonymity (Mikami et al., 2010; Valkenburg et al., 2005). Research also has supported “poor-get-poorer” effects, suggesting that social media exacerbates negative outcomes for individuals with existing behavioral problems (Mesquita, 2015; Selfhout et al., 2009).

Peer influence effects that occur via social media use also may exacerbate delinquent behaviors in adolescents. Social norms theory proposes that adolescents’ perceptions of their friends’ engagement in delinquent behaviors increase motives for these behaviors, as the need for social acceptance is of paramount importance during this developmental period (Brechwald & Prinstein, 2011; Cotter & Smokowski, 2016). Recent research suggests that these well-documented peer influence processes extend to the social media context. For example, receiving “likes” on personal photos of delinquent behaviors on social media has been found to have robust rewarding properties for adolescents, and exposure to photos of delinquent behaviors posted by peers has been shown to predict adolescents’ own delinquent behaviors (Huang et al., 2014; Sherman et al., 2016). Image-based social media apps, including Instagram and Snapchat, may therefore be especially potent social media platforms for peer influence processes to operate during adolescence.

Of note, most existing research examining the link between social media use and externalizing behaviors in youth has focused on overall social media use or specifically on Facebook. For example, in a study examining late adolescents and emerging adults, childhood conduct disorder symptoms and adolescent antisocial behaviors were significantly and positively associated with more daily social media use overall (Galica et al., 2017). In another study, narcissistic and delinquent behaviors were associated with self-promoting, grandiose, and exploitative styles of interacting on Facebook (Carpenter, 2012). Antisocial personality disorder symptoms also have been linked to more Facebook use in older adolescents and adults (Rosen et al., 2013). Few studies have examined associations between externalizing behaviors and social media apps other than Facebook. This limitation is significant, given that Facebook use has declined among youth, whereas the use of other social media apps, such as Instagram and Snapchat, has steadily increased (Associated Press-NORC, 2017; Duggan et al., 2015; Molina, 2016). It also should be noted that much of the limited research to date has focused on older adolescents and adults. The relationship between social media use and externalizing behaviors during early adolescence has been overlooked.

Gender and Race/Ethnicity Differences

Individual sociodemographic characteristics, such as gender and race/ethnicity, are likely to influence early adolescents’ social media use patterns and the nature of the relationship between social media use and externalizing behaviors (Peter & Valkenburg, 2013). Gender differences in interpersonal processes may motivate adolescent girls to use certain types of social media platforms more frequently than boys, as girls tend to place a higher value on interpersonal goals and are more likely to maintain close friendships through disclosure and communication (Rose & Rudolph, 2006). Image-based social media platforms such as Instagram and Snapchat have emerged as a primary means through which adolescents communicate with peers via posting and sharing videos and photos and making comments on others’ posts (Associated Press-NORC, 2017), and therefore, gender differences may be readily apparent in Instagram and Snapchat usage.

In contrast to social development theories (Rose & Rudolph, 2006), gender differences have not been found consistently when examining general social media use (McCrae et al., 2017; Vannucci et al., 2017). However, gender differences have been found when the use of specific social media apps has been examined. Consistent with hypotheses, recent research indicates that although older adolescent boys and girls are equally likely to use Facebook and Twitter, adolescent girls are relatively more likely to use image-centered apps such as Instagram and Snapchat (Associated Press-NORC, 2017). It is not clear whether this pattern holds for younger adolescents as well. Moreover, few studies have examined whether the associations between social media use and indicators of adolescent adjustment differ by gender. According to social role theory (Eagly, 1987; Eagly et al., 2000), the impact of peer influence processes involving externalizing behaviors may be more robust for adolescent boys than for adolescent girls because aggressive behaviors are more consistent with U.S. male gender roles. Given the amplified nature of peer influence processes that occurs in the social media context (Nesi et al., 2018), the link between social media use and externalizing behaviors may be stronger for adolescent boys than for girls.

The social media context also affords a unique, positive opportunity for racial/ethnic minority youth to express their racial/ethnic identities and connect with groups of similar peers, which may not be possible in their real-world contexts (Chan, 2017). Given that racial/ethnic identification becomes more prominent and salient during adolescence (Umaña-Taylor et al., 2014), early adolescents who identify as racial/ethnic minorities may be especially driven to use a variety of social media platforms more frequently than their non-Hispanic White peers as a means to achieve these developmental goals. However, little is known in regard to racial/ethnic differences in social media use during adolescence. In a recent study, Black adolescents reported using Snapchat and Instagram more in comparison with White adolescents (Associated Press-NORC, 2017). It is not clear whether this pattern extends to early adolescence. It also is important to note that although the U.S. Hispanic population continues to grow, with a 22% increase in Hispanic youth under age 18 between 2006 and 2016 (U.S. Census Bureau, 2010), studies examining social media use among Hispanic individuals have been scant. Moreover, few studies have examined whether there are racial/ethnic differences in the relationship between social media use and indicators of psychological adjustment. It has been hypothesized that the perceived anonymity and privacy afforded by social media heightens the expression of intergroup biases toward racial/ethnic minorities, making individuals from racial/ethnic minority groups more susceptible to discrimination in the social media context and therefore at higher risk of psychological adjustment problems (Kahn et al., 2013). In support of this notion, exposure to racial/ethnic discrimination in online settings has been shown to be common and to contribute to internalizing symptoms (symptoms associated with internalizing disorders such as depression and anxiety) in Hispanic and Black U.S. adolescents, above and beyond offline discrimination experiences (Tynes et al., 2008, 2012; Umaña-Taylor et al., 2015). However, it remains to be seen as to whether there is a stronger relationship between social media use and externalizing behaviors among Hispanic and Black U.S. adolescents relative to non-Hispanic White U.S. adolescents.

