Research Article

European Addiction Eur Addict Res Received: September 5, 2019 Research DOI: 10.1159/000506738 Accepted: February 21, 2020 Published online: March 13, 2020

Problematic Use, and Chronotype Correlations in University Students

a a b a Pavle Randjelovic Nenad Stojiljkovic Niko Radulovic Nikola Stojanovic c Ivan Ilic a b Department of Physiology, Faculty of Medicine, University of Nis, Niš, Serbia; Department of Chemistry, Faculty of c Science and Mathematics, University of Nis, Niš, Serbia; Department of Pathology, Faculty of Medicine, University of Nis, Niš, Serbia

Keywords type. Those students spent significantly more time on their Students · Chronotype · Screen time · Problematic phones compared to the non-addicted ones. There was no smartphone use statistically significant difference between the number of male and female students with problematic smartphone use. The best predictors of problematic smartphone use were Abstract longer daily smartphone screen time and evening chrono- Background: Besides numerous advantages and commodi- type personality. Conclusions: The results of our study ties offered by , there are obvious unhealthy ef- showed that a significant number of students of medicine fects. The global trend of an increase in the frequency of us- showed problematic smartphone use. There was a strong age of smartphones, that is, prolonged screen time, is close- correlation between extensive screen time and the level of ly related to problematic smartphone use. The aim of our problematic smartphone use in the studied population. study was to measure the level of problematic smartphone © 2020 S. Karger AG, Basel use in a student population through the assessment of the smartphone screen time and the determination of the stu- dent chronotype, as well as through the correlation between Introduction these variables. Methods: The participants were students of medicine of both sexes. Problematic smartphone use was In 1997, the term smartphone was introduced by the measured by the short version of the Smartphone Addiction manufacturer Ericsson to point out the advanced features Scale. Smartphone screen time was assessed by the free An- of a new generation of mobile phones [1]. Ever since, the droid application Quality Time. Chronotype was established smartphones are becoming irreplaceable in an everyday by the Morningness-Eveningness Questionnaire. Results: Al- life and are offering a wide spectrum of applications for most one quarter (22.7%) of students involved in our study news, entertainment, communications, and education. could be classified as being “smartphone-addicted”. The stu- Smartphones have touchscreens, Internet connection, dents with problematic smartphone use more frequently possibility to add new applications, and other interesting (statistical significance) belonged to the evening chrono- functions such as a digital camera, GPS navigation, and a

