Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Sep 1.
Published in final edited form as: Transl Issues Psychol Sci. 2016 Sep;2(3):227–237. doi: 10.1037/tps0000077

Adolescents’ Internalizing Symptoms as Predictors of the Content of Their Facebook Communication and Responses Received from Peers

Samuel E Ehrenreich 1, Marion K Underwood 2
PMCID: PMC5222594  NIHMSID: NIHMS804502  PMID: 28083544

Abstract

This research examined how adolescents’ internalizing symptoms, such as depression, anxiety, and loneliness, relate to the content of their Facebook communication and the responses they receive from peers on Facebook. Participants (n = 125, 56 female, age 18) reported on their internalizing symptoms in the summer following 12th grade, and downloaded an application to their Facebook account that stored the content of all of their Facebook communication to secure, online archive. Two months of participants’ status updates and comments and peers’ comments were coded for content. Relations between internalizing symptoms and Facebook communication differed for girls and boys. For girls, internalizing symptoms predicted several types of Facebook content: negative affect, somatic complaints and eliciting support. In contrast, internalizing symptoms were not related to boys’ Facebook posts. Relations between internalizing symptoms and peers’ responses on Facebook also differed by gender. For girls, internalizing symptoms positively predicted receiving more peer comments expressing negative affect, and peer responses offering support. For boys, internalizing symptoms did not predict any of the measured peer responses. These findings suggest that girls prone to internalizing symptoms use Facebook in ways that appear similar to co-rumination, by expressing problems to friends and receive possibly reinforcing feedback in return.

Adolescents enthusiastically consume social media (such as Facebook, Instagram, and Twitter) as important platforms for both social interaction and self-presentation (boyd, 2014; Lenhart et al., 2011). Adolescents report spending more time on social networking sites (SNS) than any other internet activity (Jelenchick, Eickhoff & Moreno, 2013). However communication exchanged over social media sites may be associated with depression and other internalizing symptoms (Deters & Mehl, 2013; O’Keeffe & Clarke-Pearson, 2011). Depression, anxiety, and somatic complaints increase across adolescence, particularly among girls (Angold, Costello & Worthman, 1998; Bansal, Goyal & Srivastava, 2009), thus understanding how social media communication may relate to adolescent depression and internalizing issues is vitally important.

By late adolescence, 20% of females and 7% of males exhibit at least some internalizing symptoms (Angold, Ekanli, Silberg, Eaves & Costello, 2002). Internalizing problems are comprised of symptoms that are experienced by the individual such as sadness, anxiety and loneliness (in contrast to externalizing problems, such as aggression and hyperactivity, which are aversive for others; Levesque, 2011). Adolescents may cultivate depression in each other via a depressive-contagion process, wherein youth who ruminate about their problems with peers are at risk of increasing their own and others’ internalizing symptoms (Prinstein, 2007; Rose, 2002). Given that social media is the preferred form of communication for adolescents with their peers (boyd, 2014), social media sites such as Facebook may be an important forum where adolescents exhibit and possibly promote internalizing symptoms. This study will examine how internalizing symptoms relate to adolescents’ own communication on Facebook, as well as to the responses they receive from peers.

Facebook Communication as a Predictor of Internalizing Symptoms

Facebook is the most popular social networking site in the world, boasting over 1.4 billion active users (Facebook.com). Nearly 75% of all teenagers between the ages of 13–17 report having an active Facebook account (Lenhart & Page, 2015). Although adolescents view Facebook as an important tool in their social lives, some types of Facebook activity may predict increases in depression, loneliness, and a negative self-image (Kross et al., 2013; O’Keeffe & Clark-Pearson, 2011). A recent clinical report released by the American Academy of Pediatrics identified “Facebook Depression” as a possible result of adolescents’ overexposure to Facebook (O’Keeffe & Clark-Patterson, 2011, p. 802). Facebook depression may occur as a result of adolescents viewing positive depictions of others’ lives and observing others’ social interactions in ways that may make them feel more negative about their own lives and social relationships. Although this initial report cited no empirical evidence that supported this claim (see Guernsey, 2014), subsequent research has examined the association between Facebook communication and internalizing symptoms with some conflicting results.

Some studies have not found support for a relation between Facebook usage and internalizing symptoms. Using an experience sampling method, college students reported their Facebook usage in real-time by responding to randomly sent text messages inquiring about their usage over one week (Jelenchick et al., 2013). Despite SNS’s being the most common Internet activity that participants engaged in, there was no correlation between time spent on SNS’s and depression. Another study of Filipino college students also found no significant correlation between time spent on Facebook and depression (Datu, Valdez & Datu, 2012).

