Computers in Human Behavior: Wenliang Su, Xiaoli Han, Hanlu Yu, Yiling Wu, Marc N. Potenza
Computers in Human Behavior: Wenliang Su, Xiaoli Han, Hanlu Yu, Yiling Wu, Marc N. Potenza
Computers in Human Behavior: Wenliang Su, Xiaoli Han, Hanlu Yu, Yiling Wu, Marc N. Potenza
A R T I C L E I N F O A B S T R A C T
Keywords: Males have been proposed to be more vulnerable to internet addiction (IA) than females. However, males and
Internet addiction females may differ with respect to specific patterns and types of internet usage and related IA. To investigate
Internet gaming disorder further, a meta-analysis was conducted to investigate gender-related differences in internet gaming disorder
Problematic online gaming
(IGD) and social media addiction (SMA). The current meta-analysis aimed to quantify gender-related effect sizes
Social media addiction
Gender difference
relating to IGD and SMA, examine potential moderating influences of regions and other possible confounds, and
Meta-analysis compare the findings with generalized IA at the country level. The meta-analysis comprised 53 effect sizes with
82,440 individuals from 21 countries/regions for IGD, and 41 effect sizes with 58,336 individuals from 22
countries/regions for SMA. A random-effects model confirmed important gender-specific distinctions as men
were more likely to exhibit IGD than women (g ¼ 0.479) and less likely to exhibit SMA than women (g ¼
0.202). Additional moderator analyses revealed that effect sizes of IGD and SMA were larger in Europe and the
Americas than in Asia. Further analyses indicated that the effect sizes of gender-related differences in IGD and
SMA at the country-level were significantly larger than those in generalized IA, which suggests that gender-
related differences in specific IAs may be underestimated in the “umbrella” of generalized IA. Results have
implications for explaining why males and females may become addicted to internet use through different
pathways.
* Corresponding author. Department of Applied Psychology, School of Humanities and Social Sciences, Fuzhou University, Fuzhou, 350108, China.
** Corresponding author. Department of Psychiatry, Yale University School of Medicine, New Haven, CT, United States.
E-mail addresses: suwenliang@fzu.edu.cn (W. Su), marc.potenza@yale.edu (M.N. Potenza).
https://doi.org/10.1016/j.chb.2020.106480
Received 27 February 2020; Received in revised form 29 June 2020; Accepted 3 July 2020
Available online 13 July 2020
0747-5632/© 2020 Elsevier Ltd. All rights reserved.
W. Su et al. Computers in Human Behavior 113 (2020) 106480
emerging public health problem. Psychiatric Association, 2013). Moreover, gaming disorder including
In the process of evaluating factors related to IA from a multidi online variants has been officially included in the 11th edition of the
mensional perspective, demographic factors are often first considered by International Classification of Diseases (ICD-11) (WHO, 2019).
researchers (Yen et al., 2009). Among them, “gender is a crucial factor to Males typically spend more time gaming per day than females
explain why people are addicted to the internet through different (Chang et al., 2018), and this has been observed across jurisdictions in
pathways” (Tang et al., 2017). Gender-related differences in addictive adults (Laconi et al., 2017) and adolescents (Wichstrom et al., 2019).
behaviors and their motivations have been reported, and these have Not surprisingly, males are more prone to develop IGD than females
been considered important for understanding IA (Liang et al., 2016). across age groups (Choliz & Marco, 2011; De Pasquale et al., 2018;
Males have often been reported to have higher levels of IA than females Stavropoulos, Adams, et al., 2019). In German adolescents, the preva
(Anderson et al., 2017; Chen et al., 2015), but there exist some contrary lence of IGD in boys was 5.9% and in girls was 1.0% (Wartberg et al.,
results (Rukuye Aylaz, 2015; Sun et al., 2012). To investigate further, Su 2020). In the Netherlands, boys scored higher on IGD measures than
et al. (2019) conducted a meta-analysis involving 34 global jurisdictions girls, and the effect size was moderate in magnitude (Cohen’s d ¼ 0.57)
and found that males were more likely than females to experience IA (Kokonyei et al., 2019). In adults, Amendola et al. (2019) reported that
with a small effect size of 0.145 (Hedges’ g). According to the study, the males were at greater risk of IGD than females. Similarly, according to
gender-related effect sizes of IA were larger in Asia and smaller in other the study by Bonnaire and Baptista (2019), most young adult partici
regions, which may be explained by both economic and cultural factors pants in a problematic gaming group were male (78.4%). Several
(Su et al., 2019). possible explanations have been forwarded regarding gender-related
Previous studies of generalized IA have demonstrated the impor differences in IGD. With respect to neural mechanisms, gaming cues
tance of how gender may influence online behaviors. However, gender elicit higher cravings in male subjects and exhibit different neural cor
may relate differently to specific types of IA. Young et al. (1999) clas relates (Dong et al., 2019; Dong, Zheng, et al., 2018). Competitive
sified IA into five types, including cybersex, cyber-relationships, online structures in games may be more attractive to males, who may gain
stock trading or gambling, information overload, and computer games. feelings of success and achievement from gaming (Hamlen, 2010;
Davis (2001) proposed a cognitive-behavior model to distinguish be Wartberg et al., 2020). From a sociocultural perspective, male aggres
tween “generalized pathological internet use” and “specific pathological siveness may be considered more socially acceptable, and some game
internet use.” Specific pathological internet use refers to the condition in designers set males as the target users of violent and adventurous online
which an individual pathologically uses the internet for a particular games (Barua & Barua, 2012; Bryce & Rutter, 2003). Such gaming may
purpose (e.g., online shopping or online gambling), whereas generalized then link to adverse behaviors in individuals with specific tempera
pathological internet use defines as “a general, multidimensional over mental tendencies (e.g., those high in sensation-seeking) (Zhai et al.,
use of the internet” (Davis, 2001; Montag et al., 2015). Differences be 2020).