The Present Study

Given the relative dearth of research conducted on social media use during the early adolescent period, a primary goal of this study was to examine early adolescents’ use of the most commonly used social media apps in the United States (Facebook, Twitter, Instagram, and Snapchat; Associated Press-NORC, 2017). An additional goal was to examine the relationship between social media use and externalizing behaviors during early adolescence more systematically by examining differential associations between the social media apps (Facebook, Twitter, Instagram, and Snapchat) and indicators of externalizing behaviors. More specifically, the following research questions were examined in a large, diverse, community sample of early adolescents:

Research Question 1: How often do early adolescents use social media (Facebook, Twitter, Instagram, and Snapchat)?

Research Question 2: Does social media use during early adolescence differ by gender and/or race/ethnicity? Consistent with theory and the extant literature, it was hypothesized that adolescent girls would use social media (especially Instagram and Snapchat) more than adolescent boys and that minority (Black and Hispanic/Latino) adolescents would use social media more than White adolescents.

Research Question 3: Is social media use associated with externalizing behaviors (operationalized as delinquent behaviors and poor behavioral conduct) during early adolescence, and if so, do these associations differ by gender and/or race/ethnicity? Based on theory and the limited literature, it was hypothesized that the association between social media use and externalizing behaviors would be stronger for adolescent boys than for adolescent girls, and for minority (Black and Hispanic/Latino) adolescents than for White adolescents.

Method

Participants

In the fall 2016, all the seventh- and eighth-grade students attending five middle schools in the New England region of the United States were invited to participate in the larger Predictors of ANxiety and Depression during Adolescence project (PANDA project; pandaresearchproject.org). The sample included 1,072 adolescents (51% girls; M age = 12.74 years, SD = 0.67 years, range = 12–14 years). The race/ethnicity breakdown was as follows: 64% White, 25% Hispanic/Latino, and 11% Black. Most of the parents had completed high school (94% of mothers and 94% of fathers) and some had completed 4 years of college (48% of mothers and 38% of fathers).

Measures

Social media use

The Technology Use Questionnaire (TUQ; Ohannessian, 2009) was used to assess social media use. Specifically, adolescents were asked to report how often they typically used the following apps: Facebook, Twitter, Instagram, and Snapchat. The response scale ranged from 0 = never to 8 = almost constantly. Given that some of the social media use variables were skewed, for both conceptual and empirical reasons, the social media use variables were recoded into the following categories 0 = never (no use), 1 = less than once a week to several times a week (occasional use), and 2 = at least once a day (regular use).

Delinquent behaviors

The Delinquent Behaviors Scale (DBS; American Psychiatric Association, 2013; Gault-Sherman, 2013; Prinstein et al., 2001; Sanchagrin et al., 2014) was used to assess adolescent delinquent behaviors. Adolescents were asked to report, in the past 6 months, how frequently they engaged in 17 problem behaviors, including starting physical fights, skipping school, stealing, and lying. The response scale ranged from 0 = never to 5 = about once a week or more. The DBS items were summed to reflect a delinquent behavior total score. Because this score was positively skewed, the logarithmic transformation was used. Cronbach’s alpha coefficient for the DBS total score in the PANDA sample was .95.

Behavioral conduct

The five-item behavioral conduct scale of the Self-Perception Profile for Adolescents—Revised Question Format (SPPA-R; Harter, 2012; Wichstraum, 1995) was used to assess adolescents’ reports of their behavior. This scale taps the degree to which an adolescent does the right thing, acts the way he or she is supposed to act, and avoids getting into trouble (Harter, 2012). A representative item is, “I often get in trouble for the things I do.” The response scale ranges from 1 = describes me very well to 4 = describes me very poorly. Therefore, higher scores reflect better behavioral conduct, whereas lower scores indicate poorer behavior conduct. The behavioral conduct scale has been shown to be a valid and reliable measure for adolescents (Harter, 2012). The alpha coefficient of the behavioral conduct scale in the PANDA sample was .76.

Covariates

Adolescents reported on their gender (1 = male, 2 = female), age, race/ethnicity (1 = White, 2 = Black, 3 = Hispanic) and parents’ highest level of education completed (1 = elementary school or middle school to 5 = law, medical, or graduate [Master’s, doctorate] school).

Procedures

Parents of students attending participating middle schools were mailed a letter describing the study. The letter requested that they contact the study team (by phone, mail, or email) if they did not want their adolescent to participate in the study. Only 2% of parents requested that their child not participate. Data collection took place in classrooms in the middle schools. Adolescents who had parental consent were administered a self-report survey by trained research staff (certified with human subjects training). The adolescents were reassured that the data collected were confidential, that their participation was voluntary, and that they could withdraw from the study at any time. Prior to participation, the adolescents were given an assent form. The majority of the adolescents (94%) agreed to participate and signed the assent form. After submitting the assent form, participating adolescents completed the ~60-minute survey. Immediately after completing the survey and turning it in, the adolescents were given a free movie pass. PANDA project procedures were approved by the Connecticut Children’s Medical Center Institutional Review Board.