Despite these initial null findings, several more recent studies suggest that there may indeed be a relation between internalizing symptoms and Facebook usage. In an experience sampling study, researchers randomly sent text messages inquiring about participants’ current affective state (“how do you feel”) as well as their Facebook usage. The amount of time they had spent using Facebook since the last text message-inquiry significantly predicted feeling worse and a decline in overall life satisfaction (Kross et al., 2013). Another study that examined the difference between actively engaging with Facebook versus passively viewing other people’s profiles found that actively interacting with others’ profiles on Facebook predicted depressed mood for boys, whereas passively viewing others’ Facebook profiles predicted depressed mood in girls (Frison & Eggermont, 2015). Using Facebook may also be associated with loneliness (Deters & Mehl, 2013; Kross et al., 2013), poor social adjustment (Yan & Brown, 2012), and negative affect (Reinecke & Trepte, 2014).

Internalizing Symptoms as a Predictor of Facebook Communication

Although the above studies suggest that increased Facebook usage may relate to the development or perpetuation of internalizing symptoms (Frison & Eggermont, 2015; Kross et al., 2013), they all presume that Facebook usage is the precursor to internalizing symptoms. Although this may indeed be a correct direction of effect, the possibility remains that internalizing symptoms may predict adolescents’ Facebook behaviors. Youth experiencing internalizing symptoms may use Facebook as a platform to discuss negative topics, present their depressed mood, express their anxieties, or talk about the physical symptoms such as lethargy or somatic complaints that often characterize internalizing disorders (American Psychiatric Association [APA], 2013).

This depressive Facebook communication may have a powerful effect on both the adolescents themselves and their broader peer groups. Co-rumination refers to the process of repeatedly discussing one’s problems with peers (Rose, 2002). Adolescents who co-ruminate with peers about their problems show increases in depression and anxiety over time (Schwartz-Mette & Rose, 2007). Furthermore, engaging in this co-rumination process is associated with increased depression and anxiety symptoms in the individual as well as in the co-ruminating peer (Prinstein, 2007; Rose, 2002; Schwartz-Mette & Rose, 2012).

Youth who are high on internalizing symptoms may use Facebook as a platform to engage in this co-ruminative process with peers, posting more negative content. Furthermore, if adolescents’ negative posts are acknowledged and encouraged by peers, posting negative content may be reinforced, which could exacerbate internalizing symptoms in youth already exhibiting depressive and anxious symptoms (Rose, 2002). However, examining if internalizing symptoms relates to posting negative content requires directly observing adolescents’ Facebook communication.

Very little research has examined how the actual content of Facebook relates to internalizing symptoms in adolescents. One of the few exceptions is a series of studies conducted by Moreno and her colleagues (Moreno, Christakis et al., 2011; Moreno, Jelenchick et al., 2011). By reviewing the Facebook profiles of college students whose accounts are set to be visible to the public, Moreno coded for posts containing discussion of depressive symptoms (such as sadness and lethargy.). Participants were then contacted and they provided self-report assessments of their depression symptoms. Those who posted about depressive symptoms on Facebook reported higher average depression scores than those who did not have any posts about depressive symptoms (Moreno, Christakis et al., 2011).

In a subsequent study, peers’ responses to posts about depressive symptoms significantly predicted increases in subsequent posts about depression (Moreno, Jelenchick et al., 2011). Every peer response that a depressive post received predicted a two-fold increase in subsequent posts about depression. Taken together, these findings suggest that individuals express their internalizing symptoms on Facebook, which in turn elicits a reinforcing response from the peer network. However these studies only examined the Facebook behavior of profiles that were publicly visible. Given that 60% of adolescents report setting their Facebook profiles to private (Madden et al., 2013), focusing exclusively on public content is a significant limitation. Further research is needed to examine how internalizing symptoms relate to Facebook content that may not be publicly visible.

The Current Study

The current study extends previous research in two primary ways. First, this study directly examines the content of Facebook communication sent and received by adolescents, instead of using self-reported assessments of the amount of Facebook usage (either using login frequency or number of minutes; Jelenchick et al., 2013; Kross et al, 2013; Reinecke & Trepte, 2014). Second, this study directly captured the Facebook communication of a typically developing sample of adolescents by obtaining permission to archive their Facebook communication. Facebook communication in this study was observed regardless of participants’ privacy settings.

With these methodological advancements, we hope to address the extent to which participants’ internalizing symptoms predict features of their own Facebook communication and the responses they receive from their peers. In line with previous research on depression and offline co-rumination (Rose, 2002; Schwartz-Mette & Rose, 2007), we hypothesized that high levels of internalizing symptoms would predict more negative and problem-focused Facebook communication—specifically more negative affect and less positive affect, and discussing somatic complaints and requesting support from peers more often.