tween generalized and specific IAs were supported by a cross-cultural Although multiple studies support a higher prevalence of IGD in
empirical study (Montag et al., 2015). Thus, it appears important to males, there exist some contrary findings (Scerri, Anderson, Stavro
distinguish between specific forms of IA when considering poulos, & Hu, 2019; Stavropoulos, Adams, et al., 2019; Stavropoulos,
gender-related differences. With respect to specific IAs, internet gaming Anderson, et al., 2019). In a sample from India, girls scored higher on
disorder (IGD) and social media addiction (SMA) have been described IGD measures than boys (Rajanna et al., 2016). Consequently, whether
particularly frequently (M�erelle et al., 2017). As examples of specific males are more likely than females to exhibit internet-gaming IA war
types of online behaviors, Spilkova et al. (2017) commented, “there is a rants additional systematic research.
need to distinguish between online gaming and social media use, and
perceive them as separate concepts that should not be combined into a 1.3. Gender-related differences in social media addiction
unique category of generalized internet use.” Therefore, we focused our
analytical efforts on IGD and SMA in this study. The development of the internet and the popularity of smartphones
According to research on IGD and SMA over the last decade, exten have increased the use of social media, which allows people to develop,
sive findings suggest that females are more prone to problems with initiate, and maintain existing relationships and new ones (Abbasi,
online communication and social media use, while internet gaming is 2019). The umbrella term ‘social media’ comprises multiple types of
more prevalent among males (Bouna-Pyrrou et al., 2015; Spilkova et al., websites including social networks, messengers, and blogs(van den
2017; van den Eijnden et al., 2018), and these findings may differ for Eijnden, Lemmens, & Valkenburg, 2016). SMA has been defined as an
generalized IA. For example, in the study by Tang et al. (2017) involving “inability to regulate the use of social networks, which leads to negative
a Chinese sample, the male-related effect size for generalized IA was personal outcomes” (Larose et al., 2010; Ryan et al., 2014). As there still
small at 0.15, but it was 0.67 for IGD and 0.10 for SMA; similarly, in exist ongoing debates, relatively limited research, and less direct evi
results from their United States sample, no significant male-related ef dence of clinically significant impairment, SMA as well as other
fect size was found in generalized IA (g ¼ 0.03), but there was a pos non-substance addictions (except gambling disorder and IGD) have not
itive medium effect size for IGD (g ¼ 0.58) and negative effect size for been officially included in the DSM-5 (Petry et al., 2014). However, it
SMA (g ¼ 0.42). It is possible that males consider internet gaming and has been proposed that the category of “other specified disorders due to
females consider social media when completing IA assessments; thus, addictive behaviors” may be used to diagnose SMA (Brand et al., 2020).
the effects of the two gender groups may “cancel” each other (Hawi & Multiple studies have investigated gender-related differences in SMA
Samaha, 2019). As such, it is important to distinguish gender-related (Andreassen et al., 2017), with girls spending more time on social media
differences in specific IA subtypes, especially IGD and SMA. than boys and being younger when first exposed to social media (Chae
et al., 2018). Females may be more vulnerable to developing SMA than
1.2. Gender-related differences in internet gaming disorder males (Andreassen et al., 2017; Martinez-Ferrer et al., 2018; Romer
o-Abrio et al., 2019). According to Demirciog �lu and Go €ncü Ko€se (2018),
IGD has increasingly been considered an important research topic females score significantly higher than males on measures of SMA. Fe
given the increasing prevalence of gaming and associations between IGD males in general may have higher tendencies for SMA related to inter
and psychological distress and impaired functioning (Lemos et al., personal relationship orientations (Chae et al., 2018). Females often
2016). IGD has been defined as, “persistent and recurrent use of the value relationships more than males and tend to use social media as
internet to engage in games, often with other players, leading to interaction tools (Fujimori et al., 2015; Kim et al., 2010). Gender-related
impairment or clinically significant distress” and categorized as a differences in motivation to use social media exist, as males primarily
“condition for further study” in the DSM-5 Section III (American use social media for entertainment and recreation and females use social
2
W. Su et al. Computers in Human Behavior 113 (2020) 106480
media mainly for information and interpersonal communication (Noguti 2.2. Selection of studies
et al., 2019). Females are more likely to use social media in response to
feelings of emptiness when their social needs are not fully met in real The inclusion criteria for the current meta-analysis were: a) studies
life, which may increase the likelihood of SMA (Chae et al., 2018). reporting original empirical findings in peer-reviewed academic jour
Although data suggest that SMA is more prevalent among females nals; b) the sample was assessed using IGD or SMA measures; c) studies
than males, there exist some contrary findings. For instance, males score reporting enough information to calculate the effect sizes (Hedges’ g) of
higher on SMA measures than females in some studies (Araujo Robles, gender-related differences in IGD or SMA; and, e) studies written in
2016; Cam & Isbulan, 2012). As such, systematic studies of English. Exclusion criteria included: a) data were collected from clinical
gender-related differences in SMA are warranted. samples or other samples that had been screened using IGD or SMA
measures; b) data were collected from more than one country and could
1.4. The current study not be separated; c) studies reported the same sample pool used in
previously included studies; and, d) the IGD measurement assessed the
As research on gender-related differences in IGD and SMA has offline gaming behavior rather than online gaming behavior. Finally,
increased recently, systematic evaluation of accumulated data is needed. our search resulted in 2663 records related to IGD and 1479 records
While multiple reports suggest that males more frequently experience related to SMA that were searched to identify eligible studies.
