The Development of the Problematic Series Watching Scale (PSWS)

1 The Development of the Problematic Series Watching Scale (PSWS) 2 Gábor Orosz1,2,*, Beáta Bőthe1, István Tóth-Király1 3 1 4 5 2 Eötvös Loránd University, Institute of Psychology, Budapest, Hungary MTA Research Centre for Natural Sciences, Institute of Cognitive Neuroscience and 6 Psychology, Budapest, Hungary 7 8 9 *Corresponding author 10 Dr. Gábor Orosz, Eötvös Loránd University, Institute of Psychology 11 Izabella utca 46. 12 Budapest, H-1064, Hungary 13 E-mail.: [email protected] 14 Tel.: 00 36 70 237 9471 15 16 17 Funding sources: The first author was supported by the Hungarian Research Fund: The Role 18 of Individual Differences and Situative Factors in Skill learning, (PD 106027, PD 116686). 19 20 Authors’ contribution: All authors contributed to the manuscript equally: in terms of literature 21 review, data analysis, preparation of the manuscript. 22 23 Conflict of interest: The authors declare no conflict of interest. 24 1 1 Acknowledgements: We are grateful for the students of Eötvös Loránd University for their help 2 in the questionnaire development. 3 4 The Development of the Problematic Series Watching Scale (PSWS) 5 6 ABSTRACT 7 Background and aims: The goal of the present study was to create a short Problematic Series 8 Watching Scale (PSWS). Methods: On the basis of the six components model of Griffiths 9 (2005), six items were identified covering all components of problematic series watching. 10 Confirmatory factor analyses were carried out on two independent samples (N1 = 366, N2 = 11 752). Results: The PSWS has appropriate factor structure and reliability. The amount of free 12 time was not, but the series watching time was associated with PSWS scores. Women had 13 higher scores than men. Discussion: Before PSWS, no prior scale has been created to measure 14 problematic series watching. Further research is needed to properly assess its validity and 15 reliability; and for examining whether extensive series watching can lead to health-related and 16 psychosocial problems. Conclusion: In the increasingly digitalized world there are many 17 motivational forces which encourage people watching online series. In the light of these 18 changes, research on problematic series watching will be progressively relevant. 19 20 Keywords: series watching, Problematic Series Watching Scale, PSWS, CFA, six components 21 model 22 23 INTRODUCTION 24 Since the end of the 1990s problematic Internet use has been a controversial topic: some 25 researchers and clinicians (e.g. Young, 1998 or Ko, Yen, Chen, Chen, & Yen, 2005) claim that 2 1 problematic Internet use deserves a classification as a psychiatric disorder in its own right, while 2 others (e.g. Griffiths, 1999; Griffiths, 2000 or Yellowlees & Marks, 2007) claim that people 3 who spend excessive time on the Internet are not really addicted to it, but rather to the specific 4 activities they can pursue through this medium. Recently, several screen-related online 5 activities were identified as problematic online behaviors such as Facebook use (Andreassen, 6 Torsheim, Brunborg, & Pallesen, 2012a; Ryan, Chester, Reece, & Xenos, 2014), online 7 gambling (Chóliz, 2015; Griffiths & Barnes, 2008), online gaming (Demetrovics et al., 2012; 8 Grüsser, Thaleman, & Griffiths, 2006), or online pornography use (Grubbs, Sessoms, Wheeler, 9 & Volk, 2010; Kor et al., 2014). 10 According to the research of Pontes, Szabó, and Griffiths (2015), when participants had 11 to indicate their three most preferred online activities, watching videos and movies was 12 mentioned by one-third of their respondents besides accessing general information and news, 13 social networking, e-mailing and online chatting, and gaming or gambling. Although more than 14 one-third of the examined population mentioned watching online content as their preferred non- 15 work online activity, to our best knowledge no prior scientific study examined problematic 16 series watching. 