CYBERPSYCHOLOGY, BEHAVIOR, AND SOCIAL NETWORKING

Volume 20, Number 10, 2017 ORIGINAL ARTICLES

ª Mary Ann Liebert, Inc.
DOI: 10.1089/cyber.2017.0156

Browsing, Posting, and Liking on Instagram:

The Reciprocal Relationships Between Different Types
of Instagram Use and Adolescents’ Depressed Mood
Downloaded by Gothenburg University Library from online.liebertpub.com at 10/18/17. For personal use only.

Eline Frison, PhD, and Steven Eggermont, PhD

Abstract

Although studies have shown that Instagram use and young adults’ mental health are cross-sectionally asso-
ciated, longitudinal evidence is lacking. In addition, no study thus far examined this association, or the reverse,
among adolescents. To address these gaps, we set up a longitudinal panel study among 12- to 19-year-old
Flemish adolescents to investigate the reciprocal relationships between different types of Instagram use and
depressed mood. Self-report data from 671 adolescent Instagram users (61% girls; MAge = 14.96; SD = 1.29)
were used to examine our research question and test our hypotheses. Structural equation modeling showed that
Instagram browsing at Time 1 was related to increases in adolescents’ depressed mood at Time 2. In addition,
adolescents’ depressed mood at Time 1 was related to increases in Instagram posting at Time 2. These
relationships were similar among boys and girls. Potential explanations for the study findings and suggestions
for future research are discussed.

Keywords: Instagram, browsing, posting, liking, depressed mood, adolescence

Introduction by investigating the role of Instagram as a predictor of de-

pressed mood among adolescents. As prior studies mainly
focused on Facebook6 or were correlational in nature,7 the
T he past decade, social networking sites (SNSs) in-
creasingly gained popularity among youth.1 Mobile de-
vices, such as smartphones, have particularly driven this
present study is believed to offer an important contribution to
the existing literature.
growth.2 Although Facebook is still the most popular SNS Second, although mood management theory8,9 gives us
among adolescents, it is facing more and more competition reasons to expect that depressed mood may motivate ado-
from other, fast-growing SNSs such as Instagram; recent lescents’ Instagram use, no attention in the literature has yet
numbers show that 52% of American adolescents1 and 60% been given to this reverse relationship. However, it is im-
of Flemish adolescents3 report using Instagram. portant to increase insight into the predictors of specific types
However, during adolescence, teens are not only starting of Instagram use, given the potential detrimental outcomes of
to use social media, they are also facing various develop- using SNSs.10 The second aim of this study is therefore to fill
mental changes, such as the struggle for independence and this lacuna by exploring the role of depressed mood as a
autonomy.4 As a result, adolescents’ mental health is par- predictor of adolescents’ Instagram use.
ticularly at risk during this stage of life; 13% of European SNSs, however, offer their users a wide range of activities.
teens report symptoms of anxiety and depression.5 Scholars As a result, various scholars classified users’ SNS activities
therefore agree that it is highly important to identify risk into different categories. For instance, Burke et al.11 differed
factors of depressive symptoms in adolescence. between three types of SNS use: passive consumption (e.g.,
Despite the scholarly call to detect factors that may con- browsing through SNS content), broadcasting (e.g., posting
tribute to this increase in adolescents’ depressive symptoms, SNS content), and directed communication (e.g., liking SNS
and the growing popularity of Instagram especially among content). In line with this categorization, the present study
adolescents, up until now it is unclear whether using In- differs among Instagram browsing, posting, and liking. Al-
stagram may affect adolescents’ depressed mood over time. though previous studies mainly looked at the influence of
The first aim of this study is therefore to fill this critical gap SNS browsing12 and posting,13 we believe that it is equally

Leuven School for Mass Communication Research, Faculty of Social Sciences, KU Leuven, Leuven, Belgium.

