1-s2.0-S0747563218300165-main

1. Computers in Human Behavior 82 (2018) 159e166 Contents lists available at ScienceDirect Computers in Human Behavior journal homepage: www.elsevier.com/locate/comphumbeh Full length article Examining latent classes of smartphone users: Relations with psychopathology and problematic smartphone use Jon D. Elhaia, b, *, Ateka A. Contractorc aDepartment of Psychology, University of Toledo, 2801 W. Bancroft Street, Toledo, OH, 43606, USA bDepartment of Psychiatry, University of Toledo, 3065 Arlington Avenue, Toledo, OH, 43614, USA cDepartment of Psychology, University of North Texas, 1155 Union Circle, Denton, TX, 76203, USA a r t i c l e i n f o a b s t r a c t Article history: Received 31 July 2017 Received in revised form 9 January 2018 Accepted 11 January 2018 Little is known about how people use their smartphones, and whether usage patterns are related to psychopathology and problematic smartphone use. We sampled 296 college students, and administered self-report scales to assess frequency of using various smartphone features, and problematic smartphone use. We also assessed psychopathology constructs (ruminative thinking and emotion regulation deficits), and demographics (age and gender) as potential covariates of smartphone use patterns. Using latent class analysis, we identified two distinct classes of smartphone feature use, with one class representing especially heavy use (particularly that of social networking, audio and visual entertainment, and image- and video-taking), and the other class involving light use (particularly that of social networking, audio entertainment and image- and video-taking). Higher levels of rumination, and higher levels of cognitive reappraisal as an emotion regulation strategy, were related to the heavy use class. The heavy use class scored higher on all problematic smartphone use outcome variables. Results aid in understanding the nature and relation of smartphone usage patterns to psychopathology and problematic smartphone usage. Keywords: Problematic smartphone use Latent class analyses Rumination Emotion regulation © 2018 Elsevier Ltd. All rights reserved. 1. Introduction differentially relates to PSU. Psychopathology is associated with PSU. In a recent paper, Elhai et al. (2017a) systematically reviewed the most widely studied psychopathology variables for relations with PSU: depression, anxiety, stress, and low self-esteem. The authors discovered that depression and anxiety severity demonstrated moderate and small associations (respectively) with smartphone use frequency and PSU. However, effect sizes for anxiety were small on average, and depression severity was sometimes non-significant or inversely related to PSU (Augner & Hacker, 2012; Elhai, Levine, Dvorak, & Hall, 2016, 2017c; J. Kim, Seo, & David, 2015). Studying more contemporary psychopathology constructs may aid in under- standing what drives PSU. Some contemporary psychopathology constructs that have become increasingly important in recent psychopathology research are “transdiagnostic”e that is, those constructs that cut across numerous mental disorders. Such constructs play a substantial role in the development and maintenance of mental disorders (Mansell, Harvey, Watkins, & Shafran, 2008), and are also clinical targets in psychological interventions across (Hofmann, Sawyer, Fang, & Asnaani, 2012). Two important Despite the productivity advantages of using a smartphone, its excessive use has become a concern in modern society. “Problem- atic smartphone use” (PSU) involves excessive use of a smartphone with associated symptoms observed in substance and addictive disorders, such as social or work interference, and withdrawal symptoms when unable to use one's phone (Billieux, Maurage, Lopez-Fernandez, Kuss, & Griffiths, 2015a). PSU is associated with adverse consequences such as distracted driving (Cazzulino, Burke, Muller, Arbogast, & Upperman, 2014), physical health problems such as musculoskeletal pain (Xie, Szeto, Dai, & Madeleine, 2016; _INal, Dem_Irc_I, Çet_Intürk, Akg€ onül, & Savas ¸, 2015), academic diffi- culties (Samaha & Hawi, 2016; Seo, Park, Kim, & Park, 2016), and mental health problems (reviewed in Elhai, Dvorak, Levine, & Hall, 2017a). However, little is known about the heterogeneity of ways in which people use their smartphones, and how such heterogeneity various mental disorders * Corresponding author. University of Toledo, 2801 W. Bancroft St., Mail Stop 948, Toledo, OH, 43560, USA. E-mail address: contact@jon-elhai.com (J.D. Elhai). https://doi.org/10.1016/j.chb.2018.01.010 0747-5632/© 2018 Elsevier Ltd. All rights reserved.

2. 160 J.D. Elhai, A.A. Contractor / Computers in Human Behavior 82 (2018) 159e166 assumes that differences across individuals drive them to have particularemotional needs to be gratified, and such needs influence their choices in consuming mass media. Background characteristics such as female gender, as well as psychological constructs such as depression, anxiety, negative affect, proneness to boredom, and behavioral activation have been supported in studies on UGT in explaining increased smartphone use (Dhir et al., 2015; Elhai, Levine, Dvorak, & Hall, 2017c; Elhai, Vasquez, Lustgarten, Levine, & Hall, in press; Grellhesl & Punyanunt-Carter, 2012; Park, Kim, Shon, & Shim, 2013; Wolniewicz, Tiamiyu, Weeks, & Elhai, inpress). One widely-used categorization of internet uses involves the distinction between social use vs. process (i.e., non-social) use (Song, LaRose, Eastin, & Lin, 2004), a categorization used in the smartphone use literature as well (Elhai et al., 2017c; van Deursen et al., 2015). Mixed results have been found regarding whether these broad types of smartphone use are differentially associated with levels of PSU. Some studies found that PSU was more related to social use (Lee, Chang, Lin, & Cheng, 2014; Lopez-Fernandez, Honrubia-Serrano, Freixa-Blanxart, & Gibson, 2014; Zhitomirsky- Geffet & Blau, 2016), while other studies supported non-social use (Elhai, Hall, Levine, & Dvorak, 2017b; Elhai et al., 2017c; van Deursen et al., 2015). An additional theory, Compensatory Internet Use Theory (CIUT) (Kardefelt-Winther, 2014), could be viewed as an extension of UGT. CIUT proposes that people attempt to alleviate their negative emotion by using (or overusing) technology. Thus in CIUT, negative emotion (an individual difference variable) can be viewed as the antecedent to technology (e.g., smartphone) use. And increased or excessive technology use is viewed as the compensatory behavior aimed at regulating negative emotion. Several studies thus far have supported CIUT in explaining PSU (Elhai, Tiamiyu, & Weeks, in press; Long et al., 2016; Wang, Wang, Gaskin, & Wang, 2015; Wolniewicz et al., in press; Zhitomirsky-Geffet & Blau, 2016). Thus in UGT and CIUT, psychopathology can be conceptualized as motivating increased and excessive smartphone use. And such increased