17 It might be a relevant issue for many people because accessing series by downloading 18 or streaming is (a) very cheap (or free), (b) it is available for almost everyone who has 19 broadband internet access, (c) it does not depend on a certain place and time (i.e. playing squash 20 depends on a certain place and time), (d) series have a high variety—everyone can find one 21 which fits to his/her interest, (e) they are not age- and socio-economic status-dependent, (f) it 22 does not need effort to watch them, (g) and they are constructed to be highly enjoyable and 23 often contain cliffhangers which motivate the viewer to continue. These characteristics are 24 highly similar to the ones mentioned by Cooper (1998) regarding Internet and pornography. 3 1 Based on these reasons and Pontes et al.`s (2015) findings, we assume that problematic series 2 watching deserves scientific investigation. 3 Sussman and Moran’s (2013) review on television addiction did not differentiate 4 between the types of content that can be seen on television, they define television addiction as 5 a subjective craving for watching anything on television including both classical and online 6 content. In our research, we aimed to differentiate problematic series watching from the concept 7 of television addiction as we focus on the content of the problematic use (series watching), 8 rather than on the medium through which the problematic use happens (television). In our 9 research, we observe problematic series watching which could be done either through a 10 television (i.e. classical TV series) or a screen attached to a computer (i.e. Netflix). The device 11 itself is just a medium through which the user can reach the content with the latter being in our 12 main focus. 13 According to a recent study from a leading Internet television company with over 40 14 million members, 76% of series streamers mentioned that watching several episodes of a TV 15 show is a favored escape from their busy lives and 73% of their respondents have positive 16 feelings towards binge streaming TV (Netflix Media Center, 2013). In the literature of 17 problematic behaviors, escapism and mood modification are good candidates in predicting a 18 problematic activity. We assume that if three quarter of the series streamers watch several 19 episodes in line for escaping from the problems of the everyday life it is possible to assume that 20 problematic series watching can appear among them. 21 On the basis of Griffiths’ components model (2005), we can distinguish six core 22 elements of problematic series watching: (a) salience (series watching dominates thinking and 23 behavior); (b) mood modification (series watching modifies/improves mood); (c) tolerance 24 (increasing amounts of series watching are required to achieve initial effects); (d) withdrawal 25 (occurrence of unpleasant feelings when series watching is discontinued); (e) conflict (series 4 1 watching compromises social relationships and other activities); and (f) relapse (tendency for 2 reversion to earlier patterns of series watching after abstinence or control). 3 In the present study our goal was to measure problematic series watching. Therefore we 4 created a scale with appropriate factor structure and reliability which is based on the six 5 component model of Griffiths (2005). 6 7 8 METHODS Participants and procedure 9 The research was conducted with an online questionnaire system, the filling out lasted 10 approximately six minutes. The data collection occurred in two waves: the first one in April 11 2014 (Sample 1), the second one in April-May 2015 (Sample 2). In both cases, participants 12 were informed about the goal and the content of the study, also, the anonymity and the 13 confidentiality of their answers were ensured. They were asked to check a box if they agreed to 14 continue and participate. The first part of the questionnaire contained questions regarding 15 demographic data, such as gender, age and level of education. Participants were also asked to 16 estimate the amount of free time they have on an average weekday and weekend. In the 17 subsequent part, the items of the problematic series watching were presented. 18 Sample 1 consisted of 366 Hungarian respondents (Female = 258; 70.5%) who were 19 aged between 18 and 82 (Mage = 22.83, SDage = 6.23). 250 of them (68.3%) lived in Budapest, 20 35 (9.6%) in county towns, 58 (15.8%) in towns, and 23 (6.3%) in villages. Concerning their 21 level of education, 14 (3.8%) had primary school degree, 270 (73.8%) had a high school degree, 22 82 (22.4%) of them had a degree in higher education (bachelor, masters, or doctoral). Regarding 23 their free time, on an average weekday they had 4.12 hours of free time (SD = 3.11 hours). On 24 an average weekend they had 8.38 hours of free time (SD = 5.24). No outliers were detected 25 using Mahalanobis distance test, thus all 366 cases were retained in this sample. 