© 2020 S. Karger AG, Basel Pavle Randjelovic Department of Physiology Faculty of Medicine, University of Niš E-Mail [email protected] Bulevar Dr Zorana Djindjica 81, 18000 Niš (Serbia) www.karger.com/ear E-Mail pavleus @ gmail.com Downloaded by: National Univ. of Singapore 137.132.123.69 - 3/15/2020 4:41:49 PM music player. Besides numerous advantages and commod- Chronotype refers to the preference for -wake ities offered by smartphones, there are obvious unhealthy timing: for example, morning types go to bed, get up, and effects related to their usage. Some examples of such un- experience peak alertness and performance earlier in the wanted effects are neck pain syndrome [2] or traffic acci- day than do evening types [12]. Individual circadian type dents [3]. Recent studies showed that increased screen time can be clustered into 3 categories as “morning type”, “in- is connected to the sleep disturbances and depression [4]. termediate type”, and “evening type” with the normal dis- Also, the global trend of the rise in the frequency of usage tribution around the middle [13]. of smartphones, termed as “screen time”, is closely related to problematic smartphone use [5]. Lin et al. [6] showed Aim of the Study that problematic smartphone use has similar aspects to The aim of our study was to measure the level of prob- substance addiction. The term “smartphone addiction” is lematic smartphone use in a student population and the not an official disorder and so it is recommended to use the screen time of smartphones and to determine the chro- term problematic smartphone use. The usage of the word notype of the students. Also, the significance of the cor- addiction in the context of mobile phone overuse is also relation between studied variables was assessed. problematic due to stigmatization of users [7]. When we talk about mobile phones, the problematic smartphone use could start when usual behavior, such as Material and Methods owning a mobile phone for communication or security reasons, starts to produce negative consequences and the The experimental proposal followed the principles suggested in user becomes increasingly addicted [8]. Owning a mobile the 2013 Declaration of Helsinki and was approved by the local phone for primary purposes becomes secondary and the Ethical Committee. The study was carried out at the Faculty of Med- owner starts to use his phone for entertainment reasons icine, University of Niš (Serbia) during March 2018. The partici- pants were second-year students of medicine, 20–22 years of age, of solely. Over time, a smartphone user could develop pro- both sexes. The study was based on simple random sampling. The gressively dangerous behavior, such as texting while driv- involvement in the study was strictly on a voluntary basis and anon- ing a car, and in the end, the person reaches a point where ymous. Before the commencement of the study, all participants he cannot control smartphone usage. The addiction pro- were informed about the study design. All enrolled students signed cess differentiates the need and desire. In other words, a an informed consent document. The students could drop out of the study at any moment without any consequence to them. The inclu- smartphone user starts with the desire to use his phone sion criteria were having and using an Android-based mobile and ends up feeling the need to use it. This overlapping phone. The exclusion criteria were usage of sedative medications or point is showing the transition from safe behavior, being psychoactive substances, the existence of recent stressful situations, pleasant with minor unhealthy effects, to addictive be- and other sleep-interfering reasons. Primary outcomes of the study havior, where the urge (physical and/or mental) replaced were the problematic smartphone use level, smartphone screen time, screen unlock number, and chronotype score. the desire as a motivation factor. It is believed that this Problematic smartphone use was measured by the short ver- type of problematic smartphone use uses the same neural sion of the Smartphone Addiction Scale (SAS-SV) [14]. This as- pathways involved in substance use disorder [9]. sessment tool was created in South Korea and published in the The constant use of mobile phone gives rise to the con- English language in 2013. It has 10 questions with answers on a cept of (No-Mobile-Phobia), which is the scale from 1 (completely agree) to 6 (completely disagree). The to- tal score could range from 10 to 60, with higher values indicating fear of losing contact with a mobile phone [10]. Griffiths higher addiction from a smartphone. The questionnaire showed [10] suggests that everything that produces a feeling of ex- good validity and internal consistency (Cronbach’s alpha 0.91). citement can cause an addiction. Having that in mind, the The authors of SAS-SV proposed cutoff values for the prediction cause of problematic smartphone use could be the fact that of problematic smartphone use as 31 for males and 33 for females. the usage of smartphones brings a feeling of excitement. Smartphone screen time and screen unlock number were as- sessed by the free Android application Quality Time [15]. The data So if a certain type of behavior results in a sense of excite- were collected for 14 consecutive days, and the average for 14 days ment or helps to reduce negative or unpleasant feelings, was calculated for each participant and those values were used for then such behavior is intensified and the person continues further analysis. with it for pleasure or to reduce the unpleasant situation. Chronotype was assessed by the Morningness-Eveningness Problematic smartphone use gives a sense of pleasure to Questionnaire (MEQ) [16], a self-rating questionnaire that con- sists of 19 items with a total score ranging from 16 to 86. Higher consumers as a result of usage and relieves them from anx- scores indicate greater morningness, while lower scores indicate iety or stress [11]. Therefore, it could be assumed that this more evening type. very scenario causes addiction-like processes. The average grade was self-reported by study participants.

2 Eur Addict Res Randjelovic/Stojiljkovic/Radulovic/ DOI: 10.1159/000506738 Stojanovic/Ilic Downloaded by: National Univ. of Singapore 137.132.123.69 - 3/15/2020 4:41:49 PM Table 1. Comparison of BMI, average grade, MEQ score, smartphone screen time, and screen unlocks between smartphone-addicted and non-addicted students

Addicted Non-addicted p value d

BMI 21.88±3.22 21.94±3.13 0.95 0.019 Average grade 8.65±0.88 8.64±0.62 0.97 0.013 MEQ score 44.05±6.15 48.41±10.29 0.042 0.514 Smartphone screen time, min 301.5±103.4 208.7±133.6 0.011 0.777 Screen unlocks, n 177.2±67.25 115.8±79.74 0.006 0.832

Values are represented as mean ± SD. BMI, body mass index; MEQ, Morningness-Eveningness Questionnaire.

