Computers in Human Behavior 53 (2015) 181–188

Contents lists available at ScienceDirect

Computers in Human Behavior

journal homepage: www.elsevier.com/locate/comphumbeh

Full Length Article

The role of stress and motivation in problematic smartphone use

among college students
Jin-Liang Wang a,⇑, Hai-Zhen Wang b, James Gaskin c, Li-Hui Wang a
a
Center for Mental Health Education & Center for Psychological Engineering, School of Psychology, Southwest University, Chongqing, China
b
Department of Tourism and Art for Humanity, Chongqing Youth & Vocational Technical College, China
c
Department of Information System, Brigham Young University, USA

a r t i c l e i n f o a b s t r a c t

Article history: Although numerous studies have examined factors that influence problematic smartphone use, few have
Received 19 January 2015 investigated the interactions between different types of motivation and psychosocial well-being factors
Revised 24 June 2015 in facilitating problematic smartphone use. Thus, this study analyzed the moderating role of perceived
Accepted 3 July 2015
stress on the relationship between entertainment or escapism motivation and problematic smartphone
use. Using a sample of 600 typical smartphone users (Chinese college students) we found:
(1) Perceived stress moderates the relationship between entertainment motivation and problematic
Keywords:
smartphone use, and between escapism motivation and problematic smartphone use among users scor-
Entertainment
Escapism
ing high on problematic use; (2) perceived stress moderated the relationship between entertainment
Stress motivation and problematic use also among users scoring low on problematic smartphone use, but for
Problematic smartphone use these users stress did not moderate the relationship between escapism motivation and problematic
use. Our study highlights the importance of unresolved real-life problems in facilitating problematic
smartphone use, suggesting that the resolution of these problems might be a good starting point when
researchers design interventions for people excessively relying on smartphones.
Ó 2015 Published by Elsevier Ltd.

1. Introduction smartphone can be problematic, which might be accompanied

by a diminished sense of volitional control and induce
Smartphones, distinct from standard mobile phones in terms of persistent smartphone activity (Thomee, Harenstam, & Hagberg,
the operating system and being equipped with advanced 3G or 4G 2011), leading to mental health symptoms such as sleep distur-
features and capabilities (e.g., socializing on Facebook and view- bances, depression, and psychological distress (Chesley, 2005;
ing/posting videos on YouTube), have attracted a significant num- Thomee, Eklof, Gustafsson, Nilsson, & Hagberg, 2007; Thomee
ber of users (Bian & Leung, 2014). The Statista (2013a) reported a et al., 2011).
forecasted smartphone penetration rate of 37% worldwide; this Amount of studies have been done focusing on addiction
number continues to grow. The smartphone user penetration rates (Young, 2004), problematic use (Chen & Kim, 2013; Martinotti
are estimated to reach about 65.8% of the population in the United et al., 2011), or excessive use (Bener, Al-Mahdi, Vachhani,
Kingdom (Statista, 2013b), 63.5% in the U.S. (Statista, 2013c), and Al-Nufal, & Ali, 2010; Buckner, Castille, & Sheets, 2012) of the
49% in China (Statista, 2013d) respectively by 2017. Smartphones Internet and other technologies. Young (2004) has defined
have become a perceived necessity in personal and work lives. Internet addiction as an impulse-control disorder not involving
Not only are they used for social networking purposes, but also an intoxicant, making it a ‘behavioral addiction’ akin to gambling
for features and functions like reading e-books, answering addiction and unlike alcoholism. This type of addiction has since
e-mails, sending short messages, and playing online games. been extended to incorporate other forms of technology, such as
Generally, people use smartphones for entertainment or to relieve social networking sites (Kuss & Griffiths, 2011) and smartphone
stress (Lee, Chang, Lin, & Cheng, 2014) because they can yield addictions (Bian & Leung, 2014; Kwon et al., 2013). However, con-
immediate gratification. However, spending too much time on sidering that debate exists in terms of behavioral addictions (Peters
& Malesky, 2008), and that problematic use implies that a person
⇑ Corresponding author at: Center for Mental Health Education & Center for uses their phone in such a way that it leads to problematic out-
Psychological Engineering, School of Psychology, Southwest University, Beibei, comes but without making a clinical judgment as to whether this
Chongqing 400715, China. person has the mental disorder of addiction or not, we chose to
E-mail address: [email protected] (J.-L. Wang).

http://dx.doi.org/10.1016/j.chb.2015.07.005
0747-5632/Ó 2015 Published by Elsevier Ltd.
182 J.-L. Wang et al. / Computers in Human Behavior 53 (2015) 181–188

