Addictive Behaviors Reports, volume 15, article 100426, June 2022 (online 12 April 2022)
- Open access
- 897 gamers
- 2019, 2020 and 2021
- Propensity score matching
- ANCOVA
- COVID-19
Free full text: the open access article (CC BY 4.0), re-typeset as a clean PDF with the same content as the published version. Please cite the published version.
- gaming disorder
- Internet gaming disorder
- loneliness
- family harmony
- COVID-19
- pandemic
- disordered gaming
- problematic gaming
- video games and mental health
- gamers' well-being
- lockdowns and social isolation
- family functioning
- coping through gaming
- WHO ICD-11 gaming disorder
- APA DSM-5 Internet gaming disorder
- Gaming Disorder Test (GDT)
- Internet Gaming Disorder Scale – Short-Form (IGDS9-SF)
- Family Harmony Scale (FHS-5)
- UCLA Loneliness Scale (ULS-3)
- repeated cross-sectional survey
- propensity score matching
- analysis of covariance (ANCOVA)
- Spearman partial correlation
- Steiger test
- R
- international online survey
Gamers reported more symptoms of disordered gaming during the COVID-19 pandemic than before it, while their loneliness and family harmony did not change. In three matched samples of 299 gamers each (2019, 2020 and 2021), disordered gaming also became more closely tied to loneliness and poorer family harmony, which suggests some gamers may have played to cope.
Key findings
- Disordered gaming rose during the pandemic. Gaming Disorder scores rose from 8.96 in 2019 to 10.69 in 2020 and 11.18 in 2021 (F = 30.99, p < .001, η²p = 0.065), and Internet Gaming Disorder scores from 18.29 to 21.92 and 22.80 (F = 30.00, p < .001), medium-sized increases over 2019 (Cohen’s d = 0.47 to 0.62) (Rozgonjuk et al., 2022).
- Loneliness and family harmony stayed level. Across the three matched samples of 299 gamers, loneliness (p = .342) and dysfunctional family harmony (p = .576) did not differ between years, controlling for age and gender (Rozgonjuk et al., 2022).
- Disordered gaming became more closely tied to loneliness. The correlation between loneliness and Gaming Disorder grew from ρ = 0.207 in 2019 to 0.393 in 2021 (z = 2.50, p = .013), and with Internet Gaming Disorder from 0.293 to 0.463 (z = 2.42, p = .015) (Rozgonjuk et al., 2022).
- And to poorer family harmony in 2021. Gaming Disorder correlated more strongly with dysfunctional family harmony in 2021 (ρ = 0.305) than in 2019 (0.121) or 2020 (0.122); for Internet Gaming Disorder the 2021 rise was significant only against 2020 (0.143 to 0.307, p = .035) (Rozgonjuk et al., 2022).
At a glance
- Design
- Repeated cross-sectional international online survey (Smart Gaming campaign, ESL) run in 2019, 2020 and 2021; three yearly samples matched by propensity scores; ANCOVAs with age and gender as covariates, Spearman partial correlations and Steiger tests in R 4.1.3
- Participants
- 897 gamers aged 12 to 56 (299 per year), matched from 37,394 eligible respondents on age, gender, country, education, employment and relationship status; mean age 24.4 to 25.2, 82% to 86% men; most often from France, Italy, the USA, Spain and Poland
- Disordered gaming
- Gaming Disorder (WHO framework): Gaming Disorder Test, 4 items, α = 0.81 (Pontes et al., 2019). Internet Gaming Disorder (APA framework): IGDS9-SF, 9 items, α = 0.87 (Pontes & Griffiths, 2015)
- Family harmony
- Family Harmony Scale, 5 items, scored so that higher means more dysfunctional family harmony, α = 0.90 (Kavikondala et al., 2016)
