Skip to main content

Problematic Online Behaviors Among Gamers: the Links Between Problematic Gaming, Gambling, Shopping, Pornography Use, and Social Networking

International Journal of Mental Health and Addiction, 2023

Dmitri Rozgonjuk, Bruno Schivinski, Halley M. Pontes, Christian Montag

International Journal of Mental Health and Addiction, volume 21, issue 1, pages 240–257, 2023 (online 6 October 2021)

  • 4,416 gamers
  • International sample
  • Exploratory graph analysis
  • Network analysis
  • Problematic Internet use
Download the accepted manuscript (PDF)Published version

Free full text: the authors’ accepted manuscript, identical in content to the published article. Please cite the published version.

Keywords
  • gaming disorder
  • gaming
  • gambling
  • exploratory graph analysis
  • network analysis
  • addictive behaviors
  • Internet addiction
  • problematic online gaming
  • problematic online gambling
  • problematic online shopping
  • problematic online pornography use
  • problematic social networking
  • problematic Internet use
  • specific vs generalised Internet addiction
  • co-occurring behavioural addictions
  • gamers' mental health
  • I-PACE model
  • Internet use disorders
  • ICD-11 gaming disorder
  • DSM-5 Internet gaming disorder
  • four-item problematic behaviour scales (Müller et al.)
  • Gaussian graphical model
  • Louvain community detection
  • bootstrapped EGA
  • within-subjects ANOVA
  • Spearman correlation
  • R
  • international online survey

Gamers who report problematic gaming also tend to report problems with other online activities, but each problem stands on its own. In a survey of 4,416 gamers from many countries, problematic gaming, gambling, shopping, pornography use and social networking correlated with one another yet formed five separate clusters, with gaming the most and shopping the least problematic.

Key findings

  1. Problematic gaming goes with other problematic online behaviours. Among 4,416 gamers, problematic online gaming correlated positively with problematic social networking (r = 0.389), pornography use (0.318), gambling (0.299) and shopping (0.269), all p < .001 (Rozgonjuk et al., 2021).
  2. But each behaviour is a separate problem. Exploratory graph analysis found five dimensions, one per behaviour, in 97.4% of 1,000 bootstrapped networks: every item loaded on its own scale, with only one cross-loading of note (gaming item 4 on gambling, −0.15) (Rozgonjuk et al., 2021).
  3. Gaming was the most problematic, shopping the least. On scales from 4 to 20, problematic gaming averaged 10.77, ahead of social networking (8.00), gambling (7.45), pornography (6.91) and shopping (5.77); the gap between gaming and the others was medium to large (Cohen’s d = 0.71 to 1.27; F(4, 17,660) = 1981.98, p < .001) (Rozgonjuk et al., 2021).
  4. Social networking was most closely tied to gaming. Problematic social networking had the strongest link with problematic gaming, which the authors suggest may point to unmet social needs among gamers with disordered gaming (Rozgonjuk et al., 2021).

At a glance

Design
Cross-sectional international online survey in English (Smart Gaming campaign, ESL), 2019 data; convenience sample rewarded with personalised feedback on their gaming
Participants
4,416 gamers aged 12 to 80 (mean 23.3, SD 6.7), 94% men, from many countries, most often Russia (17%), France (15%), Spain (11%), Poland (6%) and Argentina (5%); 5,905 before data cleaning
Measures
Four-item scales of problematic online gaming, gambling, shopping, pornography use and social networking (Müller et al., 2017), each item from 1 = never to 5 = very often; α and ω = 0.79 to 0.83
Analysis
Spearman correlations (Holm-adjusted); exploratory graph analysis with a Gaussian graphical model (GLASSO, EBIC) and Louvain community detection, bootstrapped 1,000 times (EGAnet); within-subjects ANOVA with Holm post hoc tests, in R 4.0.3
Ethics
As printed: approved by the Institutional Review Board of Ulm University. A correction (2 May 2022) states that ethical approval was given by the College Research Ethics Committee of Nottingham Trent University (2018/95)

Five separate problematic behaviours, linked by small to medium correlations

0.1930.2260.2310.2690.2990.3170.3180.3330.3580.389GamingM = 10.77loadings 0.25–0.32GamblingM = 7.45loadings 0.26–0.45ShoppingM = 5.77loadings 0.31–0.43PornographyM = 6.91loadings 0.34–0.45Social networkingM = 8.00loadings 0.27–0.42Line thickness and labels: Spearman correlations between summed scale scores (all p < .001; N = 4,416)
Each behaviour formed its own dimension in the exploratory graph analysis (range of each scale’s network loadings from Table 2); lines show the Spearman correlations between summed scale scores from Table 1. This is a summary drawn from the tables, not the published item-level network (Fig. 1 of the article, included in the manuscript PDF).