5 1 Sample 2 consisted of 754 Hungarian persons (Female = 550; 72.9%). They were aged 2 between 18 and 67 years (Mage = 27.25, SDage = 8.70). 315 (41.8%) lived in Budapest, 151 3 (20.0%) in county towns, 199 (26.4%) in towns and 89 (11.8%) of them lived in villages. 48 4 (6.4%) people from this sample had elementary degree, 420 (55.7%) had a high school degree, 5 286 (37.9%) had a degree in higher education. Regarding their free time, on an average weekday 6 they had 4.82 hours of free time (SD = 3). On an average weekend they had 9.10 hours of free 7 time (SD = 4.44). In this case we did not measured the time they spent on series watching. After 8 examining the data with the Mahalanobis distance test, two outliers were found in this sample. 9 These cases were removed from further analyses, resulting in a final number of 752 valid cases. 10 11 Measures 12 To our best knowledge, currently there is no measurement that can assess problematic 13 series watching. For this purpose, on the basis of the Bergen Work Addiction Scale (BWAS – 14 Andreassen, Griffiths, Hetland, & Pallesen, 2012b), we created a new scale—Problematic 15 Series Watching Scale (PSWS)—to measure the six core elements of problematic series 16 watching in terms of (a) salience, (b) tolerance, (c) mood modification, (d) relapse, (e) 17 withdrawal, and (f) conflict. We chose this measure as a basis because it grasps each possible 18 element of problematic behaviors and the basis of this questionnaire is widely used to measure 19 other online and offline problematic behaviors. Besides Work Addiction, Andreassen et al. 20 (2012a) used this set of items to measure Facebook addiction and also shopping addiction 21 (Andreassen et al., 2015). We chose to use the label of Problematic Series Watching instead of 22 Series Watching Addiction because we have no solid evidence regarding health-related 23 negative consequences of this behavior and this might be a too specific and peripheral behavior 24 to call it as an addiction (Billieux et al., 2015). 6 1 The BWAS contains seven items; however, we wanted to have a larger initial item pool 2 to ensure a better content validity regarding the framework of problematic behavior. Moreover, 3 the original scale contains an additional dimension (health problems), but our aim was to create 4 a measure using the above-mentioned six dimensions. Therefore, those 12 items have been 5 chosen that were the initial items of Andreassen et al. (2012 b) and belonged to one of the six 6 core components. First, all 12 items were translated following the protocol of Beaton, 7 Bombardier, Guillemin and Ferraz (2000). Then, the items were modified to reflect the 8 individual’s series watching habits by replacing the subject “work” in each item with the word 9 “series watching”. Grammatical errors were also corrected. Respondents had to answer using a 10 5-point scale (1 = Never; 2 = Rarely; 3 = Sometimes; 4 = Often; 5 = Always). 11 12 Statistical analysis 13 Preliminary statistical analysis comprised of the descriptive analysis in SPSS 22 such as 14 means, standard deviations, frequencies, and skewness-kurtosis values. Later, estimation of 15 Cronbach’s alpha values, correlations, t-tests and ANOVAs were performed with this software 16 as well. All cases had complete data in both samples1. 17 Participants were asked to estimate the amount of free time they have on an average 18 weekday and weekend. According to the survey of the Hungarian Central Statistical Office 19 (2011), on an average day, a Hungarian individual (aged between 15 and 74) spends 712 20 minutes with satisfying his/her physiological needs (i.e. sleeping, eating, hygiene) and the 21 remaining 728 minutes is for other activities. Based on these results, we set a threshold for the 22 amount of possible maximum free time one individual can have in our sample in order to 23 minimize bias. Reported amount of free time more than 728 minutes were recoded as missing The questions within the utilized questionnaire system were set as “required” in order to minimize the amount of missing data. If participants did not finish the filling out and click on the “submit” button, then their responses were not received, thus they were not part of the sample. 