use can be conceptualized as a strategy for compensating for and regulating psychopathology. transdiagnostic constructs are emotion regulation and rumination. Emotion regulation is the process by which people regulate their emotions to adapt to their environment. Adaptive emotion regulation involves two distinct processes - higher cognitive reappraisal, and lower expressive suppression (Gross, 1998). Cognitive reappraisal involves reinterpreting negative emotional stimuli in a non-emotional, cognitive manner, while expressive suppression involves inhibiting negative emotion (Tull & Aldao, 2015). Elhai et al. (2016) found in a community sample that expressive suppression was related to higher levels of PSU, and cognitive reappraisal was related to smartphone use frequency. Another transdiagnostic construct, rumination, is a maladaptive method of regulating negative emotion by overemphasizing negative thoughts of oneself, rather than engaging in processing the negative emotion (Mennin & Fresco, 2013). Rumination is highly related to numerous types of psychopathology (Aldao, Nolen-Hoeksema, & Schweizer, 2010). Rumination has been conceptualized to contribute to problematic internet use (Davis, 2001). Finally, Elhai, Tiamiyu, and Weeks (in press) discovered in a college student sample that after adjusting for age, gender, and other psychopathology variables, rumination was related to increased smartphone use and PSU. Our interest was to examine types of smartphone use and their correlates with psychopathology at a more granular level, based on specific features of use. One method of such granular analysis is to empirically explore subgroups of research participants based on their reported patterns of use, through mixture modeling. Mixture modeling is a statistical method for empirically uncovering het- erogeneity among a sample of participants based on a set of observed variables (Masyn, 2013; McLachlan & Peel, 2000). One type of mixture model is latent class analysis (LCA), involving ex- amination of cross-sectional, ordinal/categorical variables for distinct subgroups of individuals based on their responses (Muth? en, 2008). One previous paper used mixture modeling to study smart- phone usage patterns. Hamka, Bouwman, de Reuver, and Kroesen (2014) recruited a community sample of adults and inquired about types of smartphone uses (albeit, without assessing psy- chopathology or PSU). The authors found six latent classes of smartphone users including a) very minimal smartphone users, b) basic users with limited smartphone app use, c) average users with moderate amounts of app and webuse, d) information seekers with heavy amounts of web searching, but low app use, e) users with extensive app use, and f) users with high use of web utilities but low use of installed apps. However, the authors did not examine contemporary psychopathology constructs as predictor variables, nor was PSU included in the analyses. 1.2. Research model Prior studies have used research models that would fit with UGT and CIUT in explaining PSU. Such models have used demographic and psychopathology variables to explain increased frequency of smartphone use. In turn, smartphone frequency is modeled to explain PSU, because increased use is conceptualized to translate into problematic use (Oulasvirta, Rattenbury, Ma, & Raita, 2012; van Deursen et al., 2015). Such integrated models of de- mographics, psychopathology, smartphone use frequency and problematic use have been tested and supported in recent papers (Elhai et al., 2017c; Elhai, Tiamiyu, & Weeks, in press; J. Kim et al., 2015; J.-H. Kim, 2017; van Deursen et al., 2015). We used a similar model in the present paper, though with different statistical methods (See Fig.1). We measured smartphone use frequency with LCA, to uncover latent subgroups of individuals based on their reported frequency of using various smartphone features. We modeled proximal demographic covariates of the smartphone use LCA, including age and gender. Younger people engage in moresmartphone use compared toolder people (Lu et al., 2011; van Deursen et al., 2015). Furthermore, women use smart- phones more frequently than men do (Jeong, Kim, Yum, & Hwang, 2016; Wang et al., 2015). We also included psychopathology- related proximal covariates of the smartphone use LCA, including emotion regulation deficits and rumination. Finally, we modeled PSU using several PSU subscales as the dependent variables asso- ciated with the smartphone use LCA. 1.1. Theory There are several reasons and motives for using a smartphone. Some reasons include enhancing productivity (e.g., with reminders and scheduling), information seeking (e.g., web searching and news browsing), social networking, relaxation (e.g., music), entertainment (e.g., movies), monetary compensation (e.g., locating consumer deals), and personal status enhancement (Dhir, Chen, & Nieminen, 2015; van Deursen, Bolle, Hegner, & Kommers, 2015). Additionally, several factors influence the acceptance and use of a smartphone, including perceptions of value, and quality with the following di- mensions: performance/system, use and access, content, service, and experience (Shin, 2015; Shin, Shin, Choo, & Beom, 2011). Uses and Gratifications Theory (UGT) is a mass communications theory aimed at explaining motivations for consuming mass media (Blumler, 1979), adapted to explain use of particular smartphone features (Grellhesl & Punyanunt-Carter, 2012; Wei & Lu, 2014). UGT

3. 161 J.D. Elhai, A.A. Contractor / Computers in Human Behavior 82 (2018) 159e166 prevailing models of PSU (Elhai et al., 2017c; Elhai, Tiamiyu, & Weeks, in press; J. Kim et al., 2015; J.-H. Kim, 2017; van Deursen et al., 2015). Elhai et al. (2017b) found in a community sample that all tested subscales of PSU were related to increased social and/ or process smartphone use in bivariate analyses; subscales involving functional disturbance, excitement from and excessive use of a smartphone were significant in multivariate analyses. 2. Method 2.1. Procedure We recruited 305 participants from a Midwestern university's undergraduate psychology research pool in 2016. Participants signed up for the study based on viewing it on the university's Sona Systems web portal for available research studies. We advertised this study as a 20- to 30-min internet survey regarding “use of electronic devices, and your emotions.” Signing up for the study routed participants to an online consent statement and web survey. We removed two participants from this study for not indicating gender (a covariate), and seven participants for missing more than half the items on at least one questionnaire used in analyses. The resulting 296 participants served as the effective sample. Fig. 1. Structural model of demographic and psychopathology covariates, latent class analysis, and problematic smartphone use outcomes. 