Adolescent Internet Addiction: Testing the Association Between Self-Esteem,the Perception of Internet Attributes,and Preference for Online Social Interactions

Abstract

There is a general consensus that Internet addiction (IA) is mainly related to social aspects of the Web, especially among adolescents. The empirical link between poor social skills and IA is well documented; however, theoretical explanations for this relationship are less developed. One possibility is that people with poor social skills are especially prone to develop a preference for online social interaction (POSI), which, in turn, predicts problematic usage. This hypothesis has been tested for loneliness and social anxiety, but not for self-esteem (SE; one of the main antecedents of IA). Furthermore, the mediating role of the perceived relevance of some Internet features (e.g., anonymity) in the relationship between SE and POSI has never been investigated. A cross-sectional study was conducted with 257 adolescents. Using mediation analyses, we found evidence among females for the mediating role of (a) POSI in the relationship between SE and IA, and (b) the subjective relevance of some Internet features in the association between SE and POSI. No significant effects were found for males.

Introduction

In recent years, there has been an increasing interest in understanding the negative potential of the Internet. Despite a lack of consensus regarding whether problematic usage can be classified as a disorder, there is no doubt that some Internet users have developed problematic behaviors.^1,2 The prevalence of Internet addiction (IA) among adolescents has been estimated to be about 8 percent in Greece,³ 11 percent in Taiwan and China,⁴ and 1.4 percent in Southern Italy.⁵ Despite some conflicting findings,⁶ emerging evidence suggests that IA is mainly related to social aspects of the Web. This could be particularly true for adolescents, who predominantly use the Internet for interpersonal communication.^7–9 However, even though computer-mediated communication (CMC) is central to the social lives of adolescents, only a small proportion develops problematic usage.

Factors relating to the development of a pathological Internet use (PIU) for some and not others need to be better understood. According to Davis,¹⁰ psychosocial problems such as low self-esteem (SE) and loneliness predispose some users to exhibit maladaptive cognitions and behaviors that result in negative outcomes. A number of studies have found evidence to support an association between IA and depression,^11,12 loneliness,^13–15 low SE,^1,16–18 and shyness.¹⁹ However, the majority of studies have investigated direct linear relationships, without a priori explanatory hypotheses regarding why some aspects, such as loneliness or shyness, are significant predictors of IA.²⁰

Caplan²¹ suggests that a person's preference for an online social interaction (POSI) could be a mechanism that accounts, in part, for why deficient social skill increases one's vulnerability to developing PIU. He defines POSI as “a cognitive individual-difference construct characterized by beliefs that one is safer, more efficacious, and more confident with online interpersonal interactions than with face-to-face interactions.”²² In fact, Caplan²² found that depression and loneliness predicted levels of POSI, which, in turn, predicted negative outcomes associated with the problematic use. Some years later, Caplan concluded that self-presentational social skill deficit was a strong negative predictor of POSI.²¹ Recently, Caplan¹⁷ found that the association between social anxiety and negative outcomes resulting from one's Internet use is indirect and is mediated by POSI. Surprisingly, the mediating role of POSI in the relationship between SE and IA has never been investigated.

Many authors have stated that low SE is central to the development of IA^1,10,17,18 because it is implicated in lapses of self-control, with the problematic aspects of heavy Internet use acting as direct expressions of such lapses.^23–25 Low SE is conceptually distinct from social anxiety, depression, and loneliness, although some evidence exists for moderate correlations among these variables.^18,26 The dispositional aspect of SE may explain its potentially central role, whereas loneliness is more situational (e.g., job relocation) and has nothing to do with one's perceived social skills.¹⁷ Moreover, the study of SE in relation to PIU is relevant because traditionally, low SE has been linked to addictive behaviors.²⁷ Low SE might lead some individuals to seek temporary relief through mind-altering activities.²⁸ A number of studies have actually found a significant correlation between SE and the Internet use.^{1,14,16,27,29}

The present study aims to extend previous findings regarding the association between poor social skills, POSI, and IA by focusing primarily on SE. We hypothesized that among a group of adolescents, SE will predict POSI, which, in turn, will predict IA(H₁).

