Abstract
Background and aims
Problematic smartphone use (PSU) has gained attention, but its definition remains debated. This study aimed to develop and validate a new scale measuring PSU-the Smartphone Use Problems Identification Questionnaire (SUPIQ).
Methods
Using two separate samples, a university community sample (N = 292) and a general population sample (N = 397), we investigated: (1) the construct validity of the SUPIQ through exploratory and confirmatory factor analyses; (2) the convergent validity of the SUPIQ with correlation analyses and the visualized partial correlation network analyses; (3) the psychometric equivalence of the SUPIQ across two samples through multigroup confirmatory factor analyses; (4) the explanatory power of the SUPIQ over the Short Version of Smartphone Addiction Scale (SAS-SV) with hierarchical multiple regressions.
Results
The results showed that the SUPIQ included 26 items and 7 factors (i.e., Craving, Coping, Habitual Use, Social Conflicts, Risky Use, Withdrawal, and Tolerance), with good construct and convergent validity. The configural measurement invariance across samples was established. The SUPIQ also explained more variances in mental health problems than the SAS-SV.
Discussion and conclusions
The findings suggest that the SUPIQ shows promise as a tool for assessing PSU. Further research is needed to enhance and refine the SUPIQ as well as to investigate its clinical utility.
Introduction
The global smartphone user population has nearly doubled in 7 years, reaching an estimated 6,841 million in 2023 (O’Dea, 2023), raising concerns about the potential health, social, and economic impacts of excessive smartphone use (Elhai, Levine, Dvorak, & Hall, 2017; Jannusch, Shannon, Völler, Murphy, & Mullins, 2021; Olson et al., 2022). “Problematic smartphone use (PSU)” describes the persistent and excessive smartphone use patterns related to daily-life malfunctioning (e.g., Elhai & Contractor, 2018; Kardefelt-Winther et al., 2017). PSU shares features with addictive disorders (Haug et al., 2015; Lin et al., 2014), but the concept of “smartphone addiction” remains debated (e.g., Horvath et al., 2020; Larsen, Wiers, Su, & Cousijn, 2023). This study aimed to develop and validate a new PSU questionnaire.
PSU is typically assessed with various scales, including the Mobile Phone Problem Use Scale (MPPUS; Bianchi & Phillips, 2005), the Problematic Mobile Phone Use Questionnaire (PMPUQ; Billieux, der Linden, & Rochat, 2008), the Smartphone Addiction Scale (SAS; Kwon, Lee, et al., 2013), the Short Version of Smartphone Addiction Scale (SAS-SV; Kwon, Kim, Cho, & Yang, 2013), the Smartphone Addiction Inventory (SPAI; Lin et al., 2014) and the Short-Form of Smartphone Addiction Inventory (SPAI-SF; Lin, Pan, Lin, & Chen, 2017), which measure factors like tolerance, withdrawal, overuse, loss of control and dangerous/prohibited use. While cross-cultural validation studies have shown satisfactory psychometric properties for these scales (e.g., Agus, Mascia, Bonfiglio, & Penna, 2022; Andrade et al., 2022; Khoury et al., 2017; Lopez-Fernandez, Honrubia-Serrano, Freixa-Blanxart, & Gibson, 2014, 2018; Sfendla et al., 2018; Wang, Long, Liu, Liu, & Billieux, 2020; Zhao, Rafik-Galea, Fitriana, & Song, 2022), we would argue that there is room for improvement.
While previous scales were developed based on existing “addiction” criteria (see Supplementary Table S1), they may not fully capture the current problems people face in relation to PSU (see a review, Nawaz, 2023), particularly issues specific to smartphone use like distracted driving and impaired productivity (e.g., D. Ding & Li, 2017; Fitch, Hanowski, & Guo, 2015; Jannusch et al., 2021; Oviedo-Trespalacios, Haque, King, & Washington, 2016). They also may not address general addiction-like features like compromised relationships and cognitive impairments (e.g., D. Ding & Li, 2017; Gutiérrez, de Fonseca, & Rubio, 2016; Wilmer, Sherman, & Chein, 2017). Therefore, PSU measurement should consider the everyday smartphone use experiences to avoid over-pathologizing normal behaviors.
To optimize PSU assessment, person-centered and process-based qualitative research can provide valuable insights into individuals experiencing functional impairment and emotional distress due to smartphone use (Billieux, Maurage, Lopez-Fernandez, Kuss, & Griffiths, 2015; Billieux, Philippot, et al., 2015; Billieux, Schimmenti, Khazaal, Maurage, & Heeren, 2015; Flayelle, Schimmenti, Starcevic, & Billieux, 2022; Kardefelt-Winther et al., 2017). However, few in-depth qualitative studies (see Supplementary Table S1) have been conducted to identify PSU assessment themes (e.g., Yildirim & Correia, 2015). Instead, several studies used relatively informal interviews to generate scale items, without formal coding and analyses (e.g., Cho & Lee, 2015; J. E. Ding et al., 2019; Lee et al., 2017; Merlo, Stone, & Bibbey, 2013). A thorough qualitative approach with deductive and inductive thematic analyses can offer a richer understanding of PSU-related experiences in everyday life, leading to more accurate and valid measurements.
In this study, we have developed the Smartphone Use Problems Identification Questionnaire (SUPIQ) based on an in-depth semi-structured interview protocol (Su et al., 2023) that incorporated DSM-5 (American Psychiatric Association, 2013a) and ICD-11 (World Health Organization, 2019) “addiction” criteria, existing PSU scales (e.g., SAS, Kwon, Lee, et al., 2013; SPAI, Lin et al., 2014) and the recent research (e.g., D. Ding & Li, 2017; Fitch et al., 2015; Gutiérrez et al., 2016; Jannusch et al., 2021; Oviedo-Trespalacios et al., 2016; Wilmer et al., 2017). Our primary aim is to assess the SUPIQ's reliability and validity in measuring PSU, beginning with examining its construct validity. We will also evaluate its convergent validity by exploring the associations between SUPIQ scores and mental health problems. High levels of PSU have been related to higher levels of mental health and cognitive problems like depression and anxiety (e.g., Elhai et al., 2017, 2020; Jin et al., 2021; Kim et al., 2019), personality disorders (e.g., Pearson & Hussain, 2016), sleep problems (e.g., Cheung et al., 2019; S. Y. Sohn, Krasnoff, Rees, Kalk, & Carter, 2021), somatic symptoms (e.g., Winkler, Jeromin, Doering, & Barke, 2020), suicidal ideation (e.g., Arrivillaga, Rey, & Extremera, 2020), memory problems (e.g., Madore et al., 2020), and repetitive thoughts and behaviors (e.g., Brailovskaia, Stirnberg, Rozgonjuk, Margraf, & Elhai, 2021). To assess convergent validity, we also include the Short Version of Smartphone Addiction Scale (SAS-SV; Kwon, Kim, et al., 2013), frequently used to assess PSU (e.g., Khalily, Saleem, Bhatti, Ahmad, & Hussain, 2019; Kwon, Kim, et al., 2013; Kwon, Lee, et al., 2013; Luk et al., 2018; Zhao et al., 2022).
Besides mental health problems, we will explore how smartphone use frequency and duration relate to PSU (Harris, Regan, Schueler, & Fields, 2020; James, Dixon, Dragomir, Thirlwell, & Hitcham, 2023; Parry et al., 2021), akin to other addictive behaviors like alcohol use (AUDIT; Mattiko, Olmsted, Brown, & Bray, 2011; Saunders, Aasland, Babor, De La Fuente, & Grant, 1993) and problematic social media use (Social Media Disorder Scale; Van Den Eijnden, Lemmens, & Valkenburg, 2016). To assess the relationships, we will differentiate between screen time and unlock frequency (Ryding & Kuss, 2020), social versus nonsocial screen time (Elhai et al., 2017, 2020), active (i.e., actively interacting with others via social media on smartphone) versus passive (i.e., passively scrolling and checking via social media on smartphone) social media use (Su, Larsen, Cousijn, Wiers, & Van DenEijnden, 2022). Partial correlations between the SUPIQ factors, smartphone use statistics, and mental health problems will be tested and visualized with network analysis to further show the convergent validity. Additionally, we will compare the SUPIQ with the SAS-SV to assess its explanatory power in hierarchical multiple regression models (Kwon, Kim, et al., 2013; Olson et al., 2022). While prior PSU scale studies have primarily focused on student and adolescent populations (e.g., Andrade et al., 2022; Cheung et al., 2019; Khoury et al., 2017; Leung, 2008; Lin et al., 2014; Lopez-Fernandez et al., 2014; Pavia, Cavani, Di Blasi, & Giordano, 2016; Walsh, White, & McDYoung, 2010; Wang et al., 2020; see Supplementary Table S1), leaving the applicability of these scales to the general population largely unexplored (Busch, Hausvik, Ropstad, & Pettersen, 2021; Rosales & Fernández-Ardèvol, 2019). We will extend its applicability using two separate samples: one from the university community and a general population sample.
In summary, we aimed to develop a comprehensive and widely applicable PSU assessment tool, conducting an exploratory factor analysis (EFA), a confirmatory factor analysis (CFA), correlation analyses, and partial correlation network analyses, examining measurement invariance, and assessing explanatory power with hierarchical multiple regression models in two diverse samples.
Methods
Participants and procedures
The study involved two samples: Sample 1 (university community sample) was comprised of 323 participants (Mean age = 20.81, SD = 4.34, the range = 18–66 years; 72.76% females), recruited through the university's internal study platform, primarily composed of university students. However, note that other volunteers could also sign up through the University's recruitment platform, therefore it is not exclusively a student sample. Sample 2 (the general population sample) included 618 participants (Mean age = 28.47, SD = 7.39, the range is 18–75 years, 43.85% females), recruited on social media including LinkedIn, X (Twitter), and Facebook as well as through posters. The survey was administered online using Qualtrics XM software (Qualtrics, 2021). The general population sample had a 1 in 20 chance of winning a 20-euro gift card by providing their email, while the university community sample could earn 0.50 psychology research credits in addition to the lottery. Data collection took place from March 20 to April 12 spanned 3.5 weeks period in 2021. The survey, mainly conducted in the Netherlands, used English as the primary language. Participants confirmed their English proficiency in the consent form. Table 1 displays participants' country of residence and nationality distribution.
Measures
The quality control questions
To ensure the quality of online responses, three quality control measures were implemented (DeSimone, Harms, & DeSimone, 2015): (1) participants were instructed to select “always” from the response options “never” to “always” for the indication question; (2) they were supposed to select “false” for the statement of “I have never used a smartphone” in the bogus question; (3) participants rated their survey effort and attention on a scale of 0–100 at the survey's end.
