Associations Between Internet Attachment,Cyber Victimization,and Internalizing Symptoms Among Adolescents

Abstract

With increasing frequency of Internet use among adolescents, there are growing concerns about their risk for becoming attached to these forms of communication and increased vulnerability for negative online experiences, including cyber victimization. The effect of these experiences on adolescent mental health is not well understood. In this study, we examine how Internet attachment is related to anxiety and depression and assess the mediating effect of cyber victimization on these associations. Participants included 1,151 middle school students (51.4 percent males) aged 10 to 16 (M = 12.7, SD = 0.93). Structural equation models show that greater Internet attachment was associated with more cyber victimization and greater symptoms of anxiety and depression. Cyber victimization mediated the associations between Internet attachment and anxiety and between Internet attachment and depression. Implications for online awareness efforts are discussed.

Introduction

As technology becomes more accessible, it is not surprising that adolescents are spending more time online than ever before.¹ With limited supervision in online environments, there are growing concerns about adolescents' risk for Internet attachment and vulnerability to negative online experiences such as cyber victimization. In this study, we focus on Internet attachment as an index of the value placed on one's Internet use (i.e., the degree to which an individual desires or needs to use the Internet) rather than Internet dependency or addiction that relates to some type of dysfunction resulting from compulsive Internet use. Research with adolescents suggests that excessive preoccupations with the Internet can contribute to mental health problems.^2–4 Research also shows that cyber victimization predicts a range of mental health problems, including symptoms of anxiety and depression, poor self-esteem, and suicidal ideation.^5–7

Emerging research indicates that greater Internet attachment is associated with an increased vulnerability to cyber victimization,⁸ which may be an important precursor to the development of mental health problems. It is possible that adolescents who spend most of their time online and who place a greater value on their online presence, may not only be more susceptible to cyber victimization but may also be more likely to internalize their cyber victimization experiences compared to adolescents who are less attached to the Internet. In this study, we ask whether Internet attachment predicts cyber victimization, which in turn predicts internalizing symptoms (e.g., anxiety and depression) among adolescents. We also test the role of cyber victimization as a mediator on associations between Internet attachment and internalizing symptoms.

Internet Attachment in Adolescence

Adolescents are going online at an increasing rate for entertainment, to acquire information, and to socialize and maintain relationships.⁹ Social networking sites such as Facebook, Instagram, and Twitter have made it particularly easy for adolescents to share information with their friends quickly and effortlessly.¹⁰ However, adolescents may also be vulnerable to becoming attached to the Internet for social connections and a sense of belonging. Increasing attachment to the Internet can lead to an inability to control one's online activity, the overwhelming dominance of Internet use on one's cognitions and behaviors, and feelings of considerable uneasiness, irritability, and loneliness at times when one is offline.^4,11,12

Adolescence is a formative developmental period where identities are constructed and social feedback plays an important role in creating self-concepts.¹³ For those adolescents who rely on the Internet to enhance their self-worth (e.g., need for positive validation), receiving negative online feedback can be detrimental to their self-esteem and may increase risks for mental health problems. In a sample of adolescents aged 10 to 19, positive feedback through social networking sites increased adolescents' self-esteem and well-being, whereas negative feedback had the reverse effect.¹⁴ If adolescents become attached to the Internet, they may be more likely to internalize their negative online experiences (e.g., cyber victimization) especially if their needs are not met.

Although past research shows a positive association between Internet attachment and internalizing symptoms,^2–4 the mechanisms behind this association are unclear. Given that greater Internet attachment can increase adolescents' vulnerability for cyber victimization, the negative effects of cyber victimization may explain the association between Internet attachment and internalizing symptoms. Moreover, it has been suggested that “anxiety and depression are separate but meaningfully related constructs” (p. 823)¹⁵ and therefore may be differentially associated with Internet attachment. In this study, we examine the mediating effect of cyber victimization on associations between Internet attachment and anxiety, and between Internet attachment and depression.