Additionally, we examined whether internalizing symptoms were associated with peers’ responses via Facebook. Given that college students’ posts about depression were often met with supportive responses from the peer group (Moreno, Jelenchick et al., 2011), we predicted that internalizing symptoms would be positively associated with the overall number of comments received by peers, as well as peers’ positive affect and frequency of peers’ offering support, but negatively associated with negative affect in peers’ comments.

Finally, given that depression is more common among females during adolescence (APA, 2013; Angold et al., 2002), and females are more likely to both engage in co-rumination with their peers (Rose, 2002) and post on Facebook about depression (Moreno, Jelenchick et al., 2011), we examined gender interactions in all of the analyses. We predicted that the relation between internalizing symptoms and both adolescents’ own Facebook communication and their peers’ response would be stronger for girls.

Method

Participants

Participants were 125 18-year-old adolescents (56 female) who were part of a longer ongoing study examining the role of digital communication on peer relationships. The sample was racially diverse (2.4% Asian, 20.8% Black, 16.8% Hispanic, 52.8% White, and 7.2% of mixed race or identified as another race) and generally representative of the county in the southern United States where the sample was initially recruited (U.S. Census Bureau, n.d.). The families’ median reported income (“between $51,000 and $75,000”) was higher than the median family-income for the recruitment county ($48,000; U.S. Census Bureau, n.d.).

Procedures

During the summer following their 12th grade year, adolescents participated in research visits either in their home or in the laboratory, according to their preference. During these visits, participants completed a series of questionnaires assessing their social and psychological adjustment. Of interest to this study is the Child Behavior Checklist-Youth Self Report form (CBCL-YSR; Achenbach & Rescorla, 2001). After completing the surveys, participants were given an explanation of the observational component of the study and, if they consented, a software application was activated on the Facebook profile. Facebook communication exchanged in the fall, following the summer visits, was analyzed. Participants were compensated $100 for their time (given that this was the final scheduled visit of a longitudinal study, participants were compensated an additional $50, above their regular $50 payment, for participating in the entire 5 year-study). All procedures were reviewed and approved annually by the university’s Institutional Review Board.

Measures

Child Behavior Checklist – Youth Self Report (CBCL-YSR)

Participants’ self-reported experiences with anxiety, depression, and somatic complaints were assessed during the summer following their 12th grade year using the CBCL-YSR form (Achenbach & Rescorla, 2001). The CBCL-YSR is an empirically derived questionnaire comprised of 112 behaviors. Respondents rate each behavior on a scale ranging from 0 = “not true” to 2 = “very true or often true”. Of interest to this study is the broad band internalizing dimension which is comprised of 32 items assessing anxiety, depression and somatic complaints. Example items include “nervous, tense”, “fears doing bad”, withdrawn”, and “lacks energy”. The CBCL-YSR Internalizing dimension has been shown to correlate with parent ratings of internalizing symptoms and is correlated with clinical depression diagnoses (Achenbach & Rescorla, 2001). Internal consistency was strong in this sample (α=.86). Descriptive statistics are presented in Table 1.

Table 1.

Descriptive Statistics and Inter-Correlations for all Study Variables by Gender (values for females are above the diagonal and values for male-s are below the diagonal)

Variable 1 2 3 4 5 6 7 8 9 10
M 9.09 45.14 17.54 44.91 17.55 1.11 .34 .20 1.07 58.21
SD 7.73 84.15 31.43 64.84 26.70 3.27 .84 .48 2.63 86.10
1 Internalizing Symptoms (Participant) 9.29 7.79 - .06 .31* .12 .4** .48** .22 .17 .40** .23
2 Positive Affect (Participant)a 13.62 17.42 −.03 - .88** .94** .69** .23 .67** .70** .55** .89**
3 Negative Affect (Participant) 11.59 22.01 −.07 .81** - .92** .89** .54** .67** .76** .76** .93**
4 Positive Affect (Peer)a 21.96 27.84 .9** .74** - .83** .39** .69** .72** .71** .96**
5 Negative Affect (Peer) 14.62 23.02 .75** .85** .85** .85** - .63** .58** .64** .79** .91**
6 Somatic Complaints (Participant) .46 1.17 .03 .47** .55** .42** .49** - .33* .17 .81** .49**
7 Eliciting Support (Participant) .13 .38 −.21 .39** .62** .43** .48** .06 - .35** .57** .68**
8 Sexual Talk (Participant) .16 .72 −.14 .31** .42** .18 .21 .01 .28* - .42** .66**
9 Offering Support (Peer) .78 2.39 −.17 .60** .74** .66** .68** .40** .67** .09 - .76**
10 Comments Made by (Peer) 36.64 49.62 −.14 .78** .74** .89** .91** .42** .41** .22 .60** -
Open in a new tab

Note: Correlations for females are above the diagonal, males are below.