603
 604 FRISON AND EGGERMONT

important to examine the impact of Instagram liking for at we hypothesize that Instagram browsing (* passive con-
least two reasons. First, Instagram users can easily express a sumption) at Time 1 will relate to increases in adolescents’
positive attitude toward any content by simply clicking the depressed mood at Time 2 (H1), whereas Instagram posting
Like button. As a result, liking is a popular SNS activity.14 In (* broadcasting) and liking (* directed communication) at
addition, Lee et al.15 recently revealed that participants who Time 1 will relate to decreases in adolescents’ depressed
more frequently use the Like button reported greater bonding mood at Time 2 (H2).
social capital, suggesting that liking may be beneficial for
users’ mental health. Adolescents’ depressed mood and different types
To sum up, the present study aims to enhance our under- of Instagram use
standing of the reciprocal relationships between different
types of Instagram use (i.e., browsing, posting, and liking) Various studies have already investigated loneliness and
Downloaded by Gothenburg University Library from online.liebertpub.com at 10/18/17. For personal use only.

and adolescents’ depressed mood. depressive symptoms as predictors of adolescents’ Facebook

use.26–30 For instance, Scherr and Brunet30 found that indi-
viduals with more depressive symptoms showed more Fa-
Different types of Instagram use and adolescents’ cebook activity, that is, they posted more status updates on
depressed mood Facebook. However, while there is growing amount of in-
The majority of prior research examined the relationship terest in this relationship, no study thus far explored the
between general Facebook use (e.g., time spent on Face- predictive influence of depressed mood on the use of In-
book) and young people’s mental health. These studies, stagram, particularly in an adolescent sample. Nonetheless,
however, revealed both positive16 and negative associa- we have reasons to expect that depressed mood may predict
tions.17 Regarding the impact of general Instagram use, adolescents’ Instagram use.
similar inconsistent findings have been reported; one study The idea that individuals’ moods may drive their media
revealed a marginal positive association between Instagram use is grounded in Mood Management Theory.8,9 This theory
use and depressive symptoms in 18–29 year olds,6 whereas suggests that individuals may select certain media to distract
another study found no relationship between Instagram in- themselves from negative moods or to maintain positive
tensity and young people’s self-esteem.18 Scholars argue that moods. In line with this theory, various empirical studies
this variability of findings might be explained by the fact that revealed that individuals in negative moods seek media that
the relationship between online social networking and indi- are capable to improve their mood.31–33 Recently, however,
viduals’ mental health is complex and involves various scholars34 argue that SNSs can also be used for purposes of
factors that may mediate or moderate this relationship. mood management. More specifically, Johnson and Knobloch-
One factor that may explain this relationship is type of Westerwick34 found that participants appeared motivated to
SNS use. Various studies indeed found that directed com- repair their affective states through selective exposure to
munication and broadcasting of SNS content relates to pos- downward comparisons on SNSs that could restore their mood.
itive mental health outcomes, whereas passive consumption Against this background, this study will explore whether ad-
of SNS content relates to negative ones.6 Directed commu- olescents’ depressed mood at Time 1 is predictive of different
nication and broadcasting may be beneficial for youngsters’ types of Instagram use at Time 2 (RQ1).
mental health, because SNS use involving social interactions The current study hereby extends prior research in three
may result in social benefits, such as increases in social sup- important ways. First, this study is the first to use a longi-
port perceptions,19 relational closeness,20 and social capital.21 tudinal design to examine the reciprocal relationships be-
Passive consumption, however, may be particularly detri- tween Instagram use and adolescents’ depressed mood.
mental for young people’s mental health, because SNS users Second, this study focuses on adolescents, as this age group
tend to portray themselves in their best possible way. Re- may be particularly at risk for the impact of Instagram use,
searchers argue that exposure to overly flattering images of given Instagram’s enormous popularity among adolescents3
peers and celebrities on SNSs may increase users’ (negative) and the increase of depressive symptoms during this stage of
comparison behaviors22 which, in turn, may result in detri- life.35 Third, this study investigates the impact of different
mental outcomes for adolescents’ mental health.23 types of Instagram use, rather than general Instagram use, on
In relation to specific types of Instagram use, empirical adolescents’ depressed mood.
evidence, however, is more scarce and less consistent; one
study found that exposure to celebrity and peer images in- Method
creased women’s negative mood,24 whereas another study
Procedure
revealed that browsing on Instagram was associated with
lower loneliness among undergraduates.25 These mixed Two-wave panel data (Time 1 = March 2014; Time
findings, however, might be explained by (1) the use of 2 = October 2014) were obtained from 15 high schools in
different dependent variables, that is, one study focused on [region], [country]. These data were part of a large-scale
mood, the other on loneliness, and (2) the use of a different longitudinal panel study on the relationships between social
research design, that is, one study set up an experiment, the media use and adolescents’ mental health. During the re-
other a cross-sectional survey study. searchers’ school visits, students (ages 12–19) filled out
Building, however, on the consistent findings regarding paper-and-pencil questionnaires. Informed consent was
the impact of specific types of Facebook use on depressed obtained in accordance with the customary guidelines in
mood and the theoretical indications in the literature that [country]. Participants were informed that the study inves-
propose positive outcomes of directed communication and tigated their social media use and emotions. To ensure that
broadcasting and negative outcomes of passive consumption, participants were unable to discuss or view the answers of
 INSTAGRAM USE AND ADOLESCENTS’ DEPRESSED MOOD 605