IGD and females SMA, contrary research findings also exist, and a meta- We coded the latest data when there was more than one wave of data
analysis may help to resolve such apparent discrepancies (Gurevitch reported. Studies that only reported the prevalence of IGD/SMA were
et al., 2018). Meta-analyses provide a more powerful and less biased identified, and the authors were emailed to request the means and
means for clarifying, quantifying and disproving (or confirming) hy standard deviations of addiction measure scores for males and females.
potheses than do many other approaches (Murad & Montori, 2013). The In IGD-related records, we emailed 32 authors, and 3 provided the
primary aim of the present study is to synthesize findings from research effect-size data; for SMA-related records, we emailed 31 authors, and 7
on gender-related differences in IGD and SMA and to identify factors (e. provided the effect-size data. Ultimately, 49 studies with 53 indepen
g., geographic region) that may influence the relationships. The current dent samples met the criteria for the IGD meta-analysis (see Fig. 1A), and
study sought to investigate specific IAs (IGD and SMA), and we sus 38 studies with 41 independent samples met the criteria for the SMA
pected that in a meta-analysis investigating specific IAs that we would meta-analysis (see Fig. 1B).
observe more pronounced gender-related differences than have typically
been reported for generalized IA. The following hypotheses were 2.3. Data extraction
proposed:
Studies that met the inclusion criteria were coded for: study identi
H1. Males will demonstrate higher levels of IGD (i.e., demonstrate a
fication data (author, year of publication, title), year of data collection,
positive effect size), and females will demonstrate higher levels of SMA
effect-size data (sample size of males and females, means and standard
(i.e., demonstrate a negative effect size).
deviations of IGD/SMA scores in males and females), IGD/SMA mea
H2. Generalized IA will have a “masking effect” on gender-related surement, sample age characteristics (mean, standard deviation, range,
differences in IGD and SMA, with the effect size in generalized IA and age group), and country of the sample. Three coders (XH, HY, YW)
being significantly smaller than the absolute value of the effect sizes of independently scored all coding dimensions. All coders had adequate
IGD and SMA. research experience in the meta-analysis and were well trained in coding
IGD/SMA measurements and effect size data. The inter-rater consistency
2. Methods was 0.95.
For the meta-analysis, the databases searched included ScienceDir For the purpose of weighting effect sizes by study quality, we used a
ect, Springer, Wiley, MEDLINE, ProQuest, and Web of Science. As the Data Quality Index (DQI) developed by Su et al. (2019) through the
internet and how people use it has been changing over time, the current following two measures. First, sample representativeness, in which each
study focused on the most recent 10 years from January 2010 to August included study would have an initial score of 1, and if the sample came
2019 (including online-first publications). For IGD, search terms were: from a national distribution, with probability sampling, or with age
(“internet gaming disorder” OR “internet gaming addiction” OR “com diversity it would get one more point added to the score for each feature.
puter game addiction” OR “online game addiction” OR “problematic Thus, scores ranged from 1 to 4. Second, sample size, in which the larger
internet game use” OR “problematic online gaming” OR “pathological sample size would get higher scores, were determined logarithmically
game use” OR “gaming addiction” OR “gaming disorder” OR “prob (score ¼ lg(N)). The score of the sample size was 2.0–4.4 in the current
lematic gaming behavior” OR “pathological video game use” OR “video study. The DQI total score ¼ sample representativeness score þ sample
game addiction” OR “psychosocially problematic gaming”) AND size score.
(“gender”).
Similarly, the search terms for SMA were: (“social media addiction” 2.5. Statistical analysis
OR “social network addiction” OR “social network sites addiction” OR
“problematic social media use” OR “excessive social media use” OR To control for variations in sample sizes among studies, we calcu
“compulsive social media use” OR “problematic social networking site lated Hedge’s g coefficients for the effect-size estimates. The formula for
use” OR “Facebook addiction” OR “problematic Facebook use” OR calculating Hedges’ g is g ¼ (Mmale – Mfemale)/SD*pooled, where the
“Twitter addiction” OR “Instagram addiction”) AND (“Gender”). SD*pooled is the pooled and weighted standard deviation (Hedges, 1981).
In order to include more eligible studies for meta-analysis, a “back The effect sizes were bias-corrected for small samples (Hedges, 1981).
ward search” procedure was used to look for more relevant earlier Effect sizes of �0.80 were considered large, �0.50 were considered
studies from the references of recent studies and reviews. Furthermore, moderate, and �0.20 were small effects (Cohen, 1988). Positive effect
authors of retrieved articles reporting insufficient effect size data were sizes indicated males having higher IGD/SMA levels than females, while
emailed to request such data. negative effect sizes showed females have higher IGD/SMA levels than
males. Effect sizes were computed using random-effects models with the
method of restricted maximum likelihood (REML) (Langan et al., 2019),
3
W. Su et al. Computers in Human Behavior 113 (2020) 106480
as was evidence for heterogeneity. To test the heterogeneity of effect high levels of heterogeneity were due to real differences among the
sizes, we used the Q statistic and the I2 statistic indexes (Borenstein selected samples as opposed to sampling errors (Higgins et al., 2003).
et al., 2005). To explain observed heterogeneity in effect sizes, we per This observation warranted the use of a random-effects model. For the
formed subgroup analyses with categorical moderators (e.g., region, 53 independent samples from 21 countries/regions, the average effect
measurement, age-related population, and survey method) and size was 0.479 for the random-effects, which was a moderate effect size
meta-regression with continuous moderators (e.g., sample representa according to Cohen (1988).
tiveness, sample size, and publication year). To test for possible publi The dataset of SMA studies included 41 independent effect sizes of
cation bias, we used funnel plots and the Egger regression test with a gender-related differences in SMA tendencies from 58,336 participants
weighted regression model using the standard error as the predictor of 22 countries/regions. The sample sizes ranged from 179 to 23,532.