Cyber Victimization in Adolescence

Cyber victimization is experienced by adolescents who face repeated and intentional harassment from an individual or group of individuals through electronic forms of communication.¹⁶ These experiences are most frequent during the middle school years and are reported by both male and female adolescents who have access to technology.¹⁷ Although the reported frequency of cyber victimization varies widely across studies that operationalize and measure these experiences differently (see Ref.¹⁸ for a review), recent reviews of the literature suggest that ∼20 percent of adolescents experience cyber victimization.^16,19 Furthermore, some research finds that females are more likely to report cyber victimization than males,^16,20 others suggest males experience more cyber victimization than females,^21,22 or report no sex differences.¹⁷ Anecdotal and growing empirical evidence suggests that cyber victimization is related to similar and perhaps more harmful consequences than traditional forms of victimization, as aggressors may be harder to identify, multiple perpetrators may be involved, an unlimited audience can witness or participate, and victims can be targeted online day or night.²³ Indeed, many victims of cyberbullying report feeling sad, upset, angry, embarrassed, hopeless, and frustrated with their experience.²⁴ Growing research has also documented long-term effects of cyber victimization on increasing symptoms of anxiety and depression.^5,9,25 Furthermore, it appears that these effects may be particularly salient for females compared to males.¹⁸ In this study, we investigate the associations between cyber victimization and anxiety and depression separately for male and female adolescents.

Internet attachment has also been implicated with cyber victimization. For example, greater Internet attachment was associated with more cyber victimization in a large sample of adolescents aged 12 to 18.⁸ To our knowledge, past research has not considered whether adolescents who are more attached to the Internet are more likely to internalize their negative online experiences because they also report cyber victimization.

Current Study

This study extends past research in two important ways. First, we examine the unique associations between Internet attachment, cyber victimization, and two internalizing symptoms: anxiety and depression. Second, we test whether cyber victimization mediates the associations between Internet attachment and anxiety and depression. Moreover, we consider these associations during adolescence when cyber victimization experiences are most frequent. Based on past research, we hypothesize that greater Internet attachment will be associated with more cyber victimization and greater symptoms of both anxiety and depression. We also hypothesize that cyber victimization will mediate associations between Internet attachment and anxiety and between Internet attachment and depression. Finally, we examine the moderating effects of sex, given the inconsistencies in the literature regarding the rates of cyber victimization and limited research on the effects of cyber victimization on internalizing symptoms by sex.¹⁸

Methods

Participants and procedure

Participants included 1,151 middle school students (51.4 percent males) aged 10 to 16 (M = 12.7, SD = 0.93) who were recruited from nine middle schools in two Midwestern cities. The sample was predominantly Caucasian (62 percent) followed by Asian (14 percent), Aboriginal/Native (7 percent), Black (6 percent), Hispanic (3 percent), and other groups (8 percent).

In the fall of 2012, superintendents of school divisions within both cities were approached about participation in a larger research project.²³ Superintendents who agreed to participate in the study forwarded information to school administrators within their school division. Interested school administrators decided on an appropriate parental consent procedure for their school. Three schools utilized an active parental consent procedure (80.2 percent of parents provided their consent) and six schools used a parental assent procedure (4.2 percent of parents opted out of the study). Students who did not have parental assent/consent were unable to participate. On average, students took 25 minutes to complete the paper and pencil survey and were debriefed upon completion.

Measures

Internet attachment

Participants' perceived attachment to the Internet was measured using five items adapted from Vandebosch and Van Cleemput⁸ and Lu.²⁶ Participants rated each item (e.g., “I would rather surf the internet than do something else, I can easily go without the internet for a few days, I make a lot of new friends on the internet, if internet access was disrupted, I wouldn't miss the internet, and a life without the internet would be empty and boring”) on a 7-point Likert-type scale that ranged from 1 (strongly disagree) to 7 (strongly agree). Items were summed so that higher scores represented greater Internet attachment (α = 0.67).

Cyber victimization

Participants' self-reported experience as a victim of cyberbullying was assessed using a 9-item scale of cyber victimization adopted from Patchin and Hinduja.²⁷ Strong reliability and construct validity estimates have been found for this scale in past research.^27,28 Items ranged from less distressing experiences (e.g., “Received an upsetting email from someone you don't know”) to more distressing experiences (e.g., “Something posted online that you did not want others to see”). Participants reported the frequency of experiences in the past 2–3 months on a 5-point Likert-type scale that ranged from 0 (never) to 4 (every day). A cyber victimization score was created by summing up the responses so that higher scores reflected more cyber victimization (α = 0.88).