1 7 1 data, however, these cases were not completely deleted. To have a single indicator of free time 2 for an average day of the week, the reported values were weighted by the following formula: 3 (weekday time*5 + weekend time*2)/7. After recoding the high values, 344 cases were retained 4 where a realistic amount of time was indicated, whereas 22 cases were considered as missing 5 for this variable. 6 Prior to the confirmatory factor analyses (CFA), the data was investigated for normality. 7 Regarding univariate normality, the simulation study of Curran, West, and Finch (1996) 8 concluded that problems can arise from having a skewness value above 2.0 and a kurtosis value 9 above 7.0. Also, multivariate normality was examined in Mplus 7.3 (Muthén & Muthén, 1998- 10 2012) using two-sided test of fit for skewness and kurtosis (Wang & Wang, 2012). As these 11 tests were statistically significant in both Sample 1 and Sample 2, the assumption of 12 multivariate normality was violated. 13 In order to investigate the factor structure of this new measurement, a series of CFAs 14 were conducted using Mplus 7.3. Since the data did not have multivariate normal distribution, 15 the robust maximum-likelihood estimator (MLR) was used instead of maximum-likelihood 16 (ML). Multiple goodness of fit indices were taken into consideration based on the 17 recommendations of Brown (2015): the comparative fit index (CFI), the Tucker–Lewis index 18 (TLI), the root-mean-square error of approximation (RMSEA) and its 90% confidence interval, 19 and test of close fit (CFit), the standardized root mean square residuals (SRMR). Guided by 20 suggestions of several methodologists (Bentler, 1990; Brown, 2015; Browne & Cudeck, 1993; 21 Hu & Bentler, 1999; Schermelleh-Engel, Moosbrugger, & Müller, 2003), good or acceptable 22 model fit was defined by the following criteria: CFI (≥ .95 for good, ≥ .90 for acceptable), TLI 23 (≥ .95 for good, ≥ .90 for acceptable), RMSEA (≤ .06 for good, ≤.08 for acceptable), CFit (≥ 24 .10 for good, ≥ .05 for acceptable), and SRMR (≤ .05 for good, ≤.10 for acceptable). 8 1 Moreover, to test whether the amount of free time could have an effect on the PSWS 2 scores, a multiple indicators multiple causes (MIMIC) analysis (Brown, 2015) was carried out 3 as well. The MIMIC model consists of a measurement model (previously established in the 4 CFA) and a structural model which makes it possible to estimate the effect of indicators (spare 5 time) on the latent variable (PSWS) while controlling for other variables. 6 Finally, regarding reliability, internal consistency was measured by Cronbach’s alpha 7 taking Nunnally’s (1978) suggestions into consideration regarding the acceptability of the value 8 (.70 is acceptable, .80 is good). However, as Cronbach’s alpha value could have biases which 9 could inflate or deflate the reliability (see Osburn, 2000 or Sijtsma, 2009), two additional values 10 were calculated: factor determinacy and composite reliability. Factor determinacy refers to the 11 correlation between the estimated and the true factor scores and it describes how well the factor 12 is described by the indicators. It ranges from zero to one with higher scores indicating higher 13 levels of reliability (Muthén & Muthén, 1998-2012). Composite reliability can also be 14 considered when assessing a model. Values above .70 should be considered acceptable 15 (Nunnally & Bernstein, 1994). Additionally, inter-item correlations were also observed with 16 values between .15 and .50 considered acceptable (Clark & Watson, 1995). 17 18 Ethics 19 In case of both Sample 1 and Sample 2, the studies were conducted in accordance with 20 the Declaration of Helsinki and were approved by the Institutional Review Board of the Eötvös 21 Loránd University. All subjects were informed about the studies they participated in and all 22 provided informed consent. 23 24 RESULTS 9 1 Since we had two independent datasets, the analyses were carried out as it follows: 2 Sample 1 was used for the investigation of the factor structure and for the MIMIC analysis, 3 whereas Sample 2 was used for cross-validation and for examining differences across gender, 4 age, educational level, and place of residence. Several solutions were observed: (1) a 12-item 5 unidimensional solution, (2) a 6-item unidimensional solution based on the BWAS, (3) a 6-item 6 unidimensional solution which retained one item per factor based on modification indices as 7 suggested by Brown (2015). In a second step, using the more diverse Sample 2 for cross- 8 validating this factor structure, we carried out the analysis with the 6-item version which was 9 based on modification indices. 