Analytic Strategy

As a first step, chi-square tests were conducted to examine whether social media use differed by gender and/or race/ethnicity. Next, multivariate analysis of covariance (MANCOVA) models were conducted to examine whether social media use was related to externalizing behaviors. The fixed factors were gender, race/ethnicity, and a social media use measure (Facebook, Twitter, Instagram, or Snapchat). Age and parental education (a proxy for socioeconomic status [SES]) were included as covariates. The dependent variables were the delinquent behaviors total score and the behavioral conduct scale score. Separate models were run for each of the social media use measures. Wilks’ lambda was used to test multivariate effects. Post hoc multiple comparison of means were examined with the Bonferroni comparison test.

Results

Descriptive Statistics

As shown in Table 1, most of the adolescents reported using Instagram (72%) and Snapchat (70%). Facebook and Twitter use were much less popular in our sample of early adolescents. Only 27% of the adolescents reported using Facebook and 25% reported using Twitter. The breakdown of social media use by gender and race/ethnicity is presented in Table 1. As shown, Facebook use was found to differ by race/ethnicity, χ²(4) = 69.28, p < .001. Hispanic adolescents were more likely to use Facebook than Black adolescents, χ²(2) = 8.34, p < .05, and Black adolescents were more likely to use Facebook than White adolescents, χ²(2) = 8.28, p < .05. There were no gender differences for Facebook use. Similar to Facebook, Twitter use differed by race/ethnicity, χ²(4) = 14.52, p < .01. Hispanic adolescents were more likely to use Twitter than both Black adolescents, χ²(2) = 6.10, p < .05, and White adolescents, χ²(2) = 12.44, p < .01. Consistent with Facebook use, there were no gender differences for Twitter use. Instagram and Snapchat use did not differ by race/ethnicity. However, both Instagram and Snapchat use differed by gender, χ²(2) = 16.88, p < .001; χ²(2) = 53.06, p < .001, respectively. As shown in Table 1, girls were more likely to use Instagram and Snapchat in comparison with boys.

Table 1.

Social Media Use During Early Adolescence by Gender and Race/Ethnicity.

Social media app	Overall	Gender		Race/ethnicity
	(n = 1,072)	Boys (n = 526)	Girls (n = 546)	White (n = 686)	Black (n =115)	Hispanic (n = 271)
	% (n)	% (n)	% (n)	% (n)	% (n)	% (n)
Facebook^a
No use	72.7 (779)	72.2 (380)	73.2 (399)	79.2 (543)	67.0 (77)	58.7 (159)
Occasional use	15.8 (169)	16.9 (89)	14.7 (80)	14.4 (99)	20.9 (24)	17.0 (46)
Regular use	11.5 (124)	10.9 (57)	12.1 (67)	6.4 (44)	12.1 (14)	24.3 (66)
Twitter^a
No use	74.5 (799)	71.9 (378)	77.2 (421)	76.2 (523)	77.4 (89)	69.0 (187)
Occasional use	16.4 (176)	17.3 (91)	15.6 (85)	17.1 (117)	17.4 (20)	14.4 (39)
Regular use	9.1 (97)	10.8 (57)	7.2 (40)	6.7 (46)	5.2 (6)	16.6 (45)
Instagram^b
No use	27.6 (296)	32.5 (171)	22.9 (125)	24.9 (170)	30.5 (35)	33.6 (91)
Occasional use	16.1 (173)	16.7 (88)	15.6 (85)	16.0 (110)	21.7 (25)	14.0 (38)
Regular use	56.3 (603)	50.8 (267)	61.5 (336)	59.1 (406)	47.8 (55)	52.4 (142)
Snapchat^b
No use	29.8 (319)	38.2 (201)	21.7 (118)	29.9 (205)	30.3 (35)	29.2 (79)
Occasional use	11.4 (122)	13.5 (71)	9.4 (51)	9.0 (62)	14.7 (17)	15.9 (43)
Regular use	58.8 (631)	48.3 (254)	68.9 (377)	61.1 (419)	55.0 (63)	55.0 (149)

Significant race/ethnicity difference. ^bSignificant gender difference.

MANCOVA Models Predicting Externalizing Behaviors From Social Media Use

Facebook

The overall multivariate effect was significant, F(2, 708) = 20.47, p < .001. In addition, a significant main effect was found for gender, F(2, 708) = 4.75, p < .01. In regard to delinquent behaviors, boys (M = 4.86, SD = 10.86) had significantly higher scores than girls (M = 2.47, SD = 6.13), F(1, 709) = 7.27, p < .01. In addition, boys had significantly lower behavioral conduct scores (M = 2.99, SD = 0.63) than did girls (M = 3.21, SD = 0.60), F(1, 709) = 5.97, p < .05. A significant main effect also was observed for race/ethnicity, F(4, 1416) = 2.40, p < .05. Race/ethnicity differences were found for both the DBS score, F(2, 709) = 3.21, p < .05, and the behavioral conduct scale score, F(2, 709) = 3.49, p < .05. More specifically, in comparison with White adolescents (M = 2.31, SD = 6.42), Hispanic adolescents (M = 7.46, SD = 13.80) had significantly higher delinquent behaviors scores, mean difference = .25, p < .05. In addition, Hispanic adolescents (M = 2.87, SD = 0.67) had significantly lower behavioral conduct scores than did White adolescents (M = 3.20, SD = 0.60), mean difference = −.19, p < .05. A significant main effect also was found for Facebook use, F(4, 1416) = 4.38, p < .01, indicating that Facebook users had relatively higher delinquent behaviors scores, F(2, 709) = 8.49, p < .001. As shown in Figure 1, both regular Facebook users and occasional Facebook users had higher mean levels of delinquent behaviors in comparison with nonusers (mean difference = .46, p < .001, and .26, p < .05, respectively).