1.3. Hypotheses 2.2. Participants We posed several hypotheses based on theory and prior empirical findings. In the sample of 296 participants, average age was 19.44 (SD¼ 2.16). Most participants were women (n¼227, 76.7%). The majority identified as Caucasian (n ¼234, 79.3%), with some rep- resentation from African-American (n¼34, 11.5%), Hispanic/Latino (n¼14, 4.7%), and Asian-American (n¼ 12, 4.1%) ethnicities (en- dorsements were not mutually-exclusive). Most participants were freshman (n¼137, 49.7%) or sophomores (n ¼98, 33.1%). Employ- ment status was mostly part-time (n¼147, 49.7%), with some participants employed full-time (n¼19, 6.4%). 1) Higher levels of rumination will be associated with the heavier smartphone use class(es) discovered in LCA. Rumination is conceptualized as a maladaptive coping strategy for processing emotion (Mennin & Fresco, 2013). As such, rumi- nation can be considered an individual difference variable in UGT (Blumler, 1979), and an antecedent psychopathology construct to excess internet use in CIUT (Kardefelt-Winther, 2014), for which increased smartphone use would be the consequential compen- satory behavior. Rumination has demonstrated relations with increased smartphone use after adjusting for age, gender, and other psychopathology variables (Elhai, Tiamiyu, & Weeks, in press). 2.3. Measures 2.3.1. Demographics We inquired about such demographic characteristics as age, gender, ethnicity and employment. 2) Higher levels of expressive suppression, and lower levels of cognitive reappraisal (emotion regulation facets), will be asso- ciated with the heavier smartphone use class(es). 2.3.2. Smartphone use We used a scale measuring frequency of 11 types of smartphone features, developed by Elhai et al. (2016). We used a Likert-type scale for items, ranging from “1¼ Never” to “6 ¼Very Often.” The queried features included “Voice/video receiving),” “Texting/instant messaging (sending and receiving),” “Email (sending and receiving),” “Internet/websites,” “Music/podcasts/radio,” pictures or videos,” “"Watching video/TV/movies,” “Reading books/ magazines,” and “Maps/navigation.” Elhai et al. (2016) discovered this scale to be reliable, and with moderate correlations with Kwon et al.'s (2013) Smartphone Addiction Scale. Internal consistency in the present sample was 0.76. Using more suppression and less reappraisal is considered maladaptive in the process of regulating negative emotion (Gross, 1998). Therefore, increased suppression and decreased reap- praisal can be regarded as individual difference variables within UGT, and antecedent psychopathology constructs within CIUT, for which increased smartphone use is a compensatory behavior. Elhai et al. (2016) found in a community sample that expressive sup- pression was related to PSU and mediated relations between depression and PSU. Elhai et al. (2016) also found that higher reappraisal was related to increased smartphone use, though con- tradictory to theory. Finally, emotion regulation deficits have been related to problematic social media use (Hormes, Kearns, & Timko, 2014). calls (making and “Social networking “Games,” sites,” “Taking 2.3.3. Smartphone Addiction Scale (SAS) We used Kwon et al.'s (2013) SAS which measures severity of PSU. The SAS has 33 items ranging from “1 ¼Strongly disagree” to “6 ¼Strongly agree.” It has six subscales including “Daily life disturbance,” “Positive anticipation,” “Withdrawal,” “Cyberspace oriented relationships,” “Overuse,” and “Tolerance.” The SAS has demonstrated reliability, and convergent validity against other scales measuring PSU and (Chotpitayasunondh & Douglas, 2016; Demirci, Orhan, Demirdas, 3) Heavier smartphone use class(es) will evidence higher scores on all PSU subscales. Increased use of a smartphone is conceptualized to drive PSU (Oulasvirta et al., 2012; van Deursen et al., 2015), evident in problematic internet use

4. 162 J.D. Elhai, A.A. Contractor / Computers in Human Behavior 82 (2018) 159e166 Akpinar, & Sert, 2014; Kwon et al., 2013). In the present sample, internal consistency was 0.93 for the total scale, and ranged from 0.74 (Daily life disturbance) to 0.87 (Positive anticipation) for the subscales. Expressive Suppression, and 28.14 (SD ¼7.26) for Cognitive Reap- praisal. PSU summed scores averaged 12.32 (SD¼ 5.04) for Daily Life Disturbance, 23.80 (SD ¼7.42) for Positive Anticipation, 17.18 (SD¼ 6.19) for Withdrawal, 16.93 (SD ¼5.81) for Cyberspace Ori- ented Relationships, 14.87 (SD ¼4.83) for Overuse, and 8.43 (SD¼ 3.48) for Tolerance. The sum of smartphone frequency items averaged 48.25 (SD ¼7.63). A correlation matrix of these scale scores, along with age and gender, is presented in Table 1. Fre- quency data for using smartphone features are found inTable 2. The most frequently endorsed (as “Very Often”) smartphone features were text and instant messaging (74.7%), social networking (64.9%), internet (58.8%), music/podcasts/radio (50.7%), and taking pictures and video (40.9%). 2.3.4. Emotion regulation scale (ERQ) The ERQ (Gross& John, 2003) is a 10-itemself-report measureof emotion regulation, with subscales of cognitive reappraisal and expressive suppression. Response options range from “1¼ Strongly disagree” to “7¼ Strongly agree.” The ERQ has demonstrated good reliability for its subscales, with convergent validity against mea- sures of rumination, depression, anxiety, stress and coping (Gross & John, 2003; Spaapen, Waters, Brummer, Stopa, & Bucks, 2014). In- ternal consistency in the present sample was 0.89 for cognitive reappraisal, and 0.75 for expressive suppression. 3.2. Primary LCA results Table 3 details LCA results (without covariates) for models ranging from 1 to 4 classes. aLMR, the most objective of these indices (Nylund et al., 2007), was not statistically significant after 2 classes, indicating that model fit was not improved with more complex models. Evidence for a 2-class model was also supported by this model having the lowest BIC Value; and by a precipitous drop in aBIC after a 1-class model, with substantial plateauing after 2 classes. Entropy was 0.83, and correct classification was 96% for Class 1, and 95% for Class 2. Fig. 2 displays the latent classes, and probabilities of endorsing the most frequent use category (“Very Often”) for each smartphone use type. Class 1 appears to represent especially high users of social networking, listening and viewing entertainment, picture- and video-taking (labelled high users). Class 2 involves much less use of smartphone features across the board, but with lighter use most represented by social networking, listening entertainment, and taking pictures and videos (labelled light users). 2.3.5. Ruminative Thought Style Questionnaire (RTSQ) The RTSQ (Brinker & Dozois, 2009) is a self-report measure of ruminative thinking. It includes 20 items on a Likert-type scale of “1 ¼Does not describe me at all” to “7 ¼Describes me very well.” The RTSQ has demonstrated reliability, and convergent validity with similar scales (Brinker & Dozois, 2009; Tanner, Voon, Hasking, & Martin, 2012). Internal consistency in the present sample was 0.95. 