(Egger et al., 1997) to assess for asymmetry. The trim-and-fill analysis Approximately 40.6% of participants were male. Features, including
was adopted to estimate the number of missing studies and obtain sample characteristics, measures and effect sizes, were recorded for each
adjusted effect sizes that take into account publication bias (Duval & sample (see Table 2 for descriptive statistics and Supplementary
Tweedie, 2000). To obtain effect sizes for each country, effect sizes from Table S2 for details).
the same country’s independent samples were averaged by weighting The meta-analysis of SMA studies also showed high levels of het
the DQI. erogeneity (97.96%). The Q statistic was significant (Q(40) ¼ 1357.70,
Meta-analyses were conducted in R (version 3.4.2) using the “met p < 0.001) in our study, and I2 values (97.96%) were also over 75%. This
afor” package (version 2.0, Viechtbauer, 2010). One-way ANOVA was observation warranted the use of a random-effects model. For the 41
used for comparing the effect sizes among IA, IGD and SMA at independent samples from 22 countries/regions, the average effect size
country-levels. The effect sizes of generalized IA at country-levels used was 0.202 for the random-effects, which was a small effect size ac
in the ANOVA (n ¼ 34) and radar chart (n ¼ 12) were retrieved from our cording to Cohen (1988).
prior meta-analysis study (Su et al., 2019).
3.2. Publication bias
3. Results
The funnel plots for the IGD and SMA meta-analysis are presented in
3.1. Effect size and homogeneity tests Fig. 2A and B, respectively. There was no evidence of publication bias
according to the Egger’s regression test in both the IGD (t ¼ 0.68, df ¼
The dataset included 53 independent effect sizes of gender-related 51, p ¼ 0.501) and SMA (t ¼ 1.41, df ¼ 39, p ¼ 0.165) samples. Addi
differences in IGD tendencies from 82,440 participants of 21 coun tionally, the Trim-and-Fill procedure indicated none was missing in both
tries/regions. The sample sizes ranged from 104 to 25,573. About half samples from IGD (SE ¼ 4.201) and SMA (SE ¼ 3.867), resulting in an
(53.2%) of participants were male. Features, including sample charac adjusted mean effect size equal to the observed mean.
teristics, measures, and effect sizes, were recorded for each sample (see
Table 1 for descriptive statics and Supplementary Table S1 for details). 3.3. The moderating role of region
The results showed a high level of heterogeneity (96.84%) in the IGD
studies. The Q statistic was significant (Q(52) ¼ 1182.06, p < 0.001) in In addition to effect sizes, moderation effects were also examined.
the study, and I2 values (96.84%) were over 75%, suggesting that the We examined whether there were systematic differences in effect sizes
4
W. Su et al. Computers in Human Behavior 113 (2020) 106480
Table 1 Table 2
Characteristics of samples included in the internet gaming disorder meta- Characteristics of the sample included in the social media addiction meta-
analysis (Total k ¼ 53). analysis (Total k ¼ 41).
Variable Category K % Variable Category K %
5
W. Su et al. Computers in Human Behavior 113 (2020) 106480
Table 3
Continental regional subgroup analyses of gender-related effect sizes for internet gaming disorder and social media addiction.
Region n k Hedges’ g (95%CI) p Q p I2
IGD
Overall 86831 53 0.479 [0.392,0.565] <0.001 1182.06 <0.001 96.84%
Asia 37066 14 0.295 [0.112,0.478] 0.002 147.46 <0.001 97.32%
Europe 36755 31 0.584 [0.499,0.669] <0.001 413.48 <0.001 92.16%
North America 10860 5 0.534 [0.139,0.929] 0.008 421.29 <0.001 98.70%
Oceania 2150 3 0.062 [ 0.408,0.532] 0.796 25.13 <0.001 89.71%
SMA
Overall 58336 41 0.202 [ 0.335, 0.069] 0.003 1357.70 <0.001 97.96%
Asia 7084 11 0.125 [ 0.278,0.027] 0.109 89.19 <0.001 89.73%
Europe 35971 20 0.256 [ 0.507, 0.005] 0.045 1167.54 <0.001 98.71%
North America 14340 8 0.228 [ 0.363, 0.093] <0.001 44.67 <0.001 89.19%
South Americaa 200 1 0.193[ 0.127, 0.514] 0.236 – – –
Oceaniab 741 1 0.150[ 0.297, 0.003] 0.046 – – –
Note: n ¼ combined sample size; k ¼ number of effect size; Hedges’ g (random-effects); CI ¼ confidence interval; Q ¼ homogeneity statistic; I2 ¼ inconsistency, a second
measure of heterogeneity; IGD¼Internet gaming disorder; SMA ¼ Social media addiction.
a
only Peru was included in analysis.
b
only Australia was included in analysis.