- Loneliness
- Three-item UCLA Loneliness Scale, α = 0.83 (Hughes et al., 2004)
- Ethics
- Approved by the research team’s university ethics committee (PONTES 2018/95, Nottingham Trent University); anonymous participation with electronic informed consent
Figure 1. Estimated marginal means (with 95% CIs) across years, controlled for age and gender
Correlations of disordered gaming with loneliness and family harmony, by year
Tables
Table 1. Sample socio-demographics grouped by each year
| Variable | 2019 (N = 299) | 2020 (N = 299) | 2021 (N = 299) |
|---|---|---|---|
| Age | |||
| Mean (SD) | 24.45 (7.31) | 24.37 (7.51) | 25.23 (8.37) |
| Gender | |||
| Male | 257 (85.95%) | 251 (83.95%) | 246 (82.27%) |
| Female | 42 (14.05%) | 48 (16.05%) | 53 (17.73%) |
| Employment status | |||
| Employed | 140 (46.82%) | 134 (44.82%) | 139 (46.49%) |
| Relationship status | |||
| In a relationship | 115 (38.46%) | 109 (36.45%) | 114 (38.13%) |
| Education level | |||
| < High school | 43 (14.38%) | 46 (15.38%) | 37 (12.37%) |
| High school diploma | 150 (50.17%) | 147 (49.16%) | 157 (52.51%) |
| College/university degree | 106 (35.45%) | 106 (35.45%) | 105 (35.12%) |
| Top five countries (by total N) | |||
| France | 33 (11.04%) | 32 (10.070%) | 31 (10.37%) |
| Italy | 30 (10.03%) | 33 (11.04%) | 30 (10.03%) |
| USA | 24 (8.03%) | 29 (9.70%) | 26 (8.70%) |
| Spain | 22 (7.36%) | 32 (10.70%) | 21 (7.02%) |
| Poland | 23 (7.69%) | 22 (7.36%) | 25 (8.36%) |
Values as printed in Table 1 of the article; samples matched by propensity scores on age, gender, country, education, employment and relationship status.
Table 2. Descriptive statistics for key variables
| Variable | 2019 M | 2019 SD | 2020 M | 2020 SD | 2021 M | 2021 SD |
|---|---|---|---|---|---|---|
| GD | 8.96 | 3.30 | 10.69 | 3.81 | 11.18 | 3.95 |
| IGD | 18.29 | 6.54 | 21.92 | 8.01 | 22.80 | 8.09 |
| DFH | 12.76 | 5.24 | 13.10 | 4.96 | 12.67 | 4.90 |
| Loneliness | 6.89 | 2.80 | 7.11 | 2.82 | 7.24 | 2.83 |
M = mean; SD = standard deviation; GD = Gaming Disorder; IGD = Internet Gaming Disorder; DFH = dysfunctional family harmony.
Table 3. Results of analyses of covariance
| Variable | SS | df | MS | F | p | η²p |
|---|---|---|---|---|---|---|
| Outcome: Gaming Disorder (WHO framework) | ||||||
| Year | 837.388 | 2 | 418.694 | 30.986 | < 0.001 | 0.065 |
| Age | 152.079 | 1 | 152.079 | 11.255 | < 0.001 | 0.012 |
| Gender | 0.972 | 1 | 0.972 | 0.072 | 0.789 | < 0.001 |
| Residuals | 12053.169 | 892 | 13.513 | |||
| Outcome: Internet Gaming Disorder (APA Framework) | ||||||
| Year | 3445.648 | 2 | 1722.824 | 30.000 | < 0.001 | 0.063 |
| Age | 131.046 | 1 | 131.046 | 2.282 | 0.131 | 0.003 |
| Gender | 3.156 | 1 | 3.156 | 0.055 | 0.815 | < 0.001 |
| Residuals | 51228.114 | 892 | 57.431 | |||
| Outcome: Loneliness | ||||||
| Year | 16.922 | 2 | 8.461 | 1.074 | 0.342 | 0.002 |
| Age | 11.898 | 1 | 11.898 | 1.510 | 0.219 | 0.002 |
| Gender | 53.285 | 1 | 53.285 | 6.762 | 0.009 | 0.008 |
| Residuals | 7029.120 | 892 | 7.880 | |||
| Outcome: Dysfunctional Family Harmony | ||||||
| Year | 27.769 | 2 | 13.884 | 0.552 | 0.576 | 0.001 |
| Age | 92.869 | 1 | 92.869 | 3.689 | 0.055 | 0.004 |
| Gender | 149.283 | 1 | 149.283 | 5.930 | 0.015 | 0.007 |
| Residuals | 22454.191 | 892 | 25.173 |
SS = sum of squares; MS = mean square; η²p = partial eta squared. Each outcome was tested in a separate ANCOVA with age and gender as covariates.
