A Systematic Study on the Reliability and Validity of the Japanese Smartphone Addiction Scale among the Youth in Japan

Background: Few studies in Japan have investigated smartphone addiction utilizing an international scale. This study aimed to evaluate the reliability and validity of the Japanese version of the Smartphone Addiction Scale (SAS) in Japanese youths and to examine the factors underlying smartphone overuse. Methods: We conducted a questionnaire survey in 2018 with participants from one high school (n=1,050) and one vocational school (n=83; age range, 15-24 years; median age, 17 years) in Japan. Data from 1,037 (male: 60.8%, n=631) questionnaires were analyzed. Results: Factor analysis showed six factors with factor loading ranging from 0.42 to 0.72 for the Japanese version of the SAS. Cronbach’s alpha was 0.94 for the 33 items in this scale. The scale showed good reliability for test-retest scores ( intraclass correlation coefficient 3,1 =0.92). Multiple linear regression analyses showed that the factors related to smartphone overuse were female gender, long smartphone usage duration, poor mental health, contact with virtual friends, and being a smartphone zombie. Contact with family and real friends were not associated with smartphone overuse. Conclusions: Although smartphone differs depending cultural the Japanese of the good the other Thus, this scale was shown to be reliable and valid in the

Recently, and globally, daily internet usage has been shifting from personal computers (PC) to highlyportable smartphones; and people feel inseparable from them [6]. In Japan, smartphone ownership rates are rapidly increasing, with 14.6% in 2011 and 60.9% in 2017, and it is especially high among youth, with 79.5% ownership among 13-to 19-year-olds and over 90% among 20-to 30-year-olds [7].
A smartphone, compared to a traditional mobile phone (primarily used for telephoning/emailing purposes), is a device that provides high accessibility to the internet through the enrichment of social networking systems (SNS) and application ("app") functions. With widespread smartphone usage, internet addiction problems may be shifting from those that were PC-usage-related to those smartphone-usage-related, which can be used at any time and place. Therefore, fully evaluating this problem with an internet addiction scale created 20 years ago may not be possible.
Smartphone overuse is considered to be "the occurrence of daily life disturbance through smartphone overuse as well as withdrawal symptoms such as anxiety and impatience when smartphone cannot be used [8,9]." Further, it is correlated with problematic internet use, with many similarities in their structural concepts [8]. However, smartphone overuse is unique because it is an addiction to portable devices [8].
The adolescents of smartphone overuse is rapidly increasing, as is the seriousness of online gaming addiction [2,10,11]. Thus, in 2018, the World Health Organization International Classification of Diseases (ICD)-11 designated internet gaming disorder as a mental disease [12]; it is defined as the continuation or repetitive gaming behavior (online or offline) with the persistence of the following three symptoms for at least 1 year: impaired control over gaming; increased priority given to gaming over all other interests in life; and continuation of gaming even with the occurrence of negative consequences to their social life [13]. This disorder presents comparable symptoms to pathological gambling and similar addictions.
Additionally, smartphone overuse has detrimental effects on health [2,14,15]; it became a concern because it can generate sleep disorder, lack of exercise, and poor academic performance [16], and is also related to mental health problems such as self-control, anxiety, depression, and dysfunctional impulsivities [17,18]. Problematic smartphone users also have multiple psychiatric symptoms [19], that is, it coexists with those symptoms. Further, excessive technology use is a maladaptive coping strategy used to deal with depressive and anxious symptoms [19]. There is a bi-directional relationship between mental disorders (e.g., depression) and smartphone overuse [18].
To our knowledge, few studies in Japan have investigated internet gaming disorder or smartphone overuse utilizing an international scale. In Korea, a Smartphone Addiction Scale (SAS) was developed [8]: The questions of SAS comprise items from a modified internet addiction scale [20], and items unique to smartphone addiction devised by specialists in behavioral addiction. Further, the validity and reliability of this scale have been validated [8]. SAS or SAS-short version has been translated into various languages such as German [21], Spanish and French [22], Turkey [23], Malaysia [24], Arabic [25], and Portuguese [26]. However, the scale has not yet been translated to Japanese, so it cannot be utilized in Japan. Therefore, we planned the development of a Japanese version of SAS (SASJ). This study aimed to evaluate the reliability and validity of the SASJ among Japanese youths and investigate factors related to smartphone overuse.