The body mass index (BMI) was calculated from body classifying students to those with and without problem- weight and height measured by standard equipment by one of the atic smartphone use, 31 points for males and 33 for fe- authors. males. According to SAS-SV test, 22.7% of students in our Sample Size Calculation study population could be classified as subjects with The sample size was calculated before the beginning of the problematic smartphone use. study by the free software GPower 3.1 [17]. For predetermined We compared some parameters of the groups of stu- values for a two-tailed test, medium effect size (correlation coef- dents with and without problematic smartphone use ficient of 0.3), the statistical significance of 0.05, and the power of that could potentially be used as predictors (Table 1). study of 0.8, the sample size needed was 82. To compensate for potential dropouts, we increased sample size by 10%, giving the The results presented in Table 1 show differences in av- total number of 90 participants. Only the results for those partici- erage BMI, grade, chronotype score, smartphone screen pants who completed the study were analyzed. The number of par- time, and screen unlocks between 2 groups of subjects. ticipants who finished the study was 77. The achieved power of the There was no statistically significant difference in BMI study at the end was 0.78. and average grade between students with and without Statistical Analysis problematic smartphone use. The effect size was very Statistical analysis was done using the software GraphPad small 0.019 and 0.013, respectively. There was a statisti- Prism version 6.00 for Windows, GraphPad Software, La Jolla, CA, cally significant and medium (d = 0.51) difference in USA, www.graphpad.com. Normal distribution of continuous chronotype between the 2 groups. The students classi- data was tested with the D’Agostino-Pearson omnibus K2 test. For fied as with problematic smartphone use had signifi- normally distributed data, the results were presented as the mean and SD. For nonnormally distributed data, the median was used. cantly lower values for MEQ score, indicating a higher The correlation between outcomes was done by Pearson’s correla- level of evening chronotype. There was statistically sig- tion coefficient for normally distributed data and by Spearman’s nificant and large (d = 0.78) difference in smartphone correlation coefficient for other data. The significance of differ- screen time between the 2 groups. Students with prob- ences between means of groups of participants with and without lematic smartphone use spend significantly more time problematic smartphone use was determined by unpaired t test for normally distributed data. Welch‘s correction was used for un- on their phones compared to the ones without problem- equal variances. The multiple regression was performed by SPSS atic smartphone use. The students with problematic 21 software (IBM Corp.). smartphone use unlocked their phones significantly The level of statistical significance was set at 0.05. For interpre- more often compared to the other group. The observed tation of effect size, the recommendations by Cohen were used (0.1 difference was of high practical significance with a large small, 0.3 medium, and 0.5 large) [18]. effect size (d = 0.83). There was no statistically significant difference be- tween males and females in problematic smartphone Results use measured by SAS-SV in our study (Table 2). Even though differences were not significant, females showed The average SAS-SV score in our study was 28.16 ± a higher average problematic smartphone use score 1.69, ranging from 10 to 53 with a 95% CI from 25.12 to compared to males with a moderate effect size (d = 31.12. We used the recommended cutoff values [13] for 0.39).