2.4. Analysis Small amounts of missing item-level data were observed for the measures; approximately 7% of the sample missed an average of 1e2 items for a given scale. We used maximum likelihood (ML) estimation procedures to estimate (Graham, 2009). After estimating missing values, we summed item responses to derive scale scores. We used Mplus Version 8 to estimate the LCA based on smart- phone use feature responses. These items are on an ordinal scale, and therefore we treated them as ordinal variables in LCA. We conducted the LCA using ML with robust standard errors (Yuan & Bentler, 2000). We examined the adjusted Lo-Mendell-Rubin like- lihood ratio chi-square test (aLMR), where a model with K classes is compared to a model with K-1 classes (Nylund, Asparouhov, & Muth? en, 2007). Statistical significance in aLMR for the model with one additional class indicates better fit for that model. We also examined the Bayesian Information Criterion (BIC; and its sample size-adjusted version, aBIC), and entropy (indicating quality of correct classification) (Nylund et al., 2007). A model that is at least tenpoints loweron BIC/aBIC is the preferred model, with an odds of 150:1 (Raftery, 1995). We next added age, gender, and summed scores for cognitive reappraisal, expressive suppression and rumination as proximal covariates of latent class membership. Logistic regression analysis was used for regressing the latent class variable on the covariates. As distal dependent variables of latent class membership, we regressed SAS subscales on the LCA variable. Using the approach discussed by others (Asparouhov & Muth? en, 2014; Lanza, Tan, & Bray, 2013), we used Mplus’ three-step method to estimate the latent class variable based onposterior probabilities, regressing SAS subscales while takingintoaccount misclassification fromposterior probability estimation. missing item-level data 3.3. LCA covariate and outcome results Rumination and Cognitive Reappraisal were significant cova- riates of the latent class variable, as indicated in Fig. 3. The odds ratio was 1.02 for Rumination, indicating that each increase of one point for the Rumination scale was associated a 2% increased likelihood of being classified in Class 1 (high users). The odds ratio was 1.05 for Cognitive Reappraisal, indicating that each increased point for Cognitive Reappraisal was associated with a 5% increased likelihood of being classified in Class 1. Neither age, gender, nor Expressive Suppression were significant covariates. The two latent classes were significantly different on all PSU outcome scales. As detailed in Table 4, Class 1 scored higher on all PSU scales. 4. Discussion In this study, we empirically examined smartphone use typol- ogies in asample of college students, using a theoretical modelwith demographic and psychopathology covariates, and PSU as the outcome. We found support for two distinct latent classes of smartphone users based on their endorsements of types of smartphone use, predicted by higher levels of both rumination and cognitive reappraisal. These classes significantly differed on sub- scales measuring PSU. Very little prior research has examined typologies of smart- phone use. In contrast to Hamka et al. (2014) who examined pri- marily messaging, website and social networking features, our study examined a wider range of smartphone features e we analyzed a total of 11 features. We found two latent classes of smartphone users, including a) heavy users, especially of social networking sites, audio and visual entertainment, and b) light 3. Results 3.1. Descriptive statistics Rumination total scores averaged 91.61 (SD¼ 23.73). Emotion regulation summed scores averaged 15.20 (SD¼ 5.26) for

5. 163 J.D. Elhai, A.A. Contractor / Computers in Human Behavior 82 (2018) 159e166 Table 1 Correlation matrix of study variables. Variable 1 2 3 4 5 6 7 8 9 10 11 1. Age 2. Gender 3. Rumination 4. Suppression 5. Reappraisal 6. SUF 7. SAS DLD 8. SAS PA 9. SAS Withdrawal 10. SAS COR 11. SAS Overuse 12. SAS Tolerance ?.13* ?.07 .05 ?.10 ?.20*** ?.06 ?.14* ?.11 ?.06 .03 ?.08 .21*** ?.11 .06 .24*** .13* .09 .10 .15* .10 .14* .15* .07 .22*** .31*** .23*** .33*** .23*** .37*** .28*** .24*** .01 .07 .14* .05 .24*** .05 .20*** .23*** ?.04 .14* .03 .10 .12* .14* .17** .26*** .28*** .28*** .27*** .24*** .38*** .46*** .39*** .51*** .55*** .65*** .58*** .46*** .46*** .61*** .53*** .49*** .46*** .52*** .56*** Note: SUF¼Smartphone Use Frequency; SAS¼Smartphone Addiction Scale; DLD ¼ Daily Life Disturbance; PA¼Positive Anticipation; COR¼Cyberspace Oriented Relationships. *p < .05;**p < .01;***p < .001. Table 2 Frequency of using smartphone features. Smartphone Feature “Never” “Rarely” “Occasionally” “Somewhat Often” “Often” “Very Often” Voice/Video Calls Text/Instant Messaging Email Social Networking Internet Music/Podcasts Gaming Pictures/Video Watching TV/Movies Reading Maps/Navigation 8 0 10 3 0 5 28 1 26 81 6 29 2 36 6 4 10 98 15 45 87 27 52 10 51 20 18 24 70 27 53 61 67 74 21 68 22 22 42 36 57 38 36 78 72 42 77 53 78 65 34 75 56 15 68 61 221 54 192 174 150 30 121 78 16 50 N¼ 296. Table 3 Latent class analysis model comparisons. # of Classes BIC aBIC Entropy aLMR p 1 2 3 4 9550.90 9323.09 9426.45 9576.52 9382.82 8983.75 8915.87 8894.68 N/A .83 .88 .91 N/A 533.19 203.25 156.70 N/A <.001 .77 .85 Note: BIC¼Bayesian Information Criterion; aBIC¼ Adjusted Bayesian Information Criterion; aLMR ¼ Adjusted p ¼ aLMR's p value; N/A¼ Not Applicable (not possible to estimate for a one-class model). Lo-Mendell-Rubin Likelihood Ratio Test Value; Fig. 3. Structural Model of Demographic and Psychopathology Covariates, Latent Class Analysis, and Problematic Smartphone Use Outcomes: With Estimates for Covariates. Notes. Estimates shown are unstandardized logistic path (beta) coefficients. Estimates for the outcome (SAS) variables are displayed in Table 4.*p < .05;**p < .01;***p < .001. Fig. 2. The 2-Class LCA. Note. Class 1 (n ¼156) is denoted by the line series with cir- cles. Class 2 (n ¼140) is denoted by the line series with triangles. The Y Axis indicates the probability of endorsing the highest response option (“Very Often”) for the given smartphone feature. smartphone users. We found support for Hypothesis 1. Higher levels of rumination were related to the heavier smartphone use class. This finding fits well with UGT (Blumler, 1979) in assuming that individual differ- ence variables such as rumination would drive a particular type of media use e in this case, smartphone use. Our finding also fits well with CIUT (Kardefelt-Winther, 2014), which would conceptualize users, whose light use primarily involved social networking sites, audio entertainment, and taking pictures and videos. Our study, thus, supports the heterogeneity embedded in the amount and types of smartphone use characterized bytwo distinct subgroups of