Methods
Creating the SASJ Permission to translate the scale was obtained from its author [8]. The translation was performed based on the World Health Organization recommendations [27]. To create the SASJ, a Japanese translator translated the text from English to Japanese; then, it was examined from a cultural, linguistic, and medical perspective by the authors and specialists in two fields (i.e., addiction research and clinical practice; linguistics and translation). These specialists acted as expert judges to qualitatively assess each item.
After this examination, the text was reverse-translated by two bilingual American translators, and the resulting text was sent to the original creator for evaluation and was approved. Subsequently, the questionnaire was applied in a pilot study conducted with a few adults and junior high/high school students to verify the Japanese translation.

Study design and setting
We sent a letter to the principal of each school asking for their cooperation, along with the same number of questionnaires and return envelopes as the number of students enrolled at the school.
After agreeing, they asked each class teacher to inform students about all aspects of the survey, its purpose, that participation was voluntary, all responses would be confidential and privacy would be protected. After questionnaire completion, students returned them to our department using the provided return envelopes. The survey was conducted from November to December 2018.

Participants
In total, 1,086 participants, aged between 15-24 years (median age, 17 years), were recruited from one high school (n = 1,050) and one vocational school (n = 83) in a suburban area of central Japan.
Participants comprised high school students from 10th to 12th grade, and first-year vocational school students.

Measures
SAS. The original SAS is a self-administered scale for smartphone addiction composed of six factors (daily-life disturbance, positive anticipation, withdrawal, cyberspace-oriented relationship, overuse, and tolerance) and 33 items rated on a six-point Likert scale from 1-6 (strongly disagree-strongly agree) [8]. The total score ranges from 33-198. Patterns of smartphone usage. We asked questions on whether participant had a smartphone and on smartphone usage duration/day on weekends, primary contact person (family, real friends, virtual friends), home smartphone usage rules, and average monthly smartphone usage fees. As for being a "smartphone zombie," which meant using a smartphone while walking and not paying attention to one's surroundings, we asked "Did you use a smartphone while walking in the past 30 days?" with five response options: "never," "seldom," "sometimes," often," and "always" ("often" and "always" were considered affirmative answers). We also asked questions on smartphone use services such as internet search, email, free communication apps, SNS, blogs, online games, and videos [5].
GHQ-12. The General Health Questionnaire-12 (GHQ-12) is a self-administered questionnaire designed as a screening tool for mental diseases [28]. It consists of two elements, "depression and anxiety" and "reduction of positive emotions", and a total of 12 items (six items per element). Four answer options are provided for each item; two options represented symptomatic absence (0 points), and the other two options represented symptomatic presence (1 point). This scale ranges from 0-12 points with higher total scores indicating poorer mental health. Previous studies set the cut-off value for the GHQ-12 to 4 points, and regarded a person with a score of 4 points or higher as having poor mental health [29].

Data analyses
First, the baseline characteristics of the study participants were assessed. We described age structure, smartphone usage pattern, usage amount, and contact person by gender.
Second, exploratory factor analysis investigated construct validity and the maximum likelihood promax with Kaiser Normalization. Data factorability was assessed by the Kaiser-Meyer-Olkin test for sampling adequacy and supported by Bartlett's test of sphericity to verify the suitability of data for factor analysis. Based on the Guttman-Kaiser rule, factors with eigenvalue large than 1 are retained [30].
Third, confirmatory factor analysis was performed to evaluate the factor structure of this scale. To Fourth, to test the internal consistency of the SASJ, Cronbach's alpha correlation coefficients were calculated [31].
Fifth, to evaluate re-test reliability, the intraclass correlation coefficient (ICC 3,1 ; two-way mixed, absolute agreement, single measurement) was obtained from a subgroup of 81 students from the participating vocational school.
Finally, to evaluate associated factors of smartphone overuse, we used the Spearman correlation coefficient to investigate the relationship between smartphone addiction scores and variables such as age, gender, smartphone usage duration/day on weekends, smartphone use services, primary contact, average monthly smartphone usage fees, home smartphone usage rules, GHQ-12 scores, and smartphone zombie. All analyses were performed using SPSS version 25.0 and AMOS 23.0 for Windows with the statistical significance set at p < 0.05.

