• 沒有找到結果。

Psychometric properties and measurement invariance of the 7-item game addiction scale (GAS) among Chinese college students

N/A
N/A
Protected

Academic year: 2021

Share "Psychometric properties and measurement invariance of the 7-item game addiction scale (GAS) among Chinese college students"

Copied!
11
0
0

加載中.... (立即查看全文)

全文

(1)

R E S E A R C H A R T I C L E

Open Access

Psychometric properties and measurement

invariance of the 7-item game addiction

scale (GAS) among Chinese college

students

Yujie Liu

1

, Qian Wang

1*

, Min Jou

2

, Baohong Wang

3

, Yang An

3

and Zifan Li

4

Abstract

Background: The 7-item Gaming Addiction Scale (GAS) has been used as a screening tool for addictive game use worldwide, and this study aimed to examine its psychometric properties and measurement invariance among college students in China.

Methods: Full-time students from multiple colleges in China were recruited. A total of 1040 completed

questionnaires were used in the final analysis. Reliability of the GAS was assessed by internal consistency and split-half reliability. Validity of the GAS was assessed by structural validity, convergent validity, discriminant validity, and concurrent validity. A series of Multigroup Confirmatory Factor Analysis (MG-CFA) were conducted to test and establish measurement invariance across gender, class standing, family income and parental educational level. Results: Exploratory factor analysis revealed a unidimensional structure of the GAS. The GAS exhibited excellent internal consistency (Cronbach’s α = 0.951, theta coefficient = 0.953, omega coefficient = 0.959) and structural validity (χ2/df = 0.877 (p < 0.05), CFI = 0.999, TIL = 0.996, RMSEA =0.000). Concurrent validity of the GAS was confirmed by its correlation with problematic internet use, sleep quality, nine dimensions of psychiatric symptoms, and substance use. The GAS also demonstrated measurement invariance across father’s educational level (Δχ2 (df) = 19.128 (12), ΔCFI = − 0.009, ΔRMSEA = 0.010 for weak factorial model; Δχ2 (df) = 50.109 (42), ΔCFI = − 0.010, ΔRMSEA = 0.007 for strict factorial model.) and mother’s educational level (Δχ2 (df) = 6.679 (12), ΔCFI = 0.007, ΔRMSEA = − 0.010 for weak factorial model;Δχ2 (df) =49.131 (42), ΔCFI = − 0.009, ΔRMSEA = − 0.004 for strict factorial model), as well as partial measurement invariance across gender (except for item 2), class standing (except for item 7) and family income (except for item 5).

Conclusions: The Chinese version of the 7-item GAS can be an adequate assessment tool to assess internet gaming disorder among the college student population in China.

Keywords: Internet gaming disorder, College students, PSQI, SCL-90-R, Internet addiction, Measurement invariance

© The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visithttp://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

* Correspondence:qian.wang@shsmu.edu.cn

1School of Public Health, Shanghai Jiao Tong University School of Medicine,

No 227 S Chongqing Road, Shanghai 200025, China

(2)

Background

Internet gaming disorder (IGD) has increasingly become an internationally recognized behavioral addiction, con-stituting a growing concern worldwide including in China. Its inclusion in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-V [1];) has garnered considerable attention from researchers worldwide. The DSM-V clearly defines the diagnostic criteria for IGD, requiring at least 5 out of 9 symptoms (preoccupation, tolerance, escape, withdrawal, persist-ence, conflict, problems, deception, and displacement) to be present for at least 12 months. Prevalence of IGD var-ied across countries (ranging from 1.6% in the Netherlands to 3.0% in Germany), with higher rates con-sistently reported for adolescents residing in Asia (i.e. 10.3% in mainland China) [2]. In China, the prevalence of IGD varied widely, ranging from 3.9% for high school students in Shanghai to 15.6% for secondary school stu-dents in Hong Kong [3, 4]. The discrepancies in preva-lence rates of IGD have been largely attributed to measurement issues such as heterogeneity in assessment tools [5], or lack of measurement invariance across dif-ferent groups. Such issues may confound accurate as-sessment of IGD prevalence, affecting the screening or identification of high-risk groups. Therefore, it is im-portant for relevant instruments to be psychometrically evaluated in different populations.

The present study sought to address this aim by asses-sing the psychometric properties of the widely-used 7-item Game Addiction Scale (GAS) developed by Lem-mens et al. [6]. The GAS was created in view of the sub-stantial overlap between personality characteristics of gamblers and gaming addicts [6, 7], and has been used in various populations as an assessment tool to screen for IGD. Items on the GAS were adapted from the 7 diagnostic criteria (salience, tolerance, mood modifica-tion, withdrawal, relapse, conflict, and problems) for pathological gambling under DSM-IV-TR [6], and were confirmed as adequate to assess IGD among two inde-pendent samples of adolescents in Netherlands [6]. Con-current validity of the GAS was found to be satisfactory, indicated by the correlations between scores on the GAS and time spent on games (r = 0.576), life satisfaction (r = − 0.136), loneliness (r = 0.314), social competence (r = − 0.158) and aggression (r = 0.265) [6]. Psychometric prop-erties of the scale were later tested among gamers as well as the general population residing in France, Germany, Brazil, Spain, Iran and Italy [8–12]. Confirma-tory factor analysis (CFA) showed that the scale had a unidimensional structure [9], with a Cronbach’s alpha

value ranging from 0.85 [9] to 0.92 [10]. Although the GAS has been used to examine prevalence and corre-lates of IGD among adolescents and young adults in China [3, 13, 14], no study has focused on its

psychometric properties in this population. Understand-ing and evaluatUnderstand-ing psychometric properties of the GAS in this population is an essential aspect of scale selec-tion, it may enhance applicability of the scale and accur-acy of the scale in identifying at-risk populations.

In the current study, we chose to assess psychometric properties of the GAS among the college student popu-lation in China for the following reasons: the adolescent population has been the main focus of existing studies, whereas the young adult population has been under-researched, as it is commonly perceived that, compared to other age groups, characteristics unique to adoles-cents may make them more vulnerable to developing IGD [15–18]. However, it is critical to note that adoles-cents in China typically face intense academic pressure due to fierce competitions in the college entrance exam, or Gaokao. In comparison, once students enter college in China, they are completely relieved of the academic pressure of Gaokao, and most likely divert their atten-tion to other aspects of their college life [19]. Second, most adolescents live with their parents during their jun-ior and high school years, when close proximity to par-ents can facilitate and strengthen parental monitoring. In comparison, many students choose to attend college away from their hometown and parents, a sign of inde-pendence, which may lead to decreased parental control and monitoring. A study examining health-related be-haviors among middle school, high school and college students in China found screen time increased as educa-tional level increased [20]. A recent study examining both high school and college students in China found college students scored higher on the IGD-20 Test [13].