6. 164 J.D. Elhai, A.A. Contractor / Computers in Human Behavior 82 (2018) 159e166 Table 4 Latent class membership differences on outcome variables. SAS Scale Class 1 M SE of M Class 2 M SE of M Chi-square p Daily Life Disturbance Positive Anticipation Withdrawal Cyber Oriented Relationships Overuse Tolerance 13.08 25.58 18.69 18.20 16.36 9.31 .43 .64 .55 .48 .40 .32 1.45 21.73 15.44 15.46 13.15 7.41 .42 .59 .46 .50 .40 .26 6.75 18.06 18.79 14.54 30.44 19.30 .01 <.001 <.001 <.001 <.001 <.001 increased smartphone use as a means of tempering negative emotion e in this case, ruminative thinking. This finding also fits with recent research revealing that rumination is associated with PSU, and also associated with increased smartphone use after adjusting for age, gender, and other psychopathology variables (depression and social anxiety) (Elhai, Tiamiyu, & Weeks, in press). Rumination may drive smartphone use and/or PSU, especially among those with depression, anxiety or poor emotion regulation skills. One possible mechanism for this relationship is that rumi- nation in interpersonal relationships may be characterized by excessive reassurance seeking through checking one's phone for friends' notifications (Billieux et al., 2015b). Additionally, in- dividuals with social anxiety and who ruminate about social interaction may avoid in-person social interaction, instead exces- sively using their smartphones or other online media (Elhai, Tiamiyu, & Weeks, in press; Wolniewicz et al., in press). We failed to find support for Hypothesis 2. Contrary to this hypothesis, expressive suppression was not associated with mem- bership in the higher smartphone use class. Also in contrast to Hypothesis 2, higher levels of cognitive reappraisal were associated with membership in the higher smartphone use class. Because cognitive reappraisal is an adaptive form of emotion regulation (Gross, 1998), based on UGT and CIUT, we would expect that lower levels of cognitive reappraisal would be associated with the heavy user class. This was not borne out in our findings, however. Nonetheless, these findings parallel those from Elhai et al. (2016) who analyzed data on emotion regulation, smartphone use and PSU, using a community sample. We must note that the present study, as well as Elhai et al. (2016), used cross-sectional data, and thus causality cannot be properly assessed. Being a heavy smart- phone user could consequently drive individuals to cognitively reappraise excessive smartphone use as not being a waste of time. Longitudinal or repeated measures designs could help determine the nature of this association. We found support for Hypothesis 3. The heavier smartphone use class scored higher across all subscales of PSU. This finding fits with existing theoretical models of PSU, whereby increased and habitual smartphone use is conceptualized specifically to grow into PSU (Oulasvirta et al., 2012; van Deursen et al., 2015). This model also fits with structural equation models supporting increased smart- phone use as mediating relationships between psychopathology and PSU (Elhai et al., 2017c; Elhai, Tiamiyu, & Weeks, in press; J.; Kim et al., 2015; van Deursen et al., 2015). Thus, in attempting to understand the construct of PSU, it is important to recognize that several types of psychopathology can lead to increased or habitual use of a smartphone, as a means of regulating negative emotion with such technology use. And in turn, the habitual, increased use of a smartphone can grow into PSU, causing impairment in such areas as school, work or social interaction. Thus, psychopathology may be a risk factor for PSU, while habitual/increased smartphone use may be an important mechanism psychopathology-PSU relationship. In fact, recent research has uncovered important and novel findings in understanding the relationships among psychopathol- ogy, increased smartphone use, and PSU. Until recently, the ma- jority of relevant studies only examined levels of depression, anxiety, stress, or low self-esteem as types of psychopathology associated with PSU (reviewed in Elhai et al., 2017a). Newly pub- lished studies have examined and found support for additional types of psychopathology associated with PSU, including post- traumatic stress disorder (Contractor, Frankfurt, Weiss, & Elhai, 2017), rumination (Elhai, Tiamiyu, & Weeks, in press), emotion dysregulation (Elhai et al., 2016), self-regulation (G€ okçearslan, Mumcu, Has ¸laman, & Çevik, 2016), proneness to boredom (Elhai, Vasquez, et al., in press), behavioral activation (Elhai et al., 2016; Y.; Kim et al., 2016), and fear of social evaluation (Wolniewicz et al., in press). Many of these new studies have examined and revealed support for smartphone use frequency as a mediating variable between psychopathology and PSU. This trend in recent research expanding the types of psychopathology examined in relation to PSU is encouraging, in order to better understand PSU as a construct. The present study has several limitations. First, data collected were cross-sectional, so causality between variables cannot be determined. Future research would benefit from repeated mea- sures or longitudinal analysis to explore associations between psychopathology, smartphone use and PSU - a type of design rarely used in the literature, with few exceptions (Jun 2016; Lu, Katoh, Chen, Nagata, & Kitamura, 2014; Thome 2011). Studies would also benefit from using objectively collected smartphone data in conjunction with psychopathology variables (Elhai, Tiamiyu, Weeks, et al., in press; Faurholt-Jepsen et al., 2016), as self-reported smartphone use is not completely accurate when validated against objective use logs (Boase & Ling, 2013; Kobayashi & Boase, 2012). Finally, the present study's reliance on a college student sample is a limitation as well. Nonetheless, this study is one of the first to empirically examine the typology of smartphone use, and also integrate psychopathol- ogy correlates and PSU into the analysis. Our findings indicate the need to consider the heterogeneity embedded in the amount and types of smartphone use in future research. Clinically, it would be important to consider emotional regulation skills and PSU as intervention targets for those presenting with heavy smartphone use (particularly social networking and entertainment related uses). Future research could examine a similar research question using momentary and real-time assessments to capture more nuanced smartphone use typologies. In conclusion, distinct patterns of usage may characterize how individuals use the features of their smartphones. The pattern of such smartphone usage may be related to specific types of psy- chopathology, such as a ruminative thinking style. The pattern of use also appears related to several dimensions of PSU. e, Ha € renstam, & Hagberg, ́ that can fuel the Disclosure section No competing financial interests exist.

