Psychosocial Pathways to Smartphone Overdependence in Adolescence: A Multi-Group Path Analysis of Early vs. Late Adolescents
Article information
Abstract
Purpose
This study examined the relationship between psychosocial factors and smartphone overdependence among Korean adolescents, focusing on developmental differences between early and late adolescence and the mediating roles of loneliness and anxiety.
Methods
A cross-sectional design was used with secondary data from the 2023 Korea Youth Health Risk Behavior Survey, a nationwide stratified cluster sample. A total of 45,060 middle and high school students were included. Data were analyzed using SPSS and AMOS 28.0, employing descriptive statistics, path analysis, and multigroup analysis to assess direct and indirect effects on smartphone overdependence.
Results
Loneliness and anxiety significantly mediated the associations between psychosocial factors and smartphone overdependence. Living status, loneliness, and anxiety directly influenced smartphone overdependence in both middle and high school students, while depression had direct effects only among high school students. Stress, depression, and living status showed indirect effects. Although living status was significant, its total effect was limited among middle school students.
Conclusion
The findings underscore the importance of family support and mental health management in addressing adolescent smartphone overdependence. Developmentally tailored strategies that reflect differing psychosocial pathways and the Korean sociocultural context are essential for promoting healthy smartphone use.
INTRODUCTION
The smartphone ownership rate among teenagers in South Korea is 99.6%. Moreover, 95.5% perceive smartphones as an essential item in their daily life, with 99.5% using smartphones for at least 5 days a week [1]. In modern society, smartphones are a vital part of teenagers' daily lives and are used extensively in various aspects, including peer relationships, learning, and leisure activities [2]. In South Korea, four out of ten teenagers fall into the risk group for smartphone overdependency, a higher proportion than in other age groups [3]. Smartphone overdependency describes a condition in which smartphone use becomes overly dominant in daily life, accompanied by diminished self-regulation and resulting negative outcomes [3]. It has been associated with physical problems, such as a decreased perception of subjective health and reduced physical activity, as well as mental health issues, such as unhappiness, and sadness [4]. Mental health problems during adolescence have been linked to lower academic achievement and can lead to mental health issues in adulthood, reduced life satisfaction, diminished health-as-lescence is a crucial public health issue.
Adolescence is a developmental stage in which young people encounter stress arising from multiple domains, including peer and teacher relationships, family dynamics, concerns about appearance, academic demands, and uncertainties about future careers [6]. A high level of everyday stress increases the risk of becoming overly dependent on smartphones [6]. This overdependency can subsequently cause further problems and act as an additional source of stress [7], which can result in a negative cycle that exacerbates smartphone overdependency [8]. Moreover, depression among teenagers affects their reliance on smartphones [8]. Depressed teenagers tend to use smartphones to escape reality or relax, leading to increased smartphone usage and dependency [8].
Many teenagers in South Korea live with their parents, which can lead to behavioral moderation by family members. This living environment is also related to teenagers' smartphone dependency. Teenagers living with their parents have more frequent parental supervision or interference compared with those who do not, which can affect their smartphone usage patterns. Consequently, this impacts their level of smartphone dependency. As teenagers mainly spend their time at school and home, family intervention is essential in moderating their lifestyles. Therefore, these related factors must be considered.
Teenagers' perception of stress is associated with loneliness [9]. Stress has additionally been linked to various physical and emotional problems, such as increased anxiety [10]. In adolescence, depression, anxiety, and loneliness frequently overlap and influence one another, often emerging simultaneously [11]. Living with parents is believed to provide additional support to teenage children; parental support is further related to teenagers' loneliness and anxiety [12] . Thus, stress, depression, and living status, including cohabitation with parents, are related to loneliness and anxiety among teenagers, which may mediate their dependency on smartphones.
The academic curricula differ among South Korean teenagers in middle and high schools. High school students face a strong academic burden as they prepare for the College Scholastic Ability Test to enter university. Furthermore, adolescence is a period of rapid development that can lead to age-related differences in mental and physical levels. Therefore, studies should examine adolescents by subdividing them according to their age to reflect their characteristics appropriately. However, previous studies have rarely separately examined adolescents in middle and high school stages. This study used data from the 2023 Korea Youth Health Risk Behavior Survey (KYHRBS) [13] to explore the mediating effects of loneliness and anxiety on the influence of stress, depression, and living status on smartphone overdependency among middle and high school students. Thus, it sought to provide primary data for effective approaches to managing smartphone overdependency at the school level.