Notably, there also was a significant interaction between gender and Facebook use, F(4, 1416) = 2.86, p < .05. As shown in Figure 2, boys who used Facebook regularly had the highest delinquent behaviors scores, whereas girls who did not use Facebook had the lowest delinquent behaviors scores, F(2, 709) = 4.01, p < .05.

Twitter

The overall multivariate effect was significant, F(2, 702) = 20.06, p < .001. In addition, a significant main effect was found for Twitter use, F(4, 1404) = 4.66, p < .01, indicating that Twitter users had relatively higher delinquent behaviors scores, F(2, 703) = 7.34, p < .01, and lower behavioral conduct scores, F(2, 703) = 5.61, p < .01. More specifically, as shown in Figure 1, regular Twitter users had higher mean levels of delinquent behaviors in comparison with both occasional users (mean difference = .50, p < .05) and nonusers (mean difference = .67, p < .001). In addition, regular Twitter users (M = 2.81, SD = 0.73) had lower levels of behavioral conduct in comparison with both occasional users (M = 3.10, SD = 0.61) and nonusers (M = 3.14, SD = 0.61), mean difference = −.37, p < .05, and −.45, p < .01, respectively. Occasional Twitter users and nonusers did not differ from one another in regard to DBS scores and behavioral conduct scale scores.

Instagram

The overall multivariate effect was significant, F(2, 705) = 21.50, p < .001. A significant main effect also was found for Instagram use, F(4, 1410) = 4.64, p < .01, indicating that Instagram users had relatively higher delinquent behaviors scores, F(2, 706) = 9.24, p < .001. As shown in Figure 1, regular Instagram users had higher mean levels of delinquent behaviors in comparison with both occasional users (mean difference = .26, p < .05) and nonusers (mean difference = .39, p < .001). Occasional Instagram users did not differ from nonusers.

There also was a significant interaction between race/ethnicity and Instagram use, F(8, 1410) = 2.52, p < .05. As shown in Figure 3, Hispanic adolescents who used Instagram regularly had the highest delinquent behaviors scores, whereas White adolescents who did not use Instagram had the lowest delinquent behaviors scores, F(4, 706) = 4.01, p < .01. Among regular Instagram users, Hispanic and Black adolescents reported significantly higher delinquent behaviors scores than White adolescents (mean difference = .34 and .45, respectively, ps < .001). However, among occasional Instagram users, Hispanic adolescents reported significantly higher delinquent behaviors scores than both Black and White adolescents (mean difference = .59 and .44, respectively, ps < .01), whereas there were no racial/ethnic differences in delinquent behaviors scores among adolescents who did not use Instagram (ps > .05).

Snapchat

The overall multivariate effect was significant, F(2, 703) = 23.10, p < .001. In addition, a significant main effect was found for Snapchat use, F(4, 1406) = 2.81, p < .05, indicating that Snapchat users had relatively higher delinquent behaviors scores, F(2, 704) = 5.28, p < .01. More specifically, as shown in Figure 1, regular Snapchat users had higher mean levels of delinquent behaviors in comparison with nonusers (mean difference = .29, p < .01). Regular Snapchat users did not differ from occasional Snapchat users in regard to delinquent behaviors scores or behavioral conduct scores.

Discussion

Given the salience of social media as a developmental context in the lives of early adolescents, it is critical to elucidate the relationship between social media use and psychological adjustment. This study, therefore, examined the relationship between social media use and externalizing behaviors in a large, diverse community sample of early adolescents. Most of the adolescents reported that they did not use Facebook (73%) or Twitter (75%). In contrast, the majority of the adolescents reported using the image-based social media platforms of Instagram (72%) and Snapchat (70%), with remarkably more than half using these apps on a daily basis. These findings suggest that the trends of declining Facebook use and increasing Instagram use and Snapchat use among older adolescents and emerging adults (Associated Press-NORC, 2017; Lenhart, Duggan, et al., 2015; Molina, 2016) extend to early adolescence. In addition, it is possible that the image-centric nature of both Instagram and Snapchat, as compared to the text-focused commentary of Facebook and Twitter, appeals more to young adolescents. The many photos and video filters and retouching capabilities of Instagram and Snapchat also may serve as a particularly attractive medium for self-expression and entertainment among early adolescents exploring emerging identities and interests. The differential use of social media apps during early adolescence underscores the importance of research on early adolescents to focus on separate apps, rather than overall social media use.

In line with social developmental theories of adolescents suggesting that adolescent girls adopt greater interpersonal orientations focused on communication relative to boys (Rose & Rudolph, 2006), there were significant gender differences in the use of both Instagram and Snapchat. Early adolescent girls were more likely to report regular use of both apps relative to boys (Instagram: 62% vs. 51%; Snapchat: 69% vs. 48%), which is consistent with gender differences identified in a nationally representative sample of older adolescents (Associated Press-NORC, 2017). This pattern may stem from gender differences in social media usage motivations, as adolescent girls have been shown to use social media more than boys for the purposes of communication with peers and also are more likely to engage in self-disclosure via posting videos and photos (Barker, 2009; Valkenburg et al., 2011).