7. 165 J.D. Elhai, A.A. Contractor / Computers in Human Behavior 82 (2018) 159e166 References Faurholt-Jepsen, M., Vinberg, M., Frost, M., Debel, S., Christensen, E. M., Bardram, J. E., et al. (2016). Behavioral activities collected through smartphones and the association with illness activity in bipolar disorder. International Journal of Methods in Psychiatric Research, 25(4), 309e323. https://doi.org/10.1002/ mpr.1502. G€ okçearslan, S ¸ ., Mumcu, F. K., Has ¸laman, T., & Çevik, Y. D. (2016). Modelling smartphone addiction: The role of smartphone usage, self-regulation, general self-efficacy and cyberloafing in university students. Computers in Human Behavior, 63, 639e649. https://doi.org/10.1016/j.chb.2016.05.091. Graham, J. W. (2009). Missing data analysis: Making it work in the real world. Annual Review of Psychology, annurev.psych.58.110405.085530. Grellhesl, M., & Punyanunt-Carter, N. M. (2012). Using the uses and gratifications theory to understand gratifications sought through text messaging practices of male and female undergraduate students. Computers in Human Behavior, 28(6), 2175e2181. https://doi.org/10.1016/j.chb.2012.06.024. Gross, J. J. (1998). The emerging field of emotion regulation: An integrative review. Review of General Psychology, 2(3), 271e299. https://doi.org/10.1037/1089- 2680.2.3.271. Gross, J. J., & John, O. P. (2003). Individual differences in two emotion regulation processes: Implications for affect, relationships, and well-being. Journal of Personality and Social Psychology, 85(2), 348e362. https://doi.org/10.1037/0022- 3514.85.2.348. Hamka, F., Bouwman, H., de Reuver, M., & Kroesen, M. (2014). Mobile customer segmentation based on smartphone measurement. Telematics and Informatics, 31(2), 220e227. https://doi.org/10.1016/j.tele.2013.08.006. Hofmann, S. G., Sawyer, A. T., Fang, A., & Asnaani, A. (2012). Emotion dysregulation model of mood and anxiety disorders. Depression and Anxiety, 29(5), 409e416. https://doi.org/10.1002/da.21888. Hormes, J. M., Kearns, B., & Timko, C. A. (2014). Craving Facebook? Behavioral addiction to online social networking and its association with emotion regu- lation deficits. Addiction, 109(12), add.12713. _INal, E. E., Dem_Irc_I, K., Çet_Intürk, A., Akg€ onül, M., & Savas ¸, S. (2015). Effects of smartphone overuse on hand function, pinch strength, and the median nerve. Muscle & Nerve, 52(2), 183e188. https://doi.org/10.1002/mus.24695. Jeong, S.-H., Kim, H., Yum, J.-Y., & Hwang, Y. (2016). What type of content are smartphone users addicted to?: SNS vs. games. Computers in Human Behavior, 54, 10e17. https://doi.org/10.1016/j.chb.2015.07.035. Jun, S. (2016). The reciprocal longitudinal relationships between mobile phone addiction and depressive symptoms among Korean adolescents. Computers in Human Behavior, 58, 179e186. https://doi.org/10.1016/j.chb.2015.12.061. Kardefelt-Winther, D. (2014). A conceptual and methodological critique of internet addiction research: Towards a model of compensatory internet use. Computers in Human Behavior, 31, 351e354. https://doi.org/10.1016/j.chb.2013.10.059. Kim, J.-H. (2017). Smartphone-mediated communication vs. face-to-face interac- tion: Two routes to social support and problematic use of smartphone. Com- puters in Human Behavior, 67, j.chb.2016.11.004. Kim, Y., Jeong, J. E., Cho, H., Jung, D. J., Kwak, M., Rho, M. J., et al. (2016). Personality factors predicting smartphone addiction predisposition: Behavioral inhibition and activation systems, impulsivity, and self-control. PLoS One, 11(8), e0159788. https://doi.org/10.1371/journal.pone.0159788. Kim, J., Seo, M., & David, P. (2015). Alleviating depression only to become prob- lematic mobile phone users: Can face-to-face communication be the antidote? Computers in Human Behavior, 51(Part A), 440e447. https://doi.org/10.1016/ j.chb.2015.05.030. Kobayashi, T., & Boase, J. (2012). No such effect? The implications of measurement error in self-report measures of mobile communication use. Communication Methods and Measures, 6(2), 19312458.2012.679243. Kwon, M., Lee, J. Y., Won, W. Y., Park, J. W., Min, J. A., Hahn, C., et al. (2013). Development and validation of a smartphone addiction scale (SAS). PLoS One, 8(2), e56936. https://doi.org/10.1371/journal.pone.0056936. Lanza, S. T., Tan, X., & Bray, B. C. (2013). Latent class analysis with distal outcomes: A flexible model-based approach. Structural Equation Modeling, 20(1), 1e26. https://doi.org/10.1080/10705511.2013.742377. Lee, Y.-K., Chang, C.-T., Lin, Y., & Cheng, Z.-H. (2014). The dark side of smartphone usage: Psychological traits, compulsive behavior and technostress. Computers in Human Behavior, 31, 373e383. https://doi.org/10.1016/j.chb.2013.10.047. Long, J., Liu, T. Q., Liao, Y. H., Qi, C., He, H. Y., Chen, S. B., et al. (2016). Prevalence and correlates of problematic smartphone use in a large random sample of Chinese undergraduates. BMC Psychiatry, 16, 408. https://doi.org/10.1186/s12888-016- 1083-3. Lopez-Fernandez, O., Honrubia-Serrano, L., Freixa-Blanxart, M., & Gibson, W. (2014). Prevalence of problematic mobile phone use in British adolescents. Cyberp- sychology, Behavior, and Social Networking,17(2), 91e98. https://doi.org/10.1089/ cyber.2012.0260. Lu, X., Katoh, T., Chen, Z., Nagata, T., & Kitamura, T. (2014). Text messaging: Are dependency and excessive use discretely different for Japanese university stu- dents? Psychiatry Research, 216(2), j.psychres.2013.12.024. Lu, X., Watanabe, J., Liu, Q., Uji, M., Shono, M., & Kitamura, T. (2011). Internet and