10 11 Structural analysis 12 For the initial 12 items of the PSWS, skewness values ranged from -.24 to 1.94, and 13 kurtosis values ranged from -1.11 to 2.60 which were within the acceptability range proposed 14 by Curran et al. (1996). Next, a series of confirmatory factor analyses were conducted on 15 Sample 1 in order to test alternative models. The comparison of the examined models can be 16 seen in Table 1. The results showed that the unidimensional 6-item model showed acceptable 17 model fit in the case of both Sample 1 (CFI = .98, TLI = .97, RMSEA = .04 [90% CI .00-.08], 18 CFit = .64, SRMR = .03) and Sample 2 (CFI = .96, TLI = .93, RMSEA = .07 [90% CI .05-.09], 19 CFit = .06, SRMR =.03). 20 21 Table 1. 22 Confirmatory factor analyses results of the Problematic Series Watching Scale (PSWS) 12-item unidimensional model 10 CFI TLI RMSEA [90% CI] CFit SRMR .65 .58 .15 .00 .09 (all initial items are included) Sample 1 Sample 2 1 [.14-.16] 6-item unidimensional model of the BWAS (items 2, 3, 6, 7, 10, 11) .94 2 .89 .07 [.04-.10] .13 .64 .034 .06 .035 6-item unidimensional model of the PSWS (items 2, 3, 6, 7, 9, 12) .98 .97 .04 [.00-.08] 6-item unidimensional model of the PSWS (items 2, 3, 6, 7, 9, 12) .96 .93 .07 [.05-.09] .04 3 6 7 8 9 10 11 12 13 Notes. CFI = comparative fit index; TLI = Tucker-Lewis index; RMSEA = root mean square error of approximation; CFit = RMSEA’s test of close fit; SRMR = standardized root mean square residuals 14 15 Standardized factor loadings, reliability indices (Cronbach’s alpha, factor determinacy, 16 composite reliability, and inter-item correlations) and descriptive statistics regarding both 17 Sample 1 and 2 can been seen in Table 2. Factor loadings were acceptable (ranging from .43 to 18 .62 in Sample 1 and from .52 to .68 in Sample 2). Although the Cronbach alpha value was 19 borderline in the case of Sample 1 (α = .69), other reliability indices had adequate values and 20 inter-item correlations were also within the acceptable range for both samples. These results 21 indicate that the PSWS has good factor structure and acceptable reliability. 22 23 24 25 Table 2. 26 Series Watching Scale (PSWS) with two independent samples Standardized factor loadings, reliability indices and descriptive statistics of the Problematic 27 Standardized factor loadings Sample 1 (N= 366) Sample 2 (N = 752) salience 1 .54 .54 tolerance 2 .42 .60 mood modification 3 .41 .52 relapse 4 .62 .59 withdrawal 5 .62 .67 11 Reliability indices Descriptive statistics conflict 6 .69 .68 Cronbach’s alpha .69 .76 Factor determinacy .85 .88 Composite reliability .73 .77 Mean inter-item correlations .29 .36 13.85 (4.67) 12.62 (4.41) Skewness (SD) .62 (.13) .85 (.09) Kurtosis (SD) -.24 (.25) .67 (.18) Mean (SD) 1 2 3 MIMIC model analysis 4 In order to investigate the effect of the amount of free time on problematic series 5 watching scores, a MIMIC analysis was applied. The model fit indices showed that the model 6 remained acceptable (CFI = .99, TLI = .99, RMSEA = .02 [90% CI .00-.06], CFit = .91, SRMR 7 =.03). Daily average spare time was not significantly (β = .01, p = .93) associated with the 8 PSWS latent variable. 9 10 Gender, age, educational level and place of residence differences 11 For measuring demographic differences we used the larger sample (Sample 2) of 752 12 cases. Women (Mfemale = 12.91, SDfemale = 4.50) had higher scores [t(750) = -2.90, p <.01] than 13 men (Mmale = 11.86, SDmale = 4.08). Relatively weak negative correlations were found between 14 age and PSWS [r(750) = -.21, p = .001]. Using one-way ANOVA (with Bonferroni-corrected 15 post-hoc test), educational level-related differences were measured between the three groups 16 [F(2, 749) = 6.44, p < .01]. Those who have higher education degree (M = 11.97, SD = 4.19) 17 scored significantly lower on PSWS than those who have elementary school (M = 13.96, SD = 18 4.96, p = .011) or high school degree (M = 12.92, SD = 4.44, p = .013). Using the same method 19 no place of residence-related differences were found. 