Therefore, the purpose of this study was to assess the reliability and validity of the 7-item GAS using a sample of college students residing in China. We further assessed the association of IGD with mental health, sleep quality, substance use, problematic internet use, and so-cial media addiction in establishing the validity of the GAS. Additionally, we sought to test and establish meas-urement invariance of the GAS across socio-demographic groups. Examining measurement invari-ance is an essential aspect of instrument validation, as it reflects the extent to which a measured construct has the same meaning across all respondents regardless of their group membership [21]. Findings of this study could expand the applicability of the 7-item GAS in assessing IGD to the Chinese college student population, and lay the ground work for further analysis and comparison.

Methods Participants

A convenient sample of 1071 participants was recruited from multiple colleges in mainland China. Students who

(3)

were attending school part-time or unable to complete the questionnaire were excluded, only full-time students who were willing to complete the questionnaire were in-cluded. We only included full-time students on the grounds that part-time or non-traditional students are usually older compared to traditional-age college stu-dents, and may enter college with work experience or family situations that can predispose them to a much different pattern of internet use behavior. As a result, re-spondents indicating they were graduate students (15, 1.40%) were excluded from final analysis. We checked the remaining data for missing values, and found 16 cases (1.49%) had missing values in the variable sleep ef-ficiency, while all other cases had complete responses in every variable used in the analysis. As Schafer et al. sug-gested that a missing rate of 5% or less is commonly in-consequential [44], we performed complete case analysis. The final sample consisted of 1040 traditional-age college students, 416 of whom were males (40%) and 624 were females (60%). The maximum estimated sam-pling error of our sample was calculated to be ±3.04% with a 95% confidence probability [57].

Measures

Internet gaming addiction (IGD)

IGD was measured by the Gaming Addiction Scale (GAS) developed by Lemmens et al. The GAS consists of seven Likert-type items (1 = never, 2 = rarely, 3 = sometimes, 4 = often, 5 = very often), which all begin with a statement “During the last 6 months, how often …” For example, “During the last 6 months, how often did you think about playing a game all day long?” Total score of the GAS is between 7 and 35, with higher scores indicating higher level of gaming addiction. Chinese ver-sion of the GAS was utilized to test IGD among adoles-cents, with a Chronbach’s alpha value between 0.93 and 0.94 [3]. Concurrent validity of the GAS has been con-firmed by its correlation with Internet Addiction and hours of gaming among Italian adolescents [8]. Good in-ternal reliability was reported in the present study (Chronbach’s alpha value = 0.951).

Problematic internet use

Problematic internet use was assessed by Young’s 20-item Internet Addiction Test (IAT) [58]. The scale was developed based on the diagnostic criteria for patho-logical gambling under the DSM-IV-TR. Each item is rated using a Likert scale (1 = never, 2 = rarely, 3 = some-times, 4 = often, 5 = very often). For example,“How often do you find that you stay online longer than you intended?” The total score of the IAT is between 20 and 100. Chinese version of the IAT has demonstrated good internal consistency (Cronbach’s alpha = 0.93 [22];). Concurrent validity of the IAT has also been confirmed

by its correlation with the Revised Chen Internet Addic-tion Scale (r = 0.46 [22];), the average online time per day (r = 0.40 for weekdays, r = 0.37 for weekends [22];), and the Mobile Phone Dependence Questionnaire (r = 0.59 [23];). Good internal reliability was reported in the present study (Chronbach’s alpha value = 0.938).

Sleep quality

Sleep quality was assessed using the Pittsburgh Sleep Quality Index (PSQI). The PSQI consists of 18 items that measure seven dimensions of sleep quality over the past month [24]: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep distur-bances, use of sleep medication, and daytime dysfunc-tion. For example, “During the past month, how often have you had trouble sleeping because you have pain?” The total score of each dimension ranges from 0 to 3, with higher scores indicating poorer sleep quality. The total score of the whole scale is obtained by summing scores on each of the seven dimensions, ranging from 0 to 21. Chinese version of the PSQI has exhibited ad-equate internal consistency [25]. Consistent with values (0.62–0.66) reported in previous studies [25, 26], Cron-bach’s alpha for the scale in the present study was 0.64 and considered to be acceptable. Composite reliability for the scale was 0.78, exceeding the recommended minimum value of 0.7 [49].

Psychiatric symptoms

Psychiatric symptoms were assessed using the Symptom Checklist 90-Revised (SCL-90-R) [27]. The SCL-90-R is a widely used self-report scale consisting of 90 items that examines nine symptomatic dimensions: somatization, obsessive-compulsiveness, interpersonal sensitivity, de-pression, anxiety, hostility, phobic anxiety, paranoid ideation, and psychoticism. An example of the items would be “How much were you bothered or distressed over the past 4 weeks by headaches?” The score of each item ranges from 0 to 5 (0 = not at all, 1 = a little bit, 2 = moderately, 3 = quite a bit, 4 = extremely), and score of each item is summed up to produce a total score be-tween 0 and 360. Chinese version of the SCL-90-R ex-hibited good internal consistency (Cronbach’s α = 0.98) [28]. Scores on the nine subscales were significantly cor-related with scores on the whole scale, indicating good structural validity [28,29]. Criterion validity of the SCL-90-R has also been examined through its correlation with self-reported quality of life [28]. Cronbach’s alpha

value for the scale was 0.977 in the current study.

Substance use

Substance use, including tobacco use, binge drinking and other drug use, were assessed within a 12-month time period. Tobacco use was assessed by asking

(4)

whether respondents had used either traditional ciga-rettes or e-cigaciga-rettes. Binge drinking was assessed by asking whether respondents had at least had five drinks (including beer, wine, champagne and liquor) in one set-ting for males, and at least had four drinks in one setset-ting for females. Other drug use was assessed by asking whether respondents had used marijuana, heroin, MDMA, sedatives, or over the counter (OTC) medica-tions. Because the number of positive responses to each type of other drug use was relatively low, we combined responses to each type of other drug use into a single binary variable, comparing with those used at least one type of other drugs against those who answered“no” to all types of other drug use.