peers, they were instructed to place their questionnaire in a complete the survey at both time points, (1) to increase sta-
closed envelope after completing. Strict confidentiality of the tistical power37 and (2) to reduce the risk of problems en-
answers was assured. The institutional review board of the countered in SEM when sample sizes are small (e.g.,
host university approved the study procedures. nonconvergent solutions, incorrect standard errors, and so
on).38–40 The full information maximum likelihood proce-
Participants dure was used to estimate missing data. The chi-squared to
degrees of freedom ratio (v2/df), the root-mean-squared error
In total, 1,840 participants filled out the questionnaire at of approximation (RMSEA), and the comparative fit index
Time 1; 1,577 filled out the questionnaire at Time 2; and (CFI) were used to determine the goodness-of-fit of the
1,235 filled out the questionnaire at both time points (i.e., models.41
67% of total). More importantly, a total of 671 Instagram To investigate the reciprocal relationships between dif-
Downloaded by Gothenburg University Library from online.liebertpub.com at 10/18/17. For personal use only.

members filled out the questionnaire at Time 1 and 622 at ferent types of Instagram use and adolescents’ depressed
Time 2; 440 Instagram members filled out the questionnaire mood, we computed an autoregressive cross-lagged panel
at both time points (i.e., 66% of total). The mean age of those model.
participants with an Instagram account at Time 1 was 14.96 All variables at Time 2 were predicted by their preceding
(SD = 1.29) and 39% of this sample were boys. As this study values at Time 1 and by the value of the respective inde-
aims to investigate the impact of Instagram use, we only pendent variables at Time 1. We further allowed covariances
included those participants who had an Instagram account at between our control variables and the study variables at Time
Time 1 or Time 2 in our analyses. 1 and by estimating paths from these control variables to
Differences were explored between those who filled out each of the study variables at Time 2. In addition, we mod-
one questionnaire and those who filled out both question- eled covariances between study variables measured at the
naires with regard to all relevant Time 1 variables. Using same time point and allowed covariances between the error
Pillai’s trace, a multivariate analysis of variance showed terms of the same indicators.42
significant differences (V = 0.02, F(4, 628) = 2.57, p < 0.05, To test whether the relationships differed between boys
hp2 = 0.02). Followup univariate analyses revealed that ado- and girls, we performed a multiple group comparison test.
lescents who participated in both waves scored lower on This test compared a model in which all structural paths were
Instagram posting Time 1 (M = 3.65; SD = 0.07 vs. M = 3.92; allowed to vary across groups with a model in which all
SD = 0.10, F(1, 653) = 4.58, p < .05) and depressed mood structural paths were fixed to be equal across groups. If D v2
Time 1 (M = 1.80; SD = 0.04 vs. M = 1.94; SD = 0.05, F(1, was significant ( p < 0.05), we conducted a path-by-path
650) = 3.30, p = 0.01). analysis to examine whether the hypothesized relationships
differed significantly between the two groups.
Measures
Control variables. Participants responded to questions Results