Participants' characteristics
In total, 1,072 completed questionnaires were returned; however, 35 questionnaires were excluded owing to not owning a smartphone; therefore, data from 1,037 questionnaires were analyzed, with an effective response rate of 96.7%. Of the study population, 97% owned a smartphone (Table 1); 30.6% of male and 31.3% of female used smartphones for more than 5 hours/day on weekends, and most participants used internet search communication apps and SNS. Regarding to average SASJ score, males scored 78.0 ± 27.1 and females scored 88.7 ± 28.6. Gender differences were significant in this study (p < 0.001). Reliability and validity of SASJ Table 2 shows the exploratory factor analysis of the SASJ. The six factors corresponding to the SAS subscales were referred to as "withdrawal," "cyberspace-oriented relationship," "overuse and tolerance," "preoccupation," "daily life disturbance," and "positive anticipation." The corrected item-total correlation coefficients ranged from 0.42 (item 1) to 0.72 (item 30). Deletion of any individual items did not reveal changes in the scale internal consistency.  Table 3 shows the confirmatory factor analysis of the SASJ. Kaiser-Meyer-Olkin tests showed the sampling adequacy as 0.95 and was supported by a significant value on Bartlett's test of sphericity (p < 0.01).  Table 4 shows the reliability analysis of the SASJ.  Table 5 shows the correlation of the initial subscales and Cronbach's alpha coefficients.  Factors associated with smartphone overuse Table 6 shows the factors associated with smartphone overuse. Factors that showed significant correlations with smartphone overuse were gender (r = .173; p < 0.01), smartphone usage duration/day on weekends (r = .340; p < 0.01), virtual friends contact (r = .178; p < 0.01), mental health (r = .212 p < 0.01), and smartphone zombie (r = .262; p < 0.01). Most smartphone services such as communication apps (r = .058) and online games (r = .016) were weakly or nonsignificantly associated with smartphone overuse. Smartphone zombie: using a smartphone while walking and not paying attention to one's surroundings Gender was coded as 0 (male) and 1 (female); Age was coded as 0 (15-16 years), 1 (17-18 years), and 2 (19-24 years); Duration of smartphone use/day on weekends was coded as 0 (Less than 2 hours), 1 (2-3 hours), 2 (3-5 hours), and 3 (More than 5 hours). Average monthly smartphone usage fees (Japanese yen) was coded as 0 (Less than 3,000), 1 (3,000-4,999), 2 (5,000-6,999), 3 (7,000-9,999), 4 (10,000-14,999), and 5 (More than 15,000).