20 21 DISCUSSION 12 1 Our results show that the PSWS has an appropriate factor structure and reliability. 2 Respondents watch series more than one hour per day which is more than one fifth of their free- 3 time which indicated that series watching might be an important free-time activity. However, 4 the amount of free time one has is not associated with PSWS scores. Women had higher scores 5 on PSWS and respondents with higher education had lower scores on it. 6 We have to consider the concept of overpathologization (Billieux, Schimmenti, 7 Khazaal, Maurage, & Heeren, 2015) which argues that everyday activities are being turned into 8 behavioral addictions. Indeed, it is not obvious that problematic series watching affect a large 9 part of the population. Moreover, clinically validated diagnostic tests with adequate levels of 10 sensitivity and specificity are required for establishing an accurate diagnosis (Maráz, Király, & 11 Demetrovics, 2015). In light of these suggestions, it can be said the problematic series watching 12 belongs to a group of problematic behaviors along with for instance dance (Maráz, Urbán, 13 Griffiths, & Demetrovics, 2015) and smartphone use (Wang, Wang, Gaskin, & Wang, 2015) 14 that are not necessarily as addictive and prevalent as classical substance addictions, still it might 15 be important to consider these problematic behaviors in today’s changing era. 16 Given the lack of empirical research on series watching, we supposed that it might be 17 similar to other problematic screen-related behaviors (e.g. online gaming or Internet and 18 Facebook use). Although health-related variables were not included in this research, intense 19 screen-time can be related to various health problems, such as reduced physical and 20 psychosocial health (Tremblay et al., 2011), increased cardiovascular risk (Grontved et al. 21 2014), sleeping problems (Do, Shin, Bautista, & Foo, 2013), lower levels of life satisfaction 22 (Mentzoni et al., 2011), or interpersonal problems (Lo, Wang, & Fang, 2005). 23 Some limitations of this research need to be addressed. This was a questionnaire-based 24 study which is prone to bias. However, this limitation could be overcome by implementing 25 more objective measures that would respect the individual’s privacy as well. Although the two 13 1 samples were diverse, neither was representative which limits the generalization of the results. 2 Regarding the PSWS, the results were based on a correlational design which does not make it 3 possible to infer causality. More research is needed to examine its temporal stability, 4 convergent-, divergent-, and predictive validity in different cultures. 5 In this new research area where cross-sectional studies are rare, a longitudinal design 6 could be fruitful in examining how series watching is affected by different life events and also 7 how it might impact one’s health. In terms of clinical practice, prevalence and incidence should 8 be investigated. Further research is needed to explore whether problematic series watching and 9 other problematic online behaviors have the same roots. It is possible that they have the same 10 negative consequences. Other possible covariates could be examined in the future such as 11 loneliness or urgency. Also, further investigation is needed whether extensive series watching 12 can lead to health and psychosocial problems. 13 14 CONCLUSION 15 To our best knowledge no prior study examined problematic series watching. In the 16 present study a 6-item Problematic Series Watching Scale was created on the basis of the six 17 component model of Griffiths (2005). The scale has good factor structure and reliability. PSWS 18 scores are positively related with time spent on series watching, whereas the amount of free 19 time does not have an effect on PSWS scores. In the more and more digitalized world there are 20 many forces which encourage people watching online series. In the light of these changes, 21 research on problematic series watching will be increasingly relevant. 