Fig. 3. World distribution of gender-related differences in internet gaming disorder (A) and social media addiction (B). Note: Positive effect size reflects greater
likelihoods for males versus females.
from 22 countries/regions were identified for SMA. The main findings The results of the meta-analysis confirmed our first hypothesis that
suggest gender-related differences exist not only for generalized IA (Su males are more likely to experience IGD than females globally with a
et al., 2019), but are more pronounced for specific IAs, with IGD (mal moderate effect size (g ¼ 0.462). By contrast, females are more likely to
e-predominant) and SMA (female-predominant) showing opposite experience SMA than males globally with a small effect size (g ¼
effects. 0.202). This outcome corroborates findings from previous studies,
6
W. Su et al. Computers in Human Behavior 113 (2020) 106480
Fig. 4. Effect sizes of gender difference in IGD by the effect sizes of SMA from 16 countries/regions. Note. IGD ¼ Internet gaming disorder; SMA ¼ Social media
addiction; the circle size represents the total number of effect sizes for IGD and SMA; positive effect size means males were more likely to demonstrate the specific
type of internet addiction.
SMA.
The current meta-analysis provides stronger evidence for higher IGD
levels in males overall. As mentioned earlier, this pattern may be
explained by gender-related differences in neural mechanisms. Gaming-
related cues elicited higher cravings in males, accompanied by having
higher activations of reward-related brain regions including
corticostriatal-limbic circuitry (Dong, Wang, et al., 2018; Sayette,
2016). Overall, these factors may promote engagement in specific
addictive behaviors like gaming in IGD. From an evolutionary
perspective, males may have become more aggressive in maintaining
dominance and defending territory (Becker & Koob, 2016). Aggres
siveness is a masculine trait in many cultures, while it is not consonant
with typical social expectations for females (Barua & Barua, 2012).
Indeed, competitive elements of online games are more attractive to
men than women (Hartmann & Klimmt, 2006). Consequently, game
designers and related market investors identify males as target users of
violent and adventurous video games (Barua & Barua, 2012). As males
may experience feelings of success and achievement from engaging in
Fig. 5. Effect sizes of gender difference in IA (n ¼ 34), IGD (n ¼ 21) and SMA online games, this may lead to greater desires for gaming (Hamlen,
(n ¼ 22), with error bars (standard error). Note. IA ¼ generalized internet
2010).
addiction, IGD ¼ internet gaming disorder, SMA ¼ social media addiction.
Regarding social media, the present findings support the female
predominance of SMA. A possible explanation may involve greater fe
showing opposite directions of gender-related differences in IGD and male sensitivity to social signals and the valuation of interpersonal re
SMA (Tang et al., 2017; van den Eijnden et al., 2018), and suggests that lations (Fujimori et al., 2015; Proverbio et al., 2008). Evolutionary
people may become addicted to internet activities through different scholars propose that sex/gender differences in brain and behavior may
pathways (Tang et al., 2017). The findings indicate that in most coun relate to different demands for female’s child-care and male’s hunting
tries/regions, there is a male predominance for IGD and female pre strategies (Keverne, 2007). Compared with males, females are often
dominance for SMA, suggesting a pattern consistent across many more interpersonal (Barua & Barua, 2012; Keller & Scharff-Goldhaber,
cultures and geographic locations. Our findings also highlight 1987). Females also demonstrate more tendencies to use social media
geographic variation in gender-related effect sizes in IGD and SMA. as interactional tools, while males are inclined to access social media for
Specifically, for IGD, gender has the largest effect size in Europe and entertainment (Fujimori et al., 2015; Lin et al., 2013). SMA tendencies
North America, smaller in Asia and negligible in Oceania. Similar results are also related to psychosocial factors in females (Chae et al., 2018).
were found for SMA, although in the opposite direction. According to Specifically, females may use social media to fill voids when depressed
our prior study (Su et al., 2019), gender-related differences in IA may be and social needs are not fully met, thereby increasing the likelihood of
related to both sociocultural and economic factors. Future study is SMA (Block et al., 2014; Chae et al., 2018). Together, data suggest that
needed to examine potentially hidden factors impacting geographic gender-related differences in IGD and SMA likely involve biological,
variance in gender-related differences in effect sizes relating to IGD and
7
W. Su et al. Computers in Human Behavior 113 (2020) 106480
Fig. 6. Radar-chart of gender-related effect sizes in generalized and specific internet addiction. Note. IA ¼ generalized internet addiction, IGD ¼ internet gaming
disorder, SMA ¼ social media addiction. Positive effect size reflects greater likelihoods for males versus females.
8
W. Su et al. Computers in Human Behavior 113 (2020) 106480
9
W. Su et al. Computers in Human Behavior 113 (2020) 106480
deprivation. Journal of Behavioral Addictions, 7(4), 953–964. https://doi.org/ Liang, L., Zhou, D., Yuan, C., Shao, A., & Bian, Y. (2016). Gender differences in the
10.1556/2006.7.2018.118. relationship between internet addiction and depression: A cross-lagged study in
Duval, S., & Tweedie, R. (2000). Trim and fill: A simple funnel-plot–based method of Chinese adolescents. Computers in Human Behavior, 63, 463–470. https://doi.org/
testing and adjusting for publication bias in meta-analysis. Biometrics, 56(2), 10.1016/j.chb.2016.04.043.
