The Psychometric Properties of the Smartphone Application-Based Addiction Scale (SABAS)

The goal of the study was to validate the English version of the Smartphone Application-Based Addiction Scale (SABAS; Csibi et al. 2016), which is a short and easy-to-use tool for screening the risk of smartphone application-based addiction. Another aim was to identify the most frequently used smartphone applications and their perceived importance by the participants. Data were collected online from 240 English-speaking volunteers, aged 18 to 69 years. The instruments used were the SABAS, the Nomophobia Questionnaire (NMP-Q), the Brief Sensation Seeking Scale (BSSS), the Deprivation Sensation Scale (DSS), and the Patient Health Questionnaire (PHQ-9). Participants also ranked the importance of their most frequently used smartphone applications. The six items of the SABAS yielded one component, which accounted for 52.38% of the total variance. The internal reliability of the scale was good (Cronbach’s alpha 0.81). NMP-Q was a significant predictor of SABAS, explaining 17.6% of the total variance. The regression analysis, with SABAS score as the dependent variable and NMP-Q, DSS, PHQ-9, and BSSS scores as predictors, indicated that approximately 47% of the variance in SABAS was accounted for by the predictors (R 2 = 0.47). The English version of the SABAS appears to be a valid and reliable ultra-brief tool for a quick and easy assessment of smartphone application-based addiction symptoms.

Research has identified a heterogeneous spectrum of problematic use of smart devices for online activities and applications, such as Facebook or other forms of socializing in the virtual world (Billieux et al. 2015;Király et al. 2015;Kuss and Griffiths 2011). Specialists in behavioral addictions highlight the importance of generating valid instruments suitable for the screening and the measurement of technology-related addictive behavior (Kardefelt-Winther 2016;Lopez-Fernandez 2017). The literature features several adaptations and generations of new instruments, such as the adapted Smartphone Addiction Scale-Short Version (SAS-SV) in Spanish and Francophone Belgian countries by Lopez-Fernandez (2017). The factor analyses of both the Spanish and Francophone Belgian versions supported the scale's unidimensionality that explained 49% and 54% of the variance respectively, good construct validity, and excellent reliability. Other smartphone addiction measures have also demonstrated good reliability, including the Smartphone Addiction Inventory (SPAI) in China (Lin et al. 2014) and the Smartphone Addiction Proneness Scale (SAPS) in South Korea (Kim et al. 2014). Studies testing the factorial structure of smartphone assessment instruments (e.g., Lin et al. 2014;Lopez-Fernandez 2017;Pavia et al. 2016;Kwon et al. 2013) have identified several common variables acting as risk factors in smartphone addiction, including time distortion in using a smartphone, compulsivity with a negative impact on daily life activities and relationships, symptoms of tolerance, craving, and withdrawal symptoms.
In the current study, the primary aim was to validate an English version of the original Hungarian version of the Smartphone Application-Based Addiction Scale (SABAS; Csibi et al. 2016). In contrast to the existing instruments, the SABAS is a short and easy-to-use tool for screening the risk for smartphone addiction. A secondary aim was analyze the preferences for smartphone-based applications in relation to SABAS and Nomophobia 1 Questionnaire (NMP-Q) scores. Finally, the study also aimed to examine the ways in which these choices influence scores reported on deprivation sensation, depression, and sensation seeking variables.

Participants
The participants comprised 240 volunteers, aged between 18 and 69 years (mean age = 25.4 years), 155 males (64.6%) and 85 females (35.4%). They completed an online questionnaire following recruitment via various internet forums (such as university students' blogs, CNET, LinkedIn, Facebook), especially those related to mobile application discussion groups. By the nature of the study, self-selection was an a priori delimitation (as well as a later limitation). The research was approved by the Research Ethics Board of a large urban university. All the participants read and agreed to informed consent form prior to taking part in the study.

Instruments
In addition to a brief demographic questionnaire (e.g., gender, age), the questionnaire also included the SABAS (6 items; Csibi et al. 2016; See Appendix 1), the NMP-Q (20 items; Yildirim and Correia 2015), the Brief Sensation Seeking Scale (BSSS, 8 items;Hoyle et al. 2002), the Deprivation Sensation Scale (DSS, 9 items;Robbins and Joseph 1985), and the Patient Health Questionnaire Depression Scale (PHQ-9, 9 items; Kroenke et al. 2001). Participants were also asked about the most frequently used smartphone applications and asking them to rank by their importance (BPlease list the three most important and most used smartphone applications by yourself in a ranked order [i.e., first, second, and third most important] ).