Social media addiction

Social media addiction was assessed by the Social Media Addiction Scale - Student Form (SMA-SF) developed by Sahin [30]. The scale consists of 29 items measuring 4 di-mensions of social media addiction: virtual tolerance, vir-tual communication, virtual problem and virtual information. Total score of the SMA-SF ranges from 29 to 145, with higher scores indicating higher levels of social media addiction [30]. Each item can be rated on a 5-point Likert-type scale (1 = Definitely not appropriate, 2 = Not appropriate, 3 = Undecided, 4 = Appropriate, 5 = Quite ap-propriate) In the original study, the SMA-SF exhibited good internal reliability (Cronbach’s alpha = 0.93), split-half reliability (Guttmann Split-Half value = 0.90) and test-retest reliability (test-test-retest coefficient = 0.94). In this study, the SMA-SF demonstrated good internal reliability, the value of Cronbach’s alpha for the scale was calculated to be 0.955.

Procedure

We used a popular professional online survey platform (https://www.wjx.cn/) in China to prepare and present the survey. Recruitment occurred between June to Au-gust, 2019. The link to the survey was distributed via Wechat messages. All participants were informed that participation was completely anonymous and that their responses would be kept confidential. Upon completion of the survey, each participant was given 2 Chinese yuan (about $0.3 USD).

Statistical analysis

All analyses were conducted using SPSS 22.0 and AMOS 24.0. Reliability of the scale was assessed by internal consistency and split-half reliability. Validity of the scale was assessed by structural validity, convergent validity, discriminant validity, and concurrent validity.

Reliability

Internal consistency represents the extent to which dif-ferent items are correlated, and was assessed using Cronbach’s alpha coefficients, theta coefficient and omega coefficient [31]. A coefficient of greater than 0.7 indicates good internal consistency [32]. Split-half reli-ability indicates streli-ability of the scale, and was measured using the Spearman-Brown coefficient, with higher values representing higher stability [33].

Validity

Exploratory factor analysis (EPA) was conducted to exam-ine the factor structure of the GAS. Previous studies have found the 7-item GAS to have a unidimentional structure [6]. Confirmatory factor analysis (CFA) was used to meas-ure structural validity of the GAS. The goodness-of-fit of the model was examined using a series of indices: the χ2 to degrees of freedom ratio (χ2/ df), comparative fit index (CFI), goodness of fit index (GFI) and root mean square error of approximation (RMSEA). The assessment criteria for each index were: χ2 / df < 3, CFI > 0.9, GFI > 0.9 and RMSEA< 0.08 [34].

Convergent validity of the scale was measured by the value of average variance extracted (AVE), which was calculated using a formula

P

ðfactor loading valueÞ2 P

ðfactor loading valueÞ2þPðmeasurement errorÞ. The convergent validity of a scale is considered acceptable if the value of AVE is higher than 0.50 [35]. Concurrent validity was measured by the association between the GAS and the IAT, PSQI, SCL-90 and substance use. The Pearson product-moment correlation coefficient was used to assess their associations. The correlation coefficient ranges between − 1.0 and 1.0, an absolute value of ≥0.5 is considered large, an absolute value between 0.3 and 0.5 is considered moder-ate, an absolute value between 0.1 and 0.3 is considered small, and an absolute value less than 0.1 is considered trivial [36]. Discriminant validity refers to whether dissimilar con-structs can be differentiated, and was measured by the cor-relation between GAS and SMA-SF in the present study. A Pearson’s value of less than 0.85 indicates adequate discrim-inant validity [35].

Multigroup confirmatory factor analysis (MCFA) was conducted to test the measurement invariance of the GAS across gender, class standing, income and parental educational level. Three nested models were adopted: 1) a configural model (model 1), in which all factor param-eters were freely estimated; 2) a weak factorial invariance model (model 2), in which item loadings were con-strained to be equal across groups; and 3) a strict factor-ial invariance model (model 3), in which item residuals were constrained to be equal across groups. Chen [37] recommends that measurement invariance is not sup-ported if CFI decreases by a value greater than 0.01 or RMSEA increases by a value greater than 0.015 [37].

(5)

Because the GAS is an ordinal scale, maximum likeli-hood estimation may not be the appropriate estimate, asymptotically distribution-free estimation was used to accommodate non-normally distributed data in SEM analyses instead.

Ethics

The study procedures were carried out according to the Declaration of Helsinki. The Institutional Review Board of the [Name of the Institution] approved this study. All participants were informed about the study, and all pro-vided informed consent.

Results

Sample characteristic of the final 1040 respondents (416 male and 624 female) were shown in Table1.

Reliability

The GAS exhibited satisfactory internal consistency and split-half reliability, with Cronbach’s alpha value of 0.951, theta coefficient value of 0.953, omega coefficient value of 0.959, and a Spearman-Brown coefficient value of 0.938. All the items demonstrated good corrected item-total correlations, ranging from 0.781 to 0.867 (Table2).

Validity

Structural validity, convergent validity and discriminant validity

EFA revealed a one-factor model of the GAS, which was further confirmed by CFA. The model exhibited satisfac-tory fit indices: χ2 / df = 0.877 (p < 0.05), CFI = 0.999, GFI = 0.996, RMSEA = 0.000 (90% CI = 0.000, 0.035). In addition, no standardized factor loading was below 0.76 (Table 3). The GAS exhibited good convergent and dis-criminant validity, with the AVE value to be 0.734 and the value of Pearson’s correlation coefficient to be 0.520 (Table3).

Concurrent validity

As shown in Table4, correlation between the GAS total score and the IAT total score was large (r = 0.672). Cor-relation between the GAS total score and the SCL-90-R total score (r = 0.455) and subscale scores was moderate, somatization (r = 0.483), obsessive-compulsive symptoms (r = 0.382), interpersonal sensitivity (r = 0.390), depres-sion (r = 0.414), anxiety (r = 0.440), hostility (r = 0.457), phobic anxiety (r = 0.467), paranoid ideation (r = 0.457), and psychoticism (r = 0.427). Correlation between the GAS total score and substance use total score was also moderate (r = 0.367). However, the correlation of GAS total score with PSQI total score was small (r = 0.220).