Smartphone Addiction, Screen Time and Eur Addict Res 3 Chronotype DOI: 10.1159/000506738 Downloaded by: National Univ. of Singapore 137.132.123.69 - 3/15/2020 4:41:49 PM The correlation between problematic smartphone Table 2. Smartphone addiction score in male and female students use as measured by SAS-SV score and other variables (BMI, average grade, chronotype, smartphone screen Male Female p value d time, and screen unlocks) is presented in Table 3. The SAS-SV score 23.7±10.9 28.09±11.47 0.156 0.392 correlation between problematic smartphone use and BMI was not statistically significant and was very weak Values are reresented as mean ± SD. (–0.05). Likewise, the correlation between SAS-SV score SAS-SV, short version of the Smartphone Addiction Scale. and the average grade was very weak (0.04) and was not statistically significant. There was a statistically signifi- cant (p < 0.05) and moderately negative (–0.35) correla- Table 3. Correlation between SAS-SV score and BMI, average tion between SAS-SV and MEQ score. Lower MEQ grade, MEQ score, smartphone screen time, and screen unlocks scores (indicating an evening chronotype personality) were associated with higher SAS-SV scores. The correla- Smartphone p value tion between problematic smartphone use and smart- addiction score, r phone screen time was statistically significant (p < 0.05) and strongly positive (0.61). An increase in the time BMI –0.05 0.68 Average grade 0.04 0.76 spent on the phone was associated with problematic MEQ score –0.35 0.044 smartphone use. Even though there was a moderately Smartphone screen time 0.61 0.024 positive correlation (0.38) of screen unlocks with prob- Screen unlocks 0.38 0.08 lematic smartphone use, it was not statistically signifi- cant (p = 0.08). SAS-SV, short version of the Smartphone Addiction Scale; BMI, body mass index; MEQ, Morningness-Eveningness Ques- According to the unpaired t tests for differences be- tionnaire. tween the subjects with and without problematic smart- phone use, we tested the significantly different variables as potential predictors, in a model of multiple regression. Table 4. Multiple regrssion of the predictive factors for smart- The dependent variable was the SAS-SV score, whereas phone addiction score the independent variables were smartphone screen time and chronotype MEQ score. The multicollinearity of the Standardized β t p value independent variables was not violated because the vari- ance inflation factors (VIF) were 1.001 and the Tolerance Smartphone screen time, min 0.632 4.342 <0.001 was 0.999. The regression model statistically significant- Chronotype score –0.32 –2.2 0.038 ly predicted 51.3% of the variance in problematic smart- phone use (R2 = 0.513, p < 0.001). The significant pre- Table 5. Multiple regression beween individual SAS-SV items and dictors of problematic smartphone use were the daily screen time smartphone screen time (p < 0.001) and chronotype score (p = 0.038). Smartphone screen time was a more p value Tolerance VIF important predictive factor in this model with the high- est beta value of 0.632. The results of this analysis are Item 1 0.129 0.276 3.62 presented in Table 4. Item 2 0.05 0.577 1.735 Item 3 0.089 0.397 2.516 Since at least 5 of 10 items of the SAS-SV are related Item 4 0.372 0.498 2.007 to screen time, we analyzed how are these separate items Item 5 0.638 0.379 2.64 related to smartphone usage. We performed multiple Item 6 0.493 0.323 3.093 ­regressions with screen time as dependent variable and Item 7 0.721 0.212 4.709 10 items form SAS-SV as independent variables. The re- Item 8 0.419 0.198 5.040 Item 9 0.687 0.274 3.653 sults are presented in Table 5. There was no multicol- Item 10 0.317 0.311 3.219 linearity between SAS-SV items because VIF were below 10 for all items and Tolerance was above 0.2 as well. SAS-SV, short version of the Smartphone Addiction Scale; VIF, There was no significant correlation between items from variance inflation factor. SAS-SV. All independent variables (SAS-SV items) were without statistical significance (p > 0.05). Even though