*

p < .05,

**

p < .01

a

Indicates a significant gender difference (p < .05)

Coded Facebook Communication

All communication the participant sent and received via their Facebook profile was captured using a software application developed by the company Arkovi (operated by RegEd). This application captures all communication, regardless of the device on which the participant accesses their Facebook profile (e.g. on a laptop versus on their smartphone). The application captured all communication that the participant posted, which included status updates, wallposts, pictures, and comments posted by the participant on their own status updates or pictures (jointly be referred to as “posts” throughout this manuscript). The application also captured comments that were made by peers on these posts. The application did not capture comments that the participant made to other people’s Facebook posts. All posts and comments exchanged during September and October of 2013 were analyzed for the presence of each of the communication features described below. Six participants did not have any communication archived during these two months because of errors in registering their accounts with Arkovi. February and March of 2013 were used for three of these participants, and June and July of 2013 were used for the remaining three participants. Accordingly, the analyzed Facebook content for all but three participants occurred at least a month after they rated their internalizing symptoms.

Transcripts of participants’ Facebook communication over the two months were assigned to a team of nine graduate and undergraduate research assistants who manually reviewed the transcripts in Excel files and indicated the presence of the communication features described below. Coders were trained on the coding procedure over the course of one semester until they consistently exceeded adequate reliability standards as outlined by Landis & Koch (1975; κ’s > .6 on all variables of interest). After completing the training, 20% of all transcripts were double coded to ensure continued reliability. Kappa scores for each code of interest are presented along with the description of the code. Every post and comment was coded using all available text and photo information. The unit of analyses for coding was a full post or comment and all codes presented below were coded independently. Thus a post or comment could contain any combination of the codes below (e.g., a status update could contain both positive and negative affect, only one of these, or neither). Descriptive statistics and correlations between variables are presented in Table 1.

Positive Affect

Both the participants’ Facebook posts and peers’ comments were coded for Positive Affect. Positive Affect was coded whenever the poster discussed a positive topic (e.g. offering congratulations, discussing a success) or posted in a way that conveyed a highly positive tone (e.g. “SO YOU THINK YOU CAN DANCE FINALE TONIGHT!!!!! ☺”; κ = .74).

Negative Affect

Both the participants’ Facebook posts and peers’ comments were coded for Negative Affect. Negative Affect was coded whenever the poster discussed a negative topic (e.g. failing a test, death of a relative) or posted in a way that conveyed a highly negative tone (e.g. “made it all the way home before realizing I left my keys in class…FML1!!”; κ = .76).

Somatic Complaints

Somatic Complaints were coded for any discussion about physical discomfort, including headaches, stomach aches, and exhaustion. Only Somatic Complaints that were in the participants’ posts were included in these analyses, not peer comments (κ = .72).

Eliciting Support

Eliciting Support was coded whenever the poster requested support or encouragement from their Facebook network. Simply posting about a problem (e.g. “I’m having a terrible day”) did not constituted Eliciting Support unless a request for support was specifically made (e.g. “I’m having a terrible day, I could really use a boost from someone”). Only Eliciting Support that was in the participants’ posts were included in these analyses, not peer comments (κ = .73).

Offering Support

Offering Support was coded when a peer offered help to a participant who was having difficulty (regardless of if the assistance was requested). This could include offering financial or academic help, providing relationship advice, or simply providing encouragement when someone was having trouble. Only Offering Support that was in the peer comments were included in these analyses, not in the participants’ posts (κ = .63).

Peer Comment Frequency

The total number of comments that peers left on participants’ posts during the two month period was summed and used for the Peer Comment Frequency.

Results

Analyses began with an examination of the distributions of all dependent variables. All variables were positively skewed and violated normality assumptions. To address this, a square root transformation was conducted on all dependent variables (a constant of 1 was added to all variables that contained zero as a possible value; see Kline, 2011). All regression analyses were conducted with girls coded as zero, and then repeated with boys coded as zero. All continuous variables that were included in interaction terms were centered so their means equaled 0.

There was a large amount of variability in the number of Facebook posts and comments made by the participants across two months, ranging from zero to 784, with an average of 60.18 posts during the two month period. Girls posted and commented significantly more (M = 88.25, Mdn = 47) over two months than boys did (M = 37.41, Mdn = 18; t(123) = −2.76, p < .05). Internalizing symptoms and all Facebook variables of interest were tested for gender differences; girls expressed more positive affect in their posts (t (123) = −3.04, p < .01) and received comments from peers with more positive affect (t(123) = −2.66, p < .01). No gender differences emerged for any of the other variables. Total number of posts was controlled for in all regression analyses. Cooke’s distances were calculated to identify possible outliers. Analyses were conducted both with outliers included and excluded. Because the results followed the same pattern, the analyses presented below include all participants.

