Smartphones have become common, and problematic smartphone use (PSU) is increasing. Predictors of PSU should be identified to prevent it. Little is known about the role of content types of smartphone use as predictors of PSU. Therefore, we aimed to evaluate the predictors of two proposed concepts of PSU, namely habitual smartphone behavior (SB) and addictive SB, within the context of the application (app) categories.
We studied 1,039 smartphone users using online surveys conducted between January 2 and 31, 2019. We employed multiple regression analysis to identify the predictors of habitual and addictive SB. We controlled for sex and age (mean=39.20).
Common predictors of habitual and addictive SB were the use of social networking services, games, entertainment apps, and average weekend smartphone usage time. The predictors of habitual SB were the use of web and lifestyle apps, weekly usage frequency, and sex (female) and the predictors of addictive SB were the use of shopping apps and sleep duration.
This study revealed the need to consider habitual and addictive SB in evaluating PSU. The predictors in terms of the content types of smartphone usage can be used to develop monitoring and prevention services for PSU.
Smartphones have become a common feature around the world and an essential part of our daily lives. This global popularity has raised concerns about its negative effects associated with problematic smartphone use (PSU). Scholars define PSU as compulsive and maladaptive smartphone behavior [
Van Deursen et al. [
Previous studies have evaluated psychological predictors such as low self-esteem, depressive symptoms [
Therefore, we aimed to find out the predictors of habitual and addictive SB in relation to the role of content types of smartphone use based on application (app) usage behavior.
We received smartphone users’ survey data from a polling company between January 2 and January 31, 2019. The polling company was dataSpring. The company has Korean online panel with 379,451 people (
The study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of Catholic University (IRB number: MC18QNSI0101).
This study used 14 variables: two dependent variables, 10 independent variables, and two control variables. Furthermore, it used a multiple regression analysis to identify the predictors of habitual and addictive SB (
We used habitual SB and addictive SB as dependent variables. Habitual smartphone usage was measured using an instrument adapted from Limayem et al. [
Addictive smartphone usage was measured using the modified version 3 of the mobile phone problem use scale developed by Bianchi and Phillips [
Nine variables related to smartphone usage and sleep duration were used as independent variables. These were three smartphone usage patterns (weekday smartphone usage time, weekend smartphone usage time, and weekly frequency of smartphone use) and six content types of smartphone use according to the apps (SNS/Chatting, web, games, entertainment, shopping, and lifestyle) (
The SNS/Chatting apps are mainly used for seeking social relationships and consist of social networks, messengers, chatting, and vlogs. The web apps are primarily used for seeking information and consist of web browsers such as Naver, Google, Chrome, Daum, Nate, and Dolphin. The game apps are for game enjoyment such as simulation games, role-playing games, arcade games, action games, board games, game money, and game items. Entertainment apps are apps that have diverse content for enjoyment, such as media/videos, sports, travel, music, books, and comics. Shopping apps are mainly for consumption and buying purposes and consist of clothes, tickets, books, and used items. Finally, lifestyle apps are predominantly for ordinary life maintenance such as phone calls, text messages, e-mails, addresses, diaries, deliveries, and delivery tracking. Each item in the perceptions of app uses consisted of a five-point Likert scale from 1 (never) to 5 (always).
We also included sleep duration as an independent variable because lack of sleep and sleep disturbance was found to positively correlate with PSU and Internet addiction in previous studies [
Finally, we controlled for sex and age to focus on the 10 variables as mentioned above because previous studies indicated that women and young people engage in PSU more than men and elderly people do [
We conducted a multiple regression analysis on 1,039 smartphone users using R (version 3.5.0). Among the variable selection methods, we used the “enter” method for inputting independent variables.