Procedure
The responses to the particular question regarding the most often used applications were transformed into quantitative variables according to nine types of applications (apps): (i) traditional mobile communication and phone function, (ii) internet-based communication, (iii) social media, (iv) information, (v) entertainment, (vi) games, (vii) directions and timetable, (viii) lifestyle applications, and (ix) health-related applications. As noted above, the categories were ranked (1, 2, or 3) and scored according to the participant's responses (scoring 3 if it was the most popular, 2 if it was second, and 1 if it was ranked third).
To fulfill the secondary aim of the study, a two-step cluster analysis of the data was performed. This generated two distinct clusters, the first with 139 participants (57.9%) and the second with 101 participants (42.1%), with the ratio of sizes 1.38. The clusters were generated based on the test scores. Therefore, in the first cluster, the scores were higher on DSS, SABAS, and NMP-Q and lower on PHQ-9 and BSSS scales; in the second, the scores were lower on DSS, SABAS, and NMP-Q and higher on PHQ-9 and BSSS scales ( Table 1).
The first cluster had low values on DSS (smartphone use without sensation of deprivation) and high values on SABAS and NMP-Q (showing addiction vulnerability), low scores on depression scale (lack of negative feelings), and moderated values on BSSS. The second cluster had the opposite value on every scale, except the BSSS (which had approximately the same value).
Statistical analysis was performed with PASW 18 (SPSS, Chicago, IL). To explore the underlying factor structure of the SABAS, an exploratory factor analysis (EFA) was performed on the data. Internal consistence reliability using Cronbach's alpha coefficient was examined. A correlation coefficient was computed to assess the relationship between the scores of the participants on the SABAS, DSS, NMP-Q, PHQ-9, and BSSS. Two-step cluster analysis was used to distribute the sample according to addiction vulnerability. For the analysis of the differences between the two clusters, ANOVA and Mann-Whitney tests were performed.

Results
Theoretically, the SABAS was expected to assess a single construct (i.e., the level of smartphone addiction). Therefore, the six items of the scale were subjected to factor analyses using the principal component analysis method. Each of the six items of the SABAS was significantly correlated with all the other items in the scale (p < 0.01), supporting factorability. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was 0.84. The Bartlett's test of sphericity was significant (χ 2 (15) = 426.87, p < 0.01). Furthermore, the diagonals of antiimage correlation matrix were all over 0.80 supporting the inclusion of each item in the factor analysis. Finally, the commonalities were mostly above 0.40, further confirming that each item shared some common variance with the other items. Given these indicators, a principal axis factoring was performed with the six items of the SABAS. One factor (i.e., a single component, eigenvalue = 3.14) was identified and accounted for 52.38% of the total variance (see Table 2). The internal reliability of the scale was examined using Cronbach's alpha and demonstrated good reliability (Cronbach's alpha = 0.81; mean score = 22.16; std. deviation = 5.69). The Cronbach's alpha value did not change significantly when one item was deleted suggesting good internal consistency. The congruent validity of the test also appeared to be good. The correlation with the 20-item NMP-Q (Yildirim and Correia 2015) was statistically significant (r = 0.63, p < 0.01). A significant (negative) correlation (r = −0.60, p < 0.01) was also found with the nine-item DSS (Robbins and Joseph 1985). The other two scales applied in the present study also showed significant negative correlations with the SABAS (see Table 3).
The stepwise linear regression analysis, with SABAS score being the dependent variable and NMP-Q, DSS, PHQ, and BSSS being predictors, indicated that NMP-Q score was a significant predictor of SABAS score (b = 0.12, SE = 0.01, p < 0.01). Approximately 47% of the variance in SABAS was accounted for by all the predictors (R 2 = 0.47). Of this, the NMP-Q score explained 17.6% of the variance in SABAS (see Fig. 1).
The mean rank analysis between the two clusters of data (using a Mann-Whitney test) demonstrated significant differences in application use, such as traditional mobile communication, social media, and information gathering (see Table 4).
ANOVA results show statistically significant differences between the test scores and the application categories (Table 5).  *Correlation is significant at the 0.05 level (two-tailed)