Why might CMC be so attractive for persons with low SE? Face-to-face interaction and CMC differ in ways that may be appealing to those with interpersonal difficulties. Evaluative verbal and nonverbal cues (potentially relevant in face-to-face interactions) are not present in CMC. CMC allows greater control over self-presentation, and this greater control may create a sense of security,²⁰ allowing people to feel more free in their online interpersonal interactions than in face-to-face interactions³⁰; these liberating processes are particularly relevant to adolescents.³¹ A consensus has recently emerged concerning this explanation, as CMC has been determined to be associated with more intimate self-disclosure.³² Schouten³³ found that the subjective relevance given to both controllability and reduced nonverbal cues mediates the relationship between social anxiety and online self-disclosure. Another characteristic of CMC is the opportunity to escape from the real world. Physical isolation could lead adolescents with low SE to prefer online social interaction by heightening the focus on personal needs and private self-awareness, and by promoting their virtual and protected satisfaction.³⁴ Moreover, physical isolation could be an avoidance-coping strategy to deal with socially anxious situations.

The effect of SE on users' perception of the relevance of these Internet attributes has never been investigated. From a theoretical point of view, it is logical to suppose that people with low SE develop a preference for such attributes, which allow them to overcome their feeling of inadequacy. Moreover, no studies have investigated whether adolescents who attach more relevance to anonymity, controllability, and opportunity to escape develop a POSI. For these reasons, we hypothesized that the subjective relevance of CMC characteristics (anonymity, control, and opportunity to escape from the real world) will positively predict POSI and acts as a mediator between low SE and POSI(H₂).

Methods

Participants and procedures

Five public high schools in downtown Florence (Italy) were contacted for recruitment into the study; however, only three participated because of political demonstrations that disrupted school activities. In each of the three schools, we recruited students from the same cohort in order to constitute a homogenous sample. Fifty percent of all the freshmen classes (grade 9) were randomly selected by the principals of the three schools. In total, 264 students were contacted. Parental and student consent were obtained. During the data collection, 7 adolescents (M=1; F=6) were absent. In total, 257 adolescents (97.35 percent; M=95; F=162) participated in the study; more than half of them (54.47 percent) attend a humanities-oriented high school. The mean age was 14.57±0.69. All students received the same instructions and were told that their participation was voluntary and anonymous. Students filled out questionnaires in a classroom setting.

Instruments

Internet use

The questionnaire contained a single item regarding how many hours the participant spent online in a typical week (not counting the study-related use) for e-mail, searching for information, downloading, shopping, chat rooms, discussion forums, instant messages, blogging, online games, erotic Web sites, Facebook, and online gambling. The selection of the Internet services was based on previous studies.^15,35,36

Self-esteem

SE was assessed using the Italian adaptation³⁷ of the 10-item Rosenberg Self-Esteem Scale.³⁸ This version demonstrates good internal consistency (α=0.84). Each item was scored using a 5-point Likert scale. A sample item is, “On the whole, I am satisfied with myself.”

Preference for an online social interaction

The three items employed as indicators of POSI were based on a measure by Caplan³⁹ and were adapted for Italian students.⁴⁰ Participants rated their agreement on a scale ranging from 1=strongly disagree to 5=strongly agree. A sample item is, “Online social interaction is more comfortable for me than face-to-face interaction.” In the current study, the POSI scale was reliable (α=0.80).

Internet addiction

The Italian version⁴¹ of the Internet Addiction Test (IAT)⁴² was used. The IAT comprises 20 items on a scale ranging from 1 (never) to 5 (always). A sample item is, “How often do you find that you stay online longer than you intended?” In the current study, the IAT shows good internal consistency (α=0.90).