Table 4. Outcome variable comparisons across years (controlled for age and gender)
| Comparison | MD | SE | t | p | Cohen’s d |
|---|---|---|---|---|---|
| Outcome: Gaming Disorder | |||||
| 2019–2020 | −1.727 | 0.301 | −5.742 | < 0.001 | −0.470 |
| 2019–2021 | −2.269 | 0.301 | −7.534 | < 0.001 | −0.617 |
| 2020–2021 | −0.542 | 0.301 | −1.802 | 0.072 | −0.148 |
| Outcome: Internet Gaming Disorder | |||||
| 2019–2020 | −3.625 | 0.620 | −5.847 | < 0.001 | −0.478 |
| 2019–2021 | −4.546 | 0.621 | −7.323 | < 0.001 | −0.600 |
| 2020–2021 | −0.921 | 0.621 | −1.485 | 0.138 | −0.122 |
| Outcome: Loneliness | |||||
| 2019–2020 | −0.199 | 0.230 | −0.869 | 0.771 | −0.071 |
| 2019–2021 | −0.335 | 0.230 | −1.456 | 0.437 | −0.119 |
| 2020–2021 | −0.135 | 0.230 | −0.589 | 0.771 | −0.048 |
| Outcome: Dysfunctional Family Harmony | |||||
| 2019–2020 | −0.319 | 0.410 | −0.777 | 0.952 | −0.064 |
| 2019–2021 | 0.092 | 0.411 | 0.224 | 0.952 | 0.018 |
| 2020–2021 | 0.411 | 0.411 | 1.000 | 0.952 | 0.082 |
In all comparisons df = 892. MD = mean difference (earlier minus later year); p values adjusted with Holm’s method; significant differences in bold.
Table 5. Spearman correlation coefficients and correlation differences test results
| Variables | ρ 2019 | ρ 2020 | ρ 2021 | z 2019–2020 | p | z 2019–2021 | p | z 2020–2021 | p |
|---|---|---|---|---|---|---|---|---|---|
| GD & IGD | 0.751*** | 0.800*** | 0.802*** | 1.501 | 0.133 | 1.569 | 0.117 | 0.068 | 0.946 |
| GD & DFH | 0.121* | 0.122* | 0.305*** | 0.012 | 0.990 | 2.353 | 0.019 | 2.341 | 0.019 |
| GD & LON | 0.207 | 0.346*** | 0.393*** | 1.835 | 0.066 | 2.498 | 0.013 | 0.662 | 0.508 |
| IGD & DFH | 0.176 | 0.143* | 0.307*** | 0.412 | 0.680 | 1.696 | 0.090 | 2.108 | 0.035 |
| IGD & LON | 0.293*** | 0.427*** | 0.463*** | 1.878 | 0.060 | 2.424 | 0.015 | 0.546 | 0.585 |
| DFH & LON | 0.253*** | 0.224*** | 0.311*** | 0.374 | 0.708 | 0.767 | 0.443 | 1.141 | 0.254 |
GD = Gaming Disorder; IGD = Internet Gaming Disorder; DFH = dysfunctional family harmony; LON = loneliness. Spearman partial correlations controlling for age; p values for correlations corrected with Holm’s method for each year; z = Steiger test of the difference between two years. Significance marks as printed in the article: “*** p <.001, p <.01, * p <.01”. Significant differences in bold.
The question
Lockdowns and distancing during the COVID-19 pandemic pushed many people towards gaming, a pastime that can be enjoyed without anyone physically present, and campaigns even encouraged it. At the same time, isolation can breed loneliness and strain family life. Little was known about how gamers in particular fared. Comparing matched samples of gamers surveyed before the pandemic (2019) and during its first two years (2020 and 2021), we asked whether disordered gaming, loneliness and dysfunctional family harmony increased, and whether disordered gaming became more strongly linked with loneliness and poor family harmony.