Internalizing symptoms

Participants' self-reported symptoms of anxiety and depression were assessed using the 21-item version of the Depression, Anxiety, Stress Scale (DASS-21).²⁹ Although the DASS-21 was developed for adult populations,³⁰ strong psychometric properties were found with adolescent samples, particularly for the anxiety and depression subscales.³¹ Participants rated their symptoms of anxiety (7 items; “I felt I was close to panic”) and depression (7 items; “I felt that I had nothing to look forward to”) in the past week on a 4-point Likert-type scale that ranged from 0 (did not apply to me at all) to 3 (applied to me very much or most of the time). Items were summed to create an overall level of anxiety and depression in which higher scores represented greater levels of anxiety (α = 0.87) and depression (α = 0.92).

Data analysis

First, confirmatory factor analyses were used to assess the measurement model. Next, structural equation modeling was used to examine the hypothesized pathways between Internet attachment, cyber victimization, and anxiety and depression. Given that anxiety and depression are highly correlated (r = 0.80), mediation effects of cyber victimization on the associations between Internet attachment and anxiety and depression were examined simultaneously in one model to test for their independent effects. Mplus v7.1³² was used to conduct both the measurement and structural model. Model fit was evaluated following established guidelines^33,34 using the comparative fit indices (CFI), root-mean square error of approximation (RMSEA), and standardized root-mean of the residual (SRMR). CFI values of ≥0.95 represent an exceptional fitting model and >0.90 indicate reasonable fit; RMSEA and SRMR values ≤0.05 suggest excellent model fit and values between 0.05 and 0.08 indicate reasonable fit. Direct and indirect effects were computed using the MODEL INDIRECT option in Mplus.³² Age was controlled for in all analyses. Using the multiple group function, sex was used as a grouping variable to assess differences in the direct and indirect pathways between males and females.

Results

Descriptive statistics

Table 1 provides means and standard deviations on all study variables for both males and females and bivariate correlations for the full sample.

Table 1.

Descriptive Statistics and Bivariate Correlations

Variable	Males M (SD)	Females M (SD)	Total M (SD)	1	2	3
1. Internet attachment	16.36 (6.43)	16.17 (6.38)	16.27 (6.40)	1
2. Cyber victimization	1.78 (3.96)^a	2.51 (4.34)^a	2.14 (4.16)	0.23^***	1
3. Anxiety	2.71 (4.07)^a	3.83 (4.84)^a	3.26 (4.50)	0.22^***	0.54^***	1
4. Depression	2.80 (4.67)^a	4.49 (5.89)^a	3.64 (5.38)	0.25^***	0.51^***	0.80^***

Mean levels differ significantly (p < 0.05) between males and females.

***

p < 0.001.

Measurement model

Confirmatory factor analyses of the four first-order latent constructs (Internet attachment, cyber victimization, anxiety, and depression) showed adequate fit to the data (CFIs >0.91, RMSEAs <0.11, SRMRs <0.05). Standardized factor loadings ranged from 0.36 to 0.66 for Internet attachment, 0.53 to 0.75 for cyber victimization, 0.49 to 0.82 for anxiety, and 0.71 to 0.84 for depression.

Structural mediation model

Multiple group analysis was used to assess differences in fit related to the moderating effects of sex. A chi-square difference test was used to compare the fit of a model with unconstrained pathways (i.e., allowed to vary by sex) against the fit of a model with constrained pathways (i.e., fixed to be equal across sex). The chi-square difference was significant (Δχ² = 180.37, Δdf = 19, p < 0.001) indicating that the pathways differed by sex. Further investigation (i.e., constraining pathways sequentially) showed that two pathways differed for male and female adolescents. More cyber victimization predicted greater depressive symptoms for females (β = 0.57, p < 0.001), but not males (β = 0.41, p < 0.001), and greater Internet attachment predicted greater anxiety symptoms for males (β = 0.16, p = 0.002), but not females (β = 0.07, p = 0.12). We present the results for the full sample below.