Figure 2. Means and 95% confidence intervals for scale scores

567891011Problematic online gaming10.77Problematic online social networking8.00Problematic online gambling7.45Problematic online pornography6.91Problematic online shopping5.77Mean scale score (4 to 20) with 95% confidence interval
Redrawn from Fig. 2 of the article, with intervals computed from the means and standard deviations in Table 1 (N = 4,416). Every pairwise difference was significant (Table 3).

Tables

Table 1. Descriptive statistics and correlations for problematic Internet-based activities and age

Variable Mean SD 1 2 3 4 5
1. Problematic online gaming 10.77 3.76
2. Problematic online gambling 7.45 3.74 0.299***
3. Problematic online shopping 5.77 2.64 0.269*** 0.231***
4. Problematic online pornography 6.91 3.30 0.318*** 0.193*** 0.317***
5. Problematic online social networking 8.00 3.59 0.389*** 0.226*** 0.358*** 0.333***
6. Age 23.31 6.72 −0.044* −0.087*** 0.040* −0.008 −0.157***

N = 4,416. * p < .05, ** p < .01, *** p < .001. Spearman correlations, p values adjusted with Holm’s method. The possible score range for each scale is 4 to 20.

Table 2. Network loadings for bootstrapped EGA items
Item 1 2 3 4 5
Online Gaming 3 0.317 0.068 0.012 0.029 0.007
Online Gaming 1 0.312 0.056 0.003 0.004 0.046
Online Gaming 2 0.306 0.081 0.003 0.023 0.085
Online Gaming 4 0.253 −0.150 0.020 0.056 0.081
Online Gambling 1 0.086 0.446 0 0 0
Online Gambling 2 0.096 0.432 0.013 0 0.015
Online Gambling 4 0.116 0.395 0.013 0 0
Online Gambling 3 0.071 0.256 0.064 0.012 0.004
Online Shopping 2 0.003 0.035 0.431 0.038 0.043
Online Shopping 1 0.003 0 0.390 0.021 0.003
Online Shopping 4 0.020 0.013 0.383 0.046 0.062
Online Shopping 3 0.012 0.039 0.313 0.031 0.045
Online Pornography 2 0.022 0 0.032 0.448 0.036
Online Pornography 4 0.055 0 0.051 0.374 0.060
Online Pornography 1 0.004 0.005 0.020 0.368 0
Online Pornography 3 0.029 0.005 0.031 0.340 0.016
Social Networking 1 0.046 0 0 0 0.423
Social Networking 2 0.086 0.014 0.043 0.037 0.381
Social Networking 3 0.007 0.003 0.047 0.016 0.357
Social Networking 4 0.082 0 0.062 0.061 0.270

Each item’s highest loading in bold. Dimensions 1 to 5 correspond to gaming, gambling, shopping, pornography and social networking.

Table 3. The comparison of Internet-based problematic behaviors

Comparison Estimate SE t Cohen’s d
Online gambling – online gaming −3.321 0.059 −56.102*** −0.844
Online pornography – online gaming −3.854 0.059 −65.108*** −0.980
Online networking – online gaming −2.772 0.059 −46.836*** −0.705
Online shopping – online gaming −5.001 0.059 −84.493*** −1.271
Online pornography – online gambling −0.533 0.059 −9.006*** −0.136
Social networking – online gambling 0.548 0.059 9.266*** 0.139
Online shopping – online gambling −1.680 0.059 −28.391*** −0.427
Social networking – online pornography 1.082 0.059 18.272*** 0.275
Online shopping – online pornography −1.147 0.059 −19.385*** −0.292
Online shopping – social networking −2.229 0.059 −37.657*** −0.567

*** p < .001 (Holm-adjusted post hoc tests). SE = standard error.

The question

Researchers still debate whether "Internet addiction" is one general problem or a set of specific problems that happen to take place online. Gamers are a group in which problematic gaming is common, but little was known about whether they also struggle with other online activities, and whether those problems overlap or stay distinct. In a large international sample of gamers, we asked how problematic online gaming, gambling, shopping, pornography use and social networking relate to each other, whether their symptoms form separate clusters, and which is most prevalent.