mobile phone text-messaging dependency: Factor structure and correlation with dysphoric mood among Japanese adults. Computers in Human Behavior, Aldao, A., Nolen-Hoeksema, S., & Schweizer, S. (2010). Emotion-regulation strate- gies across psychopathology: A meta-analytic review. Clinical Psychology Re- view, 30(2), 217e237. https://doi.org/10.1016/j.cpr.2009.11.004. Asparouhov, T., & Muth? en, B. (2014). Auxiliary variables in mixture modeling: Three-step approaches Uusing Mplus. Structural Equation Modeling: A Multi- disciplinary Journal, 21(3), 10705511.2014.915181. Augner, C., & Hacker, G. W. (2012). Associations between problematic mobile phone use and psychological parameters in young adults. International Journal of Public Health, 57(2), 437e441. https://doi.org/10.1007/s00038-011-0234-z. Billieux, J., Maurage, P., Lopez-Fernandez, O., Kuss, D. J., & Griffiths, M. D. (2015a). 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), 156e162. https://doi.org/10.1007/s40429-015- 0054-y. Billieux, J., Philippot, P., Schmid, C., Maurage, P., De Mol, J., & Van der Linden, M. (2015b). Is dysfunctional use of the mobile phone a behavioural addiction? Confronting symptom-based versus process-based approaches. Clinical Psy- chology & Psychotherapy, 22(5), 460e468. https://doi.org/10.1002/cpp.1910. Blumler, J. G. (1979). The role of theory in uses and gratifications studies. Communication Research, 6(1), 009365027900600102. Boase, J., & Ling, R. (2013). Measuring mobile phone use: Self-report versus log data. Journal of Computer-mediated Communication, 18(4), 508e519. https://doi.org/ 10.1111/jcc4.12021. Brinker, J. K., & Dozois, D. J. A. (2009). Ruminative thought style and depressed mood. Journal of Clinical Psychology, 65(1), 1e19. https://doi.org/10.1002/ jclp.20542. Cazzulino, F., Burke, R. V., Muller, V., Arbogast, H., & Upperman, J. S. (2014). Cell phones and young drivers: A systematic review regarding the association be- tween psychological factors and prevention. Traffic Injury Prevention, 15(3), 234e242. https://doi.org/10.1080/15389588.2013.822075. Chotpitayasunondh, V., & Douglas, K. M. (2016). How “phubbing” becomes the norm: The antecedents and consequences of snubbing via smartphone. Com- puters in Human Behavior, 63, 9e18. https://doi.org/10.1016/j.chb.2016.05.018. Contractor, A. A., Frankfurt, S., Weiss, N. H., & Elhai, J. D. (2017). Latent-level re- lations between DSM-5 PTSD symptom clusters and problematic smartphone use. Computers in Human Behavior, 72, 170e177. https://doi.org/10.1016/ j.chb.2017.02.051. Davis, R. A. (2001). A cognitive-behavioral model of pathological Internet use. Computers in Human Behavior, 17(2), 187e195. https://doi.org/10.1016/S0747- 5632(00)00041-8. Demirci, K., Orhan, H., Demirdas, A., Akpinar, A., & Sert, H. (2014). Validity and reliability of the Turkish version of the smartphone addiction scale in a younger population. Bulletin of Clinical Psychopharmacology, 24(3), 226. https://doi.org/ 10.5455/bcp.20140710040824. 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, 411e420. https://doi.org/10.1016/ j.chb.2014.12.039. Dhir, A., Chen, S., & Nieminen, M. (2015). Predicting adolescent Internet addiction: The roles of demographics, technology accessibility, unwillingness to commu- nicate and sought Internet gratifications. Computers in Human Behavior, 51(Part A), 24e33. https://doi.org/10.1016/j.chb.2015.04.056. Elhai, J. D., Dvorak, R. D., Levine, J. C., & Hall, B. J. (2017a). Problematic smartphone use: A conceptual overview and systematic review of relations with anxiety and depression psychopathology. Journal of Affective Disorders, 207, 251e259. https://doi.org/10.1016/j.jad.2016.08.030. Elhai, J. D., Hall, B. J., Levine, J. C., & Dvorak, R. D. (2017b). Types of smartphone usage and relations with problematic smartphone behaviors: The role of con- tent consumption vs. social smartphone use. Cyberpsychology: Journal of Psy- chosocial Research on Cyberspace, 11(2), 3. https://doi.org/10.5817/CP2017-2-3. Elhai, J. D., Levine, J. C., Dvorak, R. D., & Hall, B. J. (2016). Fear of missing out, need for touch, anxiety and depression are related to problematic smartphone use. Computers in Human Behavior, j.chb.2016.05.079. Elhai, J. D., Levine, J. C., Dvorak, R. D., & Hall, B. J. (2017c). Non-social features of smartphone use are most related to depression, anxiety and problematic smartphone use. Computers in Human Behavior, 69, 75e82. https://doi.org/ 10.1016/j.chb.2016.12.023. Elhai, J. D., Tiamiyu, M. F., & Weeks, J. W. (2018a). Depression and social anxiety in relation to problematic smartphone use: The prominent role of rumination. Internet Research (in press). Elhai, J. D., Tiamiyu, M. F., Weeks, J. W., Levine, J. C., Picard, K. J., & Hall, B. J. (2018b). Depression and emotion regulation predict objective smartphone use measured over one week. Personality and Individual Differences. https://doi.org/10.1016/ j.paid.2017.04.051 (in press). Elhai, J. D., Vasquez, J. K., Lustgarten, S. D., Levine, J. C., & Hall, B. J. (2018c). Proneness to boredom mediates relationships between problematic smart- phone use with depression and anxiety severity. Social Science Computer Review. https://doi.org/10.1177/0894439317741087 (in press). 329e341. https://doi.org/10.1080/ 60, 549e576. https://doi.org/10.1146/ 9e36. https://doi.org/10.1177/ 2079e2088. https://doi.org/10.1111/ 282e291. https://doi.org/10.1016/ 126e143. https://doi.org/10.1080/ 63, 509e516. https://doi.org/10.1016/ 255e262. https://doi.org/10.1016/