4 Eur Addict Res Randjelovic/Stojiljkovic/Radulovic/ DOI: 10.1159/000506738 Stojanovic/Ilic Downloaded by: National Univ. of Singapore 137.132.123.69 - 3/15/2020 4:41:49 PM the whole model had statistical significance in multiple smartphones could have an effect on interpersonal rela- regression (p < 0.05), each individual item did not have tionships between students themselves, between students such significance. and professors and parents, as well as between work col- leagues. This phone-related problematic behavior could be an indicator of other problems in different spheres of Discussion personal functioning. Even though some studies showed a statistically significant small negative correlation be- Up to the present day, a couple of studies dealt with tween problematic smartphone use and academic perfor- the subject of problematic smartphone use worldwide; mance [24], there was no statistically significant associa- however, none of them was done in Serbia. This emerg- tion in our sample. Higher screen time was not related ing topic was researched in Taiwan [6], Korea [19], India with grades in our study. This result raise question to the [20], and the USA [21]. Most of these studies were about practical relevance of the 22% of problematic smartphone the development and validation of the instrument for users in our study. measuring problematic smartphone use. In our study, we There were reports on the phone screen time of arrived at a number of 22.7% of participants that could around 9 h per day [25]. We found that the average be classified as problematic smartphone users, based on smartphone screen time in our population was around cutoff values for the used instrument (SAS-SV). These 4 h daily, which is considerably less compared to the re- numbers are in concordance with the results from other sults from developed countries; however, it still a sig- countries. For example, the percentage of students with nificant one. There is a trend in the world of a constant problematic smartphone use in Spain was 12.8%, while increase in smartphone screen time. We could assume in Belgium it was 21.5% [22]. The highest rate of students that a similar trend is present in our country as well. The with problematic smartphone use, using the same instru- study from South Korea conducted in 2011 showed that ment, was found in China, 29.8% [23]. One could ques- the problematic smartphone use was present in 8.4% of tion the validity of used cutoff values for classifying stu- the population and was higher than Internet addiction dents into being “smartphone addicted”, since the per- (7.7%) in the same population [26]. The same study centage is relatively high (22.7%). The values are adopted showed that there was a higher level of in those from the authors of the SAS-SV instrument based on a persons with problematic smartphone use compared to Korean sample. Even though several studies used them those without. so far for same purpose and got similar values, these cut- The studies dealing with the relation between smart- off values should be interpreted with caution. Even the phone usage and addiction showed contradictory results. term used by the authors “smartphone addiction” is Some studies found that the frequency of smartphone use problematic and it should be changed to “problematic is strongly related to problematic smartphone use [19], smartphone usage”. while others showed that screen time is a better problem- However, there is very little information about the pre- atic smartphone use predictor [27]. Such opposite results dictors for problematic smartphone use. A Korean study could be a consequence of subjectivity in data collecting by [6] showed that the frequency of phone usage is more questionnaires and a tendency to report lower numbers strongly correlated with problematic smartphone use than real ones. In our study, we used an Android applica- than screen time. The same authors showed that screen tion for data collecting about smartphone usage and in this time measured by a questionnaire and by a phone appli- way eliminated possible self-report bias. Our results clear- cation was not the same. The subjects reported less time ly showed that total screen time was a better and stronger than they actually spend on their phones, where the great- indicator of problematic smartphone use compared to the est reporting error was with those having the highest simple frequency of usage. We should mention that prob- screen time. lematic smartphone use has been associated with specific Problematic smartphone use could compromise aca- behavioral addictions like gambling. A recent study indi- demic performance in a student population. These young cated that gambling apps on smartphone can cause perse- people often use their phones during classes or for cheat- verative gambling behavior [28]. However, in our study ing on exams which interfere with the study process. The we did not collect data on the type of activity on smart- negative impact of smartphone usage on academic phones, which is a limitation for inferring results. achievement goes beyond studying and could influence At least 5 of 10 items on SAS-SV are related to duration work performance in the job setting. The overuse of of smartphone use. So the results of SAS-SV could be