Measurement invariance

Model fit indices across gender, class standing, family in-come and parental educational level are presented in Table5. Results indicated that the GAS had strict measure-ment invariance across educational level of father and mother respectively, supported by the acceptance of model 2 and model 3. Model fit indices includingΔχ2 (df), ΔCFI

Table 1 Sample characteristics

Characteristics Total (n = 1040) Gender Male 416 (40%) Female 624 (60%) Class standing Freshmen 264 (25.4%) Sophomores 491 (47.2%)

Juniors & Seniors 285 (27.4%)

Family income

< 50,000 241 (23.2%)

50,000 ~ 100,000 309 (29.7%)

50,000 ~ 200,000 302 (29.0%)

> 200,000 188 (18.1%)

Father’s educational level

≤ Middle school 381 (36.3%)

High school 258 (24.8%)

≥ College 401 (38.6%)

Mother’s educational level

≤ Middle school 436 (41.9%)

High school 250 (24.0%)

≥ College 354 (34.0%)

Internet gaming disorder 16.41 (7.07)

Problematic internet use 54.09 (16.29)

Sleep quality 5.45 (2.92) Psychological symptom Interpersonal sensitivity 6.38 (7.36) Depression 8.81 (10.56) Anxiety 5.58 (7.58) Hostility 3.40 (4.64) Phobic anxiety 3.35 (5.24) Paranoid ideation 3.31 (4.59) Phychoticism 5.62 (7.62)

Socia media addiction 81.29 (22.77)

Substance use

Past-year tobacco use 179 (17.2%)

Past-year binge drinking 276 (26.5%)

Past-year substance use 304 (29.2%)

Note: Values are presented as mean (SD) or number (percentage) when appropriate

(6)

andΔRMSEA are presented in Table6. For gender, values of model fit indices wereΔ χ2 (df) = 14.910 (6), ΔCFI = − 0.013,ΔRMSEA = 0.014 for weak factorial model, and Δ χ2 (df) = 105.666 (21), ΔCFI = − 0.120, ΔRMSEA = 0.041 for strict factorial model. For class standing, values of model fit indices were Δχ2 (df) = 26.129 (12), ΔCFI = − 0.016, ΔRMSEA = 0.019 for weak factorial model, and Δχ2 (df) = 72.809 (42), Δ CFI = − 0.037, ΔRMSEA = 0.021 for strict factorial model. For family income, values of model fit indi-ces wereΔ χ2 (df) = 21.76 (12), ΔCFI = − 0.011, ΔRMSEA = 0.005 for weak factorial model, andΔ χ2 (df) = 78.121 (42), ΔCFI = − 0.042, ΔRMSEA = 0.010 for strict factorial model. Results of model 2 and 3 revealed that the GAS exhibited no weak or strict measurement invariance across gender, class standing and family income.

Considering the rejection of weak measurement invari-ance across gender, class standing and family income,

partial invariance for each item was further examined. For gender, after checking the result of measurement in-variance, item 2 had the largest value. So the loading of item 2 was set to vary and the weak measurement invari-ance was tested again. Values of model fit indices were Δ χ2 (df) = 6.027 (5), ΔCFI = − 0.002, ΔRMSEA = 0.002, indicating that partial invariance was supported for gen-der when the loading of item 2 was set to vary. The same process of setting free the loading of the item with the largest measurement invariance until |ΔCFI| < 0.01 and ΔRMSEA < 0.015 was repeated for class standing and family income. As shown in Table 7, the non-invariant factors were salience, mood modification, re-lapse, withdrawal, conflict and problems for gender; sali-ence, tolerance, mood modification, relapse, withdrawal and conflict for class standing; salience, tolerance, mood modification, relapse, conflict and problems for family income.

Discussion

This study is the first to examine the psychometric properties and measurement invariance of the 7-item GAS among Chinese college students. Consistent with results from previous studies, [9–11], we found the GAS had a unidimensional structure and exhibited excellent reliability. Our findings on measurement in-variance of the GAS across different socio-demographic groups lent support to existing studies that found measurement invariance of the GAS across linguistic groups [9], gender and groups spending dif-ferent amounts of time on gaming [11]. More specif-ically, we found the GAS had strict measurement invariance across parental educational levels, suggest-ing that scores on the GAS reflected respondents’ gaming behaviors rather than the influence of their parents’ level of education. We also found partial measurement invariance was supported for gender, class standing and family income groups. That is, all items except for tolerance were found to be operating equivalently across gender; all items except for prob-lems were found to be operating equivalently across class standing; all items except for withdrawal were found to be operating equivalently across family in-come. According to a previous study, the imprecision of the concept of withdrawal may cause unexplained variance in different groups [9]. Our study further re-vealed the relative weakness of the items of problems and tolerance when assessing internet gaming addic-tion among the Chinese college student populaaddic-tion.

Our results indicated a moderate association between the scores on the GAS and the total as well as subscale scores on the SCL-90-R. This finding was consistent with previous studies reporting the association between IGD and subscales of SCL-90-R such as depression,

Table 2 Corrected item-total correlation and reliability indices

Item Construct Corrected item-total

correlation

item1 Salience 0.849

item2 Tolerance 0.781

item3 Mood modification 0.855

item4 Relapse 0.867 item5 Withdrawal 0.835 item6 Conflict 0.830 item7 Problems 0.841 Chronbachα 0.951 Theta coefficient 0.953 Omega coefficient 0.959 Spearman-Brown Coefficient 0.938

Table 3 Standardized factor loading, goodness-of-fit indices, convergent and discriminant validity indices

Item Construct Factor loading

item1 Salience 0.86

item2 Tolerance 0.76

item3 Mood modification 0.89

item4 Relapse 0.88 item5 Withdrawal 0.87 item6 Conflict 0.87 item7 Problems 0.86 χ2 /df 0.877 CFI 0.999 GFI 0.996 RMSEA 0.000 AVE 0.734 Pearson’r 0.520

(7)

anxiety, somatization [38–40], interpersonal sensitivity, obsessive-compulsiveness, phobic anxiety, hostility [40], psychoticism and overall severity [41]. Although gaming may be a way to cope with psychological distress, yet, excessive gaming can result in elevated levels of depres-sion, anxiety and social phobia [42]. Moreover, excessive gaming may lead to increased risk of exposure to violent

games, as gaming addicts seemed to have more norma-tive beliefs about aggressions and to engage in more hos-tile behaviors [42]. The overlap between IGD and obsessive-compulsiveness may be attributed to impair-ment in inhibitory control, which may lead to repetitive dysfunctional behaviors [43]. Excessive gaming was also associated with reduced motivation in other social

Table 4 Correlations between GAS and other constructs

1 2 3 4 5 6 7 8 9 10 11 12 13 1.GAS 1.00 0.67 0.22 0.48 0.38 0.39 0.41 0.44 0.46 0.47 0.46 0.43 0.37 2.IAT 1.00 0.31 0.43 0.45 0.43 0.45 0.43 0.43 0.42 0.43 0.41 0.26 3.PSQI 1.00 0.42 0.48 0.44 0.47 0.45 0.42 0.39 0.42 0.41 0.20 4.Somatization (subscale) 1.00 0.82 0.82 0.86 0.91 0.87 0.90 0.88 0.87 0.61 5.Obsessive-compulsiveness (subscale) 1.00 0.91 0.91 0.88 0.84 0.82 0.85 0.86 0.43