Discussion
In the present study, the psychometric properties of an English language version of a short form of smartphone addiction scale were assessed. The factor analysis of the six SABAS items identified one factor (i.e., a single component). By analyzing the psychometric characteristics of the SABAS, the analysis demonstrated good internal reliability and consistency of the scale. Congruent validity was also confirmed via its significant correlation with the NMP-Q and DSS. These results are in agreement with other scale validating studies regarding smartphonerelated addiction supporting a one-factor construct alongside good reliability and validity. The regression analysis of the data demonstrated that nomophobia, deprivation, depression, and sensation seeking predicted 47% of the total variance of the SABAS. In contrast to our findings, Jeong et al. (2016) found that sensation seeking was not a significant predictor of smartphone addiction. A possible explanation provided by other authors for the aforementioned association is that addictive symptoms among healthy populations do not reach pathological levels and that the patterns of addictive behaviors are not necessarily directly related to psychological constructs such as depression or sensation seeking Van Rooij and Prause 2014).
The present study found an inverse relationship between depression and excessive smartphone use, which is in accord with other studies suggesting that the active use of social applications may be beneficial in lowering depressive symptoms (Verduyn et al. 2017;Elhai et al. 2017a, b). A possible explanation for these results might be that depressive symptoms are not necessarily associated with social withdrawal and general avoidance of social interaction. Alternatively, smartphone use may contribute to increased social engagement and decrease feelings of loneliness and isolation (Kim et al. 2016a, b). Our results are in agreement with those of a representative Hungarian population of 5961 adolescents, which reported 4.5% at risk for social media use addiction, and association with low self-esteem, and high level of depression symptoms (Banyai et al. 2017).
The present study demonstrated significant differences in application use, such as traditional mobile communication, social media, and information gathering. According to the results, it can be argued that mobile communication, social media, and information gathering were the most relevant smartphone applications in influencing the scores obtained on the SABAS, NMP-Q, DSS, PHQ, and BSSS. The results here are in line with previous research, which found that social networking was one of the most popular online activities, followed by e-mail, chat, and videos and movies (Griffiths and Szabo 2014).
A limitation of the present study is that the research was conducted on a small healthy volunteer sample, rather than on a clinical one. Therefore, it is difficult to document the presence of other psychological deficiencies present in the participants' lives. Furthermore, the analysis utilized self-reported cross-sectional data from a self-selected sample of participants, which might lead to distortions in results. The present authors cannot argue for a determinant category of symptoms manifesting along the excessive smartphone use. However, the study supports the existence of smartphone use-related addictive symptoms. Finally, the study does not provide any information regarding the harm and persistency of the problems assumed to be assessed through the new instrument. The study of the associations with other personalityrelated variables is recommended in further research.

Conclusion
By validating an English version of the Hungarian version of the SABAS (Csibi et al. 2016), we provided a short, easy-to-use, and easy-to-score tool for screening the risk of smartphone addiction, with the potential of exceeding cultural boundaries. The study adds some empirical data concerning the possible core determinants of smartphone application-related addictive behavior without establishing a clear causality in the effects. The most highly ranked preferences for smartphone-based applications were communication, social media, and information gathering in the sample with relevant impact on SABAS and NMP-Q scores. Significant associations were found between these choices and scores reported on deprivation sensations, nomophobia, and SABAS, alongside a negative significant correlation with depression and sensation seeking variables.
Author Contributions SC wrote the mojority if the first draft of the paper and did most of the data analysis. BC worked out the design and took part in writing and data analysis. ZD provided the online survey platform and took part in writing and statistical analyses. AS devised the adopted questionnaire on the basis of MDG's conceptual model of addiction. MDG wrote the final version of the paper.

Compliance with Ethical Standards
Financial Support Present work was supported by the Hungarian Scientific Research Fund (grant numbers 111938 and K109375).
Ethical Clearance Ethical permission for the study was obtained from the Research Ethics Committee of the Faculty of Education and Psychology at Eötvös Loránd University.

Conflict of Interest
The authors declare that they have no conflict of interest.

Appendix 1 Smartphone Application-Based Addiction Scale (SABAS)
Please indicate the extent to which you agree or disagree with the statements below in relation to your smartphone use habits. If I try to cut the time I use my smartphone, I manage to do so for a while, but then I end up using it as much or more than before. The total score is calculated by adding the individual scores (1-6) of each item (providing a minimum score of 6, and a maximum score of 36) Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.