about gender and age.
Preliminary analyses
Different types of Instagram use. Using a seven-point The descriptive statistics for and zero-order correlations
Likert scale (1 = never to 7 = several times per day), partici- between all relevant variables is presented in Table 1.
pants were asked (1) ‘‘How often do you look at photos
posted by other Instagram users?’’ (i.e., browsing), (2) ‘‘How Cross-lagged model
often do you post a photo on Instagram?’’ (i.e., posting), and
(3) ‘‘How often do you ‘like’ a photo on Instagram?’’ (i.e., The cross-lagged model, presented in Figure 1, showed a
liking). good fit of the data and yielded a chi-square value of 391.87
with 131 degrees of freedom, p < 0.001, RMSEA = 0.06;
Depressed mood. To assess participants’ depressive CFI = 0.96; v2/df = 2.99. Results revealed that Instagram
symptoms, we used The Center for Epidemiological Studies browsing at Time 1 positively predicted adolescents’ de-
Depression Scale for Children (CES-DC). Using a four-point pressed mood at Time 2, b = 0.16, B = 0.04, SE = 0.02,
Likert scale (1 = not at all to 4 = a lot), participants rated the p < 0.05. Instagram posting and liking at Time 1, however,
frequency of 20 statements for the past 2 weeks. Olsson and were not related to depressed mood at Time 2. Thus, results
von Knorring investigated the psychometric properties of support H1, but could not confirm H2.
this measure in a Swedish adolescent sample.36 They found In addition, our research question examined whether ad-
support for one strong underlying factor, that is, depressed olescents’ depressed mood was related to Instagram brows-
mood (e.g., ‘‘During the past week, I wasn’t able to feel ing, posting, or liking. Results showed that depressed mood
happy, even when my family or friends tried to help me feel at Time 1 positively predicted Instagram posting at Time 2,
better’’) (aTime1 = 0.87; aTime 2 = 0.88). Based on the average b = 0.08, B = 0.10, SE = 0.05, p = 0.05. Depressed mood at
of these six items, an estimate of adolescents’ depressed Time 1, however, did not predict Instagram browsing and
mood was created. liking at Time 2.
Furthermore, the predictors in the model collectively ex-
plained 28% of the variance in adolescents’ depressed mood
Data analysis
(R2 = 0.28), 35% of the variance in Instagram posting
The hypothesized relationships were tested using struc- (R2 = 0.35), 20% of the variance in Instagram browsing
tural equation modeling (SEM) in AMOS (SPSS, IBM). We (R2 = 0.20), and 24% of the variance in Instagram liking
decided to include all participants, also those that did not (R2 = 0.24).
 606 FRISON AND EGGERMONT

Table 1. Descriptive Statistics and Zero-Order Correlations

Minimum Maximum M SD 1. 2. 3. 4. 5. 6. 7. 8.
1. Instagram browsing (T1) 1 7 5.52 1.59 1 0.50** 0.57** 0.33** 0.84** 0.49** 0.11** 0.15**
2. Instagram browsing (T2) 1 7 5.88 1.45 1 0.30** 0.45** 0.49** 0.84** ns 0.14**
3. Instagram posting (T1) 1 7 3.75 1.52 1 0.57** 0.60** 0.35** 0.19** 0.15**
4. Instagram posting (T2) 1 7 3.49 1.31 1 0.35** 0.50** 0.20** 0.18**
5. Instagram liking (T1) 1 7 5.29 1.69 1 0.55** 0.16** 0.15**
6. Instagram liking (T2) 1 7 5.62 1.57 1 ns 0.14**
7. Depressed mood (T1) 1 4 1.84 0.73 1 0.46**
8. Depressed mood (T2) 1 4 1.76 0.70 1
Downloaded by Gothenburg University Library from online.liebertpub.com at 10/18/17. For personal use only.