8. 166 J.D. Elhai, A.A. Contractor / Computers in Human Behavior 82 (2018) 159e166 27(5), 1702e1709. https://doi.org/10.1016/j.chb.2011.02.009. Mansell, W., Harvey, A. G., Watkins, E. R., & Shafran, R. (2008). Cognitive behavioral processes across psychological disorders: A review of the utility and validity of the transdiagnostic approach. International Journal of Cognitive Therapy, 1(3), 181e191. https://doi.org/10.1521/ijct.2008.1.3.181. Masyn, K. E. (2013). Latent class analysis and finite mixture modeling. In T. D. Little (Ed.), The Oxford handbook of quantitative methods in psychology Vol. 2: Statistical analysis (pp. 551e611). Oxford, England: Oxford University Press. McLachlan, G. J., & Peel, D. (2000). Finite mixture models. New York: Wiley & Sons. Mennin, D. S., & Fresco, D. M. (2013). What, me worry and ruminate about DSM-5 and RDoC? The importance of targeting negative self-referential processing. Clinical Psychology: Science and Practice, 20(3), 258e267. https://doi.org/10.1111/ cpsp.12038. Muth? en, B. O. (2008). Latent variable hybrids: Overview of old and new models. In G. R. Hancock, & K. M. Samuelsen (Eds.), Advances in latent variable mixture models (pp. 1e24). Charlotte, North Carolina: Information Age Publishing, Inc. Nylund, K. L., Asparouhov, T., & Muth? en, B. O. (2007). Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study. Structural Equation Modeling, 14(4), 535e569. https://doi.org/ 10.1080/10705510701575396. Oulasvirta, A., Rattenbury, T., Ma, L., & Raita, E. (2012). Habits make smartphone use more pervasive. Personal and Ubiquitous Computing, 16(1), 105e114. https:// doi.org/10.1007/s00779-011-0412-2. Park, N., Kim, Y.-C., Shon, H. Y., & Shim, H. (2013). Factors influencing smartphone use and dependency in South Korea. Computers in Human Behavior, 29(4), 1763e1770. https://doi.org/10.1016/j.chb.2013.02.008. Raftery, A. E. (1995). Bayesian model selection in social research. Sociological Methodology, 25, 111e163. https://doi.org/10.2307/271063. Samaha, M., & Hawi, N. S. (2016). Relationships among smartphone addiction, stress, academic performance, and satisfaction with life. Computers in Human Behavior, 57, 321e325. https://doi.org/10.1016/j.chb.2015.12.045. Seo, D. G., Park, Y., Kim, M. K., & Park, J. (2016). Mobile phone dependency and its impacts on adolescents' social and academic behaviors. Computers in Human Behavior, 63, 282e292. https://doi.org/10.1016/j.chb.2016.05.026. Shin, D.-H. (2015). Effect of the customer experience on satisfaction with smart- phones: Assessing smart satisfaction index with partial least squares. Tele- communications Policy, 39(8), j.telpol.2014.10.001. Shin, D.-H., Shin, Y.-J., Choo, H., & Beom, K. (2011). Smartphones as smart peda- gogical tools: Implications for smartphones as u-learning devices. Computers in Human Behavior, 27(6), 2207e2214. https://doi.org/10.1016/j.chb.2011.06.017. Song, I., LaRose, R., Eastin, M. S., & Lin, C. A. (2004). Internet gratifications and Internet addiction: On the uses and abuses of new media. Cyberpsychology, Behavior, and Social Networking, 7(4), 384e394. cpb.2004.7.384. Spaapen, D. L., Waters, F., Brummer, L., Stopa, L., & Bucks, R. S. (2014). The emotion regulation Questionnaire: Validation of the ERQ-9 in two community samples. Psychological Assessment, 26(1), 46e54. https://doi.org/10.1037/a0034474.supp. Tanner, A., Voon, D., Hasking, P., & Martin, G. (2012). Underlying structure of ruminative thinking: Factor analysis of the ruminative thought style ques- tionnaire. Cognitive Therapy and Research, 37(3), 633e646. https://doi.org/ 10.1007/s10608-012-9492-1. Thome e, S., Ha € renstam, A., & Hagberg, M. (2011). Mobile phone use and stress, sleep disturbances, and symptoms of depression among young adults: A prospective cohort study. BMC Public Health, 11, 66. https://doi.org/10.1186/1471-2458-11- 66. Tull, M. T., & Aldao, A. (2015). New directions in the science of emotion regulation. Current Opinion in Psychology, 3. https://doi.org/10.1016/j.copsyc.2015.03.009. iv-x. Wang, J.-L., Wang, H.-Z., Gaskin, J., & Wang, L.-H. (2015). The role of stress and motivation in problematic smartphone use among college students. Computers in Human Behavior, 53, 181e188. https://doi.org/10.1016/j.chb.2015.07.005. Wei, P.-S., & Lu, H.-P. (2014). Why do people play mobile social games? An exam- ination of network externalities and of uses and gratifications. Internet Research, 24(3), 313e331. https://doi.org/10.1108/IntR-04-2013-0082. Wolniewicz, C. A., Tiamiyu, M. F., Weeks, J. W., & Elhai, J. D. (2018). Problematic smartphone use and relations with negative affect, fear of missing out, and fear of negative and positive evaluation. Psychiatry Research. https://doi.org/10.1016/ j.psychres.2017.09.058 (in press). Xie, Y., Szeto, G. P., Dai, J., & Madeleine, P. (2016). A comparison of muscle activity in using touchscreen smartphone among young people with and without chronic neck-shoulder pain. Ergonomics, 00140139.2015.1056237. Yuan, K. H., & Bentler, P. M. (2000). Three likelihood-based methods for mean and covariance structure analysis with nonnormal missing data. In M. E. Sobel, & M. P. Becker (Eds.), Sociological methodology 2000 (pp. 165e200). Washington, D.C.: Wiley-Blackwell. Zhitomirsky-Geffet, M., & Blau, M. (2016). Cross-generational analysis of predictive factors of addictive behavior in smartphone usage. Computers in Human Behavior, 64, 682e693. https://doi.org/10.1016/j.chb.2016.07.061. https://doi.org/10.1089/ 59(1), 61e72. https://doi.org/10.1080/ 627e641. https://doi.org/10.1016/ ́