Available via license: CC BY 4.0
Content may be subject to copyright.
Page 1/36
The Relationship Between Resilience and Mental
Health, Mobile Phone Addiction and Its Differences
Across Levels of Parent-Child Conict Among Left-
Behind Adolescents: A Cross-Sectional Network
Analysis
xiaoya yuan
Southwest University of Science and Technology
Yaxin Mao
Southwest University of Science and Technology
Xiaomin Xu
Southwest University of Science and Technology
Ruolan Peng
Southwest University of Science and Technology
Min Tang
Southwest University of Science and Technology
Gang Dai
Southwest University of Science and Technology
Xinyi Tang
Southwest University of Science and Technology
Haojie Fu
Tongji University
xiao Zhong
Beijing Sport University
Guanzhi zhang
Southwest University of Science and Technology
Bin Wang
Southwest University of Science and Technology
Research Article
Page 2/36
Keywords: Left-behind adolescents, Resilience, Mental health, Mobile phone addiction, Parent-child
conict, Network analysis
Posted Date: December 3rd, 2024
DOI: https://doi.org/10.21203/rs.3.rs-5063332/v1
License: This work is licensed under a Creative Commons Attribution 4.0 International License.
Read Full License
Additional Declarations: No competing interests reported.
Page 3/36
Abstract
Background
mobile phone addiction and mental health problems have become increasingly prominent among left-
behind adolescents in China. In recent years, some studies have focused on the important role of parent-
child relationship and psychological resilience. Therefore, this study aims to explore the
multidimensional relationships among resilience, mental health, and mobile phone addiction among left-
behind adolescents, and to assess the impact of parent-child conict level on these relationships.
Methods
The Brief Symptom Inventory (BSI-18), the Chinese version of the Mobile Phone Addiction Index (MPAI),
the Resilience Scale for Children and Adolescents (RSCA), and the Parent-Child Conict Scale were used
to investigate 2,100 left-behind adolescents in Sichuan Province, and R was run to make network
analysis and network comparison.
Results
(1) A structurally stable network relationship exists between left-behind adolescents' resilience, mental
health, and mobile phone addiction; (2) BSI3 (Anxiety) is the most important node of the network model,
followed by MPAI1 (the inability to control cravings subscale); (3) MPAI1 (the inability to control cravings
subscale) and RSCA4 (family support) are key to connect resilience, mental health, and smartphone
addiction in the study sample; (4) There was a signicant difference in the network structure between
the high- and low-level groups of parent-child conict, no signicant difference in the global strength of
the network, and a signicant difference in the centrality of strength and the centrality of bridge strength.
Conclusions
Chinese left-behind adolescents' resilience and mental health, mobile phone addiction are both
independent and interact with each other to some extent. Specically, high centrality dimensions such as
anxiety, the inability to control cravings, and family support can be prioritised for intervention in related
treatments, or reducing parent-child conict and enhancing resilience to mitigate mobile phone addiction
among left-behind adolescents, thus improving their mental health.
1. Introduction
With China's rapid political and economic development, a large number of people in underdeveloped
regions of the central-west have begun to move to the developed regions searching for jobs, which gives
rise to the problem of left-behind children due to family, policy, and other factors. What's more, left-
behind children are minors under the age of 16 whose parents are both working away from their
hometown, or one of whom is working away while the other is unable to take care of the children [1]. As
of 2020, China saw 66.93million left-behind children, and 138million who were affected by population
Page 4/36
mobility, accounting for 46.4% of China's total child population [2]. The children, who are left behind are
not able to stay with their parents for a long time, lack parental care, are prone to loneliness, anxiety,
depression, and other negative emotions, which not only inuence their studies and lives, but also may
have long-term adverse effects on their mental health [3]. Research has found that, from the past to the
present, migrant parents are more likely to hurt their mental health, including emotions and behaviour [4].
And the situation might become worse over time [5]. Meanwhile, these children tend to develop mobile
phone addiction [6]. As a means of communication tools provided by parents for their children,
smartphones expose children to addiction thanks to being alone for long periods and being spoiled by
their elders [7]. This may lead to their indulgence in the virtual world, which may have a negative impact
on their social skills, learning, behaviour and mental health [8–10]. Left-behind adolescents experiencing
puberty have rapid physical and mental development and face more severe academic pressure and
social environment [11]. Compared with non-left-behind adolescents, left-behind ones show higher risk
propensity and prevalence of mental health problems and mobile phone addiction [12]. We should pay
more attention to left-behind youth and provide them with more comprehensive and effective support
and guidance.