455–463. https://doi.org/10.1111/j.0006-341X.2000.00455.x. Lin, X., Li, Y., Califf, C. B., & Featherman, M. (2013). Can social role theory explain
Egger, M., Smith, G. D., Schneider, M., & Minder, C. (1997). Bias in meta-analysis gender differences in Facebook usage?. In Paper presented at the 46th Hawaii
detected by a simple, graphical test. BMJ, 315(7109), 629–634. https://doi.org/ International Conference on System Sciences, Hawaii. https://doi.org/10.1109/
10.1136/bmj.315.7109.629. hicss.2013.125.
van den Eijnden, R., Koning, I., Doornwaard, S., van Gurp, F., & Ter Bogt, T. (2018). The Lin, X., Su, W., & Potenza, M. N. (2018). Development of an online and offline
impact of heavy and disordered use of games and social media on adolescents’ integration hypothesis for healthy internet use: Theory and preliminary evidence.
psychological, social, and school functioning. Journal of Behavioral Addictions, 7(3), Frontiers in Psychology, 9, 492. https://doi.org/10.3389/fpsyg.2018.00492.
697–706. https://doi.org/10.1556/2006.7.2018.65. Malak, M. Z., Khalifeh, A. H., & Shuhaiber, A. H. (2017). Prevalence of Internet Addiction
Eijnden, van den, Regina, J. J. M., Lemmens, Jeroen S., & Valkenburg, Patti M. (2016). and associated risk factors in Jordanian school students. Computers in Human
The Social Media Disorder Scale. Computers in Human Behavior, 61, 478–487. Behavior, 70, 556–563. https://doi.org/10.1016/j.chb.2017.01.011.
https://doi.org/10.1016/j.chb.2016.03.038. Martinez-Ferrer, B., Romero-Abrio, A., Moreno-Ruiz, D., & Musitu, G. (2018). Child-to-
van den Eijnden, R. J., Meerkerk, G.-J., Vermulst, A. A., Spijkerman, R., & Engels, R. C. Parent violence and parenting styles: Its relations to problematic use of social
(2008). Online communication, compulsive internet use, and psychosocial well- networking sites, alexithymia, and attitude towards institutional authority in
being among adolescents: A longitudinal study. Developmental psychology, 44(3), adolescence. Psychosocial Intervention, 27(3), 163–171. https://doi.org/10.5093/
655. https://doi.org/10.1037/0012-1649.44.3.655. pi2018a24.
Fujimori, A., Yamazaki, T., Sato, M., Hayashi, H., Fujiwara, Y., & Matsusaka, T. (2015). Marton�cik, M., & Lok�sa, J. (2016). Do World of Warcraft (MMORPG) players experience
Study on influence of internal working models and gender differences on addiction less loneliness and social anxiety in online world (virtual environment) than in real
of social network sites in Japanese university students. Psychology, 6(14), 9. https:// world (offline)? Computers in Human Behavior, 56, 127–134. https://doi.org/
doi.org/10.4236/psych.2015.614179. 10.1016/j.chb.2015.11.035.
Griffiths, M. D., Kuss, D. J., & Demetrovics, Z. (2014). Social networking addiction: An M�erelle, S., Kleiboer, A., Schotanus, M., Cluitmans, T. L. M., Waardenburg, C. M.,
overview of preliminary findings. In K. P. Rosenberg, & L. C. Feder (Eds.), Behavioral Kramer, D., … van Rooij, A. (2017). Which health-related problems are associated
addictions (pp. 119–141). San Diego: Academic Press. with problematic video-gaming or social media use in adolescents? Clinical
Griffiths, M., & Pontes, H. (2014). Internet addiction disorder and internet gaming Neuropsychiatry: Journal of Treatments Evaluation, 14(1), 11–19.
disorder are not the same. Addiction Research and Theory, 5(4), e124. https://doi. Montag, C., Bey, K., Sha, P., Li, M., Chen, Y.-F., Liu, W.-Y., … Reuter, M. (2015). Is it
org/10.4172/2155-6105.1000e124. meaningful to distinguish between generalized and specific internet addiction?
Gurevitch, J., Koricheva, J., Nakagawa, S., & Stewart, G. (2018). Meta-analysis and the Evidence from a cross-cultural study from Germany, Sweden, Taiwan and China.
science of research synthesis. Nature, 555, 175. https://doi.org/10.1038/ Asia-Pacific Psychiatry, 7(1), 20–26. https://doi.org/10.1111/appy.12122.
nature25753. Murad, M. H., & Montori, V. M. (2013). Synthesizing evidence: Shifting the focus from
Hamlen, K. R. (2010). Re-examining gender differences in video game play: Time spent individual studies to the body of evidence. Journal of the American Medical
and feelings of success. Journal of Educational Computing Research, 43(3), 293–308. Association, 309(21), 2217–2218. https://doi.org/10.1001/jama.2013.5616.
https://doi.org/10.2190/EC.43.3.b. Noguti, V., Singh, S., & Waller, D. S. (2019). Gender differences in motivations to use
Hartmann, T., & Klimmt, C. (2006). Gender and computer games: Exploring females’ social networking sites. In information resources management association. In Gender
dislikes. Journal of Computer-Mediated Communication, 11(4), 910–931. https://doi. economics: Breakthroughs in research and practice (pp. 676–691). Hershey, PA: IGI
org/10.1111/j.1083-6101.2006.00301.x. Global. https://doi.org/10.4018/978-1-5225-7510-8.ch033.
Hawi, N., & Samaha, M. (2019). Identifying commonalities and differences in personality Petry, N. M., Rehbein, F., Gentile, D. A., Lemmens, J. S., Rumpf, H.-J., M€ oßle, T., …
characteristics of internet and social media addiction profiles: Traits, self-esteem, O’Brien, C. P. (2014). An international consensus for assessing internet gaming
and self-construal. Behaviour & Information Technology, 38(2), 110–119. https://doi. disorder using the new DSM-5 approach. Addiction, 109(9), 1399–1406. https://doi.
org/10.1080/0144929x.2018.1515984. org/10.1111/add.12457.