6.interpersonal sensitivity (subscale) 1.00 0.93 0.90 0.88 0.85 0.90 0.90 0.46

7.Depression (subscale) 1.00 0.93 0.88 0.87 0.89 0.91 0.49

8.Anxiety (subscale) 1.00 0.92 0.91 0.92 0.93 0.55

9.Hostility (subscale) 1.00 0.89 0.91 0.89 0.55

10.Phobic anxiety (subscale) 1.00 0.89 0.89 0.59

11.Paranoid ideation (subscale) 1.00 0.92 0.55

12.Psychoticism (subscale) 1.00 0.57

13.Substance use 1.00

Note:GAS Gaming Addiction Scale, IAT Internet Addiction Test, PSQI Pittsburgh Sleep Quality Index, SCL-90-R Symptom Checklist-90-Revised

Table 5 Factor loading and model fit across gender, class standing, family income and parental educational level

Factor loading Model fit

Item Item1 Item2 Item3 Item4 Item5 Item6 Item7 χ2 /df CFI GFI RMSEA

Gender Male 0.80 0.72 0.87 0.85 0.83 0.84 0.85 0.714 1.000 0.996 0.000 Female 0.90 0.8 0.91 0.90 0.89 0.87 0.88 1.411 0.99 0.983 0.026 Class standing Freshmen 0.90 0.75 0.93 0.87 0.92 0.86 0.98 1.469 0.981 0.976 0.042 Sophomores 0.85 0.76 0.89 0.88 0.84 0.89 0.83 0.537 1.000 0.994 0.000

Juniors & Seniors 0.85 0.84 0.87 0.91 0.88 0.84 0.90 0.764 1.000 0.993 0.000

Family income

< 50,000 0.856 0.773 0.93 0.931 0.89 0.919 0.897 0.898 1.000 0.987 0.000

50,000 ~ 100,000 0.859 0.785 0.887 0.834 0.914 0.85 0.874 1.179 0.995 0.985 0.024

50,000 ~ 200,000 0.871 0.825 0.881 0.878 0.854 0.844 0.873 0.939 1.000 0.986 0.000

> 200,000 0.855 0.77 0.884 0.893 0.898 0.879 0.876 1.825 0.973 0.972 0.066

Father’s educational level

≤ Middle school 0.877 0.818 0.899 0.911 0.894 0.9 0.852 1.037 0.999 0.990 0.010

High school 0.856 0.724 0.925 0.894 0.851 0.869 0.905 1.072 0.998 0.987 0.010

≥ College 0.845 0.81 0.871 0.851 0.876 0.831 0.867 0.981 1.000 0.985 0.000

Mother’s educational level

≤ Middle school 0.894 0.789 0.902 0.908 0.894 0.887 0.894 1.836 0.986 0.985 0.044

High school 0.871 0.814 0.896 0.842 0.867 0.894 0.829 1.041 0.999 0.985 0.013

(8)

activities, which could result in subsequent interpersonal problems [45].

We found a moderate association between the total score on the GAS and substance use. This finding lent support to previous findings on the positive association between IGD and alcohol, tobacco, and illicit drug use

[46–48]. Substance use has been found to be a common comorbidity of Internet addiction, as those with sub-stance use disorder seemed to exhibit similar core symp-toms of IGD [50, 51]. Both substance use disorder and IGD have been associated with deficient reward system functions, manifested as having higher responsiveness to substances and video games and lower responsiveness to other natural rewards as a result of altered dopamine levels. Another shared mechanism of these two types of addictive behaviors involves high trait impulsivity. Indi-viduals with high trait impulsivity tend to perform poorly on decision-making tasks, focusing on short-term consequences instead, thus giving priority to addictive behaviors rather than other behaviors [50,52].

In regards to IGD and sleep quality, some studies found their association to be significant [53]. It is plaus-ible that some gamers may become deprived of sleep due to significant amount of time spent playing games, or report daytime sleepiness as a consequence [54]. Some studies even showed that delayed sleep phase can improve by readjusting individual circadian rhythm with exogenous day-light cycle, thus alleviating gaming-related sleep problems [55]. However, a systematic re-view study by Lam found insufficient evidence support-ing a strong association between IGD and poor sleep quality [56], but found a stronger association between problematic internet use and sleep problems. In line with Lam’s review study, we found the association be-tween gaming addictions and sleep quality to be smaller than the association between problematic internet use and sleep quality, implying that differing mechanisms may be involved in how playing internet games or

Table 6 Measurement invariance across gender, class standing, family income and parental educational level

Model Model Fit Indices

χ2 (df) Δχ2 (Δdf) CFI ΔCFI RMSEA ΔRMSEA

Gender Configural 14.876 (14) 0.999 0.008

Weak factorial 29.786 (20) 14.910 (6) 0.986 −0.013 0.022 0.014

Strict factorial 120.542 (35) 105.666 (21) 0.879 −0.120 0.049 0.041

Class standing Configural 19.403 (21) 1.000 0.000

Weak factorial 45.532 (33) 26.129 (12) 0.984 −0.016 0.019 0.019

Strict factorial 92.212 (63) 72.809 (42) 0.963 −0.037 0.021 0.021

Family income Configural 55.048 (49) 0.993 0.011

Weak factorial 76.808 (61) 21.76 (12) 0.982 −0.011 0.016 0.005 Strict factorial 133.169 (91) 78.121 (42) 0.951 −0.042 0.021 0.010 Education(F) Configural 21.631 (21) 0.999 0.005 Weak factorial 40.759 (33) 19.128 (12) 0.990 −0.009 0.015 0.010 Strict factorial 71.74 (63) 50.109 (42) 0.989 −0.010 0.012 0.007 Education(M) Configural 29.091 (21) 0.990 0.019 Weak factorial 35.77 (33) 6.679 (12) 0.997 0.007 0.009 −0.010 Strict factorial 78.222 (63) 49.131 (42) 0.981 −0.009 0.015 −0.004

Table 7 Partial measurement invariance across gender, class standing and family income

Model Model Fit Indices Gender

χ2 (df) Δχ2 (Δdf) CFI ΔCFI RMSEA ΔRMSEA

Model 1.1 14.876 (14) 0.999 0.008 Model 1.2 29.786 (20) 14.910 (6) 0.986 −0.013 0.022 0.014 Model 1.3 20.903 (19) 6.027 (5) 0.997 −0.001 0.010 0.002 Class standing Model 2.1 19.403 (21) 1.000 0.000 Model 2.2 45.532 (33) 26.129 (12) 0.984 −0.016 0.019 0.019 Model 2.3 32.775 (31) 13.372 (10) 0.998 −0.002 0.007 0.007 Family income Model 3.1 55.048 (49) 0.993 0.011 Model 3.2 76.808 (61) 21.76 (12) 0.982 −0.011 0.016 0.005 Model 3.3 67.005 (59) 0.991 −0.002 0.011 0

Model 1.1: Unconstrained model Model 1.2: All item loading equal Model 1.3: item loadings 1,3,4,5,6,7 equal Model 2.1: Unconstrained model Model 2.2: All item loading equal Model 2.3: item loadings 1,2,3,4,5,6 equal Model 3.1: Unconstrained model Model 3.2: All item loading equal Model 3.3: item loadings 1,2,3,4,6,7 equal

(9)

engaging in excessive internet use relays to sleep quality. Although investing these mechanisms is beyond the aim of this study, future studies are needed to examine the underlying mechanisms contributing to these differences.