The Smartphone Use Problems Identification Questionnaire (SUPIQ)
The initial items of the first version of the Smartphone Use Problems Identification Questionnaire (SUPIQ) were developed from a separated qualitative study among 28 university students with smartphone use problems (Su et al., 2023). The inclusion criteria for participant recruitment were university students between 18 and 25 years old, possessed at least one European nationality, and scored above 31 (males) or 33 (females) on the Short Version of the Smartphone Addiction Scale (SAS-SV; Kwon, Kim, et al., 2013). These cut-off scores were determined based on the established thresholds for identifying PSU. The ultimate qualitative sample, consisting of 28 participants (24 males, 26 Bachelor students, 2 Master students), demonstrated an average SAS-SV score of 46.82 (SD = 3.22). The participants, with an average of 7.39 years of smartphone usage (SD = 1.47), originated from 17 distinct European countries, with six individuals holding dual nationality. The first author conducted interviews, recording, transcriptions, and coding, with the results discussed among the research team. Based on the qualitative results, the first author drafted a preliminary questionnaire comprising 126 items. This draft was then thoroughly deliberated upon with the entire research team. Based on the discussions, the initial version of SUPIQ contained 57 items (see Supplementary Table S2), including 2 negatively worded items. The SUPIQ aimed to capture 8 theory-derived factors related to PSU according to DSM-5 (American Psychiatric Association, 2013a) and ICD-11 (World Health Organization, 2019): (1) Impaired Control, (2) Preoccupation, (3) Craving, (4) Escapism/Relief/Coping, (5) Negative effects/Consequences/Risks, (6) Ignorance of Negative Effects/Consequences/Risks, (7) Tolerance, and (8) Withdrawal. Participants rated their smartphone use during the past 12 months, using a 5-point Likert scale (1 = never to 5 = always).
Mental health problems: DSM-5 Self-Rated Level 1 Cross-Cutting Symptom Measure-adult
The DSM-5 Self-Rated Level 1 Cross-Cutting Symptom Measure was used to assess participants' mental health problems over the past 12 months instead of the original 2-week version (American Psychiatric Association, 2013b). It consists of 20 questions on a 5-point Likert scale (0 = Never to 4 = Always) measuring depression, anxiety, personality functioning, sleep, somatic symptoms, suicidal ideation, memory, repetitive thoughts and behaviors, mania, psychosis, and dissociation. The total score ranges from 0 to 80, with higher scores indicating more mental health problems. See Table 4 for means, standard deviations, and internal consistencies including Cronbach's α and McDonald's ω values of the DSM-5 Self-Rated Level 1 Cross-Cutting Symptom Measure.
Smartphone use statistics
The amount of smartphone use
Participants reported total screen time, social screen time, and unlock frequency in the past 7 days using the Digital Wellbeing (Android) or Screen Time (iPhone) application on their smartphones. If participants were unable to access the data directly on their phones, they were asked to estimate these indicators to the best of their ability.
Active and passive social media use frequency on smartphone
Participants were asked to estimate the frequency of their past 7-day social media use with 6 questions. Active social media use involved active interactions with others by sending, posting, liking, and commenting on the social media apps (e.g., Estimate how many times in 7 days you send a text, message, photo or video via the social media apps (e.g., Instagram, Facebook, iMessage, TikTok, WhatsApp, LinkedIn, etc.) on your smartphone?). Passive social media use included passive scrolling behaviors like reading and browsing on these social media apps (e.g., Estimate how many times in 7 days you read messages, photos or videos from others via the social media apps (e.g., Instagram, Facebook, TikTok, WhatsApp, LinkedIn, etc.) on your smartphone?).
App usage
Participants reported the three most frequently used social media apps and three most frequently used apps after they pick up their smartphones.
The short version of smartphone addiction scale (SAS-SV)
The short version of smartphone addiction scale (SAS-SV) was developed by Kwon, Kim, and colleagues (2013), which includes 10 items and measures problematic smartphone use (Khalily et al., 2019; Kwon, Kim, et al., 2013; Kwon, Lee, et al., 2013; Luk et al., 2018; Zhao et al., 2022). Participants were asked to rate their smartphone use in the past 12 months from 1 (strongly disagree) to 6 (strongly agree), on questions like: “I use my smartphone longer than I have intended”. The means, standard deviations, and internal consistencies including Cronbach's α and McDonald's ω values of the SAS-SV were presented in Table 4.
Statistical analysis
The case-selection procedure with the Quality Control Questions and the descriptive analyses on demographic information of the two samples were performed with R (version 4.2.2; R Core Team, 2022) and RStudio (Allaire, 2022), using “janitor” (Mansley, 2021), “dplyr” (Wickham, François, & Henry, 2020), “plyr” (Wickham & Wickham, 2020), “sjmisc” (Lüdecke, 2021a), and “sjPlot” (Lüdecke, 2021b) packages. The factor analyses were conducted using Mplus 8.3 (Muthén & Muthén, 2017) and the package “MplusAutomation” (Hallquist & Wiley, 2018).
To ensure data quality in both samples, participants were excluded if they spent less than 2 s per item, reported less than 50% effort/attention, or answered quality control questions incorrectly (DeSimone et al., 2015, see more detailed explanations in Measures) were excluded from analysis. To ensure the comparability of our previous study, participants from the university community were selected based on criteria (i.e., aged 18–24 years and had completed upper secondary or bachelor's level education) in the qualitative study (Su et al., 2023).
Construct validity of the SUPIQ: exploratory factor analysis (EFA)
To examine the SUPIQ's factor structure, we conducted an EFA. Since the SUPIQ questionnaire uses 5-point Likert scales and the data deviated from the normal distribution, we employed weighted least squares estimation, accounting for the categorical and nonnormal data nature (DiStefano & Morgan, 2014; Holgado-Tello et al., 2010; Rhemtulla, Brosseau-Liard, & Savalei, 2012; Wu & Leung, 2017). To determine the number of factors underlying the data, we fitted a series of models that varied in the number of factors (starting with a one-factor model) model-based approach in which we fit a series of models that increase in their number of factors, but in which all items are allowed to load on all factors. This model-based approach is the default method in Mplus 8.3 (Muthén & Muthén, 2017). Such an approach is grounded in the benefits of a model-based strategy, which relies on an explicit statistical model instead of principal axis factoring (Brown, 2015). This approach enables a more comprehensive and objective selection of the best fitting model with diverse model fit indices: the root-mean-square-error of approximation (RMSEA), the standardized root mean square residual (SRMR), the comparative fit index (CFI), and the Tucker–Lewis index (TLI). Generally, good model fit is indicated by RMSEA <0.05, SRMR <0.08, and CFI/TLI >0.95, while acceptable fit is indicated by RMSEA <0.08, and CFI/TLI >0.90 (see Brown, 2015; Goulter et al., 2022; Little, 2013). Within the best fitting model, we retained items based on criteria: 1) no cross-factor loadings (i.e., with two factors' loadings >0.30; 2) factor loading >0.50 and <1 (William Jr, 2013); 3) alignment with the SUPIQ's intended factors according to our theoretical framework.
Construct validity of the SUPIQ and the SAS-SV: confirmatory Factor Analysis (CFA)
To investigate whether the SUPIQ factor structure found using EFA replicated in the general population sample, we used CFA with weighted least squares estimation. For the factor structure (as established in the university sample) to be replicable, the fit of the CFA model should at least be acceptable in the general population sample. In addition, we have also used CFA to test the construct validity of the SAS-SV, with residual covariances added among items 1 to 3 and items 4 to 7 since they belonged to one factor in the original version of SAS-SV (Kwon, Kim, et al., 2013; Kwon, Lee, et al., 2013; Luk et al., 2018). To assess the fit of the confirmatory model, we relied on the same model fit statistics discussed above.
Psychometric equivalence analyses of the SUPIQ: measurement invariance test
To examine whether the SUPIQ factor structure remained invariant across the university community and the general population samples, we conducted multigroup confirmatory factor analyses (MG-CFA): Three models were employed: 1) a configural model (Model 1), fitting the same factor model to the two samples with all factor model parameters freely estimated; 2) a metric measurement invariance model (Model 2), restricting item loadings to be equal across the samples while allowing differences in the factor variances; 3) a scalar measurement invariance model (Model 3), with equal factor loadings and equal item threshold parameters while allowing differences in the factor means and variances. A decrease in CFI by more than 0.01 or an increase in RMSEA by more than 0.015 indicates that measurement invariance is not established (Chen, 2007).
The analyses on descriptive statistics, reliability, and convergent validity
The descriptive statistics were done after the final version of SUPIQ was determined. For the reliability test of the SUPIQ, both Cronbach's α and McDonald's ω values were reported since they represent mean test level and general factor saturation separately (Revelle & Condon, 2019). For the convergent validity analyses, the linear correlations and partial correlation network analyses were used. The linear correlations between the SUPIQ, the SAS-SV, mental health problems, and smartphone use statistics were calculated. The SUPIQ total score was included in the analyses to provide a comprehensive assessment of the construct, enhancing understanding of participants' responses across all measured factors. Unreasonable smartphone use statistics, such as minutes exceeding 59 (as participants reported hours and minutes separately), reported screen time of 0, or nonsocial screen time less than 0, were treated as missing values and the missing values were dealt with pairwise deletion method. Partial correlation network analyses among the sum scores of different factors of the SUPIQ, mental health problems, and smartphone use statistics were performed using the packages “bootnet” (Epskamp & Fried, 2015, 2020), “qgraph” (Epskamp, Epskamp, & Rcpp, 2020) packages. We mainly focused on visualizing network results, where nodes represented the variables including mental health problems, the smartphone use statistics, and the SUPIQ factors. Edges depicted the connections between variables, with edge thickness and color saturation indicating association strength (Monteleone et al., 2022). With dealing with the missing values of the smartphone use statistics with pairwise deletion method, partial correlation networks were separately computed for the two samples (Borsboom et al., 2021; Epskamp & Fried, 2018). The stability of the edges in two networks was tested with bootstrapping (nboots = 2000). We used the “EBICglasso” algorithm, which implements the least absolute shrinkage and selection operator (LASSO) regularization method to shrink estimates and remove false positive edges. A hyperparameter of 0.5 was used to remove nonsignificant edges from the network.
The explanatory power of the SUPIQ: hierarchical multiple regression
The multivariate stepwise regression analyses used the sum score of mental health problems as the outcome variable. The first step included the SAS-SV as a predicting variable. In the second step, the SUPIQ was added to assess its additional explanatory power compared to the SAS-SV. The hierarchical multiple regression analyses were done separately in the two samples. The standardized coefficients were estimated with package “lm. beta” (Behrendt, 2022). In addition, the relative importance of the SAS-SV and the SUPIQ were compared with package “relaimpo” (Groemping, 2021).
Ethics
Participants received an information letter before starting the questionnaire and the study protocol was approved by the Ethical Review Board (ERB ID # 2021-DP-13072).