22 23 REFERENCES 14 1 Andreassen, C. S., Griffiths, M. D., Hetland, J., & Pallesen, S. (2012b). Development of a work 2 addiction scale. Scandinavian Journal of Psychology, 3(53), 265-272. doi: 3 10.1111/j.1467-9450.2012.00947.x 4 Andreassen, C. S., Torsheim, T., Brunborg, G. S., & Pallesen, S. (2012a). Development of a 5 Facebook addiction scale. Psychological 6 10.2466/02.09.18.PR0.110.2.501-517 reports, 110(2), 501-517. doi: 7 Andreassen, C. S., Griffiths, M. D., Pallesen, S., Bilder, R. M., Torsheim, T., & Aboujaoude, 8 E. (2015). The Bergen Shopping Addiction Scale: reliability and validity of a brief 9 screening test. Frontiers in psychology, 6:1374. doi: 10.3389/fpsyg.2015.01374 10 11 12 13 Beaton, D. E., Bombardier, C., Guillemin, F., & Ferraz, M. B. (2000). Guidelines for the process of cross-cultural adaptation of self-report measures. Spine, 25(24), 3186-3191. Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological bulletin, 107(2), 238-246. doi: 10.1037/0033-2909.107.2.238 14 Billieux, J., Schimmenti, A., Khazaal, Y., Maurage, P., & Heeren, A. (2015). Are we 15 overpathologizing everyday life? A tenable blueprint for behavioral addiction research. 16 Journal of Behavioral Addictions, 4(3), 119-123. doi: 10.1556/2006.4.2015.009 17 18 Brown, T. A. (2015). Confirmatory factor analysis for applied research (second edition). New York, NY: Guilford Press. 19 Browne, M. V., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen, 20 & J. S. Long (Eds.), Testing structural equation models (pp. 136-162). Newbury Park, 21 CA: Sage. 22 Chóliz, M. (2015). The Challenge of Online Gambling: The Effect of Legalization on the 23 Increase in Online Gambling Addiction. Journal of Gambling Studies, 1-8. doi: 24 10.1007/s10899-015-9558-6 15 1 Clark, L. A., & Watson, D. (1995). Construct validity: basic issues in objective scale 2 development. Psychological Assessment, 7(3), 309-319. doi: 10.1037/1040-3590.7.3.309 3 Cooper, A. (1998). Sexuality and the Internet: Surfing into the new millennium. 4 CyberPsychology & Behavior, 1(2), 187-193. doi: 10.1089/cpb.1998.1.187 5 Curran, P. J., West, S. G., & Finch, J. F. (1996). The robustness of test statistics to nonnormality 6 and specification error in confirmatory factor analysis. Psychological Methods, 1(1), 16- 7 29. doi: 10.1037/1082-989X.1.1.16 8 Demetrovics, Z., Urbán, R., Nagygyörgy, K., Farkas, J., Griffiths, M. D., Pápay, O., Kökönyei, 9 Gy., Felvinczi, K., & Oláh, A. (2012). The development of the problematic online gaming 10 questionnaire (POGQ). PLoS One, 7(5), e36417. doi: 10.1371/journal.pone.0036417 11 Do, Y. K., Shin, E., Bautista, M. A., & Foo, K. (2013). The associations between self-reported 12 sleep duration and adolescent health outcomes: What is the role of time spent on Internet 13 use?. Sleep medicine, 14(2), 195-200. doi: 0.1016/j.sleep.2012.09.004 14 15 16 17 18 19 Griffiths, M. D. (1999). Internet addiction: Internet fuels other addictions. Student British Medical Journal, 7, 428-429. Griffiths, M. (2000). Internet addiction-time to be taken seriously?. Addiction Research & Theory, 8(5), 413-418. doi: 10.3109/16066350009005587 Griffiths, M. (2005). A ‘components’ model of addiction within a biopsychosocial framework. Journal of Substance Use, 10(4), 191-197. doi: 10.1080/14659890500114359 20 Griffiths, M., & Barnes, A. (2008). Internet gambling: An online empirical study among student 21 gamblers. International Journal of Mental Health and Addiction, 6(2), 194-204. doi: 22 10.1007/s11469-007-9083-7 23 Grøntved, A., Ried-Larsen, M., Møller, N. C., Kristensen, P. L., Wedderkopp, N., Froberg, K., 24 Hu, F. B., Ekelund, U., & Andersen, L. B. (2014). Youth screen-time behaviour is 25 associated with cardiovascular risk in young adulthood: the European Youth Heart Study. 16 1 European Journal of 2 10.1177/2047487312454760 Preventive Cardiology, 21(1), 49-56. doi: 3 Grubbs, J. B., Sessoms, J., Wheeler, D. M., & Volk, F. (2010). The Cyber-Pornography Use 4 Inventory: The development of a new assessment instrument. Sexual Addiction & 5 Compulsivity, 17(2), 106-126. doi: 10.1080/10720161003776166 6 Grüsser, S. M., Thalemann, R., & Griffiths, M. D. (2006). Excessive computer game playing: 7 evidence for addiction and aggression?. CyberPsychology & Behavior, 10(2), 290-292. 8 doi: 10.1089/cpb.2006.9956 9 Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: 10 Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1-55. 11 doi: 10.1080/10705519909540118 12 Hungarian Central Statistical Office (2011). A 15-74 éves népesség napi időfelhasználása, 2010 13 [Time-use of the population aged between 15 and 74 in 2010]. Statisztikai Tükör, 5(87), 14 1-4. 