Direct pathways between Internet attachment and anxiety and between Internet attachment and depression were first examined before the addition of cyber victimization. Internet attachment was positively associated with both anxiety (β = 0.28, p < 0.001) and depression (β = 0.30, p < 0.001). Cyber victimization was subsequently entered to assess whether it served as a mediator in these associations. Acceptable goodness-of-fit indices were found: CFI = 0.927, RMSEA (90% CI) = 0.052 (0.049–0.054), and SRMR = 0.038. As shown in Figure 1, hypothesized pathways between all of the latent constructs were significant. After controlling for cyber victimization, associations between Internet attachment and anxiety (β = 0.11, p = 0.002) and between Internet attachment and depression (β = 0.15, p < 0.001) remained significant suggesting the possibility of mediation. A significant mediation effect of cyber victimization (β = 0.16, p < 0.001) confirmed that the positive association between Internet attachment and anxiety can be partially explained by cyber victimization. Similarly, a significant mediation effect of cyber victimization (β = 0.14, p < 0.001) suggests that the positive association between Internet attachment and depression can be partially explained by cyber victimization.

FIG. 1.

Structural mediational model for anxiety and depression controlling for age. All path estimates are standardized. **p < 0.01, ***p < 0.001.

Discussion

This study examined relations between Internet attachment, cyber victimization and internalizing symptoms among adolescents aged 10 to 16. As expected, greater Internet attachment was associated with more cyber victimization and greater symptoms of anxiety and depression. Our second hypothesis that cyber victimization would mediate associations between Internet attachment and anxiety and depression was also supported. Sex differences emerged for two pathways (i.e., from cyber victimization to depression and from Internet attachment to anxiety).

Consistent with past research,^2,3 Internet attachment was positively associated with both anxiety and depression. However, greater Internet attachment was associated with greater anxiety for males, but not females. One potential explanation is the differential patterns in Internet behavior among male and female adolescents. For example, male adolescents may be more likely to develop symptoms of anxiety because they tend to use the Internet for entertainment purposes compared to females who typically go online to develop and maintain interpersonal relationships.³⁵ Adolescents who are reliant on the Internet may be more likely to internalize their online experiences if their needs are not met. It is also possible that these adolescents are engaging in specific types of online activity that may be particularly detrimental for their mental health. Increased online activity has been shown to correspond with risky online behaviors such as visiting unfamiliar websites, going online at any time of the day or night, talking to strangers, and sharing personal information.³⁶ Research also shows that adolescents who engage in risky online behaviors are more likely to experience cyber victimization.³⁷ It is possible that these types of behaviors may help to explain why adolescents who were more attached to the Internet were more likely to experience cyber victimization.⁸ However, further research is needed to examine the possible mediating role of risky online behavior in the association between Internet attachment and cyber victimization.

Our findings are also consistent with past research demonstrating that adolescents who experience cyber victimization are more likely to report symptoms of anxiety and depression.^5,9,25 Demonstrating that females are more vulnerable to the effects (e.g., symptoms of depression) of cyber victimization than males also supports past research.¹⁸ We also extend past research to show that cyber victimization accounts for some of the association between Internet attachment and internalizing symptoms. Specifically, a partial mediation of cyber victimization was found on the associations between Internet attachment and anxiety and between Internet attachment and depression. In other words, greater Internet attachment increased the likelihood of cyber victimization, which in turn increased symptoms of anxiety and depression. Because these findings are based on cross-sectional data, further research with longitudinal assessments is needed to support the proposed direction of associations. Although not tested in this study, it is possible that specific types of Internet attachment (e.g., social networking, online gaming) are differentially associated with cyber victimization, anxiety, and depression. Additional research is also needed to determine potential sources of protection against the harmful effects of Internet attachment and cyber victimization.