Results
Demographic information for the final samples
The data screening process is shown in Supplementary Fig. S3: the final university community sample consisted of 292 participants, and the final general population sample consisted of 397 participants. Demographic information for the two samples is shown in Table 1.
The demographic information of the university community sample (N = 292) and the general population sample (N = 397), including the Number (n), Range, Means (M) and Standard Deviations (SD)
| University community sample | General population sample | ||||||
| n (%) | Range | M (SD) | n (%) | Range | M (SD) | ||
| Age | 18–24 | 20 (1.48) | 18–75 | 28.02 (7.66) | |||
| The onset age when starting to use smartphone | 6–18 | 12.37 (1.73) | 4–71 | 17.50 (7.44) | |||
| The years for smartphone use | 3–13 | 7.64 (1.70) | 3–33 | 10.49 (3.98) | |||
| Gender* | |||||||
| Male | 70 (23.97) | 193 (48.61) | |||||
| Female | 217 (74.32) | 203 (51.13) | |||||
| Other | 5 (1.71) | 1 (0.25) | |||||
| Highest level of education | |||||||
| Primary education | – | 2 (0.50) | |||||
| Lower secondary education | – | 3 (0.76) | |||||
| Upper secondary education | 249 (85.28) | 77 (19.40) | |||||
| Post-secondary non-tertiary education | 12 (4.11) | 19 (4.79) | |||||
| Short-cycle tertiary education | 6 (2.05) | 65 (16.37) | |||||
| Bachelor’s or equivalent level | 25 (8.56) | 156 (39.29) | |||||
| Master's or equivalent level | – | 64 (16.12) | |||||
| Doctoral or equivalent level | – | 11 (2.77) | |||||
| The marital status | |||||||
| Married | 2 (0.68) | 109 (27.46) | |||||
| Engaged | 1 (0.34) | 18 (4.53) | |||||
| Living together | 14 (4.79) | 45 (11.349) | |||||
| In a relationship/having a boyfriend/girlfriend | 90 (30.82) | 96 (24.18) | |||||
| Registered partnership | – | 3 (0.76) | |||||
| Divorced | – | 6 (1.51) | |||||
| Separated | – | 5 (1.26) | |||||
| Single | 185 (63.36) | 115 (28.97) | |||||
| Country currently live in | |||||||
| Netherlands | 214 (73.29) | United States of America | 289 (72.80) | ||||
| Germany | 27 (9.25) | Netherlands | 43 (10.83) | ||||
| Poland | 4 (1.37) | Canada | 20 (5.04) | ||||
| Other countries | 47 (16.09) | Germany | 12 (3.02) | ||||
| Other countries | 33 (8.31) | ||||||
| Nationality | |||||||
| Dutch | 100 (34.25) | American | 302 (76.07) | ||||
| German | 73 (25.00) | English | 32 (8.06) | ||||
| English | 15 (5.14) | German | 19 (4.79) | ||||
| American | 12 (4.11) | Dutch | 16 (4.03) | ||||
| Italian | 11 (3.77) | Canadian | 14 (3.53) | ||||
| Other nationalities | 81 (27.74) | Italian | 7 (1.76) | ||||
| Other nationalities | 39 (9.82) | ||||||
| Whether live in home country or abroad | |||||||
| At home | 155 (53.08) | 357 (89.92) | |||||
| Abroad | 137 (46.92) | 40 (10.08) | |||||
| The years of living abroad | 137 | 0–18 | 2.21 (4.00) | 40 | 0–21 | 4.85 (4.21) | |
Note. Here is one missing value with the general population sample in terms of “The onset age when starting to use smartphone” and “The years for smartphone use” because one participant’s years for smartphone use is “−1”.* Gender here implies the participants’ self-identified gender.
The construct validity results: exploratory factor analysis (EFA) for the SUPIQ
Based on the results of the EFA with the university community sample, we reached 24 items in a 7-factor version of the SUPIQ (see Supplementary Table S4). We restored 3 items since there were only two items for 3 of the 7 factors (i.e., craving, coping, and tolerance, see Supplementary Table S5). Item 26 was removed since the removal increased the Cronbach's α values of the corresponding subscale by 0.10 (see Supplementary Tables S6 and S7; Cronbach's alpha of the full questionnaire increased by 0.01). The final version of the SUPIQ includes 26 items with 7 factors (see Table 2). The 7 factors can be described as follows: Craving, Coping, Habitual Use, Social Conflicts, Risky Use, Withdrawal, Tolerance. The model fit indices showed a good fit: CFI = 0.993, TLI = 0.986, RMSEA = 0.030, SRMR = 0.031. The network analysis also showed the 7-factor structure well fitted the data (see Supplementary Fig. S10), the centrality indices (e.g., closeness, strength, betweenness) of the network were shown in Supplementary Figs S11 and S12.
Factor Loadings of the final 26 items (adding three extra items, item) based on EFA with the university community sample (N = 292)
| Final order | Initial order | Item | Factor 1-Craving | Factor 2-Tolerance | Factor 3-Social Conflicts | Factor 4- Withdrawal | Factor 5- Risky Use | Factor 6-Habitual Use | Factor 7-Coping |
| 2 | 2 | I feel a strong urge to check my smartphone. | 0.631 | ||||||
| 1 | 1 | I think about my smartphone when I am not using it. | 0.624 | ||||||
| 19 | 35 | I feel a strong need to be available via my smartphone. | 0.275 | 0.302 | 0.287 | ||||
| 3 | 3 | I feel empty even when I spend a lot of time on my smartphone. | 0.867 | ||||||
| 15 | 27 | I feel unsatisfied even when I spend a lot of time on my smartphone. | 0.532 | ||||||
| 7 | 14 | I need to spend more and more time on my smartphone to satisfy myself. | 0.346 | 0.298 | |||||
| 9 | 16 | I tell lies about my smartphone use. | 0.864 | ||||||
| 12 | 21 | I hide my smartphone use from others (e.g., family, partner, friend, etc.). | 0.812 | ||||||
| 10 | 19 | I have conflicts with others (e.g., family, partner, friend, etc.) due to my smartphone use. | 0.780 | ||||||
| 14 | 25 | I jeopardize important relationships (e.g., family, partner, friend, etc.) due to my smartphone use. | 0.745 | ||||||
| 25 | 55 | People around me tell me that I use my smartphone too much. | 0.674 | ||||||
| 8 | 15 | I continue using my smartphone after others ask me not to. | 0.541 | ||||||
| 6 | 10 | I feel anxious/nervous when there are internet connection problems (e.g., unstable connection, no connection, etc.) on my smartphone. | 0.851 | ||||||
| 22 | 46 | I feel anxious/nervous when I do not have access to my smartphone (e.g., exams, out of battery, etc.). | 0.801 | ||||||
| 20 | 42 | I feel angry when there are internet connection problems (e.g., unstable connection, no connection, etc.) on my smartphone. | 0.759 | ||||||
| 13 | 22 | I feel angry when I do not have access to my smartphone (e.g., exams, out of battery, etc.). | 0.654 | ||||||
| 24 | 53 | I still use my smartphone in traffic (e.g., driving a car, cycling, walking, etc.) even though I (almost) get into traffic accidents due to my smartphone use. | 0.906 | ||||||
| 4 | 5 | I (almost) get into traffic accidents (e.g., when driving a car, when cycling, when walking, etc.) due to my smartphone use. | 0.858 | ||||||
| 16 | 29 | I use my smartphone in situations that could be physically dangerous (e.g., driving a car, cycling, crossing the road, operating heavy machinery, etc.). | 0.798 | ||||||
| 17 | 30 | I automatically open apps on my smartphone. | 0.910 | ||||||
| 23 | 48 | I automatically unlock my smartphone. | 0.860 | ||||||
| 18 | 31 | I automatically check my smartphone, even when I just checked it. | 0.686 | ||||||
| 26 | 57 | I check my smartphone when I am entertaining myself in other ways (e.g., watching movies, TV series, reading, etc.). | 0.576 | ||||||
| 5 | 7 | Using my smartphone makes me feel better when I feel bad (e.g., sad, anxious, insecure, lonely, etc.). | 0.721 | ||||||
| 21 | 43 | I distract myself from negative feelings (e.g., sad, anxious, insecure, lonely, etc.) by using my smartphone. | 0.645 | ||||||
| 11 | 20 | My smartphone is the solution to my boredom. | 0.316 | 0.369 |
Note. SUPIQ = the Smartphone Use Problems Identification Questionnaire. This version contains 26 items, 7 factors. Applied rotation method is oblimin in EFA. Only factor loadings ≥ 0.250 were listed in the table. The model fits were good: CFI = 0.991, TLI = 0.983, RMSEA = 0.032, SRMR = 0.031. The three restored items were marked in bold. The 7-factor structure was also supported by parallel analyses and the scree plot (see supplementary Fig. S8). The Cronbach's α values of the three factors when excluding the restored three items are shown in supplementary Table S9.
The construct validity results: confirmatory factor analysis (CFA) for the SUPIQ and the SAS-SV
Results of the CFA (CFI = 0.970, TLI = 0.965, RMSEA = 0.072, SRMR = 0.047) for the SUPIQ indicated that there is a residual covariance between item 1 (i.e., I think about my smartphone when I am not using it.) and item 15 (i.e., I continue using my smartphone after others ask me not to.) of the SUPIQ. After adding this covariance, the model fit indices of the CFA in the general population sample indicated an acceptable to good model fit (CFI = 0.971, TLI = 0.966, RMSEA = 0.070, SRMR = 0.047). The standardized factors loadings and correlations between factors of the SUPIQ based on the CFA were presented in Supplementary Tables S13 and S14.
Results of the CFA of the SAS-SV in the two samples were as follows: CFI = 0.959, TLI = 0.932, RMSEA = 0.064, SRMR = 0.043 in the university community sample; CFI = 0.949, TLI = 0.915, RMSEA = 0.097, SRMR = 0.040 in the general population sample.
Psychometric equivalence of SUPIQ: measurement invariance test results
See Table 3 for the results concerning the measurement invariance analyses. The SUPIQ demonstrated configural invariance with acceptable model fit of Model 1. This indicated that the general structure of SUPIQ was consistent across the two samples. However, according to the criteria of changed CFI (decrease of ≥0.015) and RMSEA (increase of ≥0.015), both metric and scalar invariance were not tenable, as the model fit indices deteriorate, suggesting that some items' factor loadings and thresholds in the SUPIQ differed in the two samples (Chen, 2007). We consulted modification indices to see if there were some specific items responsible for the misfit. However, there was no clear source of misfit identifiable.