**p < 0.01.

ns, nonsignificant; T1, Time 1; T2, Time 2; NTime1 = 671.

Gender differences Different types of Instagram use and adolescents’

depressed mood
To test whether these relationships were moderated by
gender, we conducted a multiple group comparison test. In line with our expectations (i.e., H1), results revealed
Results revealed that the model was different for boys and that Instagram browsing is related to increases in adoles-
girls, D v2 (11) = 28.15, p < 0.01. However, a path-by-path cents’ depressed mood. This finding confirms previous cross-
analysis revealed that none of the hypothesized relationships sectional7,25 and experimental24 research among college
accounted for this difference, as all hypothesized relation- students and is also in line with studies linking passive
ships were similar for boys and girls. consumption of SNS content to depressive symptoms in
adolescence6 and young adulthood.12
This harmful impact of Instagram browsing on adolescents’
Discussion
depressed mood might be due to the large number of strangers
The results of this study showed that Instagram browsing that adolescents follow through Instagram. Lup et al. confirm
at Time 1 was related to greater depressed mood at Time 2. In this suggestion; their study findings revealed that at the highest
addition, depressed mood at Time 1 was related to more levels of strangers followed, more frequent Instagram use
Instagram posting at Time 2. The present study has important had direct associations with greater depressive symptoms.7 In
implications for future research. addition, similarly to Facebook use,22 Instagram use might

FIG. 1. Model examining the re-

ciprocal relationships between dif-
ferent types of Instagram use (i.e.,
browsing, posting, and liking) and
adolescents’ depressed mood.
 INSTAGRAM USE AND ADOLESCENTS’ DEPRESSED MOOD 607

stimulate negative comparison behaviors, which in turn may depressed mood were similar for boys and girls. This lack of
increase young people’s depressed mood. Future research is moderation contrasts previous cross-sectional studies that
thus needed (1) to test the moderating role of strangers fol- found support for a moderating role of gender in the asso-
lowed through Instagram and (2) to examine the mediating ciation between specific types of Facebook use and young
role of negative comparison behaviors on Instagram in the people’s depressive symptoms.44,45 This lack of gender dif-
relationship between Instagram browsing and adolescents’ ference, however, may be partly due to the fact that female
depressed mood. Instagram users (61%) were overrepresented in our sample.
Contrary to our expectations (i.e., H2), which were based Perhaps gender differences did not emerge because there
on previous Facebook research,6 no support was found for an were not enough boys to test for them. Nonetheless, this
association between Instagram posting/liking and adoles- study is the first to examine gender differences in the re-
cents’ depressed mood. This lack of significant relationships ciprocal associations between Instagram use and adoles-
Downloaded by Gothenburg University Library from online.liebertpub.com at 10/18/17. For personal use only.