With the popularity of smartphones, the addictive problem has become a new form of Internet addiction
in the mobile era, which is a behavioural addiction that causes psychological and behavioural
phenomenon of users due to the misuse of mobile phones [13]. The 44th report of China Internet
Network Information Centre [14] shows that the number of mobile phone holders in China has reached
847million, of which 17% are teenagers aged 10 to 19 [15]. Empirical studies in different countries have
found that adolescent mobile phone addiction is negatively associated with mental health [16–18].
Adolescents addicted to cell phones are more susceptible to anxiety, depression, and impulsivity at a
high level [19, 20]. Also, cross-lagged analyses show that individuals with higher depression and anxiety
are subject to developing mobile phone addiction [21]. Smartphone addiction has a strong association
with mental health, and there is even a risk of co-morbidity [22]. The cognitive-behavioural model of
pathological Internet use proposed by Davis [23] argues that psychopathologies such as depression,
anxiety, and substance dependence are distally necessary causes of pathological Internet use
symptoms. According to the compensatory Internet use model proposed by Kardefelt-Winther, negative
life situations can increase online behaviour to alleviate negative emotions [24]. Individuals with poor
mental health are more vulnerable to negative emotions and behavioural change suffering negative life
issues [25], which leads to mobile phone addiction[26].
Resilience is often described as the ability to revive or overcome certain adversity in order to extract a
positive outcome from a negative event or situation [28]. Current research has found that resilience has
an important role in the mental health and prevention of mobile phone addiction among left-behind
adolescents [29, 30]. A meta-analysis of 25 studies showed that despite differences in research
objectives and instruments, higher resilience was associated with fewer mental health problems [31]. At
the same time, resilience is also an important predictor of mobile phone addiction, and empirical studies
have found that self-resilience related to "relationships", "curiosity" and "emotional control" have been
found to moderate mobile phone use in both men and women [32]. Existing research suggests that
Page 5/36
resilience can both directly and negatively predict mobile phone addiction among Chinese adolescents
[33] and studies from different countries have found that resilience can also serve as a mediator [34, 35,
7] or an adjustment [36] to inuence mobile phone addiction tendencies. Results from a one-year
longitudinal study also indicated that problematic mobile phone use and resilience predicted
psychological disorders in college students, and that resilience mediated the association [37].
Meanwhile, mobile phone addiction depletes individuals' self-control ability [38], thus reducing their level
of resilience. The resilience process model suggests that resilience is a protective mechanism under
stress and adversity, reecting an individual's ability to adapt and prepare for challenges positively [39,
40]. Researchers have pointed out that resilience is regarded as one of the key protective elements of
Internet addiction, and that Internet addiction often stems from the individual's lack of resilience in self-
control and coping with stress and frustration [41]. Additionally, mobile phone and Internet addiction are
both behavioural addictions with similarities and may have similar addiction mechanisms [42].
According to three key factors, foreign research suggests a link between adolescent resilience,
psychological health, and mobile phone addiction [43], which has argued that the relationship between
college students' resilience and mental health is mediated by Internet addiction, and that increasing
resilience helps prevent Internet addiction and reduce the risk of depression [44]. Domestic studies have
also shown that mobile phone addiction has a direct effect on college students' physical and mental
health, and can also indirectly affect their health through resilience [45]. Internet addiction predicts
depression and anxiety in Chinese rural left-behind children, and resilience plays an independent
mediating role in the relationship between their Internet addiction and depression and anxiety symptoms
[29]. Adverse mental health conditions such as depression, anxiety, stress, and coping styles signicantly
inuence the risk of mobile phone addiction among adolescents and mediate the relationship between
resilience and mobile phone addiction among Chinese adolescents [46]. However, current research still
lacks insight into the network relationship between resilience, mental health, and mobile phone
addiction.