As shown in
We found the following to be the predictors of habitual SB. The app usages for entertainment (t=5.493, p≤0.001), SNS/chatting (t=4.983, p≤0.001), web (t=4.868, p≤0.001), lifestyle (t=4.128, p≤0.001), and games (t=2.712, p=0.007) were significantly associated with habitual SB. In addition, weekly usage frequency (t=3.041, p=0.002), average weekend smartphone usage time (t=2.169, p=0.030), and sex (female) (t=2.885, p=0.004) were significantly associated with habitual SB (F=35.601, p≤0.001). According to the standardized coefficients representing the relative contribution of the independent variables, entertainment apps had the greatest effect on habitual SB (β=0.178), followed by web apps (β=0.158), SNS/chatting apps (β=0.150), lifestyle apps (β=0.132), average weekend smartphone usage time (β=0.088), weekly usage frequency (β=0.082), sex (female) (β=0.080), and game apps (β=0.077). The coefficient of determination (R2) for this model was 0.294, indicating that 29.4% of the variation in habitual SB can be explained by these eight independent variables.
The following were found to be the predictors of addictive SB. The usage of game (t=7.719, p≤0.001), shopping (t=5.369, p≤0.001), entertainment (t=2.991, p=0.003), and SNS/chatting apps (t=2.895, p=0.004), average weekend smartphone usage time (t=2.072, p=0.039), and sleep duration (t=-2.919, p=0.004) were significantly associated with addictive SB (F=19.450, p≤0.001). Game apps had the greatest effect on addictive SB (β=0.235), followed by shopping apps (β=0.186), entertainment apps (β=0.104), SNS/chatting apps (β=0.094), average weekend smartphone usage time (β=0.090), and sleep duration (β = -0.083). The coefficient of determination (R2) for this model was 0.185, indicating that 18.5 percent of the variation in addictive SB can be explained by these five independent variables.
The predictors that were common for both habitual and addictive SB were the use of SNS, game, and entertainment apps and average weekend smartphone usage time. The predictors of habitual SB, which were not predictors of addictive SB, were the use of web and lifestyle apps, weekly usage frequency, and sex (female), while the predictors of addictive SB, which were not predictors of habitual SB, were the use of shopping apps and sleep duration.
In the present study, we identified the predictors of the two concepts of PSU, namely habitual and addictive SB, especially in relation to the content types of apps, smartphone usage pattern, and sleep duration.
First, we identified the common predictors of habitual and addictive SB.
The predictors that were common for both habitual and addictive SB were the use of SNS, games, and entertainment apps according to the content types of the apps. In other words, the problematic use of the SNS, game, and entertainment contents may have characteristics of both habitual and addictive SB. In the aspect of addictive usage, previous studies reported that SNS, game, and entertainment contents tend to be addictive [
Second, we identified the different predictors of habitual and addictive SB.
The predictors of habitual SB, which were not predictors of addictive SB in terms of content types of smartphone use were web and lifestyle apps. The web apps category consisted of web browsers, while the lifestyle app category consisted of phone calls, text messages, e-mail addresses, diaries, real estate, deliveries, and delivery tracking. The web and lifestyle apps are closely related to everyday life and currently accessed them more from smartphones than from any other types of device because of the ubiquitous accessibility nature of smartphones. A previous study reported that checking e-mails or messages was related to the strongest habitual pattern and that these kinds of brief-checking behaviors may increase overall smartphone use [
Furthermore, we found that the female sex was a predictor of habitual SB although we designed our study to control sex. In other words, women engage in more habitual SB than men do. Previous studies reported that sex is a critical factor in PSU [
The only predictor of addictive SB that was not a predictor of habitual SB in terms of application type was shopping apps. Previous studies suggested that pathological online buying shares several key characteristics with behavioral addictions and it is considered a form of Internet addiction [
Finally, we found an association between sleep duration and addictive SB. Consistent with our findings, many studies have reported a positive association between sleep deprivation and PSU or between poor sleep quality and PSU [
Overall, the present study showed the presence of different predictors of habitual and addictive SB in PSU depending on the content types of smartphone use, smartphone usage pattern, and sleep duration. However, our findings need to be interpreted cautiously because PSU cannot be explained by either habitual SB or addictive SB, just like habit formation is partly involved in the mechanism of addiction but habitual action itself does not cover all addictions in terms of sensitization or negative reinforcement [