Previous studies have indicated that gaming addiction was associated with less conscientiousness and low openness, while social networking addiction was associ-ated with high neuroticism and extraversion [14], sug-gesting that gaming addiction and social networking addiction may be associated with differing personality traits. In the present study, the relatively small correl-ation between GAS and SMA-SF scores suggested that the GAS can discriminate gamers from people with other types of Internet-related addictive behaviors such as social media addiction.

Strengths and limitations

To our knowledge, this is the first study to assess the psychometric properties of the 7-item GAS among the college student population in China. Findings of this study provided ample support for the application of the GAS as a screening tool to assess IGD among this popu-lation. However, this study also has several limitations that we would like to acknowledge along with prospect-ive directions for future research. First, this study mainly utilized self-report data such as on sleep quality and sub-stance use, reporting or recalling biases might have af-fected the accuracy of the testing results. Future studies may need to incorporate more objective measures. Sec-ond, the cross-sectional nature of our data limited us to draw tentative conclusions about the temporal sequence of IGD development. Longitudinal studies may be needed to clarify this sequence. Third, our study mainly focused on Chinese college students, our findings may not be applicable to same-age populations in other countries. More studies from other countries to corrob-orate our findings of the GAS.

Conclusions

This study entails that the 7-item GAS is a reliable and valid instrument for assessing IGD among Chinese col-lege students, ensuring researchers and clinicians that it is an adequate tool to examine problematic gaming.

Abbreviations

IGD:Internet Gaming Disorder; GAS: Gaming Addiction Scale; PIU: Problematic Internet Use; MG-CFA: Multigroup Confirmatory Factor Analysis

Acknowledgements

We would like to thank all the participants who completed our surveys used for analysis in this study.

Authors’ contributions

YL: Writing-Original draft preparation, Methodology. QW: Conceptualization, Funding acquisition, Methodology, Project Administration, Supervision,

Writing-reviewing and editing. MJ: Writing-reviewing. BW: Data curation. YA: Data curation. ZL: Methodology. All authors have read and approved the manuscript.

Funding

This work was supported by Shanghai Jiao Tong University School of Medicine under Grant 19X100040041. The funding body had no role in the design of the study, or the collection, analysis, and interpretation of data, or in the writing of the manuscript.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate

The study procedures were carried out by the Declaration of Helsinki. The Institutional Review Board of Shanghai Jiao Tong University School of Medicine College of Public Health approved this study (No. SJUPN201907). All participants were informed about the study, and all provided written informed consent.

Consent for publication Not applicable. Competing interests

The authors declare that they have no competing interests. Author details

1

School of Public Health, Shanghai Jiao Tong University School of Medicine, No 227 S Chongqing Road, Shanghai 200025, China.2Department of

Industrial Education, National Taiwan Normal University, Taipei, Taiwan.

3Shanghai Jiao Tong University School of Medicine, College of Clinical

Medicine, Shanghai, China.4School of Social Development and Public Policy, Fudan University, Shanghai, China.

Received: 5 March 2020 Accepted: 20 August 2020

References

1. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 5th Edn. Arlington, VA: Author; 2013.https://doi.org/10. 1176/appi.books.9780890425596.

2. Müller KW, Janikian M, Dreier M, et al. Regular gaming behavior and internet gaming disorder in European adolescents: results from a cross-national representative survey of prevalence, predictors, and

psychopathological correlates. Eur Child Adolesc Psychiatry. 2015;24:565–74.

https://doi.org/10.1007/s00787-014-0611-2.

3. Wang CW, Chan CL, Mak KK, Ho SY, Wong PW, Ho RT. Prevalence and correlates of video and internet gaming addiction among Hong Kong adolescents: a pilot study. Sci World J. 2014;2014:874648.https://doi.org/10. 1155/2014/874648.

4. Xu J, Shen LX, Yan CH, Hu H, Yang F, Wang L, et al. Personal characteristics related to the risk of adolescent internet addiction: a survey in Shanghai, China. BMC Public Health. 2012b;12:1106. https://doi.org/10.1186/1471-2458-12-1106.

5. King DL, Haagsma MC, Delfabbro PH, Gradisar M, Griffiths MD. Toward a consensus definition of pathological video-gaming: a systematic review of psychometric assessment tools. Clin Psychol Rev. 2013;33:331–42. 6. Lemmens JS, Valkenburg PM, Peter J. Development and validation of a

game addiction scale for adolescents. Media Psychol. 2009;12(1):77–95.

https://doi.org/10.1080/15213260802669458.

7. Walther B, Morgenstern M, Hanewinkel R. Co-occurrence of addictive behaviours: personality factors related to substance use, gambling and computer gaming. Eur Addict Res. 2012;18(4):167–74.https://doi.org/10. 1159/000335662.

8. Costa S, Barberis N, Gugliandolo MC, Liga F, Cuzzocrea F, Verrastro V. Examination of the psychometric characteristics of the Italian version of the game addiction scale for adolescents. Psychol Rep, 33294119838758. Advance online publication. 2019.https://doi.org/10.1177/ 0033294119838758.

(10)

9. Khazaal Y, Chatton A, Rothen S, Achab S, Thorens G, Zullino D, Gmel G. Psychometric properties of the 7-item game addiction scale among french and German speaking adults. BMC psychiatry. 2016;16:132.https://doi.org/ 10.1186/s12888-016-0836-3.

10. Lemos IL, Cardoso A, Sougey EB. Validity and reliability assessment of the Brazilian version of the game addiction scale (GAS). Compr Psychiatry. 2016; 67:19–25.https://doi.org/10.1016/j.comppsych.2016.01.014.