may be because Instagram is much more image driven than cents’ depressed mood. More research is thus needed to
Facebook. More specifically, because Instagram’s prime further test the role of gender in these relationships.
focus is on sharing images and videos, Instagram posting and
liking exclusively refers to posting and liking of images and Limitations
videos. Facebook posting and liking, however, can refer to
images and videos, but also status updates. Instagram may Although the present study is the first to explore important
therefore provide users less possibilities to express their reciprocal associations between different types of Instagram
emotions and seek social support and thus to decrease their use and adolescents’ depressed mood, it is not without lim-
depressed mood, than Facebook, which may potentially ex- itations. A first limitation refers to the fact that we relied on
plain why Instagram liking and posting are not capable to single-item measures to assess different types of Instagram
decrease adolescents’ depressed mood. use. Although we believe these single-item measures are
more meaningful than general measures such as time spent
Adolescents’ depressed mood and different types on Instagram, we recommend that future research should rely
of Instagram use on multi-item measures when assessing specific types of
Instagram use. Second, this study is limited by the important
Regarding the opposite relationships, our results revealed dropout of participants between Times 1 and 2. Although this
that depressed mood is related to increases in Instagram post- dropout limits the study findings, our current associations
ing. This finding is in line with prior studies that revealed between different types of Instagram use and adolescents’
positive associations between depressive symptoms and Fa- depressed mood would likely be even stronger when attrition
cebook status updating,30 but extend prior research by showing would be absent because the participants who dropped out
that depressed mood may also be predictive of broadcasting posted more on Instagram and felt more depressed. To pro-
Instagram content. In addition, this finding is also in line with vide a more correct understanding of the actual strengths of
Mood Management Theory,8,9 which argues that media are these associations, future research should try to minimize
purposively used for coping with moods. However, by re- dropout of participants.
vealing that a specific type of media use (i.e., Instagram post-
ing), instead of the choice of a certain medium, may be used to
Conclusion
manage moods, our study extends Mood Management Theory.
Based on these insights, we therefore recommend that future Despite these limitations, the present study provides in-
research that aims to test this theory should not only address the sight into the reciprocal relationships between different types
choice of media to manage moods but also how media are used of Instagram use (i.e., browsing, posting, and liking) and
for reducing negative and enhancing positive moods. adolescents’ depressed mood. Overall, we can conclude (1)
Furthermore, adolescents with greater depressed mood may that adolescents have a higher chance to develop a greater
be more likely to post images or videos on Instagram, perhaps, depressed mood when they browse more often through In-
to enhance their image in the eyes of other social network stagram and (2) that adolescents have a higher chance to post
members. Instagram is a visually centered platform that allows more on Instagram when they have higher levels of de-
their users to easily edit and filter their photos and videos; it is pressed mood.
thus an ideal medium to present oneself in their best possible These results have important implications that warrant
way, even a false way. This strategic self-presentation on In- consideration. More specifically, adolescents, but also par-
stagram may especially attract adolescents with high levels of ents and clinicians should be made aware of the harmful
depressed mood, because it not only allows them to enhance impact of passive consumption of SNS content, as looking at
their image in the eye of others but also to find encouragement photo’s on Instagram (i.e., Instagram browsing) may in-
(e.g., through likes and positive comments). In line with this crease teens’ depressed mood 6 months later. Future pre-
suggestion, Mehdizadeh found that college students lower in vention and intervention programs aimed at reducing passive
self-esteem engaged more often in self-promotional behaviors consumption on SNSs are therefore particularly needed.
on Facebook.43 Future research is however needed to further Thus, our findings not only offer important insights for teens
enhance our understanding of how depressed mood relates to and practitioners but also useful guidance for future pre-
increases in adolescents’ posting behavior on Instagram. vention and intervention programs.

Gender differences Acknowledgment

A path-by-path analysis revealed that the reciprocal relation- This research was funded by a grant from the Flemish
ships between different types of Instagram use and adolescents’ Fund for Scientific Research (FWO).
 608 FRISON AND EGGERMONT

Author Disclosure Statement 18. Stapleton P, Luiz G, Chatwin H. Generation validation: the
role of social comparison in use of Instagram among
No competing financial interests exist.
emerging adults. Cyberpsychology, Behavior, and Social
Networking 2017;20:142–149.
References
19. Li X, Chen W, Popiel P. What happens on Facebook stays
1. Lenhart A. (2015) Teens, social media, and technology on Facebook? The implications for perceived, receiving,
overview 2015. Washington, DC: Pew Internet & American and giving social support. Computers in Human Behavior
Life Project. 2015;51:106–113.
2. Anderson BYM. (2015) Technology device ownership: 20. Vitak J. (2012) Keeping connected in the Facebook age.
2015. Washington, DC: Pew Internet & American Life Doctoral dissertation, Michigan State University. https://
Project. vitak.files.wordpress.com/2009/02/vitak-dissertation-2012.
Downloaded by Gothenburg University Library from online.liebertpub.com at 10/18/17. For personal use only.