use the more conservative term ‘‘problematic use’’ rather than self-esteem) and motivation (i.e., escapism) on problematic online
‘‘addiction’’. gaming and found that both stress and self-esteem moderated the
Numerous studies on smartphone use have been conducted association between escapism and problematic online gaming. His
from the perspectives of predictive psychological factors study showed that people who were more stressed or had lower
(Beranuy, Oberst, Carbonell, & Chamarro, 2009; Bianchi & self-esteem and played primarily for escapism purposes, tended
Phillips, 2005; Chiu, 2014; Ehrenberg, Juceks, & White, 2008; Ha, to experience more problematic outcomes than those who were
Chin, Park, Ryu, & Yu, 2008; Hong, Chiu, & Huang, 2012; Jenaro, less stressed or had high self-esteem. Crucially, the direct effects
Noelia, Maria, Francisca, & Cristina, 2007; Walsh, White, Cox, & of escapism or stress alone did not have a significant impact on
Young, 2011), predictive social factors (Walsh et al., 2011), and problematic outcomes when the interaction effect between the
family environmental factors (Toda et al., 2008). However, few two variables was controlled for, which suggests that it is the com-
have examined the role played by the interactions between moti- bination of stress and escapism that leads to more problems.
vation and psychosocial well-being factors in problematic smart-
phone use. Thus, the purpose of this study is to establish a model 2.2. Gratification from smartphone usage
of interactive effect of perceived stress and motivation on problem-
atic uses of smartphone. The reason for investigating the interac- Users might derive pleasurable and rewarding feelings from
tions between motivations and psychosocial well-being variables media usage. As proposed by the uses and gratification perspective,
in problematic smartphone use is that a direct-effects only individuals’ excessive use of media may be way of satisfying their
approach does not account for how relationships between predic- psychological needs (Song, 2004). Specifically, individuals tend to
tors may impact problematic use (Kardefelt-Winther, 2014a). In actively select particular media and consume content to satisfy
other words, with a direct effects approach it remains unclear their psychological needs (Katz et al., 1974). Previous studies have
how a person’s well-being and his/her motivations for using the found that people use new media to satisfy their needs for infor-
smartphone together shape how the phone is being used (Katz, mation, convenience, entertainment, self-expression, social status,
Blumler, & Gurevitch, 1974) and when this leads to problematic and social interaction (Hawi, 2012; Ko, Cho, & Roberts, 2005;
outcomes (Kardefelt-Winther, 2014b). As a consequence, direct Papacharissi & Rubin, 2000). In relation to this, previous studies
effects models focusing only on psychosocial well-being as the sole have identified various motives influencing problematic Internet
predictor of problematic use, as has been the tendency in the field use, such as the need for information, interpersonal relationships,
so far, may be inadequate in their elucidation of its underlying entertainment, as well as the need to kill time and find an escape
mechanisms. Additional research is therefore needed to accurately outlet (Khang, Kim, & Kim, 2013; Lowry, Gaskin, & Moody, in
simulate the realistically more complex and interactive nature of press). Regarding mobile phone use, users’ needs for entertain-
the antecedents of problematic use. ment, stress relief, instantaneousness, mobility, interest, informa-
One theory of problematic technology use suggests that prob- tion, and social status serve as primary motives for using the
lematic smartphone use occurs when users repeatedly go online medium; this could lead to excessive reliance on mobile phones
as a means to cope with life problems (Kardefelt-Winther, 2014a, (Lee et al., 2014). In other studies, it has been revealed that relax-
2014b, 2014c). According to the Compensatory Internet Use theory ing, relieving stress (Ho & Syu, 2010), strengthening users’ family
(see Kardefelt-Winther, 2014b for a detailed account), a user’s bonds, expanding their psychological neighborhoods, facilitating
needs and motivation to cope with life problems via Internet use symbolic proximity to the people they call (Wei & Lo, 2006), as well
is determined by the degree of life problems experienced, which as pass time (Wei, 2008), are motives for using mobile phone.
implies that the relationship between motivations and problematic Furthermore, according to a more recent study, problematic media
smartphone use should be moderated by life problems. use may occur when individuals use media to alleviate psycholog-
Specifically, people may use smartphones for entertainment or ical distress arising from social and interpersonal confrontation
escapism purposes, as this can help them cope with low psychoso- (Khang et al., 2013).
cial well-being such as stress. Therefore, as compared to people
with high psychosocial well-being, those with low psychosocial 2.3. Perceived stress and problematic smartphone use
well-being should be more likely to use smartphones to counteract
their negative feelings which might lead to a higher degree of Perceived stress is defined as a situation that is appraised as
problematic outcomes. Following this line of thought, this study threatening or otherwise demanding, with no sufficient resources
examines whether perceived stress moderates the association to enable one to cope with the situation (Cohen, Kamarck, &
between two types of motivation (i.e., entertainment and escap- Mermelstein, 1983). Stressful environments and life events could
ism) and problematic smartphone use. provoke interacting somatic and psychological states that trans-
form stress responsiveness and, thus, render individuals vulnerable
to addiction (Hawkins, Catalano, & Miller, 1992; Rhodes & Jason,
2. Literature review 1990). In terms of the association between stress and problematic
new media usage, Jie et al. (2014) reported that stressors stemming
2.1. Compensatory Internet Use theory from interpersonal and school-related problems and anxiety symp-
toms were significantly associated with Internet addiction, after
Based on the Compensatory Internet Use theory, low psychoso- controlling for demographic characteristics. In addition, family
cial well-being may be regarded as underlying causes of some and general emotional stress, respectively, have been found to
of the motivations for Internet use, and in certain cases, this have a positive predictive power for smartphone addiction (Chiu,
interplay may lead some people to spend so much time with 2014). When confronted with excessive life stress, young people
their smartphone that it eventually results in problematic are inclined to increase their Internet use to manage moods, com-
consequences for their life. This theory can be used to explain pensate for social interaction, and escape from reality (Leung,
why some people persist in using the Internet or smartphones 2007). Therefore, researchers have argued that some individuals
despite experiencing problematic outcomes, which is by some may use the new media (e.g. Internet and smartphone) primarily
scholars referred to as addictive behaviors (Kardefelt-Winther, to cope with stress (Anderson & Collin, 1996). This notion has been
2014b). Kardefelt-Winther (2014c) investigated the interaction corroborated in a study by Whang, Lee, and Chang (2003), which
effects of psychosocial well-being variables (i.e., stress and found that work-related stress is positively related to Internet
 J.-L. Wang et al. / Computers in Human Behavior 53 (2015) 181–188 183