Measurement invariance analysis of the SUPIQ: Multi-group CFA
| Model test | Overall model fit constrained model | Changes in model fit | |||||||||
| χ2 | df | BIC | CFI | TLI | RMSEA | SRMR | ∆χ2 | ∆df | ∆CFI | ∆RMSEA | |
| Group equivalence | 1,036.20 | 278 | 44,478.73 | 0.923 | 0.911 | 0.063 | 0.061 | – | – | – | – |
| Configural (Model 1) | 1,153.52 | 556 | 43,708.69 | 0.932 | 0.920 | 0.056 | 0.053 | – | – | – | – |
| Metric (Model 2) | 1,348.63 | 575 | 43,779.63 | 0.912 | 0.900 | 0.062 | 0.076 | 195.112*** | 19 | −0.020 | 0.006 |
| Scalar (Model 3) | 1,723.33 | 594 | 44,030.16 | 0.871 | 0.859 | 0.074 | 0.082 | 374.699*** | 19 | −0.041 | 0.012 |
Note. SUPIQ = the Smartphone Use Problems Identification Questionnaire; BIC = Bayesian information criterion; CFI = comparative fit index; TLI = Tucker–Lewis index; RMSEA = root-mean-square error of approximation; Difference tests are displayed for configural versus metric and metric versus scalar, respectively. ***χ2 test was statistically significant at p < 0.001 level.
The descriptive statistics, reliability, and convergent validity results
The descriptive statistics including Means (M), Standard Deviations (SD), Range, and Reliability Coefficients including Cronbach's α and McDonald's ω values for mental health problems, the SAS-SV, the sum scores of SUPIQ, and its Seven Factors are shown in Table 4. The reliability coefficients indicated that SUPIQ has good reliability.
Summary of Means (M), Standard Deviations (SD), Range, and Reliability Coefficients for mental health problems, the SAS-SV, the SUPIQ and its Seven Factors in the university community sample (N = 292) and general population sample (N = 397)
| University community sample (N = 292) | General population sample (N = 397) | |||||||||
| M (SD) Sum score | M (SD) Item average | Range Sum score | Cronbach's α | McDonald's ω | M (SD) Sum score | M (SD) Item average | Range Sum score | Cronbach's α | McDonald's ω | |
| Mental health problems | 18.70 (10.14) | 0.94 (0.51) | 0–55 | 0.89 | 0.90 | 25.83 (17.22) | 1.29 (0.86) | 0–63 | 0.97 | 0.97 |
| SAS-SV | 27.54 (8.06) | 2.75 (0.81) | 10–50 | 0.83 | 0.86 | 33.16 (10.32) | 3.32 (1.03) | 10–57 | 0.91 | 0.93 |
| SUPIQ-Total | 56.65 (13.21) | 2.18 (0.51) | 34–103 | 0.90 | 0.92 | 66.85 (18.30) | 2.57 (0.70) | 26–104 | 0.94 | 0.96 |
| SUPIQ-Craving | 8.09 (2.44) | 2.70 (0.81) | 3–14 | 0.73 | 0.77 | 8.93 (2.48) | 2.98 (0.83) | 3–15 | 0.67 | 0.68 |
| SUPIQ-Coping | 7.76 (2.31) | 2.59 (0.77) | 3–14 | 0.68 | 0.71 | 8.31 (2.54) | 2.77 (0.85) | 3–15 | 0.69 | 0.70 |
| SUPIQ-Habitual Use | 12.60 (3.68) | 3.15 (0.92) | 5–20 | 0.85 | 0.88 | 11.52 (3.22) | 2.88 (0.80) | 4–20 | 0.75 | 0.78 |
| SUPIQ-Social Conflicts | 7.98 (2.99) | 1.33 (0.50) | 6–25 | 0.85 | 0.90 | 13.47 (5.52) | 2.25 (0.92) | 6–26 | 0.91 | 0.94 |
| SUPIQ-Risky Use | 4.34 (1.70) | 1.45 (0.57) | 3–12 | 0.76 | 0.79 | 6.44 (2.94) | 2.15 (0.98) | 3–14 | 0.83 | 0.84 |
| SUPIQ-Withdrawal | 8.28 (3.41) | 2.07 (0.85) | 4–20 | 0.83 | 0.87 | 10.27 (3.41) | 2.57 (0.85) | 4–20 | 0.81 | 0.84 |
| SUPIQ-Tolerance | 7.59 (2.70) | 2.53 (0.90) | 3–15 | 0.69 | 0.70 | 7.91 (2.55) | 2.64 (0.85) | 3–14 | 0.68 | 0.68 |
Note. SUPIQ = the Smartphone Use Problems Identification Questionnaire, SAS-SV = the short version of smartphone addiction scale. For t values, *p < 0.05, **p < 0.01, ***p < 0.001.
The descriptive statistics of the smartphone use statistics across two samples are presented in Table 5.
Summary of Means (M), Standard Deviations (SD) and Range for the smartphone use statistics in the university community sample (N = 292) and the general population sample (N = 397)
| University community sample (N = 275/292) | General population sample (N = 315/397) | t | |||
| M (SD) | Range | M (SD) | Range | ||
| Total screen time per day (hours) | 4.72 (2.30) | 0.43–11.88 | 7.19 (3.89) | 0.08–19.36 | −11.37*** |
| Non-social screen time per day (hours) | 2.24 (1.74) | 0.06–10.55 | 3.14 (2.19) | 0–12.62 | −6.45*** |
| Active social media use frequency per day | 64.38 (49.50) | 0–268.14 | 94.53 (72.49) | 0–314.71 | −7.03*** |
| Passive social media use frequency per day | 92.93 (71.13) | 0–357.14 | 94.79 (70.62) | 0–361.71 | −0.36 |
| Pickup frequency per day | 71.34 (40.94) | 0.43–142.86 | 52.82 (28.44) | 4.14–142.86 | 6.99*** |
| Top 3 social media apps (frequency) | WhatsApp (51), Instagram (47), YouTube (16), SnapChat (13), Reddit (9), TikTok (6), Twitter (6) | Facebook (211), Instagram (206), TikTok (171), WhatsApp (125), LinkedIn (62), Twitter (30), YouTube (14), SnapChat (13), Messages (11), Spotify (8) | |||
| Top 3 pickup apps (frequency) | WhatsApp (57), Instagram (34), YouTube (19), Chrome (17), SnapChat (12), Spotify (10) | Instagram (191), WhatsApp (150), Facebook (120), TikTok (92), Safari (75), Messages (25), LinkedIn (24), Spotify (16), Twitter (16), Chrome (14) | |||
Note. For the numeric data, the missing values have been considered. For the top 3 apps, we have used the whole samples to do the data analyses. For t values, *p < 0.05, **p < 0.01, ***p < 0.001.
Regarding convergent validity results, the Correlation Coefficients for mental health problems, the smartphone use statistics, the SAS-SV, the SUPIQ, and its Seven Factors are displayed in Table 6. Based on the results, all the factors and scales were highly positively correlated with each other in the general population sample (p < 0.001). However, in the university community sample, Risky Use was not correlated with Craving and Coping, and the correlation coefficients between Risky Use and Habitual Use, Social Conflicts, Withdrawal, and Tolerance were relatively lower than those between other factors of the SUPIQ. The total scores of the SUPIQ were positively correlated with mental health problems, SAS-SV, active and passive social media use on smartphone in both samples. Mental health problems were positively correlated with all the SUPIQ factors in both samples. In the university community sample, the SUPIQ factors including Craving, Habitual Use, Social Conflicts, Withdrawal, and Tolerance were positively correlated with active social media use on smartphone, while Craving, Coping, Habitual Use, and Withdrawal were positively correlated with passive social media use on smartphone. In the general population sample, the SUPIQ factors including Social Conflicts, Risky Use, and Tolerance were positively correlated with active social media use on smartphone, while Coping, Social Conflicts, Risky Use, Withdrawal, and Tolerance were positively correlated with passive social media use on smartphone. In the university community sample, the total score of the SUPIQ, Craving, Coping and Social Conflicts were positively correlated with non-social screen time per day. The results indicated that the SUPIQ has good convergent validity generally.
Correlation Coefficients between Mental health problems, the SAS-SV, the SUPIQ and its Seven Factors, and smartphone use statistics in the university community sample (N = 292) and the general population sample (N = 397, in bold)
| Variables | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 |
| 1. Mental health problems | 1 | 0.71*** | 0.78*** | 0.51*** | 0.59*** | 0.47*** | 0.73*** | 0.72*** | 0.62*** | 0.65*** | −0.04 | 0.04 | 0.22** | 0.21** | −0.10 |
| 2. SAS-SV | 0.46*** | 1 | 0.77*** | 0.58*** | 0.62*** | 0.53*** | 0.69*** | 0.55*** | 0.69*** | 0.64*** | −0.03 | 0.04 | 0.19** | 0.17* | −0.10 |
| 3. SUPIQ-Total | 0.52*** | 0.77*** | 1 | 0.74*** | 0.77*** | 0.72*** | 0.87*** | 0.78*** | 0.86*** | 0.83*** | −0.10 | 0.02 | 0.22** | 0.22** | −0.14 |
| 4. SUPIQ-Craving | 0.33*** | 0.67*** | 0.78*** | 1 | 0.60*** | 0.64*** | 0.49*** | 0.38*** | 0.63*** | 0.58*** | −0.13 | −0.06 | 0.12 | 0.14 | −0.08 |
| 5. SUPIQ-Coping | 0.39*** | 0.48*** | 0.67*** | 0.49*** | 1 | 0.66*** | 0.55*** | 0.44*** | 0.61*** | 0.64*** | −0.11 | 0.02 | 0.15 | 0.18* | −0.04 |
| 6. SUPIQ-Habitual Use | 0.30*** | 0.60*** | 0.78*** | 0.64*** | 0.49*** | 1 | 0.41*** | 0.37*** | 0.57*** | 0.57*** | −0.05 | 0.02 | −0.01 | 0.09 | −0.12 |
| 7. SUPIQ-Social Conflicts | 0.31*** | 0.53*** | 0.66*** | 0.39*** | 0.25*** | 0.31*** | 1 | 0.82*** | 0.71*** | 0.68*** | −0.09 | 0.04 | 0.28*** | 0.22*** | −0.16 |
| 8. SUPIQ-Risky Use | 0.20** | 0.19** | 0.34*** | 0.11 | 0.05 | 0.18** | 0.28*** | 1 | 0.59*** | 0.60*** | −0.06 | 0.001 | 0.24*** | 0.20** | −0.19 |
| 9. SUPIQ-Withdrawal | 0.46*** | 0.60*** | 0.75*** | 0.54*** | 0.50*** | 0.43*** | 0.41*** | 0.15* | 1 | 0.68*** | −0.07 | 0.02 | 0.14 | 0.17* | −0.08 |
| 10. SUPIQ-Tolerance | 0.45*** | 0.49*** | 0.67*** | 0.44*** | 0.37*** | 0.43*** | 0.41*** | 0.14* | 0.36*** | 1 | −0.07 | 0.02 | 0.20** | 0.19** | −0.07 |
| 11. Total screen time per day (hours) | 0.03 | 0.07 | 0.11 | 0.11 | 0.09 | 0.06 | 0.07 | 0.04 | 0.08 | 0.07 | 1 | 0.58*** | 0.002 | −0.002 | 0.20** |
| 12. Non-social screen time per day (hours) | 0.08 | 0.14* | 0.16** | 0.12* | 0.16** | 0.10 | 0.12* | 0.04 | 0.11 | 0.11 | 0.66*** | 1 | 0.08 | 0.05 | 0.04 |
| 13. Active social media use frequency per day | 0.12* | 0.28*** | 0.25*** | 0.27*** | 0.10 | 0.19** | 0.15* | −0.06 | 0.27*** | 0.14* | 0.04 | 0.03 | 1 | 0.84*** | 0.01 |
| 14. Passive social media use frequency per day | 0.03 | 0.27*** | 0.21*** | 0.29*** | 0.15* | 0.29*** | −0.05 | −0.08 | 0.19** | 0.10 | 0.09 | 0.07 | 0.63*** | 1 | 0.004 |
| 15. Pickup frequency per day | −0.04 | 0.05 | 0.07 | 0.04 | 0.01 | 0.03 | 0.15** | 0.08 | 0.004 | 0.01 | 0.37*** | 0.20** | 0.09 | 0.03 | 1 |
Note. SUPIQ = the Smartphone Use Problems Identification Questionnaire, SAS-SV = the short version of smartphone addiction scale. *p < 0.05, **p < 0.01, ***p < 0.001.