Perceived relevance of CMC attributes

Participants were asked to rate (on a scale ranging from 1=not at all to 4=a lot) the extent to which they like CMC for its (a) anonymity, (b) controllability, and (c) opportunity to escape from the real world. The item was, “I like to communicate by the Internet because of its …”

Results

In all, 100 percent of the respondents reported using the Internet, with an average of 12.63 hours spent online per week (SD=13.22). Almost all of the participants (96.9 percent) reported communicating online, predominantly using services such as Facebook (89.1 percent), chat rooms (64.6 percent), and instant messaging (52.1 percent). The most frequent noncommunicative uses included gathering information (90.0 percent) and downloading (80.0 percent). Forty-two percent of the respondents played online games, 24.5 percent had a blog, and 20 percent visited erotic Web sites. Discussion forums, shopping, and online gambling were the least reported activities (14.8 percent, 11.3 percent, and 5.8 percent, respectively). Table 1 shows descriptive statistics for the time spent online for each service.

Table 1.

Descriptive Statistics and Gender Differences

	Total sample (n=257)		Boys (n=95)		Girls (n=162)
	M	SD	M	SD	M	SD	F (1, 255)
Internet services (h/week)
Facebook	7.98	9.27	6.44	0.84	8.89	0.77	4.23^*
Chat rooms	3.77	5.76	2.46	0.44	4.54	0.50	8.06^**
Instant messaging	2.11	4.65	1.33	0.36	2.57	0.41	4.28^*
Discussion Forums	0.39	1.33	0.45	0.17	0.35	0.09	0.29
Blogging	0.55	1.73	0.47	0.22	0.59	0.11	0.32
E-mail	0.70	1.75	0.69	0.22	0.71	0.12	0.01
Information	2.20	2.78	2.33	0.34	2.13	0.19	0.30
Downloading	3.15	11.08	4.04	1.81	2.63	0.29	0.96
Shopping	0.14	0.54	0.16	0.07	0.13	0.03	0.27
Online gaming	1.12	2.62	2.15	0.40	0.51	0.08	25.69^**
Erotic Web sites	0.61	4.41	1.50	0.73	0.09	0.04	6.23^**
Online gambling	0.07	0.41	0.16	0.06	0.02	0.01	7.28^**
Psychological variables
Self-esteem	28.74	4.28	29.61	3.76	28.22	4.48	6.43^**
POSI	2.14	1.36	2.06	1.17	2.18	1.45	0.47
Anonymity	1.99	0.84	2.02	0.84	1.97	0.83	0.23
Controllability	2.31	0.99	2.09	0.93	2.43	1.00	6.04^**
Escape from the world	1.92	0.99	1.73	0.88	2.04	1.03	7.02^**

p<0.05; ^**p<0.01.

M, mean; SD, standard deviation; POSI, preference for online social interaction.

A one-way analysis of variance revealed that boys spent significantly more time playing online games, visiting sex Web sites, and gambling than girls. Girls spent significantly more time using CMC (chat rooms, instant messaging, and Facebook) than boys. Moreover, girls scored significantly higher than boys on the IAT (respectively, 40.43±13.09 and 35.78±9.29; F=9.23, p<0.01). The Pearson correlation analysis showed that IAT scores were correlated with time spent using some communicative services: Facebook (r=0.42, p<0.01), chat (r=0.30, p<0.01), instant messaging (r=0.22, p<0.01), and blogging (r=0.12, p<0.05). No significant correlations were found for playing online games (r=0.08, p=0.17), visiting sex Web sites (r=0.04, p=0.49), and gambling (r=0.00, p=0.99).

Girls reported significantly higher scores on the levels of perceived relevance of controllability and the opportunity to escape from the real world. A significant difference was also found for SE levels, with girls scoring lower than boys. No gender differences were found for the POSI scores.

Hypotheses testing

In accordance with Baron and Kenny,⁴³ we used a correlation analysis before testing our hypotheses to ensure that the independent, dependent, and mediator variables correlated with one another. Given the gender differences found, correlations were computed separately (Table 2).

Table 2.