use amongst people displaying symptoms of Internet addiction, An important methodological point proposed by
while those not using the Internet excessively reported higher Kardefelt-Winther (2014c) is that some of the relationships that
levels of drinking when stressed than the former. Additionally, in illustrate why problematic outcomes occur may only be observed
a study by Snodgrass et al. (2014), it was suggested that among individuals who experience a high degree of problematic
problematic online gaming in the World of Warcraft can be outcomes. This hypothesis was also confirmed empirically in his
conceptualized as a response to pre-existing life stress, which, work. The argument is that researchers should not expect scores
among highly stressed individuals, magnifies rather than relieves of problematic outcomes to be normally distributed; rather, we
stress levels. may expect that more people will receive low rather than high
scores on such a measure (Kardefelt-Winther, 2015). Therefore,
3. Hypotheses studies that are concerned with understanding behaviors of people
at the high end of this continuum need to account for this in their
As suggested by Kardefelt-Winther (2014c), it would be prob- sampling procedures. If this is not taken into account, problem
lematic to suggest that everyone who uses popular new media to users may not be adequately represented in samples where a
relieve stress or entertain themselves is at risk of also overusing majority of respondents are non-problem users, which means that
the media. To entertain oneself and to escape reality are two of explanatory factors pertaining to this specific part of the popula-
the primary functions of smartphone use, and may help people tion may remain hidden (Kardefelt-Winther, 2014c, 2015). In
relax (Lowry et al., in press). Since only a minority of all users expe- Kardefelt-Winther’s studies, analysis of the full study sample and
rience problematic outcomes, it is important to focus on that the low-problem sample revealed no significant interaction effects
minority when seeking to understand why only this particular between escapism and stress or low self-esteem on problematic
group experiences problematic outcomes (Kardefelt-Winther, online gaming or online gambling. Significant interaction effects
2014c). This would be helpful in understanding the differences of psychosocial problems and motivation on the extent of prob-
between those who report many problematic outcomes and those lematic gaming and gambling were found only for the
who report few, and might serve as a useful way of exploring the high-problems groups. In other words, such interactions may be
factors influencing problematic Internet use (Kardefelt-Winther, one factor that explains why some people experience more prob-
2014c). Statistical analysis on the full sample would reveal general lematic outcomes than others (and thus are found in the
tendencies within that sample, but interaction effects that may high-problem group) (Kardefelt-Winther, 2014c, 2015). One way
explain problematic smartphone use by some respondents may to explain this phenomenon is that the high-problem group use
be more observable if these respondents are examined separately online games or gambling as a coping strategy to relieve stress or
(Kardefelt-Winther, 2014c, p. 70). Therefore, in the current study, combat low self-esteem, therefore those who experience more
respondents were categorized according to their problematic problematic outcomes do so because their motivation for use is
smartphone use scores, so as to enable such a comparative partially based on a need to cope with psychosocial problems
analysis. which makes it difficult to cut back on the time spent gaming or
Based on the aforementioned Compensatory Internet Use gambling. In the current study, we follow Kardefelt-Winther’s
theory, our hypotheses are as follows: (2014c, 2015) logic and hypothesize that the interaction effects
between motivations (i.e. escapism and entertainment) and
H1. The relationship between entertainment motivation and psychosocial problems (e.g., perceived stress) will be observable
problematic smartphone use is moderated by perceived stress only among those scoring high on excessive smartphone use, as
among users scoring high on problematic smartphone use. indicative of a coping strategy.
Additionally, previous studies have suggested that gender
(Jenaro et al., 2007; Martinotti et al., 2011; Takao, Takahashi, &
Kitamura, 2009) and age (e.g., Walsh et al., 2011) may influence
H2. The relationship between escapism and problematic smart- mobile phone usage. For example, female were reported to score
phone use is moderated by perceived stress among users scoring higher on the problematic mobile phone use scale (Takao et al.,
high on problematic smartphone use. 2009). Walsh et al. (2011) found that age (younger) influence the
frequency of and involvement with users’ mobile use. Hence,
gender and age were controlled in our model.
H3. The relationship between entertainment motivation and
problematic smartphone use is not moderated by perceived stress
among users scoring low on problematic smartphone use. 4.2. Participants

H4. The relationship between escapism and problematic smart- A total of 600 college students were recruited from two colleges
phone use is not moderated by perceived stress among users in Chongqing China, from whom 549 (Mage = 18.39, SDage = 1.92,
scoring low on problematic smartphone use. 34.8% male) valid questionnaires were obtained, with a response
rate of 90%. The whole sample consisted of 62.7% in the first grade,
13.3% in the second, 23.7% in the third grade, and 0.4% in the
4. Methods fourth. Over half (52.5%) of the participants majored in natural
science (math, physics, chemistry, biology, life sciences, and com-
4.1. Analytical strategy puter sciences), and 46.7% indicated majoring in social science
(economics, business administration, education, philology, law,
SPSS 18.0 and AMOS 18.0 software were used for our data anal- politics, and philosophy). All of them indicated that they are smart-
yses. Before conducting regression analyses, we conducted a con- phone users. The participants were recruited during the Spring
firmatory factor analysis to assess the goodness of fit of the Semester, 2014, through announcements in various lecture halls.
measurement model by using AMOS 18.0, as suggested by Kline These advertisements included a brief description of the nature
(2005). Descriptive analyses were performed on all variables. of the study. Participants were compensated with ¥10, which is
Pearson correlation coefficients were used to assess the strengths equivalent to USD $1.61. The participants completed the surveys
of linear relationships between pairs of interest variables. in a large lecture hall in one of four 30-min sessions.
184 J.-L. Wang et al. / Computers in Human Behavior 53 (2015) 181–188

4.3. Measures Table 1

Means, standard deviations, and zero-order correlations of variables for the
problematic smartphone-use group (N = 264).
4.3.1. Smartphone Addiction Scale for College Students (SAS-C)
The SAS-C is a 22-item self-report scale designed by Su et al. Variable M SD Smartphone Entertainment Escapism
(2014). It measures six factors relating to problematic smartphone use