Numbers of missing values in smartphone use statistics are separately 17 in the university community sample and 82 in the general population sample, the method to deal with the missing values is pairwise deletion.
The partial correlation network with the university community sample demonstrated that Tolerance and Withdrawal were positively correlated with mental health problems (Fig. 1). In addition, passive social media use was negatively associated with Social Conflicts while positively associated with Habitual Use. The correlation stability coefficient was 0.671, indicating sufficient stability (see Supplementary Fig. S15 for bootstrapping results). The centrality indices of the network were shown in Supplementary Figs S16 and S17. The estimation of significant differences between edge weights and bootstrapped results were shown in Supplementary Fig. S18.
The network analysis including all the sum scores of the SUPIQ factors, mental health problems and all smartphone use indicators (regularized model) with the university community sample (N = 292)
Note. The color (blue = positive, red = negative) and thickness of the edges indicate the strength of association. The nodes have been colored according to the domain that they belong to. SUPIQ = The Smartphone Use Problems Identification Questionnaire.
Citation: Journal of Behavioral Addictions 13, 2; 10.1556/2006.2024.00010
The partial correlation network with the general population sample showed Craving, Social Conflicts, Risky Use, Tolerance, and Withdrawal were positively correlated with mental health problems (Fig. 2). Active social media use was negatively associated with Habitual Use and positively associated with Social Conflicts. Moreover, passive social media use was positively associated with Habitual Use. The correlation stability coefficient was 0.751, indicating sufficient stability (see Supplementary Fig. S19 for bootstrapping results). The centrality indices of the network were shown in Supplementary Figs S20 and S21. The estimation of significant differences between edge weights and bootstrapped results were shown in Supplementary Fig. S22.
The network analysis including all the sum scores of the SUPIQ factors, mental health problems and all smartphone use indicators (regularized model) with the general population sample (N = 397)
Note. The color (blue = positive, red = negative) and thickness of the edges indicate the strength of association. The nodes have been colored according to the domain that they belong to. SUPIQ = The Smartphone Use Problems Identification Questionnaire.
Citation: Journal of Behavioral Addictions 13, 2; 10.1556/2006.2024.00010
The explanatory power of the SUPIQ: hierarchical multiple regression results
Hierarchical multiple regression analyses showed that the SUPIQ generally outperformed the SAS-SV in the regression models of step 2 when the SUPIQ and the SAS-SV were added to the regression model as predictors for mental health problems at the same time (Table 7). Based on the results of relative importance analyses (Table 8), the SUPIQ played a more important role than the SAS-SV when explaining the variance of mental health problems with step 2 models.
Hierarchical multiple regression results with the two samples
| Independent variables | University community sample (N = 292) | General population sample (N = 397) | ||||||
| B | β | R2 | SE | B | β | R2 | SE | |
| Step 1 | 0.210 | |||||||
| SAS-SV | 0.576*** | 0.458*** | 0.066 | 1.192*** | 0.714*** | 0.510 | 0.059 | |
| Step 2 | 0.277 | 0.642 | ||||||
| SAS-SV | 0.179 | 0.142 | 0.099 | 0.453*** | 0.272*** | 0.079 | ||
| SUPIQ | 0.314*** | 0.409*** | 0.060 | 0.539*** | 0.573*** | 0.045 | ||
Note. SUPIQ = the Smartphone Use Problems Identification Questionnaire; SAS-SV = the short version of smartphone addiction scale. Dependent variable = mental health problems. The possible multicollinearity problems have been checked by VIF (variance inflation factor), the VIF values of SAS-SV and SUPIQ were 2.474 for the university community sample and 2.481 for the general population sample. These values are lower than the universal criterion of 10 when detecting multicollinearity (Vatcheva, Lee, McCormick, & Rahbar, 2016). *p < 0.05, **p < 0.01, ***p < 0.001.
Relative importance and 95% confidence interval indicated by relative contribution percentages (%) of the SAS-SV and the SUPIQ in the regression model of step 2
| Attribute | University community sample (N = 292) | General population sample (N = 397) | ||||||
| LMG | LAST | FIRST | PRATT | LMG | LAST | FIRST | PRATT | |
| SAS-SV | 39.30 [27.42, 52.53] | 10.82 [0.04,61.91] | 43.80 [35.19, 51.44] | 23.52 [−5.75, 56.70] | 42.02 [36.87, 47.06] | 18.36 [6.30,36.32] | 45.44 [42.24, 48.35] | 30.20 [18.12, 42.20] |
| SUPIQ | 60.70 [47.47, 72.58] | 89.18 [38.09, 99.96] | 56.20 [48.56, 64.81] | 76.48 [43.30, 105.75] | 57.98 [52.94, 63.13] | 81.64 [63.68, 93.70] | 54.56 [51.65, 57.76] | 69.80 [57.80, 81.88] |
Note. SUPIQ = the Smartphone Use Problems Identification Questionnaire; SAS-SV = the short version of smartphone addiction scale. Dependent variable = mental health problems, n of bootstrap = 2000. LMG, LAST, FIRST, PTATT are the four different estimation methods offered by package "relaimpo". 27.7% and 64.24% of mental health problems' variances could be explained in university community and general population samples separately.
Discussion
Derived from a qualitative study involving problematic smartphone users from diverse European countries, we have gained contemporary and comprehensive perspectives on PSU in everyday life, which moves beyond existing “addiction” criteria. Based on such perspectives, we developed and tested the comprehensive Smartphone Use Problems Identification Questionnaire (SUPIQ) to assess PSU. The SUPIQ exhibited good construct and convergent validity, reliability, and explanatory power in both samples. The initial version of the questionnaire comprised 8 factors and 57 items based on DSM-5 (American Psychiatric Association, 2013a) and ICD-11 (World Health Organization, 2019) criteria, after exploratory and confirmatory factor analyses, the final SUPIQ consists of 7 correlated factors: Craving, Coping, Habitual Use, Social Conflicts, Risky Use, Withdrawal, Tolerance. The final SUPIQ showed good construct validity and reliability, while the structure is different from the initial version.
The final inclusion of Craving and Coping factors aligns with the established scales like MPPUS (Bianchi & Phillips, 2005) and SPAI (Lin et al., 2014), with adaptions based on the descriptions from the qualitative study on problematic smartphone users. Impaired Control items were included in the Habitual Use factor. This aligns with previous studies (e.g., Van Deursen, Bolle, Hegner, & Kommers, 2015), indicating that Habitual Use may be a more critical factor than control-related problems in measuring PSU. This factor is distinctive compared to existing questionnaires like MPPUS (Bianchi & Phillips, 2005), PMPUQ (Billieux et al., 2008), SAS (Kwon, Lee, et al., 2013), and SPAI (Lin et al., 2014), which emphasize control issues. In the SUPIQ, the Social Conflicts and Risky Use factors retained as the main negative consequences. Risky Use measurement has been expanded to include items on frequency of real-life accidents and reluctance to change after, accidents deviating from the Dangerous Use factor of PMPUQ (Billieux et al., 2008). The retainment of social and physical risks is different from other questionnaires that encompass broader negative consequences, including MPPUS (Bianchi & Phillips, 2005), PMPUQ (Billieux et al., 2008), SAS (Kwon, Lee, et al., 2013), and SPAI. This difference may imply the significance of the two aspects to differentiate PSU from normal smartphone use. However, it is worth noting that Risky Use showed weaker correlations with other factors in the university community sample, similar to the findings in problematic cannabis use (see a review, Casajuana et al., 2016). This suggests that the Risky Use items may need optimization for different contexts, aligning with the findings related to problem drinking, where the specific scale has been designed for adolescents (White & Labouvie, 1989; White, Labouvie, & Papadaratsakis, 2005). While the inclusion of Tolerance and Withdrawal in defining behavioral addictions remains debated (e.g., Starcevic, 2016), both factors were incorporated in the final SUPIQ based on participants' everyday experiences. The inclusion of the two factors are consistent with the existing questionnaires (Bianchi & Phillips, 2005; Kwon, Lee, et al., 2013; Lin et al., 2014), expands their measurement. Tolerance of the SUPIQ involves feeling “empty” after spending much time on smartphones, while Withdrawal covers both issues with the internet connection and smartphone access. In terms of the covariances between item 1 and item 15, the correlation can be explained by the notion that individuals who genuinely contemplate and crave for their smartphones are more likely to overlook or disregard requests from others to stop using their smartphones.
While the configural invariance of the SUPIQ was established across the university community and general population samples, other forms of measurement invariances were not. As noted, it is not uncommon for addictive behaviors to show different indicators of problems at different ages (c.f., White & Labouvie, 1989). Results showed that the strength of the relationships between items and the underlying construct like the factor loadings of the SUPIQ, differed across the two samples.
Regarding convergent validity, the linear correlation analyses showed that the SUPIQ and its factors were positively correlated with mental health problems, the SAS-SV, indicating the good convergent validity of the SUPIQ. The partial correlation network analyses showed that mental health problems were positively correlated with Withdrawal and Tolerance in both samples and positively correlated with Coping, Social Conflicts, and Risky Use in the general population sample. Such results are consistent with previous studies (see a review, S. Sohn, Rees, Wildridge, Kalk, & Carter, 2019). These results suggest that Habitual Use and Craving might not be key factors in defining PSU or central symptoms of PSU (e.g., Fournier et al., 2023). Total screen time, nonsocial screen time, and pickup frequency were not correlated with the SUPIQ factors, contrary to prior research (see a review, Ryding & Kuss, 2020). This discrepancy may be due to our inclusion of mental health problems in the analyses, which was not common in previous studies (see a review, Ryding & Kuss, 2020). Active social media use was positively correlated with Social Conflicts in both samples and negatively correlated with Habitual Use in the general population sample. Passive social media use was positively correlated with Habitual Use in both samples and negatively correlated with Social Conflicts in the university community sample. These results align with former studies (e.g., Su et al., 2022), suggesting different roles of active and passive social media use in relation to the PSU. Regarding explanatory power, the SUPIQ outperformed the SAS-SV in the regression models where the mental health problems were the outcome variables for both samples.