Pearson's Correlations Between the Study Measures in Boys and Girls

	1	2	3	4	5	6
Self-esteem (1)	—	−0.18	0.15	0.01	−0.05	−0.20
POSI (2)	−0.21^**	—	−0.02	0.04	0.13	0.43^**
Anonymity (3)	−0.01	0.04	—	0.28^**	0.21	0.01
Controllability (4)	−0.19^*	0.24^**	0.08	—	0.23^**	0.12
Escape (5)	−0.25^**	0.24^**	0.10	0.33^**	—	0.24^**
IAT (6)	−0.24^**	0.43^**	0.02	0.45^**	0.40^**	—

p<0.05; ^**p<0.01.

Correlations for boys (n=95) are presented above the diagonal and correlations for girls (n=162) are presented below the diagonal.

IAT, Internet Addiction Test.

Because there were no significant correlations between the independent, dependent, and mediator variables for boys, the mediation analyses were performed only for girls.

H1: POSI as mediator of the relationship between SE and IAT

First, POSI was regressed on SE; second, IAT was regressed on SE; in the third equation, IAT was simultaneously regressed on both SE and POSI. As shown in Table 3, the effect of SE on IAT is less in the third equation than in the second equation (β dropped from −0.24 to −0.16 when POSI was added into the equation). The significance of this indirect effect was tested with a formula by Sobel. If the test leads to the critical z-value of 1.96, the mediator carries the influence of the independent variable to the dependent variable. The z-value for the mediation analysis was −2.64, p<0.01, indicating that POSI significantly mediated the relationship between SE and IAT. The mediation through POSI is partial, since SE still has a significant effect on IAT when POSI is controlled.

Table 3.

Regression Analysis Summary for Mediation Effect of Preference for Online Social Interaction on the Relationship Between Self-Esteem and Internet Addiction Test

	B	SE	β
First step
DV: POSI
IV: self-esteem	−0.068	0.025	−0.210^**
Second step
DV: IAT
IV: self-esteem	−0.711	0.224	−0.244^**
Third step
DV: IAT
IV: self-esteem	−0.467	0.210	−0.160^*
MV: POSI	3.591	0.649	0.399^**

p<0.05; ^**p<0.01.

DV, dependent variable; IV, independent variable; MV, mediating variable; SE, standard error.

H2: Perceived relevance of CMC attributes as mediators of the relationship between SE and POSI

Table 4 shows the results of the regression analyses in which the mediation effects of perceived relevance of controllability were tested. The effect of SE on POSI is less in the third equation than in the second equation (β dropped from −0.21 to −0.17 when controllability was added into the equation). However, the z-value was −1.61, p=0.10, indicating that the mediation effect was not statistically significant.

Table 4.

Regression Analysis Summary for Mediation Effect of Controllability on the Relationship Between Self-Esteem and Preference for Online Social Interaction Among Female Respondents

	B	SE	β
First step
DV: controllability
IV: self-esteem	−0.044	0.018	−0.197^*
Second step
DV: POSI
IV: self-esteem	−0.068	0.025	−0.210^**
Third step
DV: POSI
IV: self-esteem	−0.056	0.025	−0.174^*
MV: controllability	0.303	0.113	0.210^**

p<0.05; ^**p<0.01.

Table 5 shows the results of regression analyses in which the mediation effects of perceived relevance of the chance to escape from the real world were tested. The effect of SE on POSI is less in the third equation than in the second equation (β dropped from −0.21 to −0.16 when escape was added into the equation). The z-value was −2.00, p=0.04, revealing that escape significantly mediated the relationship between SE and POSI. The mediation is partial, since SE still has a significant effect on POSI when escape is controlled.

Table 5.

Regression Analysis Summary for Mediation Effect of Escape on the Relationship Between Self-Esteem and Preference for Online Social Interaction Among Female Respondents

	B	SE	β
First-step regression
DV: escape from the real world
IV: self-esteem	−0.057	0.018	−0.251^**
Second-step regression
DV: POSI
IV: self-esteem	−0.068	0.025	−0.210^**
Third-step regression
DV: POSI
IV: self-esteem	−0.051	0.025	−0.159^*
MV: escape from the real world	0.286	0.111	0.202^*

p<0.05; ^**p<0.01.