11. Lin CY, Imani V, Brostrom A, Arestedt K, Pakpour AH, Griffiths MD. Evaluating the psychometric properties of the 7-item Persian game addiction scale for Iranian adolescents. Front Psychol. 2019;10:149.https://doi.org/10.3389/ fpsyg.2019.00149.

12. Lloret Irles D, Morell Gomis R, Marzo Campos JC, Tirado González S. Validación española de la Escala de Adicción a Videojuegos para Adolescentes (GASA) [Spanish validation of Game Addiction Scale for Adolescents (GASA)]. Aten Primaria. 2018;50(6):350–8.

13. Shu MY, Ivan Jacob AP, Meng Xuan Z, Anise MSW. Psychometric validation of the internet gaming Disorder-20 test among Chinese middle school and university students. J Behav Addict. 2019;8(2):295–305.https://doi.org/10. 1556/2006.8.2019.18.

14. Wang CW, Ho RT, Chan CL, Tse S. Exploring personality characteristics of Chinese adolescents with internet-related addictive behaviors: trait differences for gaming addiction and social networking addiction. Addict Behav. 2015;42:32–5.https://doi.org/10.1016/j.addbeh.2014.10.039. 15. Bonnaire C, Liddle HA, Har A, Nielsen P, Phan O. Why and how to include

parents in the treatment of adolescents presenting internet gaming disorder? J Behav Addict. 2019;8(2):201–12.https://doi.org/10.1556/2006.8. 2019.27.

16. Festl R, Scharkow M, Quandt T. Problematic computer game use among adolescents, younger and older adults. Addiction. 2013;108(3):592–9.https:// doi.org/10.1111/add.12016.

17. Vollmer C, Randler C, Horzum MB, Ayas T. Computer game addiction in adolescents and its relationship to chronotype and personality. SAGE Open. 2014;4(1):2158244013518054.https://doi.org/10.1177/2158244013518054. 18. Xu Z, Turel O, Yuan Y. Online game addiction among adolescents:

motivation and prevention factors. Eur J Inf Syst. 2012a;21(3).https://doi. org/10.1057/ejis.2011.56.

19. Ye YL, Wang PG, Qu GC, Yuan S, Phongsavan P, He QQ. Associations between multiple health risk behaviors and mental health among Chinese college students. Psychol Health Med. 2016;21(3):377–85.https://doi.org/10. 1080/13548506.2015.1070955.

20. Liu W, He MZ, Wang Y, Wang Y, Zhou Y, Wu M, Tang Z, Dai Y, Yuan B, Zhen S, Cheskin LJ. Differences in health-related behaviors between middle school, high school, and college students in Jiangsu province, China. Asia Pac J Clin Nutr. 2017;26(4):731–7.https://doi.org/10.6133/apjcn.072016.06. 21. Meredith W. Measurement invariance, factor analysis and factorial

invariance. Psychometrika. 1993;58:525–43.https://doi.org/10.1007/ BF02294825.

22. Lai CM, Mak KK, Watanabe H, Ang RP, Pang JS, Ho RC. Psychometric properties of the internet addiction test in Chinese adolescents. J Pediatr Psychol. 2013;38(7):794–807.https://doi.org/10.1093/jpepsy/jst022. 23. Chin F, Leung CH. The concurrent validity of the internet addiction test

(IAT) and the Mobile phone dependence questionnaire (MPDQ). PLoS One. 2018;13(6):e0197562.https://doi.org/10.1371/journal.pone.0197562. 24. Buysse DJ, Iii CFR, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh sleep

quality index: a new instrument for psychiatric practice and research. Psychiatry Res. 1989;28(2):193–213. https://doi.org/10.1016/0165-1781(89)90047-4.

25. Guo S, Sun W, Liu C, Wu S. Structural validity of the Pittsburgh sleep quality index in Chinese undergraduate students. Front Psychol. 2016;7:1126.

https://doi.org/10.3389/fpsyg.2016.01126.

26. Lau J, Walden DL, Wu A, Cheng KM, Lau M, Mo P. Bidirectional predictions between internet addiction and probable depression among Chinese adolescents. J Behav Addict. 2018;7(3):633–43.https://doi.org/10.1556/2006. 7.2018.87.

27. Derogatis LR. SCL-90-R: administration, scoring & procedures manual-II, for the R (revised) version and other instruments of the psychopathology rating scale series. 2nd ed. Towson: Clinical Psychometric Research; 1992. 28. Yu Y, Wan C, Huebner ES, Zhao X, Zeng W, Shang L. Psychometric

properties of the symptom check list 90 (SCL-90) for Chinese undergraduate students. J Ment Health. 2019;28(2):213–9.https://doi.org/10.1080/09638237. 2018.1521939.

29. Zhang J, Zhang X. Chinese college students' SCL-90 scores and their relations to the college performance. Asian J Psychiatr. 2013;6(2):134–40.

https://doi.org/10.1016/j.ajp.2012.09.009.

30. Sahin C. Social media addiction scale-student form: the reliability and validity study. Turkish Online J Educ Technol - TOJET. 2018;17:169–82. 31. Wu X, Ni Z. The application of factor analysis in reliability and validity

assessment. Chinese J Prev Control Chron Dis. 1998;6(1):28–31http://doi. org/CNKI:SUN:ZMXB.0.1998-01-008.

32. Kwon M, Kim DJ, Cho H, Yang S. The smartphone addiction scale: development and validation of a short version for adolescents. PLoS One. 2013;8(12):e83558.https://doi.org/10.1371/journal.pone.0083558. 33. Basol G, Kaya AB. Motives and consequences of online game addiction: a

scale development study. Noro Psikiyatr Ars. 2018;55(3):225–32.https://doi. org/10.5152/npa.2017.17017.

34. McDonald RP, Ho MH. Principles and practice in reporting structural equation analyses. Psychol Methods. 2002;7(1):64–82.https://doi.org/10. 1037/1082-989x.7.1.64.

35. Lee D. The convergent, discriminant, and nomological validity of the depression anxiety stress Scales-21 (DASS-21). J Affect Disord. 2019;259:136 42.https://doi.org/10.1016/j.jad.2019.06.036.

36. Cohen J. Statistical power analysis for the behavioral sciences. J Am Stat Assoc ,2nd. 1988;334:499–500.https://doi.org/10.1016/B978-0-12-179060-8.50012-8. 37. Chen FF. Sensitivity of goodness of fit indexes to lack of measurement

invariance. Struct Equ Model Multidiscip J. 2007;14(3):464–504.https://doi. org/10.1080/10705510701301834.