Hypotheses regarding racial/ethnic differences in social media use were partially supported. Neither Instagram nor Snapchat use differed by racial/ethnic group, perhaps because the salience of these platforms for communication and connection with peers is ubiquitous across all early adolescents, regardless of their racial/ethnic background. However, consistent with hypotheses regarding racial/ethnic identity expression and connection via social media (Chan, 2017), there were significant race/ethnicity differences found for Facebook and Twitter use. Specifically, a greater proportion of Hispanic adolescents used Facebook than Black adolescents, both of whom were more likely to use Facebook than non-Hispanic White adolescents. Hispanic adolescents also were more likely to use Twitter than non-Hispanic White adolescents. These findings contrast prior work focusing on older adolescents suggesting that Hispanic, Black, and White youth are equally likely to use Facebook and Twitter (Lenhart, Duggan, et al., 2015). However, it is important to note that the Lenhart, Duggan, et al. (2015) study assessed social media usage dichotmously (yes/no), which may have obscured racial/ethnic differences among those who frequently use Facebook and Twitter. Furthermore, it also is possible that racial/ethnic differences in Facebook and Twitter use may have emerged in the current study due to recent sociopolitical climate shifts within the United States that have spurred mass civic engagement and political participation of racial/ethnic minority youth via social media (Middaugh et al., 2017). In addition, racial/ethnic differences in social connectedness also may in part contribute to the observed patterns in our study, as Hispanic adolescents have been found to be more likely than White adolescents to feel more connected to their friends’ feelings and what is happening with friends through using social media (Lenhart, Smith, et al., 2015).

Consistent with prior work on older adolescents and adults examining overall social media use and Facebook use (Carpenter, 2012; Galica et al., 2017; Rosen et al., 2013), in our sample of early adolescents, regular use (defined as at least once a day) of Facebook, Twitter, Instagram, and Snapchat each was associated with significantly higher delinquent behaviors scores in comparison with nonuse of these social media apps. These findings support hypotheses made by the transformational framework of social media and peer experiences (Nesi et al., 2018), which proposes that peer influence processes are amplified in the social media context, and therefore, social media use may confer heightened risk for externalizing behaviors across all social media apps (Brechwald & Prinstein, 2011; Cotter & Smokowski, 2016). In addition, they are consistent with social compensation theory (Campbell et al., 2006), which suggests that youth with preexisting externalizing behavior problems tend to have poor social skills and are attracted to social media and use it more frequently because it allows for anonymity (Mikami et al., 2010; Valkenburg et al., 2005). Alternatively, according to the “poor-get-poorer” hypothesis (Mesquita, 2015; Selfhout et al., 2009), daily social media use may exacerbate preexisting externalizing behavior problems in vulnerable youth when social media–based interactions remain unsatisfying and isolating.

Of note, gender differences in the relationship between Facebook use and externalizing behaviors were observed, with boys who used Facebook on a daily basis reporting the highest levels of delinquent behaviors and girls who did not use Facebook reporting the lowest levels of delinquent behaviors. These findings support the tenets of social role theory (Eagly, 1987; Eagly et al., 2000), which proposes that adolescent boys may be more vulnerable than adolescent girls to peer influence processes regarding delinquent and aggressive, externalizing behaviors because these behaviors align with male gender roles. As such, exposure to peers’ content related to delinquent behaviors and poor behavioral conduct on Facebook may be especially rewarding to boys and increase their motivations to engage in externalizing behaviors to adhere to the perceived norms of their male peers. Of note, adolescent boys prone to externalizing problems tend to have more friends who have externalizing behavior problems and are more likely to be rejected by nondeviant peers (Chen et al., 2015; Van Lier et al., 2005). Therefore, it also is plausible that boys with preexisting externalizing behavior problems, who are regular Facebook users, have more online exposure to peer externalizing behavioral problems and peer rejection such as cyberbullying, which in turn may promote further externalizing behaviors in a maladaptive cycle.

Notably, the relationship between Instagram use and externalizing behaviors differed by race/ethnicity among early adolescents in this study. Hispanic adolescents who used Instagram regularly (daily) reported the highest levels of delinquent behaviors, whereas White adolescents who did not use Instagram reported the lowest levels of delinquent behaviors. Among regular Instagram users, Hispanic and Black adolescents reported significantly higher levels of delinquent behaviors than White adolescents. These findings are consistent with theories suggesting that Hispanic and Black youth may be exposed to online racial/ethnic discrimination via social media use that subsequently contributes to distress and delinquent behaviors (Kahn et al., 2013; Tynes et al., 2008, 2012; Umaña-Taylor et al., 2015). Interestingly, among occasional Instagram users, Hispanic adolescents reported substantially higher DBS scores than both Black and White adolescents, whereas there were no racial/ethnic differences in delinquent behaviors among adolescents who did not use Instagram. It is possible that Instagram use may exacerbate bicultural stress, intergenerational conflict, and family disconnection among Hispanic adolescents, which may contribute to worsening delinquent behaviors relative to Black and White adolescents (Forster et al., 2015). Viewing politically oriented Instagram posts promoting anti-immigration sentiments also may lead Hispanic adolescents to believe that their communities are hostile and unsupportive of their families, which has been shown to uniquely promote delinquent behaviors in Hispanic adolescents (Forster et al., 2015).