The family is a direct and dominant subsystem inuencing adolescent development [47], so a
harmonious family atmosphere is essential for the healthy physical and mental development of
adolescents. During puberty, there is an increase in conict and a decrease in interaction in parent-child
relationships [48]. A Comparative Study of Parent-Child Relationships in the Internet Age in China, the
United States, Japan, and Korea showed that 82.1% of Chinese primary and secondary school students
had conicts with their parents, and 25.2% of these conicts were focused on Internet access. Studies
have shown that the parent-child relationship is an important mediating mechanism in the family system
that inuences individual development and adaptation [49]. Conict is an important part of the parent-
child relationship, and adolescents with higher parent-child conict are more likely to develop mobile
phone addiction [50, 51]. Substantial empirical studies have also demonstrated that parent-child conict
can negatively predict adolescent mental health [52–54]. In addition, there is an association mechanism
between parent-child relationships (including parental support and parent-child conict) and adolescent
resilience [55, 56]. According to the individual-situation interaction theory, situational factors may interact
with an individual's characteristics [57]. The situational factors of parent-child conict may interact with
Page 6/36
the resilience of individuals' psychological traits to inuence individuals' psychological behavioural
states. Therefore, this study will also study the relationship between parent-child conict and left-behind
adolescents' resilience, mental health, and mobile phone addiction.
At present, most domestic and international studies related to resilience, mental health, mobile phone
addiction, and parent-child conict adopt cross-sectional empirical research methods to explore the
predictive mechanisms by constructing structural equation modelling [45, 46, 44]. While this approach
can also deal with relationships between multivariate variables, it is mainly applied to validation factor
analyses, focusing on verifying the pre-determined model structure [58], may have limitations for highly
complex and dynamic systems, and may be insensitive to the discovery of new structures and patterns.
Resilience [59] and good parent-child relationships [60] are protective factors in children's growth.
"protective model" of adolescent development proposed by Fergus et al. [61], suggests that different
protective factors may interact in predicting developmental outcomes, i.e., the predictive effect of one
protective factor (e.g., resilience) on outcome variables (e.g., mobile phone addiction, mental health)
may be inuenced by another protective factor (e.g., parent-child relationship).
In recent years, the network analysis model has rapidly emerged as a new method of describing
individual psychological traits as a complement to latent variable models, providing new ideas for
understanding human psychological phenomena [62]. In response to the neglect of symptom
interactions in latent variable models of traditional psychological perspectives [63], Borsboom proposed
a network theory of psychopathology, which suggests that symptoms are an integral part of mental
disorders, and that the onset and persistence of mental disorders are driven by tightly intertwined causal
relationships between symptoms and mutually reinforcing feedback mechanisms [64]. Based on this
theory, the study by Cramer et al. used a Gaussian graph theory model to analyse the relational network
of symptoms [65]. Subsequently, this model became the foundational method for employing network
analysis to process transect data. The method refers to symptoms as nodes of a network graph, and
links between symptoms as edges connected between nodes, with the weights of the edges
representing the strength of the association between the nodes, which is usually visualised as the
thickness of the edges in the network graph. The network analysis method can deal with complex
interactions and dynamic relationships between variables, and can reveal the underlying structures and
patterns in the system. Moreover, by displaying the associations between variables through graphical
visualisation, it is possible to see which variables are closely related to each other and how these
associations affect the whole system, which makes the results of the study more concise and easy to
understand.
Therefore, this study takes the group of left-behind adolescents in Sichuan Province, China, as the
research object, and explores the multidimensional relationship between resilience, psychological health,
and mobile phone addiction through network analysis, and assesses the characteristics of the network
structure under different levels of parent-child conict. This study is geographical and population-
specic, combining psychology, sociology, and complex network analysis, which provides a novel
theoretical framework and methodological tool for the study of the relationship between resilience,
Page 7/36
psychological health, and mobile phone addiction among left-behind adolescents, which can help to
provide a scientic basis for the development of precise social intervention strategies.
2. Methodology
2.1 Participants
This study used an online questionnaire platform to conduct a survey in 28 secondary schools in
Sichuan Province for those who met the following criteria: (1) students in their rst to third year of high
school; (2) fullled the condition of being left behind, "neither parent can supervise or take care of me";
(3) gave informed consent and voluntarily took part in this research. A total of 2,824 questionnaires were
distributed, excluding duplicates, missing questions, and consecutive cases with the same answers, and
removing outliers according to the standard deviation of three times, resulting in 2,100 valid
questionnaires, with an effective recovery rate of 74.4%.
2.2 Measurement tools
2.2.1 Mental health
This study used the Brief Symptom Inventory 18 (BSI-18) prepared by Derogatis [66] to measures mental
health. The scale consists of 18 questions with three dimensions, somatization, depression, and anxiety,
and three subscales with six items each. All questions are scored on a 5-point scale (1 = never, 2 = mild,
3 = moderate, 4 = quite severe, 5 = severe), with higher scores indicating higher levels of psychological
distress and lower levels of mental health. In this study, Cronbach's alpha coecient for this scale was
0.926, with 0.845 for the somatization subscale, 0.850 for the depression subscale, and 0.846 for the
anxiety subscale.