use, including withdrawal behavior, salience behavior, social com- Smartphone use 3.29 .42 –
fort, negative effects, use of applications (APPs), and APP renewal. Entertainment 3.40 .80 .29a –
Escapism 2.63 1.08 .22a .26a –
Responses were given through the use of a 5-point Likert scale Stress 2.96 .42 .33a .01 .15b
ranging from 1 (strongly disagree) to 5 (strongly agree), with high
a
scores representing high levels of addiction (M = 2.76, SD = .65, p < .001.
b
p < .01.
a = .91). Sample items included, ‘‘My friends have told me that I
spend too much time using my smartphone’’ and ‘‘I think I need
to spend more time using my smartphone to get satisfied.’’
Table 1. The results show that entertainment and escapism, as
4.3.2. Perceived Stress Scale (PSS) types of motivation, as well as perceived stress, significantly corre-
The participants’ perceived stress was assessed with a 14-item lated with problematic smartphone use (ps < .01).
scale developed by Cohen et al. (1983). Also, using a 5-point Table 2 shows that escapism motivation and the moderator
Likert scale, the scale measures participants’ perceived level of con- variable (i.e., perceived stress) significantly correlated with smart-
trol over their lives in the past month, with a high score indicating phone use (ps < .001) among users with few problematic outcomes.
perceived lack of control and high stress levels (M = 2.84, SD = .41, However, there was no significant correlation between entertain-
a = 0.74). Sample items included, ‘‘In the last month, how often ment motivation and problematic smartphone use for these
have you been upset because of something that happened respondents.
unexpectedly?’’ and ‘‘How often have you felt that you were unable We ran a regression analysis regarding H1 and included an
to control the important things in your life?’’ interaction term of entertainment motivation and perceived stress
in the regression model. We followed Cohen, Cohen, West, and
4.3.3. Motivation for smartphone use Aiken (2003) recommendations for plotting the results as an aid
Entertainment and escapism, representing types of motivation, in interpreting the nature of the interaction between the two pre-
were measured through items adopted from Wang, Jackson, Wang, dictor variables. Specifically, the relationship between the first pre-
and Gaskin’s (2015) Social Media Usage Motivation Scale. The dictor variable (i.e., entertainment motivation) and the dependent
items were revised to fit the context of smartphone use. variable (i.e., problematic smartphone use) was plotted when
Specifically, entertainment motivation was measured through four levels of the second predictor variable (i.e., the moderator) were
items, including the sample item, ‘‘I use my smartphone to relax’’ one standard deviation below and one standard deviation above
(M = 3.25, SD = .79, a = .83). Escapism motivation was measured the mean for that variable (i.e. the moderator). We also tested
through three items, including the sample item, ‘‘I use the smart- the statistical significance of each of these two slopes (Aiken &
phone to get away from what I am doing’’ (M = 2.31, SD = 1.02, West, 1991), which represented the simple effect of the predictor
a = .85). All the response options were in the form of a 5-point variable on problematic smartphone use at two levels of the
Likert scale. moderator variable.
As shown in Table 3, entertainment, perceived stress, and the
interaction term of entertainment and perceived stress were signif-
5. Results icantly related to problematic smartphone use, with b = .29,
p < .001; b = .31, p < .001; and b = .21, p < .001, respectively.
5.1. Measurement model Regarding the correspondence of the slopes with the two-level
moderator (i.e. perceived stress), the results of a simple-effect
Result of the confirmatory factor analysis showed that the analysis indicated that the positive relationship between
measurement model yielded an adequate fit, with CFI = .96, entertainment motivation and problematic smartphone use was
TLI = .95, RMSEA = .07. Standardized factor loadings ranged from significant when perceived stress levels were high (b = .58,
.41 to .84 and were all significant at the p < .001 level. p < .001), while the association between entertainment motivation
and problematic smartphone use was not significant when per-
5.2. Descriptive and regression analyses ceived stress levels were low (b = .01, p > .05). Therefore, the
results supported H1.
In order to facilitate a comparative analysis, the sample was A similar procedure to the one for H1 was conducted for H2. As
divided into the high problematic smartphone-use group and the shown in Table 4, escapism motivation, perceived stress, and the
non-problematic smartphone-use group, based on the respon- interaction term of escapism and perceived stress were signifi-
dents’ scores on the SAS-C (Moveruse = 2.81, SDoveruse = .41). The cantly related to problematic smartphone use, with b = .17,
respondents with Moveruse = 2.81 or higher were referred to as the p < .01; b = .29, p < .001; and b = .12, p < .05, respectively. With
problematic smartphone-use group, while those with
Moveruse < 2.81 were referred to as non-problematic users. This
resulted in a final sample size of N = 264 (Mage = 18, SDage = 3.85, Table 2
Moveruse = 3.288, SDoveruse = .42, male 41.3%) for the problematic Means, standard deviations, and zero-order correlations of variables for non-
problematic smartphone-use group (N = 285).
smartphone users, and N = 281 (Mage = 18.8, SDage = 3.24,
Moveruse = 2.27, SDoveruse = .40, male 28.8%) for the Variable M SD Smartphone use Entertainment Escapism
non-problematic users. The scores obtained by the two groups on Smartphone use 2.27 .39 –
SAS-C proved significantly different (t = 28.99, p < .001). Entertainment 3.11 .75 .01 –
Means, standard deviations, and zero-order correlations Escapism 2.01 .87 .35a .08 –
Stress 2.73 .38 .30a .28a .29a
obtained on the variables among respondents with high problem-
a
atic outcomes (i.e. problematic smartphone use) are presented in p < .001.
 J.-L. Wang et al. / Computers in Human Behavior 53 (2015) 181–188 185

Table 3 Table 5
Multiple regression model on mean SAS-C scores with interaction term of stress and Multiple regression model on mean SAS-C scores with interaction term of stress and
entertainment motivation among respondents in the problematic smartphone-use entertainment motivation for respondents in the non-problematic smartphone-use
group, N = 264. group, N = 285.