Since no correlations between anonymity, SE, and POSI were found, the mediation effect of the relevance of this CMC attribute was not tested.

Discussion

In recent years, POSI has become a central variable that could at least partially explain IA related to communicative services. Even though encouraging results have been provided,^17,22 previous investigations of POSI predictors have not considered the role of SE and users' perception of the relevance of Internet features. Both of these aspects must be investigated for two reasons. First, low SE is considered one of the main predictors of IA. Second, Internet characteristics, such as anonymity and controllability, are usually seen as structural, fixed aspects of CMC³³; however, the actual effects of these attributes depend on Internet users' perceptions of the relevance of these features.⁴⁴

In accordance with several other studies,⁸ we found that communicative services are the most utilized Internet applications among both males and females. Females use CMC more than males and are more Internet-addicted. Moreover, a significant association between usage of communicative services and IAT was found. These results confirm that IA is mainly associated with social aspects of the Web among adolescents. Accordingly, a rather strong association between POSI and IAT was confirmed.²² However, the association between low SE and POSI was significant only for females. For females, POSI partially mediated the relationship between SE and IAT. It is possible that low SE has a direct effect on IA levels in the sense that it leads some individuals to seek temporary relief through mind-altering activities such as the ones offered by the Web.²⁸

Some important gender differences emerged. We did not find a significant difference in POSI levels, but significant differences emerged for SE and IAT scores. One possible explanation is that there are some gender differences in the risk factors for POSI: for females, POSI may be based on the importance given to emotional proximity and the need for sharing.⁴⁵ Females with low SE may be likely to develop a POSI as a result of this need to share. For males, low SE could lead to solitary behaviors rather than affiliation behaviors. Among males, POSI may be independent of their SE levels. It is possible that these significant gender differences are linked to CMC features that were not investigated in the present study, such as the opportunity to talk with many people simultaneously or the opportunity to perform multiple tasks. These issues need to be explored in future studies. Alternatively, these gender differences could also be explained by the 2:1 female to male ratio in the sample. Correlations of 0.20 were statistically significant for females, but not for males because of the sample size. For this reason, the mediation analyses were performed only for girls.⁴³ Nonetheless, these associations need to be further explored among males in order to verify the size of the differences noted in the present study.

Our results contrast with the widespread expectation that, as a result of the developmental tasks of adolescence, anonymity and controllability represent the two most appealing characteristics of the Web in terms of the development of POSI.^20,33 Our findings indicate that this expectation is true only for females, and only regarding the role played by the perceived relevance of controllability rather than anonymity. Furthermore, among females, the CMC characteristic that emerged as the most important mediator between SE and POSI is the opportunity to escape from the real world. This result is interesting because the opportunity to escape is a characteristic that is often overlooked in favor of anonymity and controllability. The lack of an association between the perceived relevance attached to anonymity and POSI could be explained by the fact that we did not distinguish between source anonymity and audiovisual anonymity. We cannot assume that the respondents' understanding of anonymity is based on being unknown rather than a lack of audiovisual cues, and that adolescents usually use the Internet to communicate with already-known friends rather than to meet new people. Furthermore, future research should explore the association between users' perceptions of the relevance of the opportunity to control the richness of the cues they wish to convey and POSI levels.

Beyond this methodological consideration, several limitations of the present study should be noted. First, given the correlational design, inferences about causal directions are not warranted. Second, generalizability may be compromised by the recruitment strategies that were used. Students were recruited from general public high schools in Italy, which are oriented to prepare students for college. These results are not generalizable to students who attend vocational schools. Hence, further explorations of potential differences among adolescents from different types of secondary schools are needed. Moreover, students were recruited in schools from the urban center of a major Italian city. Adolescents from the suburbs or from smaller cities are not represented in our results. The sample also consisted entirely of younger students who were in the first year of high school. Our results cannot be extended to older students. Third, this study is limited by the use of a single item to measure each aspect of the perceived relevance of various features of the Internet.

Despite these limitations, the present findings build upon and confirm previous studies regarding the association between poor social skills and IA. Specifically, we extended findings regarding the role played by loneliness and social anxiety to SE, highlighting its effect on the preference of online social interaction.