3. Van Waeg S, Van Hoecke L, Demeulenaere A, et al. (2016) pdf (accessed July 25, 2017.)
Onderzoeksrapport Apenstaartjaren 5. Gent, Belgium: 21. Aubrey JS, Rill L. Investigating relations between Face-
Mediaraven & Linc. book use and social capital among college undergraduates.
4. Steinberg L. (2005) Adolescence. New York: McGraw-Hill. Communication Quarterly 2013;61:479–496.
5. Skrove M, Romundstad P, Indredavik MS. Resilience, 22. Fardouly J, Vartanian LR. Negative comparisons about
lifestyle and symptoms of anxiety and depression in ado- one’s appearance mediate the relationship between Face-
lescence: the Young-HUNT study. Social Psychiatry and book usage and body image concerns. Body Image 2015;
Psychiatric Epidemiology 2013;48:407–416. 12:82–88.
6. Frison E, Eggermont S. Toward an integrated and differential 23. Frison E, Eggermont S. ‘‘Harder, Better, Faster, Stronger’’:
approach to the relationships between loneliness, different types negative comparison on Facebook and adolescents’ life
of Facebook use and adolescents’ depressed mood. Commu- satisfaction are reciprocally related. Cyberpsychology,
nication Research 2015;1–28 [Epub ahead of print] DOI: Behavior, and Social Networking 2016;19:158–164.
10.1177/0093650215617506. 24. Brown Z, Tiggeman M. Attractive celebrity and peer im-
7. Lup K, Trub L, Rosenthal L. Instagram #Instasad?: explor- ages on Instagram: effect on women’s mood and body
ing associations among Instagram use, depressive symptoms, image. Body Image 2016;19:37–43.
negative social comparison, and strangers. Cyberpsychology, 25. Yang C. Instagram use, loneliness, and social comparison
Behavior, and Social Networking 2015;18:247–252. orientation: interact and browse on social media, but don’t
8. Zillmann D. Mood management through communication compare. Cyberpsychology, Behavior, and Social Net-
choices. American Behavioral Scientist 1988;31:327–340. working 2016;19:703–708.
9. Zillmann D. (2000) Mood management in the context of se- 26. Clayton RB, Osborne RE, Miller BK, et al. Loneliness,
lective exposure theory. In Roloff ME, ed. Communication anxiousness, and substance use as predictors of Facebook
yearbook 23. Thousand Oaks, CA: Sage, pp. 103–123. use. Computers in Human Behavior 2013;29:687–693.
10. Tsitsika AK, Tzavela EC, Janikian M, et al. Online social 27. Lou LL, Yan Z. An examination of the reciprocal re-
networking in adolescence: patterns of use in six European laionship of loneliness and Facebook use among first-year
countries and links with psychosocial functioning. Journal college students. Journal of Educational Computing Re-
of Adolescent Health 2014;55:141–147. search 2012;46:105–117.
11. Burke M, Kraut R, Marlow C. (2011) Social capital on 28. Ryan T, Xenos S. Who uses Facebook? An investigation
Facebook: differentiating uses and users. In Proceedings of into the relationship between the Big Five, shyness, nar-
the 2011 ACM Conference on Human Factors in Comput- cissism, loneliness, and Facebook usage. Computers in
ing Systems. Vancouver, British Columbia, Canada: Asso- Human Behavior 2011;27:1658–1664.
ciation for Computing Machinery (ACM) SIGCHI, 29. Sheldon KM, Abad N, Hinsch C. A two-process view
pp. 571–580. of Facebook use and relatedness need-satisfaction:
12. Verduyn P, Lee DS, Park J, et al. Passive Facebook usage disconnection drives use, and connection rewards it.
undermines affective well-being: experimental and longi- Journal of Personality and Social Psychology 2011;100:
tudinal evidence. Journal of Experimental Psychology 766–775.
2015;144:480–488. 30. Scherr S, Brunet, A. Differential influences of depression
13. Deters FG, Mehl MR. Does posting Facebook status up- and personality traits on the use of Facebook. Social Media
dates increase or decrease loneliness? An online social + Society 2017;3:1–14.
networking experiment. Social Psychological and Person- 31. Knobloch S, Zillmann D. Mood management via the digi-
ality Science 2013;4:1–8. tal jukebox. Journal of Communication 2002;52:351–366.
14. Hampton KN, Goulet LS, Rainie L, et al. (2011) Social 32. Knobloch S. Mood adjustment via mass communication.
networking sites and our lives. Washington, DC: Pew In- Journal of Communication, 2003;53:233–250.
ternet & American Life Project. 33. Dillman Carpentier FR, Brown JD, Bertocci M, et al. Sad
15. Lee E, Kim YJ, Ahn J. How do people use Facebook fea- kids, sad media? Applying mood management theory to
tures to manage social capital? Computers in Human Be- depressed adolescents’ use of media. Media Psychology
havior 2014;36:440–445. 2008;11:143–166.
16. Rae JR, Lonborg SD. Do motivations for using Facebook 34. Johnson BK, Knobloch-Westerwick S. Glancing up or
moderate the association between Facebook use and psy- down: mood management and selective comparison on
chological well-being? Frontiers in Psychology 2015;6:1–9. social networking sites. Computers in Human Behavior
17. Steers MLN, Wickham RE, Acitelli LK. Seeing everyone 2014;41:33–39.
else’s highlight reels: how Facebook usage is linked to 35. Costello EJ, Copeland W, Angold A. Trends in psychopa-
depressive symptoms. Journal of Social and Clinical Psy- thology across the adolescent years: what changes when
chology 2014;33:701–731. children become adolescents, and when adolescents be-
 INSTAGRAM USE AND ADOLESCENTS’ DEPRESSED MOOD 609