Nevertheless, the current study is significant. The study proposes that there are habitual and addictive SB aspects in PSU and clarifies which types of smartphone use are related to PSU in the habitual and addictive SB aspects. The results show that the types of productivity enhancement, such as e-mail, and information seeking, such as browsing the news, are related more with habitual SB, while the types of consumption and buying, such as shopping are more related with addictive SB. The types of social information and relationships such as social network, gaming, and entertainment (e.g., viewing movies) are related to both habitual and addictive SB. These results suggest that the categorization with a strong association between stimuli (e.g., e-mail alarm or headlines) and response (e.g., checking) occurs in habitual SB and that the maladaptive and aberrant recruitment of habit process leads to PSU [
We developed the smartphone addiction risk rating score for a smartphone addiction management application based on habitual SB and addictive SB [
In future research, we intend to collect actual usage by app category usage, weekly usage frequency, weekend usage time, and sleep duration through updated app. By collecting the actual data, it is possible to calculate the habitual SB score and addictive SB score without users’ response of questionnaires. We can provide interventions for each based on the derived scores. Currently, the smartphone management app has been developed as the first version. In the future, our results will be verified with actual data.
The limitations of this study and suggestions for future studies are as follows. First, we used smartphone usage data from a self-reporting questionnaire. Some studies have reported that respondents’ estimates of their smartphone use do not necessarily relate to actual use [
The present study suggests that there is a need to consider the habitual and addictive SB aspects to evaluate PSU and that habitual and addictive SB have not only similar features but also different features with respect to the content types of smartphone usage. In addition, these common and different predictors, including using apps and sleep duration, can be used to develop monitoring and prevention services for PSU. Our findings may also be able to take a new approach to future investigations about PSU.
This work was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (NRF2018R1C1B6007750).
The authors have no potential conflicts of interest to disclose.
Conceptualization: Mi Jung Rho. Data curation: Jihwan Park. Formal analysis: Jihwan Park. Funding acquisition: Mi Jung Rho. Investigation: Mi Jung Rho. Methodology: Mi Jung Rho, Jihwan Park. Project administration: Mi Jung Rho. Resources: Mi Jung Rho. Software: Jihwan Park. Supervision: Mi Jung Rho. Validation: Jo-Eun Jeong. Visualization: Jihwan Park. Writing—original draft: Jihwan Park, Mi Jung Rho. Writing—review & editing: Mi Jung Rho, Jo-Eun Jeong.
Research process.
Predictors of habitual and addictive SB.
Six smartphone app categories
Category | Purpose of use | Applications |
---|---|---|
SNS/chatting | Social relationship seeking | Social networks, messengers, chatting, and vlogs |
Web | Information seeking | Naver, Google, Chrome, Daum, Nate, and Dolphin |
Game | Game enjoyment | Simulation games, role-playing games, arcade games, action games, board games, game money, and game items |
Entertainment | Content enjoyment | Media/videos, sports, travel, music, books, and comics |
Shopping | Consumption seeking and buying | Clothes, tickets, books, and used items |
Lifestyle | Ordinary life maintenance | Phone calls, text messages, e-mails, addresses, diaries, deliveries, and delivery tracking |
Demographic characteristics
Variables | N | Percentage |
---|---|---|
Sex | ||
Male | 520 | 50.0 |
Female | 519 | 50.0 |
Age (mean=39.20) | ||
20–29 years | 258 | 24.8 |
30–39 years | 261 | 25.1 |
40–49 years | 263 | 25.3 |
50–59 years | 257 | 24.8 |
Marital status | ||
Single |
477 | 45.9 |
Married or living with a partner | 562 | 54.1 |
Occupation | ||
Office worker, etc. |
695 | 66.9 |
Student | 165 | 15.9 |
Housewife, unemployed and other | 179 | 17.2 |
Monthly income | ||
Under $1,792.11 | 111 | 10.7 |
$1,792.11–$3,584.23 | 331 | 31.8 |
$3,584.23–$5,376.34 | 354 | 34.1 |
Over $5,376.34 | 243 | 23.4 |
Residential area | ||
Capital area (including Seoul) | 657 | 63.2 |
Noncapital area | 382 | 36.8 |
Smartphone device type | ||
Android | 861 | 82.9 |
Apple iOS | 178 | 17.1 |
Total | 1,039 | 100.0 |
The exchange rate for the Korean won to the U.S. dollar is 1,116.00 won (buy and sell base rate on January 31, 2019).