Model Variables b T p Model Model Variables b T p Model

R2 R2
Problematic .24 Problematic .138
smartphone-use smartphone-
Gender .06 .94 .348 use
Age .01 .12 .908 Gender .17 2.64 .009
Entertainment .29 5.31 .000 Age .06 1.68 .095
Stress .31 5.45 .000 Stress .31 5.07 .000
Entertainment ⁄ Stress .21 3.71 .000 Entertainment .10 .10 .095
Entertainment ⁄ Stress .13 2.27 .024

regard to the slopes’ correspondence with the two-level moderator

(i.e. perceived stress), the result of a simple-effect analysis indi- Table 6
cated that the positive relationship between escapism motivation Multiple regression model on mean SAS-C scores with interaction term of stress and
and problematic smartphone use was significant when scores on escapism motivation for respondents in the non-problematic smartphone-use group,
perceived stress were high (b = .42, p < .001), while the association N = 285.

between entertainment motivation and problematic smartphone Model Variables b T P Model

use was not significant when scores on perceived stress were R2
low (b = .09, p > .05). Therefore, H2 was supported (see Table 5). Problematic 0.198
We ran another regression analysis on the non-problematic smartphones-use
smartphone-use group, to test H3. The results showed that per- Gender .12 2.00 .046
Age .08 1.30 .194
ceived stress and the interaction term of entertainment and per-
Escapism .28 4.82 .000
ceived stress were significantly related to problematic Stress .19 3.33 .001
smartphone use, with b = .31, p < .001, and b = .13, p < .05, respec- Escapism ⁄ Stress .06 1.09 .277
tively. However, entertainment was not related to problematic
smartphone use, with b = .10, p > .05. With regard to the slopes cor-
respondence with the two-level moderator (i.e. perceived stress), whether perceived stress moderates the association between two
the results of a simple-effect analysis indicated that the relation- types of motivation and the extent of problematic smartphone
ship between entertainment motivation and problematic smart- usage, using a procedure proposed by Kardefelt-Winther (2014b,
phone use was not significant neither when scores on perceived 2014c, 2015). Furthermore, Kardefelt-Winther (2014c) suggested
stress were high (b = .10, p > .05), nor when scores on perceived that quantitative studies focusing on a particular group of people
stress were low (b = -.001, p > .05). with certain characteristics must isolate this group from the wider
We ran the last regression analysis on respondents in sample, or significant effects could remain hidden. Therefore, in the
non-problematic smartphone-use group, to test H4. This model current study, we divided the respondents into two groups, based
included an interaction term of escapism and perceived stress. In on their scores on the SAS-C, and separately investigated the asso-
support of H4, the model found no significant interaction effect ciations between perceived stress, motivation, and problematic
of escapism and stress on problematic smartphone use, since we smartphone use.
found a non-significant interaction term with b = .06, p > .05 The findings supported three of our four hypotheses. H1 stated
(see Table 6). that the relationship between entertainment and problematic
smartphone use would be moderated by perceived stress, and H2
6. Discussion stated that the association between escapism motivation and prob-
lematic smartphone use would also be moderated by perceived
Our contribution towards literature is that we tested the mod- stress. The two interaction terms were both significant (b = .21,
erating role of a factor related to psychosocial well-being (i.e., per- p < .001; and b = .12, p < .05, respectively). Further, we found that,
ceived stress) on the association between smartphone use among participants with high scores on the SAS-C, which indicated
motivations and problematic smartphone use. Our findings high- the extent of problematic smartphone use, entertainment and
light the vital role played by real-life needs in explaining problem- escapism motivation, as well as perceived stress were positively
atic smartphone use. Our findings also demonstrated the related to problematic smartphone use. This indicated that individ-
adherence of problematic smartphone use to the tenets of the uals with high escapism and entertainment motivations for their
Compensatory Internet Use theory. Specifically, we investigated smartphone use who are also more stressed, were more likely to
experience problematic outcomes from their smartphone use than
Table 4 those who are less stressed. These findings are in line with previ-
Multiple regression model on mean SAS-C scores with interaction term of stress and ous studies (Chiu, 2014; Jie et al., 2014; Khang et al., 2013; Lee,
escapism motivation for respondents in the problematic smartphone-use group,
Song et al., 2014). For example, Lee, Song et al. (2014) reported that
N = 264.
users’ need for entertainment and stress relief were two of the pri-
Model Variables b T p Model mary motives for problematic use of mobile phones. Khang et al.
R2
(2013) found that entertainment and finding an outlet for escape
Problematic .15 were predictors of Internet addiction. In a study by Chiu (2014), life
smartphone-use
stress was found to be positively related to smartphone addiction.
Gender .02 .22 .825
Age .04 .62 .530 The findings by these earlier studies may be further explained by
Escapism .17 2.93 .004 the results obtained in this study, as it illustrates how it is the
Stress .29 4.75 .000 interplay between entertainment (e.g., Khang et al., 2013; Lee,
Escapism ⁄ Stress .12 2.10 .036 Song et al., 2014), escapism (e.g., Khang et al., 2013) and stress
186 J.-L. Wang et al. / Computers in Human Behavior 53 (2015) 181–188