Hedges, L. V. (1981). Distribution theory for glass’s estimator of effect size and related Proverbio, A. M., Zani, A., & Adorni, R. (2008). Neural markers of a greater female
estimators. Journal of Educational Statistics, 6(2), 107–128. https://doi.org/10.2307/ responsiveness to social stimuli. BMC Neuroscience, 9, 56. https://doi.org/10.1186/
1164588. 1471-2202-9-56.
Higgins, J. P. T., Thompson, S. G., Deeks, J. J., & Altman, D. G. (2003). Measuring Rajanna, S. H., Sharma, M. K., & Palanichamy, T. S. (2016). Exploration of technology
inconsistency in meta-analyses. BMJ, 327(7414), 557–560. https://doi.org/ use pattern among teenagers and its relationship with psychological variables.
10.1136/bmj.327.7414.557. ASEAN Journal of Psychiatry, 17(2), 239–249.
Internet World Stats. (2019). Internet usage statistics, the internet big picture. https://www. Romero-Abrio, A., Martinez-Ferrer, B., Musitu-Ferrer, D., Leon-Moreno, C., Villarreal-
internetworldstats.com/stats.htm. Gonzalez, M. E., & Callejas-Jeronimo, J. E. (2019). Family communication problems,
Kandell, J. J. (1998). Internet addiction on campus: The vulnerability of college students. psychosocial adjustment and cyberbullying. International Journal of Environmental
CyberPsychology and Behavior, 1(1), 11–17. https://doi.org/10.1089/cpb.1998.1.11. Research and Public Health, 16(13). https://doi.org/10.3390/ijerph16132417.
Keller, E. F., & Scharff-Goldhaber, G. (1987). Reflections on gender and science. American Rukuye Aylaz, G. G., Günaydın, Y., Kocaer, M., & Pehlivan, E. (2015). Problematic
Journal of Physics, 55(3), 284–286. https://doi.org/10.1119/1.15186. internet usage among high school students and the relevant factors. Turkish Journal
Keverne, E. B. (2007). Genomic Imprinting and the evolution of sex differences in mammalian of Public Health, 13(3), 184–192.
reproductive strategies advances in genetics (Vol. 59, pp. 217–243). Academic Press. Ryan, T., Chester, A., Reece, J., & Xenos, S. (2014). The uses and abuses of Facebook: A
Kim, J. H., Kim, M.-S., & Nam, Y. (2010). An analysis of self-construals, motivations, review of Facebook addiction. Journal of Behavioral Addictions, 3(3), 133–148.
Facebook use, and user satisfaction. International Journal of Human-Computer https://doi.org/10.1556/JBA.3.2014.016.
Interaction, 26(11–12), 1077–1099. https://doi.org/10.1080/ Sayette, M. A. (2016). The role of craving in substance use disorders: Theoretical and
10447318.2010.516726. methodological issues. Annual Review of Clinical Psychology, 12(1), 407–433. https://
Kokonyei, G., Kocsel, N., Kiraly, O., Griffiths, M. D., Galambos, A., Magi, A., … doi.org/10.1146/annurev-clinpsy-021815-093351.
Demetrovics, Z. (2019). The role of cognitive emotion regulation strategies in Scerri, M., Anderson, A., Stavropoulos, V., & Hu, E. (2019). Need fulfilment and internet
problem gaming among adolescents: A nationally representative survey study. gaming disorder: A preliminary integrative model. Addictive Behaviors Reports, 9,
Frontiers in Psychiatry, 10. https://doi.org/10.3389/fpsyt.2019.00273. 100144. https://doi.org/10.1016/j.abrep.2018.100144.
Laconi, S., Pires, S., & Chabrol, H. (2017). Internet gaming disorder, motives, game Shaw, M., & Black, D. W. (2008). Internet addiction: Definition, assessment,
genres and psychopathology. Computers in Human Behavior, 75, 652–659. https:// epidemiology and clinical management. CNS Drugs, 22(5), 353–365. https://doi.
doi.org/10.1016/j.chb.2017.06.012. org/10.2165/00023210-200822050-00001.
Laconi, S., Tricard, N., & Chabrol, H. (2015). Differences between specific and Snodgrass, J. G., Lacy, M. G., Dengah, H. J. F., II, & Fagan, J. (2011). Enhancing one life
generalized problematic Internet uses according to gender, age, time spent online rather than living two: Playing MMOs with offline friends. Computers in Human
and psychopathological symptoms. Computers in Human Behavior, 48, 236–244. Behavior, 27(3), 1211–1222. https://doi.org/10.1016/j.chb.2011.01.001.
https://doi.org/10.1016/j.chb.2015.02.006. Spilkova, J., Chomynova, P., & Csemy, L. (2017). Predictors of excessive use of social
Lam, L. T. (2014). Internet gaming addiction, problematic use of the internet, and sleep media and excessive online gaming in Czech teenagers. Journal of Behavioral
problems: A systematic review. Current Psychiatry Reports, 16(4), 444. https://doi. Addictions, 6(4), 611–619. https://doi.org/10.1556/2006.6.2017.064.
org/10.1007/s11920-014-0444-1. Stavropoulos, V., Adams, B. L. M., Beard, C. L., Dumble, E., Trawley, S., Gomez, R., &
Langan, D., Higgins, J. P. T., Jackson, D., Bowden, J., Veroniki, A. A., Kontopantelis, E., Pontes, H. M. (2019). Associations between attention deficit hyperactivity and
… Simmonds, M. (2019). A comparison of heterogeneity variance estimators in internet gaming disorder symptoms: Is there consistency across types of symptoms,
simulated random-effects meta-analyses. Research Synthesis Methods, 10(1), 83–98. gender and countries? Addictive Behaviors Reports, 9, 100158. https://doi.org/
https://doi.org/10.1002/jrsm.1316. 10.1016/j.abrep.2018.100158.