single: never married, divorced, separated, or widowed,
office worker, etc.: office worker, administrative professional, service industry professional, professional technician, freelancer, or production employee
Multiple regression analysis results
Dependent variables | Habitual SB (mean: 22.41, range: 6–30, Cronbach’s α=0.860) |
Addictive SB (mean: 65.72, range: 26–130, Cronbach’s α=0.946) |
|||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Independent variables | Nonstandardized coefficients |
Standardized coefficients |
t value | Sig. | Nonstandardized coefficients |
Standardized coefficients |
t value | Sig. | |||
B | SE | β | B | SE | β | ||||||
(Constant) | 11.945 | 0.822 | - | 14.531 | <0.000 |
43.073 | 3.872 | 11.125 | <0.000 |
||
App category | |||||||||||
SNS/chatting | 0.548 | 0.110 | 0.150 | 4.983 | <0.000 |
1.500 | 0.518 | 0.094 | 2.895 | <0.004 |
|
Games | 0.235 | 0.087 | 0.077 | 2.712 | <0.007 |
3.153 | 0.408 | 0.235 | 7.719 | <0.000 |
|
Entertainment | 0.706 | 0.129 | 0.178 | 5.493 | <0.000 |
1.810 | 0.605 | 0.104 | 2.991 | 0.003 |
|
Web | 0.709 | 0.146 | 0.158 | 4.868 | <0.000 |
-0.426 | 0.686 | -0.022 | -0.621 | 0.535 | |
Lifestyle | 0.556 | 0.135 | 0.132 | 4.128 | <0.000 |
-0.727 | 0.635 | -0.039 | -1.145 | 0.252 | |
Shopping | 0.027 | 0.125 | 0.007 | 0.217 | 0.829 | 3.165 | 0.589 | 0.186 | 5.369 | <0.000 |
|
Sleep duration | -0.002 | 0.001 | -0.038 | -1.442 | 0.150 | -0.017 | 0.006 | -0.083 | -2.919 | 0.004 |
|
Weekly usage frequency | 0.002 | 0.001 | 0.082 | 3.041 | 0.002 |
0.003 | 0.004 | 0.024 | 0.814 | 0.416 | |
Average weekend smartphone usage time | 0.002 | 0.001 | 0.088 | 2.169 | 0.030 |
0.007 | 0.003 | 0.090 | 2.072 | 0.039 |
|
Average weekday smartphone usage time | 0.000 | 0.001 | 0.000 | -0.001 | 0.999 | 0.000 | 0.003 | 0.001 | 0.026 | 0.979 | |
Sex (0:male, 1:female) | 0.662 | 0.230 | 0.080 | 2.885 | 0.004 |
-1.657 | 1.082 | -0.046 | -1.532 | 0.126 | |
Age | 0.117 | 0.107 | 0.032 | 1.093 | 0.275 | 0.790 | 0.502 | 0.049 | 1.573 | 0.116 |
Habitual SB: R2 (adjusted R2)=0.294 (0.286), F change=35.601, significance of F change≤0.001. Addictive SB: R2 (adjusted R2)=0.185 (0.176), F change=19.450, significance of F change≤0.001.
p<0.05,
p<0.01,
p<0.001.
Duration unit: minute. SE: standard error