Limitations and future study directions
While our study provides valuable insights into the initial psychometric features of the SUPIQ, there is a need for further investigation to deepen our understanding of PSU. We comprehensively examined construct and convergent validity, reliability, measurement invariance, and explanatory power using samples primarily from Western, Educated, Industrialized, Rich, and Democratic (WEIRD) countries. However, the volunteerism of participants and the focus on WEIRD countries with English as the primary survey language impacted the representativeness of the samples. To address this limitation, future research should target larger and more diverse samples, including clinical samples, while considering the different languages to assess the broader applicability of the SUPIQ globally (Billieux, Philippot, et al., 2015; Billieux, Schimmenti, et al., 2015; Flayelle et al., 2022; Lopez-Fernandez, 2017). Further refinements are warranted, particularly in the items of Craving, Coping, and Tolerance. The tailoring of the SUPIQ for different populations is crucial, and assessing the relative significance of different factors in defining PSU remains an important avenue for future research. Additionally, the test-retest reliability can also be further investigated with future studies and a short form of the SUPIQ should be generated to ensure the feasibility of usage and compare the short form with the SAS-SV in future studies (c.f. Kwon, Kim, et al., 2013).
In our study, we tested the SUPIQ's convergent validity with smartphone use statistics, mental health problems, and the SAS-SV. Future research could explore the overlap between the SUPIQ and other behavioral and substance-related addictions (e.g., Andrade et al., 2022; Kwon, Lee, et al., 2013) or the possible correlations between the SUPIQ and other variables like personality traits (Giustiniani et al., 2022) and childhood trauma (Fan et al., 2023). We have also tested the explanatory power of the SUPIQ in comparison to the SAS-SV (Kwon, Kim, et al., 2013), while future studies would benefit from employing the full versions of the PMPUQ (Billieux et al., 2008) and SAS (Kwon, Lee, et al., 2013) to confirm the predictive power of the complete SUPIQ. Additionally, individual differences in the SUPIQ warrant investigation through person-centered analyses, such as latent profile analysis with cross-sectional data (e.g., Yue et al., 2021) and growth mixture modeling with longitudinal data (e.g., Lai et al., 2022), to identify the problematic subgroups.
Conclusions
To conclude, we tested the psychometric quality of a newly developed instrument to measure problematic smartphone use, namely the Smartphone Use Problems Identification Questionnaire (SUPIQ). Throughout this process, the SUPIQ underwent refinements in different aspects, including the adaptation of items within factors like Craving and Coping, identification of pivotal negative consequences related to Social Conflicts and Risky Use, expanded measurements for Risky Use, Withdrawal, and Tolerance factors, and the introduction of a novel factor-Habitual Use. The SUPIQ shows good reliability and validity, and it is a valuable tool for evaluating contemporary and severe smartphone use problems rooted in users' everyday behaviors. The SUPIQ stands as an updated and robust tool, aiming to contribute significantly to the nuanced assessment of problematic smartphone use.
Funding sources
The first author is funded by China Scholarship Council (CSC) from the Ministry of Education of P.R. China.
Authors' contribution
SS: conceptualization, methodology, software, formal analysis, investigation, data curation, writing - original draft, writing - review & editing, visualization, project administration; JC: conceptualization, methodology, investigation, writing - review & editing, supervision; DM: methodology, validation, formal analysis, writing - review & editing; RF: methodology, formal analysis, writing - review & editing; HL: conceptualization, methodology, investigation, writing - review & editing, supervision; RWW: conceptualization, methodology, formal analysis, writing - review & editing, supervision.
Conflict of interest
Authors declare that they have no conflicts of interest.
Acknowledgements
We want to thank all the participants who contributed their experiences in current study, especially those who left some inspiring comments after completion. Thanks for the help of Margaret O’Day and Johanna Gerstein during data collection.
Supplementary material
Supplementary data to this article can be found online at https://doi.org/10.1556/2006.2024.00010.
References
Agus, M., Mascia, M. L., Bonfiglio, N. S., & Penna, M. P. (2022). The Italian version of the mobile phone problematic use scale for adults (MPPUS): A validation study. Heliyon, 8(12), e12209. https://doi.org/10.1016/j.heliyon.2022.e12209.
Allaire, J. (2022). RStudio: Integrated development environment for R [Computer software]. Citeseer.
American Psychiatric Association (2013a). Diagnostic and statistical manual of mental disorders: DSM-5. American Psychological Association.
American Psychiatric Association (2013b). DSM-5 self-rated level 1 cross-cutting symptom measure—Adult. In Diagnostic and statistical manual of mental disorders: DSM-5 (Vol. 5, pp. 734–739).
Andrade, A. L. M., Spritzer, D. T., Scatena, A., de Oliveira Pinheiro, B., da Silva, G. T., Kim, H. S., … De Micheli, D. (2022). Psychometric properties of the smartphone addiction inventory-short form (SPAI-SF) in Brazilian adolescents. Psychiatry Research, 115001. https://doi.org/10.1016/j.psychres.2022.115001.
Arrivillaga, C., Rey, L., & Extremera, N. (2020). Adolescents’ problematic internet and smartphone use is related to suicide ideation: Does emotional intelligence make a difference? Computers in Human Behavior, 110(March), 106375. https://doi.org/10.1016/j.chb.2020.106375.
Behrendt, S. (2022). Lm. Beta: Add standardized regression coefficients to lm-objects (1.6-2) [Computer software].
Bianchi, A., & Phillips, J. G. (2005). Psychological predictors of problem mobile phone use. CyberPsychology & Behavior, 8(1), 39–51. https://doi.org/10.1089/cpb.2005.8.39.
Billieux, J., der Linden, M., & Rochat, L. (2008). The role of impulsivity in actual and problematic use of the mobile phone. Applied Cognitive Psychology: The Official Journal of the Society for Applied Research in Memory and Cognition, 22(9), 1195–1210. https://doi.org/10.1002/acp.1429.
Billieux, J., Maurage, P., Lopez-Fernandez, O., Kuss, D. J., & Griffiths, M. D. (2015). Can disordered mobile phone use be considered a behavioral addiction? An update on current evidence and a comprehensive model for future research. Current Addiction Reports, 2(2), 156–162. https://doi.org/10.1007/s40429-015-0054-y.
Billieux, J., Philippot, P., Schmid, C., Maurage, P., De Mol, J., & Van der Linden, M. (2015). Is dysfunctional use of the mobile phone a behavioural addiction? Confronting symptom-based versus process-based approaches. Clinical Psychology and Psychotherapy, 22(5), 460–468. https://doi.org/10.1002/cpp.1910.
Billieux, J., Schimmenti, A., Khazaal, Y., Maurage, P., & Heeren, A. (2015). Are we overpathologizing everyday life? A tenable blueprint for behavioral addiction research. Journal of Behavioral Addictions, 4(3), 119–123. https://doi.org/10.1556/2006.4.2015.009.
Borsboom, D., Deserno, M. K., Rhemtulla, M., Epskamp, S., Fried, E. I., McNally, R. J., … Waldorp, L. J. (2021). Network analysis of multivariate data in psychological science. Nature Reviews Methods Primers, 1(1), 58. https://doi.org/10.1038/s43586-021-00055-w.
Brailovskaia, J., Stirnberg, J., Rozgonjuk, D., Margraf, J., & Elhai, J. D. (2021). From low sense of control to problematic smartphone use severity during Covid-19 outbreak: The mediating role of fear of missing out and the moderating role of repetitive negative thinking. Plos One, 16(12 December), 1–17. https://doi.org/10.1371/journal.pone.0261023.
Brown, T. A. (2015). Confirmatory factor analysis for applied research. Guilford publications.
Busch, P. A., Hausvik, G. I., Ropstad, O. K., & Pettersen, D. (2021). Smartphone usage among older adults. Computers in Human Behavior, 121, 106783. https://doi.org/10.1016/j.chb.2021.106783.
Casajuana, C., López-Pelayo, H., Balcells, M. M., Miquel, L., Colom, J., & Gual, A. (2016). Definitions of Risky and problematic cannabis use: A systematic review. Substance Use & Misuse, 51(13), 1760–1770. https://doi.org/10.1080/10826084.2016.1197266.
Chen, F. F. (2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal, 14(3), 464–504. https://doi.org/10.1080/10705510701301834.
Cheung, T., Lee, R. L. T., Tse, A. C. Y., Do, C. W., So, B. C. L., Szeto, G. P. Y., & Lee, P. H. (2019). Psychometric properties and demographic correlates of the smartphone addiction scale-short version among Chinese children and adolescents in Hong Kong. Cyberpsychology, Behavior, and Social Networking, 22(11), 714–723. https://doi.org/10.1089/cyber.2019.0325.
Cho, S., & Lee, E. (2015). Development of a brief instrument to measure smartphone addiction among nursing students. CIN - Computers Informatics Nursing, 33(5), 216–224. https://doi.org/10.1097/CIN.0000000000000132.
DeSimone, J. A., Harms, P. D., & DeSimone, A. J. (2015). Best practice recommendations for data screening: Data screening. Journal of Organizational Behavior, 36(2), 171–181. https://doi.org/10.1002/job.1962.
Ding, D., & Li, J. (2017). Smartphone overuse – A growing public health issue. Journal of Psychology & Psychotherapy, 7(1). https://doi.org/10.4172/2161-0487.1000289.
Ding, J. E., Liu, W., Wang, X., Lan, Y., Hu, D., Xu, Y., … Fu, H. (2019). Development of A smartphone overuse classification scale. Addiction Research and Theory, 27(2), 150–155. https://doi.org/10.1080/16066359.2018.1474204.
DiStefano, C., & Morgan, G. B. (2014). A comparison of diagonal weighted least squares robust estimation techniques for ordinal data. Structural Equation Modeling: A Multidisciplinary Journal, 21(3), 425–438. https://doi.org/10.1080/10705511.2014.915373.
Elhai, J. D., & Contractor, A. A. (2018). Examining latent classes of smartphone users: Relations with psychopathology and problematic smartphone use. Computers in Human Behavior, 82, 159–166. https://doi.org/10.1016/j.chb.2018.01.010.