38. Na E, Lee H, Choi I, Kim DJ. Comorbidity of internet gaming disorder and alcohol use disorder: a focus on clinical characteristics and gaming patterns. Am J Addict. 2017;26(4):326–34.https://doi.org/10.1111/ajad.12528. 39. Rho MJ, Lee H, Lee TH, Cho H, Jung DJ, Kim DJ, Choi IY. Risk factors for

internet gaming disorder: psychological factors and internet gaming characteristics. Int J Environ Res Public Health. 2017;15(1).https://doi.org/10. 3390/ijerph15010040.

40. Torres-Rodriguez A, Griffiths MD, Carbonell X, Oberst U. Internet gaming disorder in adolescence: psychological characteristics of a clinical sample. J Behav Addict. 2018;7(3):707–18.https://doi.org/10.1556/2006.7.2018.75. 41. Gonzalez-Bueso V, Santamaria JJ, Fernandez D, Merino L, Montero E,

Jimenez-Murcia S, et al. Internet gaming disorder in adolescents: personality, psychopathology and evaluation of a psychological intervention combined with parent Psychoeducation. Front Psychol. 2018;9:787.https://doi.org/10. 3389/fpsyg.2018.00787.

42. Gentile DA, Choo H, Liau A, Sim T, Li D, Fung D, Khoo A. Pathological video game use among youths: a two-year longitudinal study. Pediatrics. 2011; 127(2):e319–29.https://doi.org/10.1542/peds.2010-1353.

43. Kim YJ, Lim JA, Lee JY, Oh S, Kim SN, Kim DJ, et al. Impulsivity and compulsivity in internet gaming disorder: a comparison with obsessive-compulsive disorder and alcohol use disorder. J Behav Addict. 2017;6(4): 545–53.https://doi.org/10.1556/2006.6.2017.069.

44. Schafer JL & Graham JW. Missing Data Our View of the State of the Art. Psychological Methods 2002;7: 147-177.

45. Skarupova K, Blinka L. Interpersonal dependency and online gaming addiction. J Behav Addict. 2016;5(1):108–14.https://doi.org/10.1556/2006.5.2016.002. 46. Konkoly Thege B, Hodgins DC, Wild TC. Co-occurring substance-related and

behavioral addiction problems: a person-centered, lay epidemiology approach. J Behav Addict. 2016;5(4):614–22.https://doi.org/10.1556/2006.5.2016.079. 47. Lam LT. Risk factors of internet addiction and the health effect of internet

addiction on adolescents: a systematic review of longitudinal and prospective studies. Curr Psychiatry Rep. 2014b;16(11):508.https://doi.org/10. 1007/s11920-014-0508-2.

48. Van Rooij AJ, Kuss DJ, Griffiths MD, Shorter GW, Schoenmakers TM, Dike VDM. The (co-)occurrence of problematic video gaming, substance use, and psychosocial problems in adolescents. J Behav Addict. 2014;3(3):157–65.

https://doi.org/10.1556/JBA.3.2014.013.

49. Fornell C & Larcker DF (1981). Evaluating Structural Equation Models with Unobservable Variables and Measurement Error. Journal of Marketing Research, 18, 39-50.

50. Jorgenson AG, Hsiao RC, Yen CF. Internet addiction and other behavioral addictions. Child Adolesc Psychiatr Clin N Am. 2016;25(3):509–20.https://doi. org/10.1016/j.chc.2016.03.004.

51. Kuss DJ, Griffiths MD, Karila L, Billieux J. Internet addiction: a systematic review of epidemiological research for the last decade. Curr Pharm Des. 2014;20(25):4026–52.https://doi.org/10.2174/13816128113199990617.

(11)

52. Hammond CJ, Mayes LC, Potenza MN. Neurobiology of adolescent substance use and addictive behaviors: treatment implications. Adolescent Med. 2014;25(1):15–32.https://doi.org/10.1189/jlb.0112042.

53. Hawi NS, Samaha M, Griffiths MD. Internet gaming disorder in Lebanon: relationships with age, sleep habits, and academic achievement. J Behav Addict. 2018;7(1):70–8.https://doi.org/10.1556/2006.7.2018.16. 54. Achab S, Nicolier M, Mauny F, Monnin J, Trojak B, Vandel P, Sechter D,

Gorwood P, Haffen E. Massively multiplayer online role-playing games: comparing characteristics of addict vs non-addict online recruited gamers in a French adult population. BMC Psychiatry. 2011;11:144.https://doi.org/ 10.1186/1471-244X-11-144.

55. Lemola S, Brand S, Vogler N, Perkinson-Gloor N, Allemand M, Grob A. Habitual computer game playing at night is related to depressive symptoms. Pers Individ Differ. 2011;51(2):117–22.https://doi.org/10.1016/j. paid.2011.03.024.

56. Lam LT. Internet gaming addiction, problematic use of the internet, and sleep problems: a systematic review. Curr Psychiatry Rep. 2014a;16(4):444.

https://doi.org/10.1007/s11920-014-0444-1.

57. Wimmer RD & Dominick JR (2014). Mass Media Research: An Introduction. 10th Edition, Wadsworth, Cengage Learning, Boston.

58. Young KS. Internet addiction: the emergence of a new clinical disorder. CyberPsychol Behav. 1996;1(3):237–44.https://doi.org/10.1089/cpb.1998.1.237.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

數據

Table 1 Sample characteristics
Table 3 Standardized factor loading, goodness-of-fit indices, convergent and discriminant validity indices
Table 5 Factor loading and model fit across gender, class standing, family income and parental educational level
Table 6 Measurement invariance across gender, class standing, family income and parental educational level

參考文獻

相關文件

Then, we tested the influence of θ for the rate of convergence of Algorithm 4.1, by using this algorithm with α = 15 and four different θ to solve a test ex- ample generated as

Particularly, combining the numerical results of the two papers, we may obtain such a conclusion that the merit function method based on ϕ p has a better a global convergence and

Then, it is easy to see that there are 9 problems for which the iterative numbers of the algorithm using ψ α,θ,p in the case of θ = 1 and p = 3 are less than the one of the

volume suppressed mass: (TeV) 2 /M P ∼ 10 −4 eV → mm range can be experimentally tested for any number of extra dimensions - Light U(1) gauge bosons: no derivative couplings. =&gt;

Define instead the imaginary.. potential, magnetic field, lattice…) Dirac-BdG Hamiltonian:. with small, and matrix

• Formation of massive primordial stars as origin of objects in the early universe. • Supernova explosions might be visible to the most

We investigate some properties related to the generalized Newton method for the Fischer-Burmeister (FB) function over second-order cones, which allows us to reformulate the

(The Emotional and Mental Health Needs of Gifted Students and the Main Categories of Emotional and Mental