Larose, R., Kim, J.-H., & Peng, W. (2010). Social networking: Addictive, compulsive, Stavropoulos, V., Anderson, E. E., Beard, C., Latifi, M. Q., Kuss, D., & Griffiths, M. (2019).
problematic or just another media habit? In Z. Papacharissi (Ed.), A networked self: A preliminary cross-cultural study of Hikikomori and Internet Gaming Disorder: The
Identity, community, and culture on social network sites (pp. 59–81). New York: moderating effects of game-playing time and living with parents. Addictive Behaviors
Routledge. Reports, 9. https://doi.org/10.1016/j.abrep.2018.10.001.
Lemos, I. L., Cardoso, A., & Sougey, E. B. (2016). Cross-cultural adaptation and Su, W., Han, X., Jin, C., Yan, Y., & Potenza, M. N. (2019). Are males more likely to be
evaluation of the psychometric properties of the Brazilian version of the Video Game addicted to the internet than females? A meta-analysis involving 34 global
Addiction Test. Computers in Human Behavior, 55, 207–213. https://doi.org/ jurisdictions. Computers in Human Behavior, 99, 86–100. https://doi.org/10.1016/j.
10.1016/j.chb.2015.09.019. chb.2019.04.021.
10
W. Su et al. Computers in Human Behavior 113 (2020) 106480
Sun, P., Johnson, C. A., Palmer, P., Arpawong, T. E., Unger, J. B., Xie, B., … Sussman, S. Wichstrom, L., Stenseng, F., Belsky, J., von Soest, T., & Hygen, B. W. (2019). Symptoms
(2012). Concurrent and predictive relationships between compulsive internet use of internet gaming disorder in youth: Predictors and comorbidity. Journal of
and substance use: Findings from vocational high school students in China and the Abnormal Child Psychology, 47(1), 71–83. https://doi.org/10.1007/s10802-018-
USA. International Journal of Environmental Research and Public Health. https://doi. 0422-x.
org/10.3390/ijerph9030660. Widyanto, L., & Griffiths, M. (2006). “Internet addiction”: A critical review. International
Tang, C. S., Koh, Y. W., & Gan, Y. (2017). Addiction to internet use, online gaming, and Journal of Mental Health and Addiction, 4(1), 31–51. https://doi.org/10.1007/
online social networking among young adults in China, Singapore, and the United s11469-006-9009-9.
States. Asia-Pacific Journal of Public Health, 29(8), 673–682. https://doi.org/ Winkler, A., D€orsing, B., Rief, W., Shen, Y., & Glombiewski, J. A. (2013). Treatment of
10.1177/1010539517739558. internet addiction: A meta-analysis. Clinical Psychology Review, 33(2), 317–329.
Valkenburg, P. M., & Peter, J. (2007). Online communication and adolescent well-being: https://doi.org/10.1016/j.cpr.2012.12.005.
Testing the stimulation versus the displacement hypothesis. Journal of Computer- Yen, C. F., Ko, C. H., Yen, J. Y., Chang, Y. P., & Cheng, C. P. (2009). Multi-dimensional
Mediated Communication, 12(4), 1169–1182. discriminative factors for Internet addiction among adolescents regarding gender
Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package. and age. Psychiatry and Clinical Neuroscience, 63(3), 357–364. https://doi.org/
Journal of Statistical Software, 36(3), 1–48. https://doi.org/10.18637/jss.v036.i03. 10.1111/j.1440-1819.2009.01969.x.
Wartberg, L., Kriston, L., & Thomasius, R. (2020). Internet gaming disorder and Young, K., Pistner, M., O’Mara, J., & Buchanan, J. (1999). Cyber disorders: The mental
problematic social media use in a representative sample of German adolescents: health concern for the new millennium. CyberPsychology and Behavior, 2(5),
Prevalence estimates, comorbid depressive symptoms and related psychosocial 475–479. https://doi.org/10.1089/cpb.1999.2.475.
aspects. Computers in Human Behavior, 103, 31–36. https://doi.org/10.1016/j. Zajac, K., Ginley, M. K., Chang, R., & Petry, N. M. (2017). Treatments for internet gaming
chb.2019.09.014. disorder and internet addiction: A systematic review. Psychology of Addictive
Weinstein, A., & Lejoyeux, M. (2010). Internet addiction or excessive internet use. The Behaviors, 31(8), 979–994. https://doi.org/10.1037/adb0000315.
American Journal of Drug and Alcohol Abuse, 36(5), 277–283. https://doi.org/ Zhai, Z. W., Hoff, R. A., Howell, J. C., Wampler, J., Krishnan-Sarin, S., & Potenza, M. N.
10.3109/00952990.2010.491880. (2020). Differences in associations between problematic video-gaming, video-
WHO. (2019). ICD-11: 6C51.0 gaming disorder, predominantly online. Retrieved from gaming duration, and weapon-related and physically violent behaviors in
https://icd.who.int/browse11/l-m/en#/http://id.who.int/icd/entity/338347362. adolescents. Journal of Psychiatric Research, 121, 47–55. https://doi.org/10.1016/j.
jpsychires.2019.11.005.
11