2.2.2 Mobile phone addiction
The Chinese version of the Mobile Phone Dependence Index (MPAI) developed by Leung et al. [67] to
measure mobile phone addiction. The scale consists of 17 questions, including four dimensions, namely,
the inability to control cravings subscale, the feeling anxious and lost subscale, the withdrawal and
escape subscale, and the productivity loss subscale, with the number of questions in each dimension
ranging from 3–7. A 5-point scoring system was applied, with higher scores indicating higher levels of
individual cell phone addiction. In this study, the Cronbach's alpha coecient of the scale was 0.892, with
0.849 for the inability to control cravings subscale, 0.786 for the withdrawal and escape subscale, 0.764
for the feeling anxious and lost subscale, and 0.755 for the productivity loss subscale.
2.2.3 Resilience
The Resilience Scale for Chinese Adolescents (RSCA) developed by Yue-Qin Hu and Yi-Qun Gan was
used in this study [68] to measure resilience. The scale consists of 27 questions, including ve
dimensions: goal planning, emotional control, positive thinking, family support, and interpersonal
Page 8/36
assistance, with the number of questions in each dimension ranging from four to six. A 5-point scoring
system was used, with higher scores indicating higher levels of resilience. In this study, the Cronbach's
alpha coecient of the scale was 0.874, with 0.763 for the goal-focused subscale, 0. 747 for the
emotional control subscale, 0.762 for the positive thinking subscale, 0.793 for the family support
subscale, and 0.745 for the interpersonal assistance subscale.
2.2.4 Parent-child conict
Based on Nelissen [69] 's study, the Parent-Child Conict Scale consists of 6 questions on a 5-point
scale, with higher scores indicating higher levels of parent-child conict. The Cronbach's alpha
coecient for the scale in this study was 0.797.
2.3 Data analysis
In this study, SPSS23.0 was applied for total score calculation, common method bias test, and
descriptive statistical analysis, R (4.3.2) was used for network analysis, and R packages qgraph (1.9.8),
mgm (1.2–14), networktools (1.5.2), and bootnet (1.5.6) were used for network estimation and
visualisation, network centrality estimation and stability tests [70]. The top 27% of the total parent-child
conict score was taken as the high level of the parent-child conict group, and the bottom 27% of the
total score was taken as the low level of the parent-child conict group [71], constructed the networks
separately and compared them using the R package NetworkComparisonTest (2.2.2) [72].
2.3.1 Data pre-processing
Invalid questionnaires were ltered according to the following steps: rstly, questionnaires with less than
300s of response time were excluded, then questionnaires with missing items were deleted to facilitate
subsequent data analysis, then duplicate cases were identied and deleted based on information such
as IP, time of submission, age, school, etc., and those with more than 15 consecutive questions with the
same response within the same scale were considered as invalid data were deleted, and nally, Z-scores
for the respective scale and its dimensions were calculated, and extreme case data with Z-scores
exceeding plus or minus 3 were removed to make the results more stable and reliable. Finally, the Z-
scores of each scale and its dimension scores were calculated, and the data of extreme cases with Z-
scores exceeding plus or minus 3 were deleted to make the results more stable and reliable.
2.3.2 Network estimation and visualisation
In this study, each dimension of the Brief Symptom Scale 18, the Mobile Phone Dependence Index, and
the Resilience Scale for Chinese Adolescents was used as a node, and the correlation between the
dimensions were used to generate the edges of the network, and the partial correlation structured
network was constructed and visualised using the R package qgraph (1.9.8) [73]. Applying the least
absolute shrinkage and selection operator (LASSO) [74] and the extended Bayesian information criterion
(EBIC) [75]. Regularisation was performed with a tuning parameter of 0.5 to prevent overtting and
obtain a concise and interpretable structure. The predictability of each node was calculated using the R
Page 9/36
package mgm(1.2–14) [76]. The more predictable a node is, the more it can be predicted or determined
by other connected nodes in the network; conversely, if the predictability value is low, we need to
intervene directly on the node or look for markers outside the network [62, 77]. In addition, using the
spinglass algorithm [78] for modular analysis of node clustering to reveal and optimise the structure of
associations in the network.