(e.g., Chiu, 2014; Khang et al., 2013; Lee, Song et al., 2014) that lead negative interaction term between entertainment motivation and
some users to experience more problematic outcomes than others. stress also for the low-problem group. This finding can be
This behavior is reminiscent of a coping strategy for life problems explained as users in the low-problem group are average smart-
(e.g., Kardefelt-Winther, 2014b). phone users. Average users, who are not using the smartphone to
Also important for this explanation for problematic smartphone cope with life problems, will adapt appropriately and use the
use and its association with entertainment and escapism motiva- smartphone less often for entertainment when their life stress
tions (e.g., Tables 3 and 4) is that individuals are active and increases. This is because they simply do not have the time to keep
goal-directed in their media use and intentionally choose media using the smartphone as often, which is why the association is
and content to gratify psychological needs or certain types of negative. Hypothetically, when these users are less stressed, they
motivation (Katz et al., 1974). Several studies have showed the might start using the phone more often again for entertainment,
important role played by media use motivations in shaping users’ and therefore they could experience slightly more problematic
online behaviors (Chou & Hsiao, 2000; Park, Kee, & Valenzuela, outcomes but not at dangerous levels. Again, this mirrors some
2009; Park et al., 2009; Raacke & Bonds-Raacke, 2008; Smock, of the findings by Kardefelt-Winther (2015) in his study of online
Ellison, Lampe, & Wohn, 2011). As smartphones are used for a wide gamers and gamblers (see Figs. 1–3).
range of activities, such as gaming, communication, playing videos,
visiting social networking sites, and reading, among other uses,
which can all satisfy users’ needs for entertainment and stress 7. Conclusion
relief, it is not surprising that user tend to spend more time on
smartphones when these types of motivation are induced. Thus far, this study is one of the first to investigate the associ-
However, as this study has shown, such motivations may also ation between types of motivation and problematic smartphone
put users at risk of experiencing some problematic outcomes when use, in conjunction with an indicator of psychosocial well-being
they are more stressed in real life. This lends some support to (i.e., perceived stress). This was done from the perspective of the
Kardefelt-Winther’s Compensatory Internet Use theory (2014b, Compensatory Internet Use theory (Kardefelt-Winther, 2014b)
2014c) in the study of problematic smartphone use. It would be which was supported by the current study. We found that, for par-
interesting to explore whether these findings also applies to a ticipants with many problematic outcomes, entertainment and
wider range of life problems. escapism motivation were positively related to problematic smart-
The positive association between perceived stress and excessive phone use for individuals with higher levels of perceived stress,
smartphone use might be attributable to the fact that problematic and that these relationships were non-significant for individuals
smartphone use can provide a distraction from stressful events and with low levels of perceived stress. For participants with few prob-
serves as a way of coping with stress (Snodgrass et al., 2014). As lematic outcomes, there were no significant difference in terms of
Young (1998) suggested, the impulsive behaviors of individuals the effect of escapism on problematic smartphone use between
with Internet addiction can be regarded as rewards that reduce high and low stress groups, and only minor difference was found
emotional tension and facilitate future Internet usage behavior. regarding the effect of entertainment motivation on problematic
Internet use can distract gamers’ attention from offline problems; smartphone use between high and low stress groups. However,
so, online gaming can serve a stress-management function for those in the problematic group, the differences were quite
(Snodgrass et al., 2014). Hence, players with more stress in their large. Our study highlights the important role played by unre-
lives may overuse that cognitive diversion, with attendant solved real-life problems in facilitating problematic smartphone
problematic consequences. Furthermore, Snodgrass et al. (2014) use, indicating that the resolution of these problems might be a
concluded that online gaming may yield different outcomes for good starting point when researchers design interventions for
users with different levels of perceived stress. Specifically, users
with low perceived stress manage to play online games so as to
enhance their offline lives, while users perceiving high stress
further magnify the stress and suffering in their lives. This is
because highly stressed and distressed individuals are less able
to control their online pleasures and experiences, seeming to need
them more (Snodgrass et al., 2014). As the present study suggests,
when users’ motivation to use smartphones is grounded in stress,
the smartphone use may on the one hand alleviate the stress, but
may also result in excessive usage with problematic outcomes.
Important for purposes of methodology and future studies, H3
and H4 stated that among users with low scores on problematic
smartphone use, there would be no interactions between
motivations and perceived stress in predicting problematic smart-
phone use. Our findings confirmed H4 and partly supported
Kardefelt-Winther’s (2014c, 2015) proposed methodological pro-
cedure. The reason why some people experience more problematic
outcomes than others might thus be because participants with
many problematic outcomes tend to use smartphones as a coping
strategy to deal with stress, while those with fewer problematic
outcomes may only be using it for purposes of entertainment
and escapism – unrelated to a negative real life situation. This
lends some support to the proposal that those who experience
many problematic outcomes from smartphone use, and perhaps
also Internet use in general, may be compensating for some sort
of real life issue (Kardefelt-Winther, 2014b, 2014c). It is important Fig. 1. Two-way interaction effect of stress and entertainment motivation on extent
to note that H3 was not supported since we found a significant of use in the problematic smartphone-use group, N = 264.
 J.-L. Wang et al. / Computers in Human Behavior 53 (2015) 181–188 187

smartphone use were examined in this study, although Lee, Song

et al. (2014) has indicated other types of motivation, including
instantaneousness, mobility, interest, information, and social
status. Future research should examine these types of motivation,
along with other psychosocial well-being variables (e.g.,
self-esteem and life satisfaction) in predicting problematic smart-
phone use. Finally, we suggest that future studies specifically
consider the functions of smartphones when investigating the
associations between motivation, psychological well-being vari-
ables, and smartphone use. Some users may rely on reading novels
on their smartphones, while others may problematically use social
networking sites on theirs. A refined investigation would enable a
more thorough understanding of problematic smartphone use.

Acknowledgement

This study has been funded by the Planning Projects for the
National Education Science: Grant No. CBA140146. Grant title is
The Influence of Social Networking Sites Use on Adolescents’ and
Young Adults’ Mental Health and its Interventions.