Study Limitations

The current study extends the extant literature on social media and externalizing behaviors in several important ways, including the focus on early adolescents, the examination of four popular social media apps, and the investigation of racial/ethnic differences. Additional study strengths include the large, diverse community sample of early adolescent girls and boys, and the use of well-validated measures of delinquent behaviors and behavioral conduct. However, the cross-sectional nature of the design limits inferences that may be made. Future longitudinal studies are needed so that the directionality of the relationship between social media use and externalizing behaviors may be examined. Longitudinal research also would allow for the examination of underlying mechanisms involved. Notably, the reliance on adolescent self-reports may have contributed to shared method variance, potentially biasing the findings. However, research has shown that adolescents are accurate reporters of their own behaviors (Deković et al., 2006). Nonetheless, it would be informative for future studies to include other reporters (e.g., teachers, parents, peers). Importantly, the focus on time spent using social media represents an important first step toward understanding the links between social media use and externalizing behaviors. However, future research is needed to capture other salient aspects of social media use that may be related to psychological adjustment, such as motivations, characteristics of adolescents’ online social networks (e.g., number of followers), and how adolescents engage with social media apps (e.g., passive viewing, actively posting content, making plans with friends). Moreover, additional research using other research designs would be informative (e.g., laboratory experiments that manipulate attributes of social media apps, ecological momentary assessment studies that provide next-to-real-time assessments).

Conclusion

Findings from this study indicate that regular, daily social media use is positively associated with externalizing behaviors during early adolescence. In addition, findings suggest the need to take gender and race/ethnicity into account when understanding the relationship between social media use and externalizing behavior. Results from this study suggest that early adolescents prone to externalizing behavior problems may benefit from reducing their social media use. Moreover, the findings indicate that it would be important for intervention programs to target social media use and psychosocial factors related to at-risk use.

Footnotes

Acknowledgements

The involvement of all the students who participated in the study is greatly appreciated. Special thanks go to members of the Predictors of ANxiety and Depression during Adolescence project staff, especially Sonja Gagnon, Courtney Lincoln, Emily Simpson, Melanie Klinck, Victoria Galica, and Kaitlin Flannery.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the Alvord Foundation.

ORCID iD

Christine McCauley Ohannessian

Author Biographies

Christine McCauley Ohannessian is a professor of pediatrics and psychiatry at the University of Connecticut School of Medicine and the director of the Center for Behavioral Health at Connecticut Children’s Medical Center. Dr. Ohannessian’s research interests focus on psychological problems, substance use, and social media use during adolescence and emerging adulthood.

Anna Vannucci is a doctoral student at Columbia University. Her research interests primarily focus on understanding how interactions among biology, behavior, and the environment influence the development of disordered eating and related psychiatric and medical comorbidities across childhood and adolescence.

References

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (DSM-5®). American Psychiatric Publishing.

Associated Press-NORC. (2017). Summary of findings: Snapchat and Instagram are the most popular social network platforms among American teens. http://www.apnorc.org/projects/Pages/HTML%20Reports/instagram-and-snapchat-are-most-popular-social-networks-for-teens.aspx#snapchat-and-instagram-are-the-most-popular

Barker

(2009). Older adolescents’ motivations for social network site use: The influence of gender, group identity, and collective self-esteem. CyberPsychology & Behavior, 12(2), 209–213.

Barry

C. T.

Sidoti

C. L.

Briggs

S. M.

Reiter

S. R.

Lindsey

R. A.

(2017). Adolescent social media use and mental health from adolescent and parent perspectives. Journal of Adolescence, 61, 1–11.

Brechwald

W. A.

Prinstein

M. J.

(2011). Beyond homophily: A decade of advances in understanding peer influence processes. Journal of Research on Adolescence, 21(1), 166–179.

Bronfenbrenner

(2005). Making human beings human: Bioecological perspectives on human development. SAGE.

Bronfenbrenner

Morris

P. A.

(2006). The bioecological model of human development. In Lerner

R. M.

Damon

(Eds.), Handbook of child psychology (6th ed., Vol. 1, pp. 793–828). John Wiley.

Campbell

A. J.

Cumming

S. R.

Hughes

(2006). Internet use by the socially fearful: Addiction or therapy? CyberPsychology & Behavior, 9(1), 69–81.

Carpenter

C. J.

(2012). Narcissism on Facebook: Self-promotional and anti-social behavior. Personality and Individual Differences, 52(4), 482–486.

10.

Chan

(2017). Racial identity in online spaces: Social media’s impact on students of color. Journal of Student Affairs Research and Practice, 54(2), 163–174.

11.

Chen

Drabick

D. A.

Burgers

D. E.

(2015). A developmental perspective on peer rejection, deviant peer affiliation, and conduct problems among youth. Child Psychiatry & Human Development, 46(6), 823–838.

12.

Common Sense Media. (2015). The Common Sense Census: Media used by tweens and teens. Common Sense. https://www.commonsensemedia.org/sites/default/files/uploads/research/census_researchreport.pdf

13.

Common Sense Media. (2019). The Common Sense Census: Media used by tweens and teens. Commons Sense. https://www.commonsensemedia.org/Media-use-by-tweens-and-teens-2019-infographic

14.

Cotter

K. L.

Smokowski

P. R.

(2016). Perceived peer delinquency and externalizing behavior among rural youth: The role of descriptive norms and internalizing symptoms. Journal of Youth and Adolescence, 45(3), 520–531.

15.

Deković

ten Have

Vollebergh

W. A.

Pels

Oosterwegel

Wissink

I. B.

. . . Ormel

(2006). The cross-cultural equivalence of parental rearing measure: EMBU-C. European Journal of Psychological Assessment, 22(2), 85–91.