Footnotes

Disclosure Statement

No competing financial interests exist.

References

Douglas

, Mills

, Niang

et al. Internet addiction: meta-synthesis of qualitative research for the decade 1996–2006. Computers in Human Behavior, 2008; 24:3027–3044.

Yellowlees

, Marks

. Problematic Internet use or Internet addiction? Computers in Human Behavior, 2007; 23:1447–1453.

Siomons

, Dafouli

, Braimiotis

et al. Internet addiction among Greek adolescent students. Cyberpsychology & Behavior, 2008; 11:653–657.

Liu

, Potenza

. Problematic Internet use: clinical implications. CNS Spectrums, 2007; 12:453–466.

Villella

, Martinotti

, Di Nicola

et al. Behavioral addictions in adolescents and young adults: Results from a prevalence study. Journal of Gambling Studies, 2011; 2:203–214.

Kim

, LaRose

, Peng

. Loneliness as the cause and the effect of problematic Internet use: the relationship between Internet use and psychological well-being. Cyberpsychology & Behavior, 2009; 12:451–455.

Desjarlais

, Willoughbya

. A longitudinal study of the relation between adolescent boys and girls' computer use with friends and friendship quality: support for the social compensation or the rich-get-richer hypothesis? Computers in Human Behavior, 2010; 26:896–905.

Gross

. Adolescent Internet use: What we expect, what teens report. Applied Developmental Psychology, 2004; 25:633–649.

van Rooij

, Schoenmakers

, van de Eijnden

RJJM

, van de Mheen

. Compulsive Internet use: the role of online gaming and other Internet applications. Journal of Adolescent Health, 2010; 47:51–57.

10.

Davis

. A cognitive-behavioral model of pathological Internet use. Computers in Human Behavior, 2001; 17:187–195.

11.

Kim

, Ryu

, Chon

et al. Internet addiction in Korean adolescents and its relation to depression and suicidal ideation: a questionnaire survey. International Journal of Nursing Studies, 2006; 43:185–192.

12.

Caplan

, Williams

, Yee

. Problematic Internet use and psychosocial wellbeing among MMO players. Computers in Human Behavior, 2009; 25:1312–1319.

13.

Ceyhan

, Ceyhan

. Loneliness, depression, and computer self-efficacy as predictors of problematic Internet use. Cyberpsychology & Behavior, 2008; 11:699–701.

14.

Ghassemzadeh

, Mehrnaz Shahraray

et al. Prevalence of Internet addiction and comparison of Internet addicts and non-addicts in Iranian high schools. Cyberpsychology & Behavior, 2008; 11:731–733.

15.

Casale

, Fioravanti

. Psychosocial correlates of Internet use among Italian students. International Journal of Psychology, 2011; 46:288–298.

16.

Yang

, Tung

. Comparison of Internet addicts and non-addicts in Taiwanese high school. Computers in Human Behavior, 2007; 23:79–96.

17.

Caplan

. Relation among loneliness, social anxiety, and problematic Internet use. Cyberpsychology & Behavior, 2007; 10:234–242.

18.

Kim

, Davis

. Toward a comprehensive theory of problematic Internet use: Evaluating the role of self-esteem, anxiety, flow, and the self-rated importance of Internet activities. Computers in Human Behavior, 2009; 25:490–500.

19.

Chak

, Leung

. Shyness and locus of control as predictors of Internet addiction and Internet use. Cyberpsychology & Behavior, 2004; 7:559–570.

20.

Valkenburg

, Peter

. Online communication among adolescents: an integrated model of its attraction, opportunities, and risks. Journal of Adolescent Health, 2011; 48:121–127.

21.

Caplan

. A social skill account of problematic Internet use. Journal of Communication, 2005; 55:721–736.

22.

Caplan

. Preference for online social interaction: a theory of problematic Internet use and psychosocial well-being. Communication Research, 2003; 30:625–648.

23.

La Rose

, Lin

, Eastin

. Unregulated Internet usage: addiction, habit, or deficient self-regulation? Media Psychology, 2003; 5:225–253.