Do internalizing symptoms predict posted Facebook content?

To examine whether internalizing symptoms relate to the Facebook content produced by adolescents and if this relation is moderated by gender, a series of regression models was conducted. Total Facebook posts, internalizing symptoms, gender, and the gender-by-internalizing symptoms interaction were all included as predictors of the four Facebook communication variables (positive affect, negative affect, somatic complaints, and eliciting support) in four separate models; results are presented in Table 2a. The total number of posts was a significant predictor of all outcomes: posting more frequently was associated with more positive and negative affect, and both somatic complaints and eliciting support. The main effect of gender significantly predicted both positive and negative affect: girls were more likely to include positive affect and less likely to include negative affect in their Facebook posts. The interaction between gender and internalizing symptoms significantly predicted somatic complaints and eliciting support, and approached significance for negative affect. The slopes for these three interactions are graphed in Figure 1. For girls, elevated internalizing symptoms were associated with posting more frequently about somatic complaints, requests for support, and posts containing negative affect (trend). In contrast, internalizing symptoms was not a significant predictor of these features in boys’ Facebook posts.

Table 2.

Table 2a. Gender, Internalizing Symptoms, and their Interaction as Predictors of Features of Adolescents’ Facebook Posts
Features of Adolescent Facebook Posts
Positive Affect Negative Affect Somatic Complaints Eliciting Support
b SE β b SE β b SE β b SE β
Total Posts .61** .03 .87 .43** .03 .83 .04** .01 .34 .02** .01 .36
Gender −.76** .25 −.12 .53* .24 .11 −.03 .09 −.03 −.02 .03 −.05
Internalizing symptoms .02 .16 .01 .47** .15 .23 .21** .05 .46 .05* .02 .30
Internalizing x Gender .15 .21 .04 −.39t .21 −.14 −.18* .07 −.29 −.07* .03 −.31
n 125 125 125 124
Table 2b. Gender, Internalizing Symptoms, and their Interaction as Predictors of Peers’ Facebook Responses
Features of Peer’s Facebook Comment Responses
# of Peers’ Comments Peers’ Positive Affect Peers’ Negative Affect Peers’ Offering Support
b SE β b SE β b SE β b SE β
Total Posts .65** .05 .79 .5** .04 .78 .36** .03 .71 .07** .01 .43
Gender .27 .44 .04 −.44 .33 −.08 .61* .30 .13 .05 .12 .04
Internalizing symptoms .27 .27 .08 −.08 .21 −.03 .42* .19 .21 .19** .07 .32
Internalizing x Gender −.49 .37 −.11 .02 .28 .01 −.52* .25 −.19 −.27** .10 −.33
n 125 125 125 125
Open in a new tab

Note: Gender is dummy coded (0 = Female, 1= Male). The table presenting effects for males is available by emailing the first author directly.

t

< .1,

*

p < .05,

**

p < .01

Figure 1.

Figure 1

Open in a new tab

Gender differences in the association between internalizing and features of adolescents Facebook posts

Do internalizing symptoms predict peers’ Facebook responses?

Next, to examine if internalizing symptoms were related to the responses that adolescents receive from their peers, four additional regression analyses were conducted. Total Facebook posts, gender, internalizing symptoms, and the gender-by-internalizing symptoms interaction term were all regressed on the number of comments made by peers, peers’ positive and negative affect, and peers’ offers of support. The results are presented in Table 2b.

The number of posts was a significant predictor of all peer responses. The main effect of gender was only a significant predictor of peers’ negative affect; boys elicited more negative affect from their peers than girls. There was a significant interaction between internalizing symptoms and gender predicting peers negative affect and peers’ offering support. The slope of each of these relations were graphed by gender (Figure 2). For girls, higher levels of internalizing symptoms were positively associated with their peers’ expressing more negative affect and more frequently offering support. In contrast, internalizing symptoms were not associated with any of these peer behaviors for boys.

Figure 2.

Figure 2

Open in a new tab

Gender differences in the association between internalizing symptoms and features of Peers’ Facebook Comments

Discussion

The results of this study demonstrated that internalizing symptoms related to Facebook communication differently for boys and girls. For girls, internalizing symptoms positively predicted posts containing negative affect, somatic complaints, and requests for support. In contrast, internalizing symptoms did not predict any of these features of boys’ Facebook posts. Girls with high levels of internalizing symptoms also received comments from peers with more negative affect and offers of support, whereas internalizing symptoms did not predict any particular peer response for boys. Although girls posted on Facebook more frequently than boys did, these findings held while controlling for total number of posts.