1. Hypothetical Model
Based on the stress-coping-adaptation theory [14] and a literature review [4,15], we developed a hypothetical model of adolescents' smartphone overdependence (Figure 1). It included variables that influence smartphone overdependence among adolescents, such as stress, depression, loneliness, and anxiety, and considered the impact of living status. Based on this model, we proposed that loneliness and anxiety act as pathways through which stress, depression, and living status relate to adolescents' smartphone use. Additionally, we hypothesized that these relationships would differ between middle and high school students.
Hypothetical model.
METHODS
1. Study Design
In this cross-sectional study, we performed a secondary analysis of data from the 19th KYHRBS carried out by the Korea Centers for Disease Control and Prevention in 2023 via path analysis. In this analysis, we explored how loneliness and anxiety operate as mediating variables and delineated the direct and indirect pathways through which various factors contribute to adolescents' smartphone overdependence.
2. Sample
We analyzed raw data from the 19th KYHRBS, an anonymous, self-administered online survey carried out between August 28 and October 19, 2023. To ensure that the sample accurately reflected middle and high school students across the country, we employed the following sampling process, which included population stratification, sample distribution, and sample extraction. Population stratification divided the population into 117 strata, with local districts and schools as the stratification variables. Sample distribution was conducted via the proportional allocation method. For sample extraction, we used a stratified cluster sampling approach, in which schools served as the primary sampling units and grades were selected as the secondary units.
Raw data comprised data of 52,880 students from 399 middle schools and 400 high schools nationwide, with a participation rate of 92.9%. Among these, only students who used their phones on weekdays and weekends (n=50,980) were included in this study. Additionally, students who did not provide information regarding their family members (n=7,770) or reported living with biological and stepparents (n=50) were excluded. Hence, the final sample comprised 45,060 students.
3. Measures
The KYHRBS assessed adolescents' health status and covered various topics, such as physical activity, dietary habits, mental health, alcohol consumption, and smoking. In this study we examined the variables related to the psychological factors of adolescents and their smartphone overdependence.
1) Sociodemographic characteristics
For this study, we included gender and living status as sociodemographic variables. The sociodemographic characteristic of living status was included as an independent variable.
2) Stress
A single-item measure was used to assess perceived stress in daily life. Participants responded on a 5-point Likert scale (1=very much, 5=not at all), and the responses were subsequently recoded into five categorical groups for analysis.
3) Depression
Depression was assessed using a single question asking whether, at any point in the previous 12 months, participants had experienced sadness or hopelessness severe enough to interfere with their usual daily activities. Responses were coded as 1="yes" and 0="no," with "no" designated as the reference group.
4) Living Status
The original categories (e.g., living with father/stepfather, mother/stepmother, grandparents, siblings, or no family members) were recoded into two groups for analysis: living with both parents (including biological or stepparents) versus living with a single parent or without parents. The reference group was set as "living with a single parent or without parents."
5) Loneliness
A single question measured loneliness by asking how often participants had experienced loneliness during the previous year, rated on a 5-point Likert scale (1=never, 5=always).
6) Anxiety
Anxiety was evaluated using the Generalized Anxiety Disorder 7-item scale (GAD-7), developed by Spitzer et al. in 2006 [16], and employed in the Korea Centers for Disease Control and Prevention's in-depth surveys. The scale consists of seven items assessing core anxiety symptoms— such as feeling nervous or having difficulty controlling worry—with responses rated from 0 (not at all) to 3 (nearly every day). In this study, the Cronbach's ⍺ of the GAD-7 was .90, indicating high internal consistency.
7) Smartphone Overdependence
Smartphone overdependence was assessed using the adolescent version of the Smartphone Overdependence Scale developed by the National Information Society Agency (NIA) of Korea in 2011 [17]. The scale evaluates three components—self-control failure, salience, and serious consequences—and total scores are calculated by summing these domains. It includes 10 items, such as "I fail to reduce my smartphone usage time," "I have experienced health problems due to smartphone use," and "I have had severe conflicts with my family because of smartphone use." Participants rated each item on a 4-point Likert scale, ranging from 1 (strongly disagree) to 4 (strongly agree). In this study, the Cronbach's ⍺ for this scale was .90.
4. Data Analysis
AMOS 28.0 and SPSS 28.0 were used, and the statistical significance was set at 5% with a two-tailed test.