Elhai, J. D., Gallinari, E. F., Rozgonjuk, D., & Yang, H. (2020). Depression, anxiety and fear of missing out as correlates of social, non-social and problematic smartphone use. Addictive Behaviors, 105(February), 106335. https://doi.org/10.1016/j.addbeh.2020.106335.
Elhai, J. D., Levine, J. C., Dvorak, R. D., & Hall, B. J. (2017). Non-social features of smartphone use are most related to depression, anxiety and problematic smartphone use. Computers in Human Behavior, 69, 75–82. https://doi.org/10.1016/j.chb.2016.12.023.
Epskamp, S., Epskamp, M. S., & Rcpp, L. (2020). Package ‘psychonetrics’ [Computer software].
Epskamp, S., & Fried, E. (2015). Package “bootnet”: Bootstrap methods for various network estimation routines. 5, 0–1.
Epskamp, S., & Fried, E. I. (2018). A tutorial on regularized partial correlation networks. Psychological Methods, 23(4), 617–634. https://doi.org/10.1037/met0000167.
Epskamp, S., & Fried, E. I. (2020). Package ‘bootnet’ [Computer software].
Fan, T., Twayigira, M., Song, L., Luo, X., Huang, C., Gao, X., & Shen, Y. (2023). Prevalence and associated factors of internet addiction among Chinese adolescents: Association with childhood trauma. Frontiers in Public Health, 11, 1172109. https://doi.org/10.3389/fpubh.2023.1172109.
Fitch, G. M., Hanowski, R. J., & Guo, F. (2015). The risk of a safety-critical event associated with mobile device use in specific driving contexts. Traffic Injury Prevention, 16(2), 124–132. https://doi.org/10.1080/15389588.2014.923566.
Flayelle, M., Schimmenti, A., Starcevic, V., & Billieux, J. (2022). The pitfalls of recycling substance-use disorder criteria to diagnose behavioral addictions. In N. Heather, M. Field, A. C. Moss, & S. Satel (Eds.), Evaluating the brain disease model of addiction (1st ed., pp. 339–349). Routledge. https://doi.org/10.4324/9781003032762-34.
Fournier, L., Schimmenti, A., Musetti, A., Boursier, V., Flayelle, M., Cataldo, I., … Billieux, J. (2023). Deconstructing the components model of addiction: An illustration through “addictive” use of social media. Addictive Behaviors, 143, 107694. https://doi.org/10.1016/j.addbeh.2023.107694.
Giustiniani, J., Nicolier, M., Pascard, M., Masse, C., Vandel, P., Bennabi, D., … Haffen, E. (2022). Do individuals with internet gaming disorder share personality traits with substance-dependent individuals? International Journal of Environmental Research and Public Health, 19(15), 9536. https://doi.org/10.3390/ijerph19159536.
Goulter, N., McMahon, R. J., Lansford, J. E., Bates, J. E., Dodge, K. A., Max Crowley, D., & Pettit, G. S. (2022). Externalizing psychopathology from childhood to early adolescence: Psychometric evaluation using latent variable and network modeling. Psychological Assessment, 34(11), 1008–1021. https://doi.org/10.1037/pas0001163.
Groemping, U. (2021). Relative importance and value for linear models: The “relaimpo” package. R Package Version 2.2-5 [Computer Software]. https://CRAN.R-project.org/package=relaimpo.
Gutiérrez, J. D. S., de Fonseca, F. R., & Rubio, G. (2016). Cell-phone addiction: A review. Frontiers in Psychiatry, 7(OCT). https://doi.org/10.3389/fpsyt.2016.00175.
Hallquist, M. N., & Wiley, J. F. (2018). MplusAutomation: An R package for facilitating large-scale latent variable analyses in M plus. Structural Equation Modeling: A Multidisciplinary Journal, 25(4), 621–638. https://doi.org/10.1080/10705511.2017.1402334.
Harris, B., Regan, T., Schueler, J., & Fields, S. A. (2020). Problematic mobile phone and smartphone use scales: A systematic review. Frontiers in Psychology, 11. https://doi.org/10.3389/fpsyg.2020.00672.
Haug, S., Paz Castro, R., Kwon, M., Filler, A., Kowatsch, T., & Schaub, M. P. (2015). Smartphone use and smartphone addiction among young people in Switzerland. Journal of Behavioral Addictions, 4(4), 299–307. https://doi.org/10.1556/2006.4.2015.037.
Holgado–Tello, F. P., Chacón–Moscoso, S., Barbero–García, I., & Vila–Abad, E. (2010). Polychoric versus Pearson correlations in exploratory and confirmatory factor analysis of ordinal variables. Quality & Quantity, 44(1), 153–166. https://doi.org/10.1007/s11135-008-9190-y.
Horvath, J., Mundinger, C., Schmitgen, M. M., Wolf, N. D., Sambataro, F., Hirjak, D., … Christian Wolf, R. (2020). Structural and functional correlates of smartphone addiction. Addictive Behaviors, 105(November 2019), 106334. https://doi.org/10.1016/j.addbeh.2020.106334.
James, R. J. E., Dixon, G., Dragomir, M.-G., Thirlwell, E., & Hitcham, L. (2023). Understanding the construction of ‘behavior’ in smartphone addiction: A scoping review. Addictive Behaviors, 137, 107503. https://doi.org/10.1016/j.addbeh.2022.107503.
Jannusch, T., Shannon, D., Völler, M., Murphy, F., & Mullins, M. (2021). Smartphone use while driving: An investigation of young novice driver (YND) behaviour. Transportation Research Part F: Traffic Psychology and Behaviour, 77, 209–220. https://doi.org/10.1016/j.trf.2020.12.013.
Jin, L., Hao, Z., Huang, J., Akram, H. R., Saeed, M. F., & Ma, H. (2021). Depression and anxiety symptoms are associated with problematic smartphone use under the COVID-19 epidemic: The mediation models. Children and Youth Services Review, 121(December 2020), 105875. https://doi.org/10.1016/j.childyouth.2020.105875.
Kardefelt-Winther, D., Heeren, A., Schimmenti, A., van Rooij, A., Maurage, P., Carras, M., … Billieux, J. (2017). How can we conceptualize behavioural addiction without pathologizing common behaviours? Addiction, 112(10), 1709–1715. https://doi.org/10.1111/add.13763.
Khalily, M. T., Saleem, T., Bhatti, M. M., Ahmad, I., & Hussain, B. (2019). An Urdu adaptation of smartphone addiction scale-short version (SAS-SV). JPMA. the Journal of the Pakistan Medical Association, 69(5), 700–704.
Khoury, J. M., De Freitas, A. A. C., Roque, M. A. V., Albuquerque, M. R., Das Neves, M. D. C. L., & Garcia, F. D. (2017). Assessment of the accuracy of a new tool for the screening of smartphone addiction. Plos One, 12(5), 1–13. https://doi.org/10.1371/journal.pone.0176924.
Kim, S. G., Park, J., Kim, H. T., Pan, Z., Lee, Y., & McIntyre, R. S. (2019). The relationship between smartphone addiction and symptoms of depression, anxiety, and attention-deficit/hyperactivity in South Korean adolescents. Annals of General Psychiatry, 18(1), 1. https://doi.org/10.1186/s12991-019-0224-8.
Kwon, M., Kim, D. J., Cho, H., & Yang, S. (2013). The smartphone addiction scale: Development and validation of a short version for adolescents. Plos One, 8(12), 1–7. https://doi.org/10.1371/journal.pone.0083558.
Kwon, M., Lee, J. Y., Won, W. Y., Park, J. W., Min, J. A., Hahn, C., … Kim, D. J. (2013). Development and validation of a smartphone addiction scale (SAS). Plos One, 8(2). https://doi.org/10.1371/journal.pone.0056936.
Lai, X., Huang, S., Nie, C., Yan, J. J., Li, Y., Wang, Y., & Luo, Y. (2022). Trajectory of problematic smartphone use among adolescents aged 10–18 years: The roles of childhood family environment and concurrent parent–child relationships. Journal of Behavioral Addictions, 11(2), 577–587. https://doi.org/10.1556/2006.2022.00047.
Larsen, H., Wiers, R. W., Su, S., & Cousijn, J. (2023). Excessive smartphone use and addiction: When harms start outweighing benefits. Addiction, 118(4), 586–588. https://doi.org/10.1111/add.16060.
Lee, H.-K., Kim, J.-H., Fava, M., Mischoulon, D., Park, J.-H., Shim, E.-J., … Jeon, H. J. (2017). Development and validation study of the smartphone overuse screening questionnaire. Psychiatry Research, 257, 352–357. https://doi.org/10.1016/j.psychres.2017.07.074.
Leung, L. (2008). Leisure boredom, sensation seeking, self-esteem, and addiction: Symptoms and patterns of cell phone use. In Mediated interpersonal communication (pp. 373–396). Routledge.
Lin, Y. H., Chang, L. R., Lee, Y. H., Tseng, H. W., Kuo, T. B. J., & Chen, S. H. (2014). Development and validation of the smartphone addiction inventory (SPAI). Plos One, 9(6). https://doi.org/10.1371/journal.pone.0098312.
Lin, Y. H., Pan, Y. C., Lin, S. H., & Chen, S. H. (2017). Development of short-form and screening cutoff point of the Smartphone Addiction Inventory (SPAI-SF). International Journal of Methods in Psychiatric Research, 26(2). https://doi.org/10.1002/mpr.1525.
Little, T. D. (2013). Longitudinal structural equation modeling. Guilford press.
Lopez-Fernandez, O. (2017). Short version of the Smartphone Addiction Scale adapted to Spanish and French: Towards a cross-cultural research in problematic mobile phone use. Addictive Behaviors, 64, 275–280. https://doi.org/10.1016/j.addbeh.2015.11.013.
Lopez-Fernandez, O., Honrubia-Serrano, L., Freixa-Blanxart, M., & Gibson, W. (2014). Prevalence of problematic mobile phone use in British adolescents. Cyberpsychology, Behavior, and Social Networking, 17(2), 91–98. https://doi.org/10.1089/cyber.2012.0260.
Lopez-Fernandez, O., Kuss, D. J., Pontes, H. M., Griffiths, M. D., Dawes, C., Justice, L. V., … Billieux, J. (2018). Measurement invariance of the short version of the problematic mobile phone use questionnaire (PMPUQ-SV) across eight languages. International Journal of Environmental Research and Public Health, 15(6), 1213. https://doi.org/10.3390/ijerph15061213.
Lüdecke, D. (2021a). Sjmisc: Miscellaneous functions for data handling. R Package Version 2.11.3 [Computer Software]. https://CRAN.R-project.org/package=sjmisc.
Lüdecke, D. (2021b). sjPlot: Data Visualization for Statistics in social science. R Package Version 2.11.3 [Computer Software]. https://CRAN.R-project.org/package=sjPlot.