Girls’ internalizing symptoms predicted expressing negative affect on Facebook, and receiving potentially reinforcing responses from peers: more offers of support, along with more negative affect in the peer comments. Girls may use Facebook to engage in a process akin to co-rumination with their full digital peer networks (Rose, 2002). Although it is potentially positive that girls with internalizing symptoms express negative feelings on Facebook and ask and receive support, it is also possible that the affirmation and support that girls receive from peers on Facebook may reinforce depressive communication in a way that could lead to further increases in internalizing symptoms, just as co-rumination in offline peer interactions predicts increasing in internalizing symptoms for girls (Rose, 2002).

Whereas girls’ internalizing symptoms predicted features of their Facebook communication, internalizing symptoms did not predict features of boys’ Facebook communication. This distinction may be the result of boys’ attempt to conform to gender norms. Males in general are less likely to self-disclose to their peer group and to strangers (Dindia & Allen, 1992) and report using ruminative coping strategies less than girls (Broderick, 1998), thus it is in some ways not surprising that they may not be sharing their depressive feelings on Facebook. Another reason that boys may not express depressive feelings on Facebook is that they may not receive any response from their peers when they do. For boys, internalizing symptoms was not related to any of the measured peer responses. For a boy, being lonely, anxious and depressed does not seem to result in much peer response, support, and encouragement from peers on Facebook. Boys’ friends do not join them in Facebook co-rumination.

Taken together, these findings suggest that for girls, a process akin to co-rumination may occur on Facebook. Girls with high levels of internalizing symptoms both discuss negative and problem-focused topics on Facebook and also elicit negative, problem-focused communication from their peers. Although this may in fact strengthen the interpersonal bond between the adolescent and their peers through this interpersonal disclosure (such as peers offering support; Rose, 2002), peers’ responses likely also reinforce communication about these topics. Adolescents are often preoccupied with peer feedback on social media (boyd, 2014), and this may be particularly salient for peers’ response to posts expressing internalizing symptoms (Moreno, Jelenchick et al., 2011). Thus the increased responses from girls’ peers may come with some immediate comfort and perhaps solidify friendships, but these responses may also encourage posting about these topics in the future and thus actually increase internalizing symptoms over the long run (Rose, 2002). This may lead to a reciprocal relation between Facebook and internalizing symptoms; adolescents with internalizing symptoms co-ruminate about them with their peers on Facebook, which in turn results in increases in internalizing symptoms. Future research should examine internalizing symptoms as both a predictor and outcome of negative Facebook communication, and examine what specific features of peer responses may contribute to the development (or desistence) of internalizing symptoms.

All of these findings must be interpreted in light of methodological limitations. Although this sample was representative of the region of the United States where data collection occurred, these findings would not necessarily generalize to all adolescents. Furthermore, all participants had just finished their senior year of high school; it is possible that younger or older adolescents may communicate with peers on Facebook in different ways. Finally, the archiving application was not able to capture the number of “Likes” that adolescents’ received on their posts. Given the importance of “Likes” as reinforcement for adolescents on Facebook (boyd, 2014; Marwick & boyd, 2011), this is likely an important measure of peers’ reinforcement that cannot be examined in these data. Despite these shortcomings, this is one of the first studies to examine the relation between internalizing symptoms and the content of adolescents’ public and private Facebook communication. Given the increasing popularity of Facebook and other social networking sites in adolescents’ social lives, it is important that future research investigate how adolescents’ express their emotions on Facebook, how peers respond, and how psychological adjustment and Facebook communication are reciprocally related.

These findings are in line with a growing body of evidence that depression and negative emotional states can spread through social networks (Hill, Rand, Nowak & Christakis, 2010; Kramer, Guillory, & Hancock, 2014). Because co-rumination about problems predicts increases in not only one’s own depression, but also their peers’ as well (Rose, 2002) it is possible that negative and problem-focused communication on social media may in fact be associated with increases in depression and internalizing symptoms among the broader peer network (Kramer et al., 2014).

These findings have meaningful implications for a parents, educators, counselors, and policy makers. These results suggest possible policy changes for social media platforms. For example, Facebook has recently added features that allow individuals to report statuses indicating that a peer may be suicidal, and provide resources for how to offer that person help (Isaac, 2016). These findings are also relevant to parents of adolescents, suggesting parents can gain insight into adolescents’ emotional state by monitoring these communication platforms.