Smartphone Addiction, Screen Time and Eur Addict Res 5 Chronotype DOI: 10.1159/000506738 Downloaded by: National Univ. of Singapore 137.132.123.69 - 3/15/2020 4:41:49 PM largely influenced by screen time. Therefore, it was not a different instrument (Mobile Phone Dependence Ques- clear if problematic smartphone usage was a result of tionnaire) in that study. One can speculate about the rela- merely spending too much time using the device or some tion between screen time and chronotype. Since chrono- additional problems were present. To investigate this type has a genetic load [36], it is more likely that it will question, we performed multiple regression analysis be- influence screen time and not the other way around. Our tween individual items from SAS-SV and screen time. The study was observational one giving only correlation data. results showed that although there was strong and signifi- So no causation could be drawn from it. cant positive correlation between SAS-SV score and screen In our model of multiple regression, the best predic- time, individual items did not significantly predicted the tors for problematic smartphone use were shown to be model (Table 5). Because one of the pre-assumptions of the screen time and the evening chronotype score. Since multiple regression is no multicollinearity between inde- these parameters are easily measured, one could be con- pendent variables, we showed this by calculation Collin- scious about it to prevent and reduce problematic earity statistics: Tolerance and VIF. Since all the items smartphone use. This is especially important for the have Tolerance values > 0.2 and VIF < 10, we can state that evening chronotype since, more or less, it cannot be there was no collinearity between individual items from changed, leaving the person the only option to reduce SAS-SV. In other words, each item added value to overall screen time. regression model but without individual correlation. We can conclude that problematic overuse of smartphones in Strengths and Limitations of the Study our sample was not only due to excessive screen time. The main strength would be the objectivity coming In our study, there was no significant difference in the from measuring smartphone screen time by a free and frequency of problematic smartphone use occurrence be- readily available Android application. The strength was tween males and females. Up to now, published results also a predetermined sample size needed to achieve the related to gender differences in problematic smartphone desired power of the study. use are not clear. There were studies where males were Our study has several limitations. The main one is the significantly more affected than females [29]. There were very design of the study, being of a cross-sectional type. other studies showing the opposite results, with females The next limitation is a correlation analysis that does not being more affected [30]. And, finally, there were studies show causation, but only the strength of association. The where results were in agreement with our data, with no following limitation is the usage of the questionnaire for statistical differences between the sexes [31]. However, measuring problematic smartphone use due to the nature the moderate effect size of the difference in favor of fe- and limitations of the instrument. Also, our results are not males in our study should be noted. The lack of statistical easily generalized because the population only included significance could be interpreted as coming from the medical students. There is no information about the type small sample size. of activities engaged on smartphone. We do not have in- Some studies indicate that the evening chronotype is formation is gaming or gambling was present. Thus, it is tied to Internet addiction or problematic Internet use [32] not possible to infer how problematic in clinically relevant and to computer game addiction [33]. Other findings in- way the current results are. At the end, the relatively small dicate that the morning types are associated with the us- sample size could be treated as a limitation, although it age of traditional media, while the evening types showed was precalculated and aimed to find a medium effect size. a significantly stronger preference for and the use of new media [34]. Our results are in accordance with this trend. We found a moderate and statistically significant correla- Conclusion tion between problematic smartphone use and the eve- ning chronotype. Also, problematic smartphone users The results of our study showed that among medical had significantly lower MEQ scores, which mean a great- students there was a significant number of those with er tendency toward the evening chronotype. Similar re- problematic smartphone use. There was a strong cor- sults were found in the study of Japanese students, where relation between screen time and the level of problem- also weak (r = –0.16) statistically significant correlation atic smartphone use in our population. The best predic- was found between problematic smartphone use and the tors of problematic smartphone use in our study were evening chronotype [35]. However, it should be men- screen time and evening chronotype. Having in mind tioned that problematic smartphone use was assessed by that there is a trend of increase in smartphone screen

6 Eur Addict Res Randjelovic/Stojiljkovic/Radulovic/ DOI: 10.1159/000506738 Stojanovic/Ilic Downloaded by: National Univ. of Singapore 137.132.123.69 - 3/15/2020 4:41:49 PM time and unhealthy consequences coming from that, human participants were in accordance with the ethical standards more attention should be directed to this growing prob- of the institutional and/or national research committee and with the 2013 Helsinki declaration. lem among students.

Acknowledgement Disclosure Statement

The authors want to thank all study participants for taking part The authors have no conflicts of interest to declare. in this study on volunteer basis.

Statement of Ethics Funding Sources

Subjects have given their written informed consent. The study This work was supported by the Ministry of Education, Science protocol has been approved by the research institute’s committee and Technological Development of the Republic of Serbia (Grants on human research. All procedures performed in studies involving 43012 and 172061).