Discussion
This was the first study to investigate smartphone overuse in Japanese youth using the SASJ. The SASJ showed good reliability and validity, so it is an appropriate questionnaire to examine smartphone overuse in youths; female gender, long smartphone usage duration, poor mental health, contact with virtual friends, and being a smartphone zombie were associated with smartphone overuse.
The SASJ exhibited good internal consistency; Cronbach's alpha coefficient for the total scale was 0.94, and the respective coefficients for the six factors were above 0.83. The test-retest reliability values were found to be similar to those of the original, Malay, and the Arabian SAS versions [8,24,25]. Thus, the SASJ is as good as the other language versions.
The six dominant components that explained a large proportion of the variability of the SASJ were similar to those of the original SAS. The components in the original version were "daily life disturbance," "positive anticipation," "withdrawal," "cyberspace-oriented relationship," "overuse," and "tolerance." In this study, the changes were: "overuse and tolerance" became one component Our results showed that SASJ scores were associated with the female gender. Corroborating, the risk of engaging in smartphone overuse is higher among females than males [8,[32][33][34]. However, males have higher risks of developing smartphone overuse than females [35]. Thus, gender differences may owe to sociocultural factors, potential confounding factors, and study methodology.
Results showed that smartphone usage duration was strongly associated with smartphone overuse, consistent with previous studies [21,[36][37][38]. Correlating, smartphone usage duration presented the strongest direct effects on smartphone overuse [37]. Further, adolescents with higher smartphone usage and for longer durations were aware of their addictive behavior and did not consider it problematic [21,38]. Adolescents' awareness over their problematic behaviors may a key determinant regarding smartphone overuse prevention.
Our study revealed that smartphone overuse was positively related to online friendship and unrelated to real friendship; corroborating, excessive chat room users constitute one of the Internet addiction groups [39]. For them, the Internet is a major source of social/interpersonal rewards, implying potential addiction to virtual relationships and communication [39]. Smartphone overuse and preference to communicate with virtual friends may demonstrate youths' tendency to seek virtual friendships, and this association between the first and the latter may be bidirectional: Youth who prefer online friendships may experience more conflicts in the offline world; they may ignore offline friends and cultivate their tendency to overuse the virtual world through the smartphone. Besides, youth who are prone to smartphone overuse may have some reasons to spend a lot of time online; they may be looking for virtual friends to get the social support they do not find in the real world.
Longitudinal research is needed to investigate these associations.
Youth studies show an important association between smartphone overuse and psychological and psychiatric problems [17,18]. In this study, as expected, poor mental health status was positively associated with smartphone overuse. There may be a bi-directional relationship between mental disorders and smartphone overuse [18]. Thus, smartphones may worsen the status of -and be particularly attractive to -people with poor mental health.
Being a smartphone zombie was moderately related to smartphone overuse, implying that individuals get so absorbed in using their smartphones that, eventually, they are unable to cease their behavior while in transit. Being a smartphone zombie is caused by the need to continuously look at the screen to be able to instantly respond to messages or to check the transit progress on a map app. Moreover, smartphones require screen touch to input information, which may also be a cause of smartphone zombies.
Our results showed that smartphone applications were weakly related to smartphone overuse, unlike previous studies [14,40]. Regarding the reason, many youths used the services listed in the questionnaire, and the differences between service usage in relation to smartphone overuse could not be extracted because the questions did not survey which service was mostly used. Future research should investigate the association between smartphone overuse and the usage duration of different app types by using separate apps that allow for the identification of usage duration by smartphone service type.
Our study has several limitations. First, it was conducted in a specific area of central Japan and the demographic characteristics of the study population were unbalanced. Therefore, it is difficult to generalize the results. Future studies should include developed and semi-developed demography and areas to obtain an insight on smartphone overuse and addictions led by socio-financial situations.
Second, we adopted a self-administered questionnaire format. Thus, for questions such as smartphone usage time, we could have yielded more robust findings if we had chosen an approach that could validate the responses, such as an application to confirm usage time. Third, we did not take lifestyle factors (smoking or drinking alcohol) or sociological data (socioeconomic status or parents' academic history) into consideration. Such factors may have shaped participants' smartphone overuse, so their influence should be verified. Fourth, there is no diagnostic criteria for smartphone overuse, so it is difficult to determine the cut-off value; very few medical professionals in Japan can diagnose smartphone overuse, and the reliability/validity of the Japanese version of an internet addiction scale have yet to be validated. Therefore, we could not report the prevalence of smartphone overuse.
Despite the above limitations, our study provides important information on smartphone usage in Japanese youth. The SASJ is a reliable and valid questionnaire that can be used to screen for high risk of smartphone overuse in Japanese youth.

Conclusions
This was the first survey on smartphone overuse to utilize the SASJ. We found that female gender, long smartphone usage duration, poor mental health, contact with virtual friends, and being a smartphone zombie are related to smartphone overuse in the Japanese youth. Future studies should extend the survey area to include the nationwide population, other age groups, examine additional factors that may be related to smartphone overuse and consider intervention programs to prevent it. The data collection was also approved by the headmasters of the schools selected for this study.
Participants were informed that the survey was totally voluntary.

Consent for publication
Not applicable.