Given that adolescents with internalizing symptoms communicate on Facebook about more negative and problem-focused topics, adolescents’ social media communication may be an important warning sign for the development of clinical levels of depression. Therapists would be wise to probe adolescents’ social media behaviors for indications of depression, and develop therapies that address how problematic social media habits might lead to increased internalizing symptoms. Furthermore, evidence suggests that older adolescents are accepting of encouragement to seek help and counseling for their depressive posts, and may be open to being approached by counseling professionals based on their social media communication (Whitehill, Brockman, & Moreno, 2013). Thus although the findings presented here are cautionary, they also indicate that Facebook may be a meaningful method for identifying and contacting youth who are at risk for depression and other internalizing symptoms.

Acknowledgments

The preparation of this manuscript was supported by two grants from the Eunice Kennedy Schriver National Institute of Child Health and Human Development: R01 HD060995 and R21 HD072165.

Footnotes

1

“FML” is an acronym for “Fuck my life”

Contributor Information

Samuel E. Ehrenreich, The University of Texas at Dallas

Marion K. Underwood, The University of Texas at Dallas

References

  1. Achenbach TM, Rescorla LA. Manual for the ASEBA school-age forms and profiles. Burlington: University of Vermont, Research Center for Children, Youth, & Families; 2001. [Google Scholar]
  2. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders; DSM-5. Washington, DC: American Psychiatric Association; 2013. [Google Scholar]
  3. Angold A, Costello EJ, Worthman CM. Puberty and depression: The roles of age, pubertal status and pubertal timing. Psychological Medicine. 1998;28:51–61. doi: 10.1017/s003329179700593x. [DOI] [PubMed] [Google Scholar]
  4. Angold A, Erkanli A, Silberg J, Eaves L, Costello EJ. Depression scale scores in 8–17-year olds: Effects of age and gender. Journal of Child Psychology and Psychiatry. 2002;43:1052–1063. doi: 10.1111/1469-7610.00232. [DOI] [PubMed] [Google Scholar]
  5. Bansal V, Goyal S, Srivastava K. Study of prevalence of depression in adolescent students of a public school. Industrial Psychiatry Journal. 2009;18:43–46. doi: 10.4103/0972-6748.57859. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. boyd d. It’s Complicated: The social lives of networked teens. New Haven, CT: Yale University Press; 2014. [Google Scholar]
  7. Broderick PC. Early adolescent gender differences in the use of ruminative and distracting coping strategies. Journal of Early Adolescence. 1998;18:173–191. doi: 10.1177/0272431698018002003. [DOI] [Google Scholar]
  8. Datu JA, Valdez JP, Datu N. Does Facebook make us sad? Hunting relationship between Facebook use and depression among Filipino adolescents. International Journal of Research Studies in Educational Technology. 2012;1:83–91. [Google Scholar]
  9. Deteres FG, Mehl M. Does posting Facebook status updates increase or decrease loneliness? An online social networking experiment. Social Psychological & Personality Science. 2013;4:579–586. doi: 10.1177/1948550612469233. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Dindia K, Allen M. Sex differences in self-disclosure: A meta-analysis. Psychological Bulletin. 1998;112:106–124. doi: 10.1037/0033-2909.112.1.106. [DOI] [PubMed] [Google Scholar]
  11. Facebook.com. Company Info: Stats. Facebook Newsroom. n.d http://newsroom.fb.com/company-info/
  12. Frison E, Eggermont S. Exploring the relationships between different types of Facebook use, perceived online social support, and adolescents’ depressed mood. Social Science Computer Review. 2015:1–19. doi: 10.1177/0894439314567449. [DOI] [Google Scholar]
  13. Guernsey L. Garbled in translation: Getting media research to the press and public. Journal of Children and Media. 2014;8:871–97. doi: 10.1080/17482798.2014.863486. [DOI] [Google Scholar]
  14. Hill AL, Rand DG, Nowak MA, Christakis NA. Emotions as infectious diseases in a large social network: The SISa model. Proceedings of the Royal Soceity of Biological Sciences. 2010;277:3827–3835. doi: 10.1098/rspb.2010.1217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Issac M. Facebook offers tools for those who fear a friend may be suicidal. The New York Times. 2016 Jun 14; Retrieved from: http://nyti.ms/1XUsoUn.
  16. Jelenchick LA, Eickhoff JC, Moreno MA. “Facebook Depression?” Social Networking Site use and depression in older adolescents. Journal of Adolescent Health. 2013;52:128–130. doi: 10.1016/j.jadohealth.2012.05.008. [DOI] [PubMed] [Google Scholar]