References 1 Smartphone S. 2010. History. [Accessed in adolescence: Neurobiological, psychosocial 20 Davey S, Davey A. Assessment of Smartphone March 11, 2019]. Available from: http://www. and clinical issues. Neurosci Biobehav Rev. Addiction in Indian Adolescents: A Mixed

stockholmsmartphone.org/history/. 2017 May; 76 Pt A: 174–84. Method Study by Systematic-review and Me- 2 Lee J, Cho B, Kim Y, Noh J. Smartphone Ad- 10 Griffiths M. Internet gambling: issues, con- ta-analysis Approach. Int J Prev Med. 2014

diction in University Students and Its Impli- cerns, and recommendations. Cyberpsychol Dec; 5(12): 1500–11.

cation for Learning, in: Chen G, Kumar V, Behav. 2003 Dec; 6(6): 557–68. 21 Roberts JA, Yaya LH, Manolis C. The invisible Kinshuk R, Huang S. (editors), Emerging Is- 11 Grant JE, Potenza MN, Weinstein A, Gorelick addiction: cell-phone activities and addiction sues in Smart Learning. Lecture Notes in Edu- DA. Introduction to behavioral addictions. among male and female college students. J Be-

cational Technology. Berlin: Springer; 2015. Am J Drug Alcohol Abuse. 2010 Sep; 36(5): hav Addict. 2014 Dec; 3(4): 254–65. 3 Klauer SG, Guo F, Simons-Morton BG, Oui- 233–41. 22 Lopez-Fernandez O. Short version of the met MC, Lee SE, Dingus TA. Distracted driv- 12 Adan A, Archer SN, Hidalgo MP, Di Milia L, Smartphone Addiction Scale adapted to ing and risk of road crashes among novice and Natale V, Randler C. Circadian typology: a Spanish and French: towards a cross-cultural

experienced drivers. N Engl J Med. 2014 Jan; comprehensive review. Chronobiol Int. 2012 research in problematic mobile phone use.

370(1): 54–9. Nov; 29(9): 1153–75. Addict Behav. 2017 Jan; 64: 275–80. 4 Lemola S, Perkinson-Gloor N, Brand S, De- 13 Natale V, Cicogna P. Morningness–evening- 23 Chen B, Liu F, Ding S, Ying X, Wang L, Wen wald-Kaufmann JF, Grob A. Adolescents’ ness dimension: is it really a continuum? Pers Y. Gender differences in factors associated

electronic media use at night, sleep distur- Individ Dif. 2002; 32(5): 809–16. with smartphone addiction: a cross-sectional bance, and depressive symptoms in the smart- 14 Kwon M, Kim DJ, Cho H, Yang S. The smart- study among medical college students. BMC

phone age. J Youth Adolesc. 2015 Feb; 44(2): phone addiction scale: development and vali- . 2017 Oct; 17(1): 341. 405–18. dation of a short version for adolescents. PLoS 24 Samaha M, Hawi N. Relationships among

5 Lee U, Lee J, Ko M, Lee C, Kim Y, Yang S, et One. 2013 Dec; 8(12):e83558. smartphone addiction, stress, academic per- al. Hooked on smartphones: An exploratory 15 Randjelović P, Stojiljković N, Radulović N, formance, and satisfaction with life. Comput

study on smartphone overuse among college Ilić I, Stojanović N, Ilić S. The association of Human Behav. 2016; 57: 321–5. students. In CHI 2014: One of a CHInd - Con- smartphone usage with subjective sleep 25 Report L. U.S., 2015. Teens Use an Average of ference Proceedings, 32nd Annual ACM quality and daytime sleepiness among medi- Nine Hours of Media Per Day, Tweens Use

Conference on Human Factors in Computing cal students. Biol Rhythm Res. 2019; 50: 857– Six Hours | Common Sense Media. [Accessed Systems. Association for Computing Machin- 65. March 21, 2019]. Available from: http://www. ery. 2014. pp. 2327–36. 16 Horne JA, Östberg O. A self-assessment ques- commonsensemedia.org/about-us/news/ 6 Lin YH, Lin YC, Lee YH, Lin PH, Lin SH, tionnaire to determine morningness-eve- press-releases/landmark-report-us-teens- Chang LR, et al. Time distortion associated ningness in human circadian rhythms. Int J use-an-average-of-nine-hours-of-media-

with smartphone addiction: identifying Chronobiol. 1976; 4(2): 97–110. per-day. smartphone addiction via a mobile applica- 17 Faul F, Erdfelder E, Lang AG, Buchner A. 26 South Korea National Information Society

tion (App). J Psychiatr Res. 2015 Jun; 65: 139– G*Power 3: a flexible statistical power analysis Agency. Internet Addiction Survey. 2011;