Appendix A. Supplementary material

Fig. 2. Two-way interaction effect of stress and escapism motivation on extent of
use in the problematic smartphone-use group, N = 264.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.chb.2015.07.005.
people who use smartphones to the extent that it causes problems
in real life. References
Our study is not without limitations. First, convenience sam-
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting
pling was used; therefore, caution should be exercised when the
interactions. Newbury Park, CA: Sage.
conclusions are generalized to other populations. However, the Anderson, D. R., & Collin, P. A. (1996). Stressful life events and television viewing.
population from which we drew our sample represents the most Communication Research, 23, 243–260.
Bener, A., Al-Mahdi, H. S., Vachhani, P. J., Al-Nufal, M., & Ali, A. I. (2010). Do excessive
common mobile user demographic – young adults in China.
internet use, television viewing and poor lifestyle habits affect low vision in
Therefore, despite being a convenience sample, it is a sample that school children? Journal of Child Health Care, 14(4), 375–385.
is perfectly representative of the largest smartphone user group. Beranuy, M., Oberst, U., Carbonell, X., & Chamarro, A. (2009). Problematic internet
Second, a cross-sectional design cannot enable conclusions about and mobile phone use and clinical symptoms in college students: The role of
emotional intelligence. Computers in Human Behavior, 25(5), 1182–1187.
causal relationships between the variables in this study. Future Bian, M., & Leung, L. (2014). Linking loneliness, shyness, smartphone addiction
studies should address these limitations by using experimental symptoms, and patterns of smartphone use to social capital. Social Science
and longitudinal designs that can yield causal inferences about Computer Review, 33(1), 61–79.
Bianchi, A., & Phillips, J. G. (2005). Psychological predictors of problem mobile
the relationships. Additionally, only two types of motivation for phone use. Cyberpsychology & Behavior: The Impact of the Internet, Multimedia
and Virtual Reality on Behavior and Society, 8(1), 39–51.
Buckner, J. E., Castille, C. M., & Sheets, T. L. (2012). The Five Factor Model of
personality and employees’ excessive use of technology. Computers in Human
Behavior, 28(5), 1947–1953.
Chen, H.-T., & Kim, Y. (2013). Problematic use of social network sites: The
interactive relationship between gratifications sought and privacy concerns.
Cyberpsychology, Behavior and Social Networking, 16(11), 806–812.
Chesley, N. (2005). Blurring boundaries? Linking technology use, spillover,
individual distress, and family satisfaction. Journal of Marriage and Family, 67,
1237–1248.
Chiu, S.-I. (2014). The relationship between life stress and smartphone addiction on
taiwanese university student: A mediation model of learning self-Efficacy and
social self-Efficacy. Computers in Human Behavior, 34, 49–57.
Chou, C., & Hsiao, M.-C. (2000). Internet addiction, usage, gratification, and pleasure
experience: The Taiwan college students’ case. Computers & Education, 35(1),
65–80.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple
regression/correlation analysis for behavioral sciences. Mahwah, NJ: Lawrence
Erlbaum.
Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived
stress. Journal of Health and Social Behavior, 24(4), 385–396.
Ehrenberg, A., Juceks, S., & White, K. M. (2008). Personality and self-esteem as
predictors of young peoples’ technology use. CyberPsychology & Behavior, 11(6),
739–741.
Ha, J. H., Chin, B., Park, D.-H., Ryu, S.-H., & Yu, J. (2008). Characteristics of excessive
cellular phone use in Korean adolescents. Cyberpsychology & Behavior: The
Impact of the Internet, Multimedia and Virtual Reality on Behavior and Society,
11(6), 783–784.
Hawi, N. S. (2012). Internet addiction among adolescents in Lebanon. Computers in
Human Behavior, 28(3), 1044–1053.
Hawkins, J. D., Catalano, R. F., & Miller, J. Y. (1992). Risk and protective factors for
alcohol and other drug problems in adolescence and early adulthood:
Fig. 3. Two-way interaction effect of stress and entertainment motivation on extent Implications for substance abuse prevention. Psychological Bulletin, 112(1),
of use in the non-problematic smartphone-use group, N = 285. 64–105.
188 J.-L. Wang et al. / Computers in Human Behavior 53 (2015) 181–188