16.

Dredge

Gleeson

J. F.

de la Piedad Garcia

(2014). Risk factors associated with impact severity of cyberbullying victimization: A qualitative study of adolescent online social networking. Cyberpsychology, Behavior, and Social Networking, 17(5), 287–291.

17.

Duggan

Ellison

N. B.

Lampe

Lenhart

Madden

(2015). Social media update 2014. Pew Research Center. http://www.pewinternet.org/2015/01/09/social-media-update-2014/

18.

Eagly

A. H.

(1987). Sex differences in social behavior: A social-role interpretation. Lawrence Erlbaum.

19.

Eagly

A. H.

Wood

Diekman

A. B.

(2000). Social role theory of sex differences and similarities: A current appraisal. In Eckes

Trautner

H. M.

(Eds.), Developmental social psychology of gender (pp. 123–174). Lawrence Erlbaum.

20.

Elsaesser

Russell

Ohannessian

C. M.

Patton

(2017). Parenting in a digital age: A review of parents’ role in preventing adolescent cyberbullying. Aggression and Violent Behavior, 35, 62–72.

21.

Forster

Grigsby

Soto

D. W.

Schwartz

S. J.

Unger

J. B.

(2015). The role of bicultural stress and perceived context of reception in the expression of aggression and rule breaking behaviors among recent-immigrant Hispanic youth. Journal of Interpersonal Violence, 30(11), 1807–1827.

22.

Galica

V. L.

Vannucci

Flannery

K. M.

Ohannessian

C. M.

(2017). Social media use and conduct problems in emerging adults. Cyberpsychology, Behavior, and Social Networking, 20(7), 448–452.

23.

Gault-Sherman

(2013). The gender gap in delinquency: Does SES matter? Deviant Behavior, 34(4), 255–273.

24.

Harter

(2012). Self-perception profile for adolescents: Manual and questionnaires. University of Denver.

25.

Huang

G. C.

Unger

J. B.

Soto

Fujimoto

Pentz

M. A.

Jordan-Marsh

Valente

T. W.

(2014). Peer influences: The impact of online and offline friendship networks on adolescent smoking and alcohol use. Journal of Adolescent Health, 54(5), 508–514.

26.

Ilakkuvan

Johnson

Villanti

A. C.

Evans

W. D.

Turner

(2019). Patterns of social media use and their relationship to health risks among young adults. Journal of Adolescent Health, 64(2), 158–164.

27.

Kahn

K. B.

Spencer

Glaser

(2013). Online prejudice and discrimination: From dating to hating. In Amichai-Hamburger

(Ed.), The social net: Understanding our online behavior (Vol. 2, pp. 201–219). Oxford University Press.

28.

Kaplan

A. M.

Haenlein

(2010). Users of the world, unite! The challenges and opportunities of social media. Business Horizons, 53(1), 59–68.

29.

Lenhart

Duggan

Perrin

Stepler

Rainie

Parker

(2015). Teens, social media & technology overview 2015. http://www.pewinternet.org/2015/04/09/teens-social-media-technology-2015/

30.

Lenhart

Smith

Anderson

Duggan

Perrin

(2015). Teens, technology and friendships. http://www.pewinternet.org/2015/08/06/teens-technology-and-friendships/

31.

Lerner

R. M.

Agans

J. P.

DeSouza

L. M.

Hershberg

R. M.

(2014). Developmental science in 2025: A predictive review. Research in Human Development, 11(4), 255–272.

32.

Lerner

R. M.

Johnson

S. K.

Buckingham

M. H.

(2015). Relational developmental systems-based theories and the study of children and families: Lerner and Spanier (1978) revisited. Journal of Family Theory & Review, 7(2), 83–104.

33.

Liu

(2004). Childhood externalizing behavior: Theory and implications. Journal of Child and Adolescent Psychiatric Nursing, 17(3), 93–103.

34.

Madden

Lenhart

Cortesi

Gasser

Duggan

Smith

Beaton

(2013). Teens, social media, and privacy. https://www.pewresearch.org/internet/2013/05/21/teens-social-media-and-privacy/

35.

McCrae

Gettings

Purssell

(2017). Social media and depressive symptoms in childhood and adolescence: A systematic review. Adolescent Research Review, 2(4), 315–330.

36.

Mesquita

(2015). Human behavior, psychology, and social interaction in the digital era. IGI Global.

37.

Middaugh

Clark

L. S.

Ballard

P. J.

(2017). Digital media, participatory politics, and positive youth development. Pediatrics, 140(Suppl. 2), S127–S131.

38.

Mikami

A. Y.

Szwedo

D. E.

Allen

J. P.

Evans

M. A.

Hare

A. L.

(2010). Adolescent peer relationships and behavior problems predict young adults’ communication on social networking websites. Developmental Psychology, 46(1), 46–56.

39.

Molina

(2016, April 14). Teens love Snapchat. Also Instagram. USA Today. https://www.usatoday.com/story/tech/news/2016/04/14/survey-snapchat-most-popular-app-among-teens/83021810/

40.

Nesi

Choukas-Bradley

Prinstein

M. J.

(2018). Transformation of adolescent peer relations in the social media context: Part 2—Application to peer group processes and future directions for research. Clinical Child and Family Psychology Review, 21, 295–319. https://doi.org/10.1007/s10567-018-0262-0

41.