45. program for the social, behavioral, and bio- 2011: 118–9. 7 Panova T, Carbonell X. Is smartphone addic- medical sciences. Behav Res Methods. 2007 27 Haug S, Castro RP, Kwon M, Filler A,

tion really an addiction? J Behav Addict. 2018 May; 39(2): 175–91. Kowatsch T, Schaub MP. Smartphone use

Jun; 7(2): 252–9. 18 Cohen J. Statistical Power Analysis for the Be- and smartphone addiction among young peo-

8 Demirci K, Akgönül M, Akpinar A. Relation- havioral Sciences. 2nd ed. Hillsdale: Lawrence ple in Switzerland. J Behav Addict. 2015 Dec;

ship of smartphone use severity with sleep Erlbaum, NJ; 1988. 4(4): 299–307. quality, depression, and anxiety in university 19 Cho S, Lee E. Development of a brief instru- 28 James RJ, O’Malley C, Tunney RJ. Gambling

students. J Behav Addict. 2015 Jun; 4(2): 85–92. ment to measure smartphone addiction on Smartphones: A Study of a Potentially Ad- 9 Cerniglia L, Zoratto F, Cimino S, Laviola G, among nursing students. Comput Inform dictive Behaviour in a Naturalistic Setting.

Ammaniti M, Adriani W. Internet Addiction Nurs. 2015 May; 33(5): 216–24. Eur Addict Res. 2019; 25(1): 30–40.

Smartphone Addiction, Screen Time and Eur Addict Res 7 Chronotype DOI: 10.1159/000506738 Downloaded by: National Univ. of Singapore 137.132.123.69 - 3/15/2020 4:41:49 PM 29 Öztunç M. Analysis of problematic mobile 32 Lin YH, Gau SS. Association between morn- 34 Nimrod G. Early birds and night owls: differ- phone use, feelings of shyness and loneliness ingness-eveningness and the severity of com- ences in media preferences, usages, and envi-

in accordance with several variables. Procedia pulsive Internet use: the moderating role of ronments. Int J Commun. 2015; 9: 133–53.

Soc Behav Sci. 2013; 106: 456–66. gender and parenting style. Sleep Med. 2013 35 Toda M, Nishio N, Ezoe S, Takeshita T. Chro-

30 Augner C, Hacker GW. Associations between Dec; 14(12): 1398–404. notype and Smartphone Use among Japanese problematic mobile phone use and psycho- 33 Vollmer C, Randler C, Horzum MB, Ayas T. Medical Students. Int J Cyber Behav Psychol

logical parameters in young adults. Int J Pub- Computer game addiction in adolescents and Learn. 2015; 5(2): 75–80.

lic Health. 2012 Apr; 57(2): 437–41. its relationship to chronotype and personali- 36 Jones SE, Lane JM, Wood AR, van Hees VT,

31 Demirci K, Orhan H, Demirdas A, Akpınar A, ty. SAGE Open. 2014; 4(1): 1–9. Tyrrell J, Beaumont RN, et al. Genome-wide Sert H. Validity and reliability of the Turkish association analyses of chronotype in Version of the Smartphone Addiction Scale in 697,828 individuals provides insights into

a younger population. Klinik Psikofarmakol circadian rhythms. Nat Commun. 2019 Jan;

Bülteni. 2014; 24(3): 226–34. 10(1): 343.

8 Eur Addict Res Randjelovic/Stojiljkovic/Radulovic/ DOI: 10.1159/000506738 Stojanovic/Ilic Downloaded by: National Univ. of Singapore 137.132.123.69 - 3/15/2020 4:41:49 PM