This study had some limitations that should be considered. The cross-sectional nature of the study limited our ability to make causal inferences. Although there is growing empirical research^8,9 that supports the direction of our proposed pathways, longitudinal research is needed to further support our meditation model. Our sample was limited to adolescents in middle schools; so it is unclear how the findings would generalize to younger children in elementary schools who are just beginning to develop an online presence.¹⁶ There was some variability in the items assessing Internet attachment that resulted in a lower, but acceptable factor loading for one item (i.e., “I make a lot of new friends online”). Finally, data were based exclusively on self-report assessments and may reflect some bias due to shared reporter variance. However, it may be difficult for parents and teachers to accurately assess adolescents' online experiences as adolescents are connected online at any time of the day or night and are likely unsupervised.³⁸

Despite these limitations, our findings show that male and female adolescents who are more attached to the Internet are more likely to report anxiety and depression because they also experience cyber victimization. Greater efforts are needed to educate adolescents, parents, and teachers on the dangers of Internet attachment and also to determine why some adolescents are particularly susceptible to these online preoccupations that can contribute to cyber victimization and mental health concerns. It is important to address the online habits and behaviors of adolescents, particularly at an early age before an attachment can develop, which negatively impacts their development.

It is also important for parents and teachers to recognize the potential consequences of adolescents engaging in excessive online behavior. Although it can be difficult for parents to monitor the amount of time their child goes online, they can attempt to develop an open dialogue about the frequency and types of interactions their child has online.¹⁶ Discussing netiquette (e.g., rules for engaging in online communication) and setting limits on screen time (e.g., no phone at the dinner table or before bed), particularly when the child is at home may also help to reduce their risk for Internet attachment and cyber victimization. Teachers can also help parents identify potential warning signs of problematic online behavior such as whether a child is going online obsessively.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Lenhart

. (2015) Teen, social media and technology overview 2015 (Pew Research Center Report 202.419.4372). www.pewinternet.org/2015/04/09/teens-social-media-technology-2015 (accessed March 1, 2016).

Cho

, Sung

, Shin

, et al. Does psychopathology in childhood predict internet addiction in male adolescents?. Child Psychiatry & Human Development, 2013; 44:549–555.

Kelleci

, Inal

. Psychiatric symptoms in adolescents with Internet use: comparison without Internet use. Cyberpsychology, Behavior, and Social Networking, 2010; 13:191–194.

Van Den Eijnden

, Meerkerk

, Vermulst

, et al. Online communication, compulsive internet use, and psychosocial well-being among adolescents: a longitudinal study. Developmental Psychology, 2008; 44:655–665.

Hemphill

, Kotevksi

, Heerde

. Longitudinal associations between cyber-bullying perpetration and victimization and problem behavior and mental health problems in young Australians. Int J Public Health, 2015; 60:227–237.

Modecki

, Barber

, Vernon

, et al. Mapping developmental precursors of cyber-aggression: trajectories of risk predict perpetration and victimization. Journal of Youth and Adolescence, 2013; 42:651–661.

Nixon

. Current perspectives: the impact of cyberbullying on adolescent health. Adolescent Health, Medicine and Therapeutics, 2014; 5:143–158.

Vandebosch

, Van Cleemput

. Cyberbullying among youngsters: profiles of bullies and victims. New Media & Society, 2009; 11:1349–1371.

Rose

, Tynes

. Longitudinal associations between cybervictimization and mental health among U.S. adolescents. Journal of Adolescent Health, 2015; 57:1–8.

10.

Peluchette

, Karl

, Wood

, et al. Cyberbullying victimization: do victims' personality and risky social network behaviors contribute to the problem?. Computers in Human Behavior, 2015; 52:424–435.

11.

Guan

SSA

, Subrahmanyam

. Youth internet use: risks and opportunities. Current Opinion in Psychiatry, 2009; 22:351–356.

12.

Meerkerk

, Van Den Eijnden

, Vermulst

, et al. The compulsive internet use scale (CIUS): some psychometric properties. CyberPsychology & Behavior, 2009; 12:1–6.

13.

Côté

, Levine

. (2014) Identity, formation, agency, and culture: a social psychological synthesis. Psychology Press.

14.

Valkenburg

, Peter

, Schouten

. Friend networking sites and their relationship to adolescents' well-being and social self-esteem. CyberPsychology & Behavior, 2006; 9:584–590.

15.

Cummings

, Caporino

, Kendall

. Comorbidity of anxiety and depression in children and adolescents: 20 years after. Psychological Bulletin, 2014; 140:816–845.