Implications

Practical implications

Clinicians and prevention programmes working with gamers should screen for other problematic online behaviours, since they often co-occur with disordered gaming, and especially for problematic social networking, which may reflect unmet social needs. Because each behaviour is distinct, treatment and prevention may need to target the specific activity rather than "Internet use" in general.

Theoretical implications

The results support the view of specific Internet use disorders: the Internet is the medium that fuels several distinct problematic behaviours rather than a single general addiction. They are consistent with frameworks such as I-PACE and with treating gaming disorder as one facet of a wider set of Internet-based problematic behaviours.

Limitations and future research

  • The sample was self-selected, mostly young and male, and recruited by convenience, so the results may not generalise.
  • Self-reports may not match actual, objectively measured gaming, gambling, shopping, pornography and social media use.
  • The cross-sectional design cannot show how these problems develop or influence one another over time.
  • Future work could study female gamers, older cohorts and relationship satisfaction in the context of gaming.

Questions this paper answers

Do gamers have other problematic online behaviours?

Often, yes. Among 4,416 gamers, problematic gaming correlated with problematic social networking, pornography use, gambling and shopping (r = 0.27 to 0.39), although gaming was by far the most problematic of the five (Rozgonjuk et al., 2021).

Is Internet addiction one disorder or several?

This study supports several: in a network analysis of gamers’ symptoms, problematic gaming, gambling, shopping, pornography use and social networking each formed a separate cluster, even though they were correlated (Rozgonjuk et al., 2021).

Which online behaviour is most linked to problematic gaming?

Problematic social networking (r = 0.389), followed by problematic pornography use (0.318) and gambling (0.299) (Rozgonjuk et al., 2021).

What is exploratory graph analysis?

A network method that estimates the partial correlations between questionnaire items and uses a community detection algorithm to find how many clusters, or dimensions, the items form.

Key concepts

Problematic Internet use
Excessive or poorly controlled use of the Internet that causes problems in daily life; also called Internet addiction or Internet use disorder.
Specific vs generalised problematic Internet use
Whether problems concern particular online activities, such as gaming or shopping, or the Internet in general.
Gaming disorder
A pattern of gaming with impaired control, increasing priority over other activities and continuation despite harm (WHO, ICD-11).
Exploratory graph analysis (EGA)
A network psychometrics method that estimates item networks and detects clusters of items as dimensions.
Network loading
The sum of an item’s connections to the items of a dimension; values above 0.15, 0.25 and 0.35 are small, medium and large.

Open science and materials

Study materials

Scale items
For each behaviour (gaming, gambling, shopping, pornography, social networking) the four items were: “How often do you find that you spend more time [doing the behavior] than you intended?”; “How often do you neglect household chores to spend more time [doing the behavior]?”; “How often do you feel preoccupied with [doing the behavior] when offline or fantasize about [doing the behavior]?”; “How often do you choose to spend more time [doing the behavior] over going out with others?” (1 = never to 5 = very often; Müller et al., 2017).
Exclusion criteria
Respondents were excluded if they were outside 12 to 80 years, had not played video games in the past 12 months, reported playing a non-existent game (attention check), reported more than 119 hours of gaming a week or more than 48 hours at weekends, declared themselves professional gamers, or lacked sufficient English.
Correction
A correction published on 2 May 2022 replaced the third sentence of the second paragraph of the Sample and Procedure section with: “The study has received ethical approval by the research team’s University Ethics Committee (College Research Ethics Committee of the Nottingham Trent University [2018/95]).”

How to cite

Rozgonjuk, D., Schivinski, B., Pontes, H. M., & Montag, C. (2023). Problematic Online Behaviors Among Gamers: the Links Between Problematic Gaming, Gambling, Shopping, Pornography Use, and Social Networking. International Journal of Mental Health and Addiction, 21(1), 240–257. https://doi.org/10.1007/s11469-021-00590-3

Related publications

All publications