come adults? Journal of Child Psychology and Psychiatry equation modeling. Journal of Abnormal Psychology
2011;52:1015–1025. 2003;112:558–577.
36. Olsson G, von Knorring A-L. Depression among Swedish 43. Mehdizadeh S. Self-presentation 2.0: narcissism and self-
adolescents measured by the self-rating scale Center for esteem on Facebook. Cyberpsychology, Behavior, and
Epidemiology Studies-Depression Child (CES-DC). Euro- Social Networking 2010;13:357–364.
pean Child & Adolescent Psychiatry 1997;6:81–87. 44. Frison E, Eggermont S. Exploring the relationships be-
37. Raaijmakers QA. Effectiveness of different missing data tween different types of Facebook use, perceived online
treatments in surveys with Likert-type data: introducing the social support, and adolescents’ depressed mood. Social
relative mean substitution approach. Educational and Psy- Science Computer Review 2015;45:1–19.
chological Measurement 1999;59:725–748. 45. Simoncic TE, Kuhlman KR, Vargas I, et al. Facebook use
38. Anderson JC, Gerbing DW. The effect of sampling error on and depressive symptomatology: investigating the role of
Downloaded by Gothenburg University Library from online.liebertpub.com at 10/18/17. For personal use only.

convergence, improper solutions, and goodness-of-fit indi- neuroticism and extraversion in youth. Computers in Hu-
ces for maximum likelihood confirmatory factor analysis. man Behaviour 2014;40:1–5.
Psychometrika 1984;49:155–173.
39. Marsh HW, Balla JR. Goodness-of-fit indices in confir-
matory factor analysis: the effect of sample size and model Address correspondence to:
complexity. Quality and Quantity 1994;28:185–217. Dr. Eline Frison
40. Marsh HW, Balla JR, McDonald RP. Goodness-of-fit in- Leuven School for Mass Communication Research
dexes in confirmatory factor analysis: the effect of sample Faculty of Social Sciences
size. Psychological Bulletin 1988;103:391–410. KU Leuven
41. Byrne BM. (2010) Structural Equation Modeling with Parkstraat 45 (PO box 3603)
AMOS: Basic Concepts, Applications and Programming. Leuven B-3000
Mahwah: Lawrence Erlbaum. Belgium
42. Cole DA, Maxwell SE. Testing mediational models with
longitudinal data: questions and tips in the use of structural E-mail: [email protected]