16.

Holfeld

, Leadbeater

. The nature and frequency of cyber bullying behaviors and victimization experiences among young Canadian children. Canadian Journal of School Psychology, 2015; 30:116–135.

17.

Tokunaga

. Following you home from school: a critical review and synthesis of research on cyberbullying victimization. Computers in Human Behavior, 2010; 26:277–287.

18.

Kowalski

, Giumetti

, Schroeder

, et al. Bullying in the digital age: a critical review and meta-analysis of cyberbullying research among youth. Psychological Bulletin, 2014; 140:1073–1137.

19.

Patchin

, Hinduja

. (2012). Cyberbullying: an update and synthesis on the research. In Patchin

, Hinduja

, eds. Cyberbullying prevention and response: expert perspectives. New York: Routledge, pp. 13–36.

20.

Mishna

, Khoury-Kassabri

, Gadalla

, et al. Risk factors for involvement in cyber bullying: victims, bullies, and bully/victims. Children and Youth Services Review, 2012; 34:63–70.

21.

Africak

, Siyahhan

, Uzunhasanoglu

, et al. Cyberbullying among Turkish adolescents. Cyberpsychology & Behavior, 2008; 11:253–261.

22.

Erdur-Baker

. Cyberbullying and its correlation to traditional bullying, gender and frequent and risky usage of internet-mediated communication tools. New Media & Society, 2010; 12:109–125.

23.

Holfeld

. Perceptions and attributions of bystanders to cyber bullying. Computers in Human Behavior, 2014; 38:1–7.

24.

Mishna

, Cook

, Gadalla

, et al. Cyber bullying behaviors among middle and high school students. American Journal of Orthopsychiatry, 2010; 80:362–374.

25.

Gámez-Guadix

, Orue

, Smith

, et al. Longitudinal and reciprocal relations of cyberbullying with depression, substance use, and problematic internet use among adolescents. Journal of Adolescent Health, 2013; 53:446–452.

26.

. Sensation-seeking, Internet dependency, and online interpersonal deception. CyberPsychology & Behavior, 2008; 11:227–231.

27.

Patchin

, Hinduja

. Cyberbullying and self-esteem. Journal of School Health, 2010; 80:614–621.

28.

Hinduja

, Patchin

. (2009) Bullying beyond the schoolyard: preventing and responding to cyber bullying. Thousand Oaks, CA: Corwin Press.

29.

Lovibond

, Lovibond

. (1995) Manual for the depression anxiety stress scales. Sydney: Psychology Foundation Monograph.

30.

Henry

, Crawford

. The short-form version of the Depression Anxiety Stress scales (DASS-21): construct validity and normative data in a large non-clinical sample. British Journal of Clinical Psychology, 2005; 44:227–239.

31.

Campbell

, Slee

, Spears

, et al. Do cyberbullies suffer too? Cyberbullies' perceptions of the harm they cause to others and their own mental health. School Psychology International, 2013; 34:613–629.

32.

Muthen

, Muthen

. (1998–2012). Mplus user's guide. Seventh Edition. Los Angeles, CA: Muthen & Muthen.

33.

, Bentler

. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Structural Equation Modeling, 1999; 6:1–55.

34.

Kline

. (2005) Principles and practices of structural equation modeling. 2nd ed. New York, NY: Guilford Press.

35.

Weiser

. Gender differences in Internet use patterns and internet application preferences: a two-sample comparison. CyberPsychology and Behavior, 2000; 3:167–178.

36.

Meter

, Bauman

. When sharing is a bad idea: the effects of online social network engagement and sharing passwords with friends on cyberbullying involvement. Cyberpsychology, Behavior, and Social Networking, 2015; 18:437–442.

37.

Gámez-Guadix

, Borrajo

, Almendros

. Risky online behaviors among adolescents: longitudinal relations among problematic internet use, cyberbullying perpetration, and meeting strangers online. Journal of Behavioral Addictions, 2016; 5:100–107.

38.

Shapka

, Law

. Does one size fit all? Ethnic differences in parenting behaviors and motivations for adolescent engagement in cyberbullying. Journal of Youth and Adolescence, 2013; 42:723–738.