Ohannessian

C. M.

(2009). Media use and adolescent adjustment: An examination of gender differences. Journal of Child and Family Studies, 18(5), 582–593.

42.

Ohannessian

C. M.

Vannucci

Flannery

Khan

(2017). Social media use and substance use during emerging adulthood. Emerging Adulthood, 5(5), 364–370.

43.

O’Keeffe

G. S.

Clarke-Pearson

(2011). The impact of social media on children, adolescents, and families. Pediatrics, 127(4), 800–804.

44.

Patton

D. U.

Hong

J. S.

Ranney

Patel

Kelley

Eschmann

Washington

(2014). Social media as a vector for youth violence: A review of the literature. Computers in Human Behavior, 35, 548–553.

45.

Peter

Valkenburg

P. M.

(2013). The effects of internet communication on adolescents’ psychological development. In The international encyclopedia of media studies. https://doi.org/10.1002/9781444361506.wbiems136/full

46.

Pew Research Center. (2018). Teens, social media & technology 2018. http://www.pewinternet.org/2018/05/31/teens-social-media-technology-2018/

47.

Prinstein

M. J.

Boergers

Vernberg

E. M.

(2001). Overt and relational aggression in adolescents: Social-psychological adjustment of aggressors and victims. Journal of Clinical Child Psychology, 30(4), 479–491.

48.

Rose

A. J.

Rudolph

K. D.

(2006). A review of sex differences in peer relationship processes: Potential trade-offs for the emotional and behavioral development of girls and boys. Psychological Bulletin, 132(1), 98–131.

49.

Rosen

L. D.

Whaling

Rab

Carrier

L. M.

Cheever

N. A.

(2013). Is Facebook creating “iDisorders”? The link between clinical symptoms of psychiatric disorders and technology use, attitudes and anxiety. Computers in Human Behavior, 29(3), 1243–1254.

50.

Sanchagrin

Heimer

Paik

(2014). Adolescent delinquency, drinking, and smoking: Does the gender of friends matter? Youth & Society, 49(6), 805–826.

51.

Selfhout

M. H.

Branje

S. J.

Delsing

ter Bogt

T. F.

Meeus

W. H.

(2009). Different types of Internet use, depression, and social anxiety: The role of perceived friendship quality. Journal of Adolescence, 32(4), 819–833.

52.

Shapiro

L. A. S.

Margolin

(2014). Growing up wired: Social networking sites and adolescent psychosocial development. Clinical Child and Family Psychology Review, 17(1), 1–18.

53.

Sherman

L. E.

Payton

A. A.

Hernandez

L. M.

Greenfield

P. M.

Dapretto

(2016). The power of the like in adolescence: Effects of peer influence on neural and behavioral responses to social media. Psychological Science, 27(7), 1027–1035.

54.

Tynes

B. M.

Giang

M. T.

Williams

D. R.

Thompson

G. N.

(2008). Online racial discrimination and psychological adjustment among adolescents. Journal of Adolescent Health, 43(6), 565–569.

55.

Tynes

B. M.

Umana-Taylor

A. J.

Rose

C. A.

Lin

Anderson

C. J.

(2012). Online racial discrimination and the protective function of ethnic identity and self-esteem for African American adolescents. Developmental Psychology, 48(2), 343–355.

56.

Umaña-Taylor

A. J.

Quintana

S. M.

Lee

R. M.

Cross

W. E.

Rivas-Drake

Schwartz

S. J.

, . . . Ethnic and Racial Identity in the 21st Century Study Group. (2014). Ethnic and racial identity during adolescence and into young adulthood: An integrated conceptualization. Child Development, 85(1), 21–39.

57.

Umaña-Taylor

A. J.

Tynes

B. M.

Toomey

R. B.

Williams

D. R.

Mitchell

K. J.

(2015). Latino adolescents’ perceived discrimination in online and offline settings: An examination of cultural risk and protective factors. Developmental Psychology, 51(1), 87–100.

58.

U.S. Census Bureau. (2010). US Census Bureau news: Facts for features CB10-FF17. 1. https://www.pewresearch.org/fact-tank/2018/09/13/key-facts-about-young-latinos/

59.

Valkenburg

P. M.

Schouten

A. P.

Peter

(2005). Adolescents’ identity experiments on the Internet. New Media & Society, 7(3), 383–402.

60.

Valkenburg

P. M.

Sumter

S. R.

Peter

(2011). Gender differences in online and offline self-disclosure in pre-adolescence and adolescence. British Journal of Developmental Psychology, 29(2), 253–269.

61.

Van Lier

P. A.

Vitaro

Wanner

Vuijk

Crijnen

A. A

. (2005). Gender differences in developmental links among antisocial behavior, friends’ antisocial behavior, and peer rejection in childhood: Results from two cultures. Child Development, 76(4), 841–855.

62.

Vannucci

Flannery

K. M.

Ohannessian

C. M.

(2017). Social media use and anxiety in emerging adults. Journal of Affective Disorders, 207, 163–166.

63.

Vannucci

Ohannessian

C. M.

(2019). Social media use subgroups differentially predict psychosocial well-being during early adolescence. Journal of Youth and Adolescence, 48, 1469–1493. https://doi.org/10.1007/s10964-019-01060-9

64.

Wichstraum

(1995). Harter’s self-perception profile for adolescents: Reliability, validity, and evaluation of the question format. Journal of Personality Assessment, 65(1), 100–116.