24.

Stinson

, Logel

, Zanna

et al. The cost of lower self-esteem: Testing a self- and social-bonds model of health. Journal of Personality and Social Psychology, 2008; 94:412–428.

25.

Tangney

, Baumeister

, Boone

. High self-control predicts good adjustment, less pathology, better grades, and interpersonal success. Journal of Personality, 2004; 72:271–324.

26.

Leary

, MacDonald

. 2003. Individual differences in trait self-esteem: a theoretical integration. Leary

, Tangney

. Handbook of self and identity. New York: Guilford Press, 401–418.

27.

Armstrong

, Phillips

, Saling

. Potential determinants of heavier Internet usage. International Journal of Human-Computer Studies, 2000; 53:537–550.

28.

Greenberg

, Lewis

, Dodd

. Overlapping addictions and self-esteem among college men and women. Addictive Behaviors, 1999; 24:565–571.

29.

Young

. Psychology of computer use: XL. Addictive use of the Internet: a case that breaks the stereotype. Psychological Reports, 1996; 79:899–902.

30.

Bonetti

, Campbell

, Gilmore

. The Relationship of loneliness and social anxiety with children's and adolescents' online communication. Cyberpsychology, Behavior, and Social Networking, 2010; 13:279–285.

31.

Valkenburg

, Peter

. Social consequences of the Internet for adolescents. A decade of research. Current Directions in Psychological Science, 2009; 18:1–5.

32.

Tidwell

, Walther

. Computer-mediated communication effects on disclosure, impressions, and interpersonal evaluations: getting to know one another a bit at a time. Human Communication Research, 2002; 28:317–348.

33.

Schouten

, Valkenburg

, Peter

. Precursor and underlying processes of adolescents' online self-disclosure: Developing and testing an “Internet-Attribute-Perception” model. Media Psychology, 2007; 10:292–315.

34.

Sassenberg

, Jonas

. 2007. Attitude change and social influence on the net. Joinson

, McKenna

, Postmes

, Reips

. The Oxford handbook of Internet psychology. Oxford: Oxford University Press, 273–288.

35.

Amichai-Hamburger

, Ben-Artzi

. Loneliness and Internet use. Computers in Human Behavior, 2003; 19:71–80.

36.

Morahan-Martin

, Schumacher

. Incidence and correlates of pathological Internet use among college students. Computers in Human Behavior, 2000; 16:13–29.

37.

Prezza

, Trombaccia

, Armento

. La scala dell'autostima di Rosenberg: traduzione e validazione italiana [Rosenberg Self-Esteem Scale: Italian translation and validation] Bollettino di Psicologia Applicata, 1997; 223:35–44.

38.

Rosenberg

. 1965. Society and the adolescent self-image. Princeton, NJ: Princeton University Press.

39.

Caplan

. Problematic Internet use and psychosocial well-being: development of a theory based cognitive behavioral measurement instrument. Computers in Human Behavior, 2002; 18:553–575.

40.

Casale

, Giannini

, Fioravanti

. Psychometric properties of the Italian version of the Generalized Pathological Internet Use Scale. Paper Presented at the 118th Annual Convention of the American Psychological Association, San Diego, August 12–15, 2010.

41.

Ferraro

, Caci

, D'Amico

et al. Internet addiction disorder: an Italian study. Cyberpsychology & Behavior, 2007; 10:170–175.

42.

Young

. Internet addiction: The emergence of a new clinical disorder. Cyberpsychology & Behavior, 1998; 3:237–244.

43.

Baron

, Kenny

. The moderator-mediator variable distinction in social psychology research: conceptual, strategic, and statistical consideration. Journal of Personality and Social Psychology, 1986; 51:1173–1182.

44.

Peter

, Valkenburg

. Research Note: Individual differences in perceptions of Internet communication. European Journal of Communication, 2006; 21:213–226.

45.

Durell

. Gender, grade and relationship differences in emotional closeness within adolescent friendships. Adolescence, 2004; 39:243–255.