15 http://www.ksh.hu/docs/hun/xftp/stattukor/idomerleg10.pdf Retrieved on October 22, 2015 from 16 Ko, C. H., Yen, J. Y., Chen, C. C., Chen, S. H., & Yen, C. F. (2005). Proposed diagnostic 17 criteria of Internet addiction for adolescents. The Journal of Nervous and Mental Disease, 18 193(11), 728-733. 19 Kor, A., Zilcha-Mano, S., Fogel, Y. A., Mikulincer, M., Reid, R. C., & Potenza, M. N. (2014). 20 Psychometric development of the Problematic Pornography Use scale. Addictive 21 Behaviors, 39(5), 861-868. doi: 10.1016/j.addbeh.2014.01.027 22 Lo, S. K., Wang, C. C., & Fang, W. (2005). Physical interpersonal relationships and social 23 anxiety among online game players. CyberPsychology & Behavior, 8(1), 15-20. doi: 24 10.1089/cpb.2005.8.15 17 1 Maráz, A., Király, O., & Demetrovics, Z. (2015). Commentary on: Are we overpathologizing 2 everyday life? A tenable blueprint for behavioral addiction research: The diagnostic 3 pitfalls of surveys: If you score positive on a test of addiction, you still have a good chance 4 not to be addicted. Journal of Behavioral Addictions, 4(3), 151-154. doi: 5 10.1556/2006.4.2015.026 6 7 Maráz, A., Urbán, R., Griffiths, M. D., & Demetrovics, Z. (2015). An Empirical Investigation of Dance Addiction. PLoS ONE, 10(5):e0125988. doi: 10.1371/journal.pone.0125988 8 Mentzoni, R. A., Brunborg, G. S., Molde, H., Myrseth, H., Skouverøe, K. J. M., Hetland, J., & 9 Pallesen, S. (2011). Problematic video game use: estimated prevalence and associations 10 with mental and physical health. Cyberpsychology, behavior, and social networking, 11 14(10), 591-596. doi: 10.1089/cyber.2010.0260 12 13 Muthén, L. K., & Muthén, B. O. (1998-2012). Mplus User’s Guide. Seventh Edition. Los Angeles, CA: Muthén & Muthén. 14 Netflix Media Center (2013). Netflix Declares Binge Watching is the New Normal. 15 <https://pr.netflix.com/WebClient/getNewsSummary.do?newsId=496>. Retrieved on 16 June 9 2015. 17 Nunnally, J. C. (1978). Psychometric theory (2nd ed.). New York: McGraw-Hill. 18 Nunnally, J. C. & Bernstein, I.H. (1994). Psychometric theory. (3rd ed). New York: McGraw- 19 Hill. 20 Osburn, H. G. (2000). Coefficient alpha and related internal consistency reliability coefficients. 21 Psychological Methods, 5(3), 343-355. doi: 10.1037/1082-989X.5.3.343Pontes, H. M., 22 Szabó, A., & Griffiths, M. D. (2015). The impact of Internet-based specific activities on 23 the perceptions of Internet addiction, quality of life, and excessive usage: A cross- 24 sectional study. Addictive Behaviors Reports, 1, 19-25. doi: 10.1016/j.abrep.2015.03.002 18 1 Ryan, T., Chester, A., Reece, J., & Xenos, S. (2014). The uses and abuses of Facebook: A 2 review of Facebook addiction. Journal of Behavioral Addictions, 3(3), 133-148. doi: 3 10.1556/JBA.3.2014.016 4 Schermelleh-Engel, K., Moosbrugger, H., & Müller, H. (2003). Evaluating the fit of structural 5 equation models: Tests of significance and descriptive goodness-of-fit measures. Methods 6 of psychological research online, 8(2), 23-74. 7 8 9 10 Sijtsma, K. (2009). On the Use, the Misuse, and the Very Limited Usefulness of Cronbach's Alpha. Psychometrika, 74(1), 107-120. doi: 10.1007/s11336-008-9101-0 Sussman, S., & Moran, M. B. (2013). Hidden addiction: television. Journal of Behavioral Addictions, 2(3), 125-132. doi: 10.1556/JBA.2.2013.008 11 Tremblay, M. S., LeBlanc, A. G., Kho, M. E., Saunders, T. J., Larouche, R., Colley, R. C., 12 Goldfield, G., & Gorber, S. C. (2011). Systematic review of sedentary behaviour and 13 health indicators in school-aged children and youth. International Journal of Behavioral 14 Nutrition and Physical Activity, 8(1), 98. doi: 10.1186/1479-5868-8-98 15 16 Wang, J., & Wang, X. (2012). Structural equation modeling: Applications using Mplus. West Sussex, UK: John Wiley & Sons. 17 Wang, J. L., Wang, H. Z., Gaskin, J., & Wang, L. H. (2015). The role of stress and motivation 18 in problematic smartphone use among college students. Computers in Human Behavior, 19 53, 181-188. doi: 10.1016/j.chb.2015.07.005 20 Yellowlees, P. M., & Marks, S. (2007). Problematic Internet use or Internet addiction?. 21 Computers in Human Behavior, 23(3), 1447-1453. doi: 10.1016/j.chb.2005.05.004 22 Young, K. S. (1998). Internet addiction: The emergence of a new clinical disorder. 23 CyberPsychology & Behavior, 1(3), 237-244. doi: 10.1089/cpb.1998.1.237 19