• Participants' sociodemographic characteristics and variables were analyzed via descriptive statistics and frequency analysis.
• Path analysis was performed to examine whether loneliness and anxiety functioned as mediators linking depression, stress, and living with parents to smartphone overdependence. The indirect effects through these mediators were evaluated using a bootstrapping procedure.
• To confirm differences based on school level in the verified model, we conducted a multigroup path analysis to examine the direct, indirect, and total effects of each variable on smartphone overdependence.
• Model fit was evaluated using several indices, including the chi-square (x2) statistic, comparative fit index, normed fit index, goodness-of-fit index, Tucker-Lewis index, incremental fit index, root mean square error of approximation (RMSEA), and the standardized root mean square residual(SRMR).
5. Ethical Considerations
This study was granted an exemption from review by the Institutional Review Board (IRB No. withheld for review) of S University. The Korea Centers for Disease Control and Prevention provided the KYHRBS for academic purposes and allowed it to be used freely.
RESULTS
1. Descriptive Statistics
Among the middle school students, 16,617 (81.3%) were living with both parents, 3,127 (15.3%) were living with a single parent, and 688 (3.4%) were living without any parents. Among the high school students, 20,718 (84.1%), 3,416 (13.9%), and 494 (2.0%) lived with both parents, a single parent, and without any parents, respectively. Examination of the descriptive characteristics revealed that high school students experienced significantly higher levels of stress, loneliness, and smartphone overdependency compared with middle school students. Additionally, significant differences were observed in living status based on students' school level (Table 1).
Descriptive Statistics
2. Model Fit
The model fit indices—goodness of fit index, comparative fit index, normed fit index, incremental fit index, and Tucker-Lewis index—indicate a good model fit if they ex-ceed 0.9, and the middle and high school models met this criterion. RMSEA and SRMR indicate a good model fit if they are less than 0.1 and 0.05, respectively. Both models had RMSEA values <0.1 and SRMR values <0.05, indicating an adequate fit.
3. Hypothetical Model Verification
In the initial hypothesized model, the direct path from stress to smartphone overdependence was not statistically significant for either middle or high school students; therefore, this path was removed during model respecification. The final model results, presented in Figures 2 and 3, reflect the standardized path coefficients after this adjustment.
Path analysis of middle school students.
Path analysis of high school students.
Figure 2 illustrates the paths for middle school students based on the standardized path coefficients of the path model. Regarding middle school students, the paths that had significant direct effects on loneliness were stress (β=.41, p<.001), depression (β=.27, p<.001), and living status (living with both parents; β=-.03, p<.001). Furthermore, the paths that had significant direct effects on anxiety were stress (β=.44, p<.001), depression (β=.29, p<.001), and living status (β=-.02, p=.003). Additionally, the paths that had significant direct effects on smartphone overdependency were loneliness (β=.13, p<.001), living status (β=.01, p=.028), and anxiety (β=.22, p<.001).
Figure 3 illustrates the paths for high school students based on standardized path coefficients. The paths that had significant direct effects on loneliness were stress (β=.37, p<.001), depression (β=.26, p<.001), and living status (living with both parents; β=-.03, p<.001). Moreover, the paths that had significant direct effects on anxiety were stress (β=.45, p<.001), depression (β=.27, p<.001), and living status (β=-.03, p<.001). Additionally, the paths that had significant direct effects on smartphone overdependency were depression (β=-.02, p=.020), loneliness (β=.11, p<.001), living status (β=.04, p=.008), and anxiety (β=.24, p<.001).
4. Path Model Analysis
The results of bootstrapping to examine the direct and indirect relationships of the research model for verification of significance were as follows (Table 2). First, regarding the examination of the results for middle school students, the factors that influenced loneliness and anxiety were stress, depression, and living status, which were statistically significant in direct and total effects. Factors that influenced smartphone overdependency included stress, depression, living status, loneliness, and anxiety. Stress had significant indirect and total effects, whereas depression had significant effects in the indirect and total pathways. Living status had significant direct and indirect effects; however, it was not significant in total effects.
Direct, Indirect, and Total Effects of the Model
Regarding high school students, the factors that influenced loneliness and anxiety were stress, depression, and living status, which had statistically significant direct and total effects. Factors that influenced smartphone overdependency included stress, depression, living status, loneliness, and anxiety. Stress had significant indirect and total effects, whereas depression and living status were significant across all the pathways.