Ho, H. Y., & Syu, L. Y. (2010). Uses and gratifications of mobile application users. In ICEIE Peters, C. S., & Malesky, L. A. (2008). Problematic usage among highly-engaged
2010 – 2010 International Conference on Electronics and Information Engineering, players of massively multiplayer online role playing games. CyberPsychology &
Proceedings (Vol. 1). doi:http://dx.doi.org/10.1109/ICEIE.2010.5559869. Behavior, 11(4), 481–484.
Hong, F.-Y., Chiu, S.-I., & Huang, D.-H. (2012). A model of the relationship between Raacke, J., & Bonds-Raacke, J. (2008). MySpace and Facebook: Applying the uses and
psychological characteristics, mobile phone addiction and use of mobile phones gratifications theory to exploring friend-networking sites. CyberPsychology &
by Taiwanese university female students. Computers in Human Behavior, 28(6), Behavior, 11(2), 169–174.
2152–2159. Rhodes, J. E., & Jason, L. (1990). A social stress model of substance abuse. Journal of
Jenaro, C., Noelia, F., Maria, G.-V., Francisca, G.-G., & Cristina, C. (2007). Problematic Consulting and Clinical Psychology, 58, 395–401.
internet and cell-phone use: Psychological, behavioral, and health correlates. Smock, A. D., Ellison, N. B., Lampe, C., & Wohn, D. Y. (2011). Facebook as a toolkit: A
Addiction Research & Theory, 15(3), 309–320. uses and gratification approach to unbundling feature use. Computers in Human
Jie, T., Yizhen, Y., Yukai, D., Ying, M., Dongying, Z., & Jiaji, W. (2014). Addictive Behavior, 27(6), 2322–2329.
Behaviors Prevalence of internet addiction and its association with stressful life Snodgrass, J. G., Lacy, M. G., Dengah, H. J. F., Eisenhauer, S., Batchelder, G., &
events and psychological symptoms among adolescent internet users. Addictive Cookson, R. J. (2014). A vacation from your mind: Problematic online gaming is
Behaviors, 39(3), 744–747. a stress response. Computers in Human Behavior, 38, 248–260.
Kardefelt-Winther, D. (2014a). Problematizing excessive online gaming and its Song, N. (2004). Stress, self-control, and self-esteem of college students and Internet
psychological predictors. Computers in Human Behavior, 31, 118–122. usage. Korea: Younsei University.
Kardefelt-Winther, D. (2014b). A conceptual and methodological critique of internet Statista (2013a). Global smartphone penetration from 2008 to 2014 (in percent of new
addiction research: Towards a model of compensatory internet use. Computers handset sales). Retrieved January 15, 2015, from <http://www.statista.com/
in Human Behavior, 31, 351–354. statistics/218532/global-smartphone-penetration-since-2008/>.
Kardefelt-Winther, D. (2014c). The moderating role of psychosocial well-being on Statista (2013b). Smartphone penetration rate in the United Kingdom (UK) from 2010
the relationship between escapism and excessive online gaming. Computers in to 2017. Retrieved January 15, 2015, from <http://www.statista.com/statistics/
Human Behavior, 38, 68–74. 270888/smartphone-penetration-in-the-united-kingdom-uk/>.
Kardefelt-Winther, D. (2015). Excessive internet use – fascination or compulsion? Statista (2013c). Smartphone penetration rate in the United States from 2010 to 2017.
(PhD thesis). London: The London School of Economics and Political Science. Retrieved January 15, 2015, from <http://www.statista.com/statistics/
Katz, E., Blumler, J. G., & Gurevitch, M. (1974). The uses of mass communication: 201183/forecast-of-smartphone-penetration-in-the-us/>.
Current perspectives on gratifications research. Sociology the Journal of the British Statista (2013d). Share of mobile phone users that use a smartphone in China from
Sociological Association (pp. 19–32). Beverly Hills, CA: Sage. 2010 to 2017. Retrieved January 15, 2015, from <http://www.statista.com/
Khang, H., Kim, J. K., & Kim, Y. (2013). Self-traits and motivations as antecedents of statistics/257045/smartphone-user-penetration-in-china/>.
digital media flow and addiction: The internet, mobile phones, and video Su, S., Pan, T.-T., Liu, X.-Q., Chen, X.-W., Wang, Y.-J., & Li, M.-Y. (2014). Development
games. Computers in Human Behavior, 29(6), 2416–2424. of the Smartphone addiction scale for college students. Chinese Mental Health
Kline, R. B. (2005). Principles and practice of structural equation modeling (2nd ed.). Journal, 28(5), 392–397.
New York: Guilford. Takao, M., Takahashi, S., & Kitamura, M. (2009). Addictive personality and
Ko, H., Cho, C., & Roberts, M. S. (2005). Internet uses and gratifications: A structural problematic mobile phone use. CyberPsychology & Behavior, 12(5), 501–507.
equation model of internet advertising. Journal of Advertising, 34(2), 57–70. Thomee, S., Eklof, M., Gustafsson, E., Nilsson, R., & Hagberg, M. (2007). Prevalence
Kuss, D. J., & Griffiths, M. D. (2011). Online social networking and addiction – a of perceived stress, symptoms of depression and sleep disturbances in relation
review of the psychological literature. International Journal of Environmental to information and communication technology (ICT) use among young
Research and Public Health, 8(9), 3528–3552. adults – An explorative prospective study. Computers in Human Behavior,
Kwon, M., Lee, J.-Y., Won, W.-Y., Park, J.-W., Min, J.-A., Hahn, C., et al. (2013). 1300–1321.
Development and validation of a smartphone addiction scale (SAS). PLoS ONE, Thomee, S., Harenstam, A., & Hagberg, M. (2011). Mobile phone use and stress, sleep
8(2), e56936. disturbances, and symptoms of depression among young adults – A prospective
Lee, Y., Chang, C., Lin, Y., & Cheng, Z. (2014). The dark side of smartphone usage: cohort study. BMC Public Health, 11, 66.
Psychological traits, compulsive behavior and technostress. Computers in Toda, M., Ezoe, S., Nishi, A., Mukai, T., Goto, M., & Morimoto, K. (2008). Mobile phone
Human Behavior, 31, 373–383. dependence of female students and perceived parental rearing attitudes. Social
Lee, U., Song, J., Lee, J., Ko, M., Lee, C., Kim, Y., et al. (2014). Hooked on smartphones. Behavior and Personality, 36(6), 765–770.
In Proceedings of the 32nd annual ACM conference on Human factors in computing Walsh, S. P., White, K. M., Cox, S., & Young, R. M. (2011). Keeping in constant touch:
systems – CHI ’14 (pp. 2327–2336). New York, New York, USA: ACM Press. The predictors of young Australians’ mobile phone involvement. Computers in
Leung, L. (2007). Stressful life events, motives for Internet use, and social support Human Behavior, 27(1), 333–342.
among digital kids. CyberPsychology & Behavior, 10(2), 204–214. Wang, J.-L., Jackson, L. A., Wang, H.-Z., & Gaskin, J. (2015). Predicting Social
Lowry, P., Gaskin, J., & Moody, G. (in press). Proposing the multimotive information Networking Site (SNS) use: Personality, attitudes, motivation and Internet self-
systems continuance model (MISC) to better explain end-user system efficacy. Personality and Individual Differences, 80, 119–124.
evaluations and continuance intentions. Journal of the Association of Wei, R. (2008). Motivations for using the mobile phone for mass communications
Information System. and entertainment. Telematics and Informatics, 25, 36–46.
Martinotti, G., Villella, C., Di Thiene, D., Di Nicola, M., Bria, P., Conte, G., et al. (2011). Wei, R., & Lo, V.-H. (2006). Staying connected while on the move: Cell phone use
Problematic mobile phone use in adolescence: A cross-sectional study. Journal and social connectedness. New Media & Society, 8, 53–72.
of Public Health, 19(6), 545–551. Whang, L., Lee, S., & Chang, G. (2003). Internet over-users’ psychological profiles: A
Papacharissi, Z., & Rubin, A. M. (2000). Predictors of internet use. Journal of behavior sampling analysis on internet addiction. CyberPsychology & Behavior, 6,
Broadcasting & Electronic Media, 44(2), 175–196. 143–150.
Park, N., Kee, K. F., & Valenzuela, S. (2009). Being immersed in social networking Young, K. (1998). Caught in the net. New York: Wiley.
environment: Facebook groups, uses and gratifications, and social outcomes. Young, K. S. (2004). Internet addiction: A new clinical phenomenon and its
CyberPsychology & Behavior, 12(6), 729–733. consequences. American Behavioral Scientist, 48(4), 402–415.