Flow on the Internet: a longitudinal study of Internet addiction symptoms during adolescence

ABSTRACT

Internet Addiction (IA) constitutes an excessive Internet use behavior with a significant impact on the user’s well-being. Online flow describes the users’ level of being absorbed by their online activity. The present study investigated age-related, gender, and flow effects on IA in adolescence. The sample comprised 648 adolescents who were assessed twice at age 16 and 18 years. IA was assessed using the Internet Addiction Test and online flow was assessed using the Online Flow Questionnaire. A three-level hierarchical model estimated age-related, gender, and online flow effects on IA symptoms and controlled for clustered random effects. IA symptoms decreased over time (for both genders) with a slower rate in males. Online flow was associated with IA symptoms and this remained consistent over time. Findings expand upon the available literature suggesting that IA symptoms could function as a development-related manifestation at the age of 16 years, while IA-related gender differences gradually increase between 16 and 18 years. Finally, the association between online flow and IA symptoms remained stable independent of age-related effects. The study highlights individual differences and provides directions for more targeted prevention and intervention initiatives for IA.

KEYWORDS:

• Internet addiction
• online addictions
• adolescent addiction
• longitudinal study
• flow online

Acknowledgements

VS contributed to the data collection and analyses. RG contributed to the literature review and hypotheses formulation. MG contributed to the structure, sequencing of theoretical arguments, and paper editing. TB contributed to the data analyses. DK and YD contributed to the literature review.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Tyrone L. Burleigh http://orcid.org/0000-0002-3405-140X

Notes

1 The present data have been used in three more published studies that address different theoretical questions (Stavropoulos et al. 2016a, 2017a; Stavropoulos, Gentile, and Motti-Stefanidi 2016b). Instruments used in the data include the: (i) Internet Addiction Test IAT (Young 1998); (ii) Presence II questionnaire (Witmer and Singer 1998); (iii) Online Flow Questionnaire (Chen, Wigand, and Nilan 1999); (iv) Symptom Check List 90 (Derogatis and Savitz 1994); (v) Rosenberg Self-Esteem Scale (Rosenberg 1965); (vi) Five Factor Questionnaire for Children (Fünf-Faktoren-Fragebogen für Kinder) (Asendorpf and Van Aken 2003); (vii) Generalized Self-Efficacy Scale (Schwarzer 1993); (viii) Family Adherence and Cohesion Evaluation Scale (Olson 2000); (ix) Socio-metric Questionnaire (Coie, Dodge, and Coppotelli 1982); (x) Greek version of the Experience of Close Relationships Revised (Tsagarakis, Kafetsios, and Stalikas 2007); (xi) demographic and internet use questions; and (xii) school grades of the participants were retrieved from their school records.

2 The data abide with the sample size requirements suggesting: (a) a minimum ratio of 10_clusters/5_participants to test for fixed effects and cross-level interactions in models with one explanatory variable at each of the levels, and: (b) a minimum requirement of 30 clusters for testing standard errors of fixed effects (Maas and Hox 2004, 2005).

3 Conducting covariance based structural equation modeling (CBSEM) was not selected as: (a) it requires at least three or four indicators (the current study includes two time points) for every latent variable (growth) (Baumgartner and Homburg 1996); and (b) it assumes multi-normal distribution of the observed variables to ensure meaningful results-which is rarely the case in empirical research (Micceri 1989). Similarly, latent growth modeling was not chosen as it assumes that Level 1 predictors with random effects have the same distribution across all participants in each subpopulation, while HLM allows different distributions (Raudenbush and Bryk 2002). Finally, HLM was preferred over partial least square analysis, as it estimates the effects of variables on the outcome variable at one level (i.e. individual), while at the same time taking into account the effect of variables on the outcome variable at another level (i.e. classroom) (Raudenbush and Bryk 2002).

Additional information

Funding

Data collection in Greece for this study has been co-financed by the European Union (European Social Fund – ESF) and Greek national funds, under the Operational Program ‘Education and Lifelong Learning’ of the National Strategic Reference Framework (NSRF) – Research Funding Program: Heracleitus II.