DISCUSSION
We examined how stress, depression, and living status were associated with adolescents' smartphone overdependence and explored whether loneliness and anxiety served as mediating mechanisms in these associations.
First, in all the path models, stress and depression were positively associated with loneliness and anxiety. This finding was in line with earlier studies that reported mental health as a factor that directly influenced loneliness and anxiety [11]. In Korea, adolescents typically enter middle school around the age of 12 years, which coincides with the onset of pubert. Moreover, this transition period marks their first encounter with the highly significant cultural process of preparing for college entrance exams, which also leads to various psychological changes [15]. Therefore, they experience dual psychological burdens and thus need tailored interventions that reflect the unique characteristics of this developmental stage.
Second, in all the path models, living with parents was negatively associated with loneliness and anxiety. This finding aligned with that of previous research, which in-dicated that the presence of parents positively impacted adolescents' mental health through emotional support and a sense of stability and safety [18]. Adolescents who live with their parents are likely to receive consistent emotional and psychological support, which can mitigate their feelings of loneliness and anxiety. This underscores the important role of parents in adolescents' mental health and the importance of fostering strong, supportive family environments to promote adolescent psychological wellbeing [19]. However, this association should be interpreted carefully. Living status itself may not fully capture the quality or frequency of parental support, which can differ across family dynamics and communication patterns. Therefore, future research should explore how the quality and nature of parental relationships contribute to adolescents' mental health and psychological well-being.
Third, findings from the multigroup path analysis comparing middle and high school students showed that, among high school students, depression exerted a negative direct effect on smartphone overdependency. It had a positive indirect effect in middle and high school students. These results suggested that depression was accompanied by loneliness or anxiety, and that a lack of energy might lead to increased smartphone dependency as a coping mechanism. This finding aligned with those of existing studies, which suggested that loneliness and anxiety owing to depression led to smartphone dependence [20].
However, the negative direct effect of depression on smartphone overdependency observed only among high school students may reflect developmental changes in the reward system. Schreuders et al. reported that reward-related activation of the nucleus accumbens declines in late adolescence, corresponding to reduced hedonic reward pleasure [21]. Because depression is characterized by anhedonia and diminished motivation [22], these deficits may directly suppress reward-seeking behaviors, including smartphone use, independent of indirect effects through loneliness or anxiety [23]. Thus, when depressive symptoms intersect with the developmentally attenuated reward responsiveness of the nucleus accumbens in late adolescence, the rewarding value of smartphone use may diminish, leading to lower levels of overdependency. Nevertheless, this interpretation should be considered inferential, as the underlying neural mechanisms were not directly assessed. Further research using neurobiological data or longitudinal designs is needed to determine whether these reward-system changes account for the observed negative association.
In contrast, depression showed no direct effect among middle school students. Early to mid-adolescence is marked by heightened reward drive, during which activation of the nucleus accumbens is more strongly linked to approach-oriented motivation than to hedonic pleasure [21]. Consequently, smartphone use in this developmental period is shaped primarily by social-contextual factors— including peer relationships, the need for belonging, and parental monitoring—rather than by emotional states [24]. These influences, together with the buffering the influence of the parent-child relationship [25] may limit the extent to which depression directly contributes to smartphone overdependency among middle school students. Therefore, mental health screening and interventions to prevent smartphone overdependency among adolescents should differ according to their school level and developmental stage [19].
Fourth, living with a parent had a positive direct effect on smartphone overdependency; however, it also had a negative indirect effect. Specifically, adolescents who lived with a parent exhibited higher tendencies toward smartphone overdependency. Nevertheless, those with higher levels of loneliness or anxiety exhibited lower smartphone overdependency.
These findings suggest two distinct pathways rather than contradictory effects. The direct positive effect of living with parents on smartphone overdependency may reflect parental restrictions on smartphone use [19], which could heighten adolescents' desire to use smartphones as a form of psychological reactance when their autonomy feels constrained [26]. In comparison, adolescents with higher levels of loneliness or anxiety may use smartphones for emotional comfort and social connection despite parental control, resulting in lower reported overdependency [27]. The coexistence of these two pathways implies that parental involvement may exert both restrictive and supportive influences depending on the emotional context. However, this interpretation is inferential, as smartphone overdependency was measured by self-reported items, and living with parents in this study was regarded as being under parental influence rather than directly measuring parental control. Therefore, future research should directly examine these mechanisms by incorporating objective or multidimensional measures of parental behavior and adolescents' perceived autonomy.