Implications
Practical implications
Health professionals, families and the gaming industry should watch for rising disordered gaming among gamers in times of crisis, and for gaming used to cope with loneliness or a difficult home life. Prevention and support can target the link itself: gamers who feel lonely or whose family relations are strained are those for whom heavy gaming is most closely tied to problems.
Theoretical implications
The study shows that a sub-population such as gamers may not follow the same pandemic trends as the general population: loneliness and family harmony did not rise, yet disordered gaming did and became more closely tied to both. It also shows that results can differ depending on whether the WHO (Gaming Disorder) or APA (Internet Gaming Disorder) framework is used, supporting the study of both.
Limitations and future research
- Data were self-reported in a cross-sectional online survey, so changes cannot be attributed causally to the pandemic.
- Each year is a different sample of gamers; the same people were not followed over time.
- Daily-life impairment was self-assessed; clinical interviews or objective gaming data would strengthen the findings.
- The samples were mostly young men, matched on socio-demographics rather than representative of all gamers.
Questions this paper answers
Did gaming disorder increase during COVID-19?
Among gamers, yes. Matched samples of 299 gamers each showed higher Gaming Disorder and Internet Gaming Disorder scores in 2020 and 2021 than in 2019, with medium effect sizes (Rozgonjuk et al., 2022).
Were gamers lonelier during the pandemic?
Not on average: loneliness and family harmony did not change significantly between 2019, 2020 and 2021. But loneliness became more closely linked with disordered gaming, the correlation rising from about 0.2 to 0.4 (Rozgonjuk et al., 2022).
Do people game to cope with loneliness?
The stronger links between disordered gaming, loneliness and poor family harmony during the pandemic suggest that some gamers may have used gaming to cope, although the cross-sectional data cannot prove this (Rozgonjuk et al., 2022).
What is the difference between Gaming Disorder and Internet Gaming Disorder?
Gaming Disorder is the WHO’s officially recognised diagnosis (ICD-11); Internet Gaming Disorder is the APA’s research diagnosis in the DSM-5. The study measured both and found broadly similar trends.
Key concepts
- Gaming Disorder (GD)
- A pattern of gaming with impaired control, increasing priority over other activities and continuation despite harm, recognised by the World Health Organization in ICD-11.
- Internet Gaming Disorder (IGD)
- A condition for further study in the American Psychiatric Association’s DSM-5, defined by nine criteria such as preoccupation, withdrawal and loss of interest in other activities.
- Family harmony
- How peaceful, functional and harmonious a person’s family relationships are; reverse-scored here as dysfunctional family harmony.
- Loneliness
- The distressing feeling that one’s social relationships are lacking in number or quality.
- Propensity score matching
- A method that pairs respondents from different groups who have similar characteristics, so the groups can be compared fairly.
Open science and materials
- Full text (PDF) Open access article (CC BY 4.0), re-typeset
- Published version Addictive Behaviors Reports (open access), with supplementary data
Study materials
- Eligibility
- Respondents gave informed consent, were aged 12 to 80, proficient in English, had played video games in the past year, passed an attention check (denying play of a fictional game), were not professional gamers, and reported at most 119 hours of gaming a week and 48 hours at weekends.
- Matching
- From 47,503 eligible gamers, the analysis kept comparable ages (12 to 56) and countries with at least five respondents in 2020 and 2021 (37,394 gamers: 36,078 in 2019, 1,017 in 2020, 299 in 2021); the 2019 and 2020 samples were then matched to the 2021 sample by greedy nearest-neighbour propensity scores (seed 999).
- Supplementary material
- The R packages used are listed in the article’s supplementary data, available with the published version.
How to cite
Rozgonjuk, D., Pontes, H. M., Schivinski, B., & Montag, C. (2022). Disordered gaming, loneliness, and family harmony in gamers before and during the COVID-19 pandemic. Addictive Behaviors Reports, 15, Article 100426. https://doi.org/10.1016/j.abrep.2022.100426
Open access under a Creative Commons CC BY 4.0 licence.