  17. Kline RB. Principles and Practice of Structural Equation Modeling. 3. New York: Guilford Press; 2011. [Google Scholar]
  18. Kramer ADI, Guillory JE, Hancock JT. Experimental evidence of a massive-scale emotional contagion through social networks. Proceedings of the National Acadamy of Sciences of the United States of America. 2014;111:8788–8790. doi: 10.1073/pnas.1320040111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kross E, Verduyn P, Demiralp E, Park J, Lee DS, et al. Facebook use predicts declines in subjective well-being in young adults. PloS ONE. 2013;8:e69841. doi: 10.1371/journal.pone.0069841. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33:159–174. doi: http://www.jstor.org/stable/2529310. [PubMed] [Google Scholar]
  21. Lenhart A, Madden M, Smith A, Purcell K, Zickuhr K, Rainie L. Teens, Kindeness and Cruelty on Social Network Sites. 2011 Retrieved from http://www.pewinternet.org/2011/11/09/teens-kindness-and-cruelty-on-social-network-sites/
  22. Lenhart A, Page D. Teens, social media & technology overview 2015. 2015 Retrieved from http://www.pewinternet.org/2015/04/09/teens-social-media-technology-2015/
  23. Levesque RJR. Externalizing and internalizing symptoms. In: Levesque R, editor. Encyclopedia of Adolescence. 2011. pp. 903–905. [DOI] [Google Scholar]
  24. Madden M, Lenhart A, Cortesi S, Gasser U, Duggan M, Smith A, Beaton M. Teens, social media, and privacy. 2013 Retrieved from http://www.pewinternet.org/2013/05/21/teens-social-media-and-privacy/
  25. Marwick A, boyd d. The drama! Teen conflict, gossip, and bullying in networked publics. Paper presented at Oxford Internet Institute’s “A Decade in Internet Time: Symposium on the Dynamics of the Internet and Society; Oxford, UK. 2011. [Google Scholar]
  26. Moreno MA, Christakis DA, Egan KG, Jelenchick LA, Cox E, et al. A pilot evaluation of associations between displayed depression references on Facebook and self-reported depression using a clinical scale. Journal of Behavioral Health Services & Research. 2011;39:295–304. doi: 10.1007/s11414-011-9258-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Moreno MA, Jelenchick LA, Egan KG, Cox E, Young H, Gannon KE, Backer T. Feeling bad on Facebook: Depression disclosures by college students on a social networking site. Depression and Anxiety. 2011;28:447–455. doi: 10.1002/da.20805. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. O’Keeffe GS, Clarke-Pearson K. Clinical report—The impact of social media on children, adolescents, and families. Pediatrics. 2011;127:800–804. doi: 10.1542/peds.2011-0054. [DOI] [PubMed] [Google Scholar]
  29. Prinstein MJ. Moderators of peer contagion: A longitudinal examination of depression socialization between adolescents and their best friends. Journal of Clinical Child and Adolescent Psychology. 2007;36:159–170. doi: 10.1080/15374410701274934. [DOI] [PubMed] [Google Scholar]
  30. Reinecke L, Trepte S. Authenticity and well-being on social network sites: A two-wave longitudinal study on the effects of online authenticity and the positivity bias in SNS communication. Computers in Human Behavior. 2014;30:95–102. doi: 10.1016/j.chb.2013.07.030. [DOI] [Google Scholar]
  31. Schwartz-Mette RA, Rose AJ. Co-rumination mediates contagion of internalizing symptoms within youths’ friendships. Developmental Psychology. 2012;48:1355–1365. doi: 10.1037/a0027484. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Shelton AK, Skalski P. Blinded by the light: Illuminating the dark side of social network use through content analysis. Computers in Human Behavior. 2014;33:339–348. doi: 10.1016/j.chb.2013.08.017. [DOI] [Google Scholar]
  33. U.S. Census Bureau. Small Area Income and Poverty Estimates. n.d Retreived April 7, 2015, from http://www.census.gov/did/www/saipe/data/interactive/#view=StateAndCounty&utilBtn=&yLB=1&stLB=44&cLB=57&dLB=0&gLB=0&usSts_cbSelected=false&usTot_cbSelected=true&stateTot_cbSelected=true&pLB=4&multiYearSelected=false&multiYearAlertFlag=false&prStateFlag=false&invalidSDYearsFlag=false.
  34. Whilehill JM, Brockman LN, Moreno MA. “Just Talk to Me”: Communicating with college students about depression disclosures on Facebook. Journal of Adolescent Health Care. 2013;52:122–127. doi: 10.1016/j.jadohealth.2012.09.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Yang C, Brown BB. Motives for using Facebook, patterns of Facebook activities, and late adolescents’ social adjustment to college. Journal of Youth and Adolescence. 2013;42:403–416. doi: 10.1007/s10964-012-9836-x. [DOI] [PubMed] [Google Scholar]

RESOURCES