This can also be attributed to the fact that parents who live with their children are likely to educate them regarding the responsible use of smartphones or directly regulate their smartphone usage. However, Korean society has recently undergone rapid industrialization, leading to various family structures, such as single-parent households and families with single mothers. Such changes often result in fewer resources and support systems compared with traditional families, potentially worsening smartphone overdependency among adolescents [28]. Therefore, implementing diverse interventions at the school or community level is essential to prevent smartphone overdependence among adolescents living in nontraditional family structures. Additionally, efforts that provide evening education programs on smartphone overdependency and consider parents' working hours are required at the municipal level to adapt to changing family structures [29].
Nonetheless, unlike high school students, living status, especially living with parents, did not have a significant effect on smartphone overdependency among middle school students. This suggests that smartphone overdependency in middle school students was influenced by other factors, such as peer groups and social interactions [30]. Middle school students are at a developmental stage where peer influence can become increasingly significant. Furthermore, they may rely more on their friends and social circles for support and validation rather than on parental guidance [29]. This distinction underscores the need to account for students' developmental stages and surrounding social contexts when formulating approaches to adolescent smartphone overdependency.
This study is significant in various ways. First, it constructed a model of smartphone overdependency among students and considered the sociocultural context of Korea, which added to its relevance and applicability. Second, it elucidated the overall relationships between psychosocial factors related to smartphone overdependency via a path analysis. Third, these findings offer a conceptual basis for designing targeted strategies and programs to mitigate smartphone overdependence among students.
1. Limitations and Strengths
There are a few limitations to this study. First, we considered the students to have a smartphone overdependency if their total score on the 10-item screening tool exceeded 23 out of 40. However, we did not differentiate between different levels of dependency when constructing the models. Future research should replicate and extend this study to address this issue and include a further detailed classification of dependency levels. Second, some mental health measures were single-item scales, which stemmed from the utilization of data from the KYHRBS. Therefore, future research should use various mental health measures to examine how adolescent smartphone dependency is connected to mental health. Despite these limitations, this study offers important strengths. The application of the KYHRBS, a nationally representative and large-scale dataset, enhances the applicability of these findings to Korean adolescents and provides robust evidence for public health and educational policy. This study also explored the effects of stress, depression, and living status on smartphone overdependency and examined whether loneliness and anxiety act as mediators in these relationships. Importantly, its strength lies in its categorizing participants into middle and high school students; this facilitated the analysis of differences stemming from different developmental stages. This approach provides a more detailed understanding of smartphone overdependency issues among Korean adolescents, highlighting the need for diverse interventions at the school and community levels, especially for those living in nontraditional family structures.
2. Practical Implications
Pediatric nurses, particularly those working in community settings, play a crucial role in addressing smartphone overdependency among adolescents. Considering the high rates of smartphone ownership, mental health nurses should be proactive in promoting healthy smartphone usage. Nurses can educate teenagers and their families regarding the potential risks of excessive use, including its effects on mental health issues, such as depression, anxiety, and loneliness. Nurses can also build and implement interventions at the community level that target these psychosocial factors and help reduce the incidence of smartphone overdependency. Additionally, they should emphasize the need for balanced digital usage by considering adolescents' loneliness and anxiety levels and family structure. Through these efforts, pediatric nurses can help mitigate the negative consequences of smartphone overdependency, improve mental health outcomes among adolescents, and contribute to their overall well-being.
CONCLUSION
This study is significant because it presents a comprehensive model of adolescent smartphone overdependency that considers the Korean sociocultural context. Its findings provide a basis for formulating intervention programs aimed at preventing smartphone dependency among students, offering tailored insights for middle and high schoolspecific interventions.
Notes
The authors decleared no conflicts of interest.
AUTHOR CONTRIBUTIONS
Conceptualization or/and Methodology: Jeon, Y & Woo, S
Data curation or/and Analysis: Jeon, Y, Woo, S & Seo, S
Funding acquisition: Jeon, Y
Investigation: Jeon, Y, Woo, S & Seo, S
Project administration or/and Supervision: Jeon, Y
Resources or/and Software: Jeon, Y &Woo, S
Validation: Jeon, Y & Woo, S
Visualization: Jeon, Y & Woo, S
Writing: original draft or/and review & editing: Jeon, Y, Woo, S & Seo, S