Luk, T. T., Wang, M. P., Shen, C., Wan, A., Chau, P. H., Oliffe, J., … Lam, T. H. (2018). Short version of the Smartphone Addiction Scale in Chinese adults: Psychometric properties, sociodemographic, and health behavioral correlates. Journal of Behavioral Addictions, 7(4), 1157–1165. https://doi.org/10.1556/2006.7.2018.105.
Madore, K. P., Khazenzon, A. M., Backes, C. W., Jiang, J., Uncapher, M. R., Norcia, A. M., & Wagner, A. D. (2020). Memory failure predicted by attention lapsing and media multitasking. Nature, 587(7832), 87–91. https://doi.org/10.1038/s41586-020-2870-z.
Mansley, C. (2021). Janitor: Simple tools for examining and cleaning dirty data. R Package Version 0.3.5 [Computer Software]. https://CRAN.R-project.org/package=janitor.
Mattiko, M. J., Olmsted, K. L. R., Brown, J. M., & Bray, R. M. (2011). Alcohol use and negative consequences among active duty military personnel. Addictive Behaviors, 36(6), 608–614. https://doi.org/10.1016/j.addbeh.2011.01.023.
Merlo, L. J., Stone, A. M., & Bibbey, A. (2013). Measuring problematic mobile phone use: Development and preliminary psychometric properties of the PUMP scale. Journal of Addiction, 2013. https://doi.org/10.1155/2013/912807.
Monteleone, A. M., Tzischinsky, O., Cascino, G., Alon, S., Pellegrino, F., Ruzzi, V., & Latzer, Y. (2022). The connection between childhood maltreatment and eating disorder psychopathology: A network analysis study in people with bulimia nervosa and with binge eating disorder. Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity, 27(1), 253–261. https://doi.org/10.1007/s40519-021-01169-6.
Muthén, L. K., & Muthén, B. (2017). Mplus users guide. Los Angeles, CA: Muthen & Muthen.
Nawaz, S. (2023). Rethinking classifications and metrics for problematic smartphone use and dependence: Addressing the call for reassessment. Computers in Human Behavior Reports, 12, 100327. https://doi.org/10.1016/j.chbr.2023.100327.
O’Dea, S. (2023). Number of smartphone mobile network subscriptions worldwide from 2016 to 2022, with forecasts from 2023 to 2028 (in millions). Statista. https://www.statista.com/statistics/330695/number-of-smartphone-users-worldwide/.
Olson, J. A., Sandra, D. A., Colucci, É. S., Al Bikaii, A., Chmoulevitch, D., Nahas, J., … Veissière, S. P. L. (2022). Smartphone addiction is increasing across the world: A meta-analysis of 24 countries. Computers in Human Behavior, 129(November 2021). https://doi.org/10.1016/j.chb.2021.107138.
Oviedo-Trespalacios, O., Haque, M. M., King, M., & Washington, S. (2016). Understanding the impacts of mobile phone distraction on driving performance: A systematic review. Transportation Research Part C: Emerging Technologies, 72, 360–380. https://doi.org/10.1016/j.trc.2016.10.006.
Parry, D. A., Davidson, B. I., Sewall, C. J. R., Fisher, J. T., Mieczkowski, H., & Quintana, D. S. (2021). A systematic review and meta-analysis of discrepancies between logged and self-reported digital media use. Nature Human Behaviour, 1–13. https://doi.org/10.1038/s41562-021-01117-5.
Pavia, L., Cavani, P., Di Blasi, M., & Giordano, C. (2016). Smartphone Addiction Inventory (SPAI): Psychometric properties and confirmatory factor analysis. Computers in Human Behavior, 63, 170–178. https://doi.org/10.1016/j.chb.2016.05.039.
Pearson, C., & Hussain, Z. (2016). Smartphone addiction and associated psychological factors. Addicta: The Turkish Journal on Addictions, 3(2). https://doi.org/10.15805/addicta.2016.3.0103.
Qualtrics, X. (2021). Qualtrics XM-Experience management software [Computer software].
R Core Team. (2022). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria [Computer Software]. http://www.R-project.org/.
Revelle, W., & Condon, D. M. (2019). Reliability from α to ω: A tutorial. Psychological Assessment, 31(12), 1395. https://doi.org/10.1037/pas0000754.
Rhemtulla, M., Brosseau-Liard, P. É., & Savalei, V. (2012). When can categorical variables be treated as continuous? A comparison of robust continuous and categorical sem estimation methods under suboptimal conditions. Psychological Methods, 17(3), 354–373. https://doi.org/10.1037/a0029315.
Rosales, A., & Fernández-Ardèvol, M. (2019). Smartphone usage diversity among older people. In Perspectives on human-computer interaction research with older people. Springer, Cham.
Ryding, F. C., & Kuss, D. J. (2020). Passive objective measures in the assessment of problematic smartphone use: A systematic review. Addictive Behaviors Reports, 11(September 2019), 100257. https://doi.org/10.1016/j.abrep.2020.100257.
Saunders, J. B., Aasland, O. G., Babor, T. F., De La Fuente, J. R., & Grant, M. (1993). Development of the alcohol use disorders identification test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption-II. Addiction, 88(6), 791–804. https://doi.org/10.1111/j.1360-0443.1993.tb02093.x.
Sfendla, A., Laita, M., Nejjar, B., Souirti, Z., Touhami, A. A. O., & Senhaji, M. (2018). Reliability of the Arabic smartphone addiction scale and smartphone addiction scale-short version in two different Moroccan samples. Cyberpsychology, Behavior, and Social Networking, 21(5), 325–332. https://doi.org/10.1089/cyber.2017.0411.
Sohn, S. Y., Krasnoff, L., Rees, P., Kalk, N. J., & Carter, B. (2021). The association between smartphone addiction and sleep: A UK cross-sectional study of young adults. Frontiers in Psychiatry, 12(March), 1–10. https://doi.org/10.3389/fpsyt.2021.629407.
Sohn, S., Rees, P., Wildridge, B., Kalk, N. J., & Carter, B. (2019). Prevalence of problematic smartphone usage and associated mental health outcomes amongst children and young people: A systematic review, meta-analysis and grade of the evidence. BMC Psychiatry, 19(356). https://doi.org/10.1186/s12888-019-2350-x.
Starcevic, V. (2016). Tolerance and withdrawal symptoms may not be helpful to enhance understanding of behavioural addictions: Letter to the Editor. Addiction, 111(7), 1307–1308. https://doi.org/10.1111/add.13381.
Su, S., Cousijn, J., Wiers, R. W., Murray, H., Schoenmakers, T. M., & Larsen, H. (2023). The good and the bad: A qualitative investigation of students’ perspectives on their problematic smartphone use [Manuscript in preparation].
Su, S., Larsen, H., Cousijn, J., Wiers, R. W., & Van Den Eijnden, R. J. J. M. (2022). Problematic smartphone use and the quantity and quality of peer engagement among adolescents: A longitudinal study. Computers in Human Behavior, 126(March 2021), 107025. https://doi.org/10.1016/j.chb.2021.107025.
Van Den Eijnden, R. J. J. M., Lemmens, J. S., & Valkenburg, P. M. (2016). The social media disorder scale: Validity and psychometric properties. Computers in Human Behavior, 61, 478–487. https://doi.org/10.1016/j.chb.2016.03.038.
Van Deursen, A. J. A. M., Bolle, C. L., Hegner, S. M., & Kommers, P. A. M. (2015). Modeling habitual and addictive smartphone behavior: The role of smartphone usage types, emotional intelligence, social stress, self-regulation, age, and gender. Computers in Human Behavior, 45, 411–420. https://doi.org/10.1016/j.chb.2014.12.039.
Vatcheva, K. P., Lee, M., McCormick, J. B., & Rahbar, M. H. (2016). Multicollinearity in regression analyses conducted in epidemiologic studies. Epidemiology (Sunnyvale, Calif.), 6(2), 227. https://doi.org/10.4172/2161-1165.1000227.
Walsh, S. P., White, K. M., & McD Young, R. (2010). Needing to connect: The effect of self and others on young people’s involvement with their mobile phones. Australian Journal of Psychology, 62(4), 194–203. https://doi.org/10.1080/00049530903567229.
Wang, Y. Y., Long, J., Liu, Y. H., Liu, T. Q., & Billieux, J. (2020). Factor structure and measurement invariance of the problematic mobile phone use questionnaire-short version across gender in Chinese adolescents and young adults. BMC Psychiatry, 20(1), 34. https://doi.org/10.1186/s12888-020-2449-0.
White, H. R., & Labouvie, E. W. (1989). Towards the assessment of adolescent problem drinking. Journal of Studies on Alcohol, 50(1), 30–37. https://doi.org/10.15288/jsa.1989.50.30.
White, H. R., Labouvie, E. W., & Papadaratsakis, V. (2005). Changes in substance use during the transition to adulthood: A comparison of college students and their noncollege age peers. Journal of Drug Issues, 35(2), 281–306. https://doi.org/10.1177/002204260503500204.
Wickham, H., François, R., & Henry, L. (2020). dplyr: A grammar of data manipulation [Computer software].
Wickham, H., & Wickham, M. H. (2020). Package ‘plyr’ [Computer software].
William Jr, C. (2013). Multivariate data analysis-Pearson education limited.
Wilmer, H. H., Sherman, L. E., & Chein, J. M. (2017). Smartphones and cognition: A review of research exploring the links between mobile technology habits and cognitive functioning. Frontiers in Psychology, 8(APR), 1–16. https://doi.org/10.3389/fpsyg.2017.00605.
Winkler, A., Jeromin, F., Doering, B. K., & Barke, A. (2020). Problematic smartphone use has detrimental effects on mental health and somatic symptoms in a heterogeneous sample of German adults. Computers in Human Behavior, 113(November 2019), 106500. https://doi.org/10.1016/j.chb.2020.106500.
World Health Organization (2019). Frozen version of ICD11 for implementation. https://www.who.int/classifications/classification-of-diseases.
Wu, H., & Leung, S.-O. (2017). Can Likert scales be treated as interval scales?—a simulation study. Journal of Social Service Research, 43(4), 527–532. https://doi.org/10.1080/01488376.2017.1329775.
Yildirim, C., & Correia, A. P. (2015). Exploring the dimensions of nomophobia: Development and validation of a self-reported questionnaire. Computers in Human Behavior, 49, 130–137. https://doi.org/10.1016/j.chb.2015.02.059.
Yue, H., Zhang, X., Sun, J., Liu, M., Li, C., & Id, H. B. (2021). The relationships between negative emotions and latent classes of smartphone addiction. 1–13. https://doi.org/10.1371/journal.pone.0248555.
Zhao, H., Rafik-Galea, S., Fitriana, M., & Song, T.-J. (2022). Translation and psychometric evaluation of smartphone addiction scale—short version (SAS-SV) among Chinese college students. Plos One, 17(11), e0278092. https://doi.org/10.1371/journal.pone.0278092.