In the network, green edge lines represent positive correlations and red edge lines represent negative
phases, and the thickness of the edges indicates the absolute magnitude of the correlation, with thicker
edges indicating higher correlations. Using the Fruchterman-Reingold algorithm [79]. A visual network
layout was performed so that nodes with strong and numerous connections were located in the centre of
the network and nodes with weak and few connections were distributed at the periphery of the network.
When performing network comparisons between the high and low level groups of parent-child conict,
the averageLayout function in the R package qgraph (1.9.8) was used to perform network layouts,
presenting a consistent visual layout of nodes using the average position in the two networks.
2.3.3 Centrality estimates
In network analysis, the centrality metric is an important metric used to describe how central a node is in
the network. Using the centrality function in the R package qgraph (1.9.8) [73]. Calculate the strength,
betweenness and closeness of the network nodes and use the bridge function in the R package
networktools (1.5.2) [80] Calculate bridge centrality metrics for network nodes, including bridge strength,
bridge betweenness, and bridge closeness. Previous studies [81] found that strength centrality is the
most persuasive metric in psychometrics, and when the three metrics do not have the same numerical
ordering, the result of the ordering of strength centrality generally prevails. Therefore, in this study, we
chose the strength centrality and the bridge strength centrality of the node to be reported, and plotted
the normalised (z-scored) values for each node. Where strength refers to the sum of the absolute value
of the weights of all edges connecting the node, the larger its value indicates that the node is more
closely connected to other nodes and has a greater effect on the whole network [82]; Bridge Strength
refers to the sum of the absolute values of the weights of the edges of the nodes of other communities
that are connected to this node, the higher its value the more inuence this node has on the nodes of
other communities [83].
2.3. 4 Accuracy and stability test
The accuracy and stability of the constructed network were calculated and veried using the R package
bootnet (1.5.6) using the Bootstrapping method [70]. The estimation results were validated and
analysed. Firstly, the 95% condence interval (CI) of each edge weight in the network are calculated
based on the non-parametric bootstrapping method. If the 95% condence intervals of the sampling set
and the original dataset overlap more, it means that the network edge weights are estimated more
accurately. Secondly, based on removing the case-dropping subset bootstrap to assess the stability of
the central indicator, delete a certain proportion of samples, and re-estimate the network, if the network
structure of the central indicator order remains unchanged, the stability is good. And the correlation
Page 10/36
stability coecient (CS-coecient) of the network is calculated using the corStability function for
assessment, which indicates that when the maximum proportion is removed, the correlation between the
original centrality indicator and the network centrality indicator of the subset at 95% probability is higher
than 0.7, the value of the CS-coecient should not be lower than 0.25, and higher than 0.5 indicates
good centrality stability [70]. Finally, the centrality difference test was conducted to assess the
differences in centrality indicators between nodes and edges using the non-parametric bootstrap
method. p < 0.05 was considered statistically signicant.
2.3.5 Network comparison
Using the R package NetworkComparisonTest (2.2.2) [72] Network Comparison Test (NCT) was
performed on the high and low-level groups of parent-child conict. Network invariance test and global
strength invariance test were performed in 5000 permutations to assess whether the two networks
differed in weight of edges and global strength.
3. Results
3.1 Common method bias test
As all data in this study were collected using participant self-report, there may be common
methodological bias. The Harman one-way method of testing was used in this study, which showed that
there were a total of 12 factors with eigenvalues greater than 1, the rst of which had a percentage of the
variance of 23.4%, which did not exceed the critical value of 50% [85]. Therefore, common method bias
had little effect on the results of this study.
3.2 Descriptive statistics
Of the total of 2,100 adolescent secondary school students included in this study (Table1), 905 (43.1%)
were boys, and 1,195 (56.9%) were girls, with an average age of 15.60 years (age 12 to 20, SD = 1.78); the
urban population was 386 (18.4%), the township population was 528 (25.1%), and the rural population
was 1,186 (56.5%). An analysis of the respondents' left-behind status shows that 78 (3.7%) live with their
fathers but cannot be under their guardianship or care, 75 (3.6%) have fathers who are away for less than
three months a year, 415 (19.8%) have fathers who are away from three to six months a year, 722 (34.4%)
have fathers who are away for more than six months a year, and 754 (35.9%) have fathers who are
absent for almost all of the year (35.9%), 56 (2.7%) whose fathers had passed away; 135 (6.4%) who
lived with their mothers but could not be under their guardianship or care, 103 (4.9%) whose mothers
were away for less than three months a year, 391 (18.6%) whose mothers were away for three to six
months a year, 691 (32.9%) whose mothers were away for more than six months a year and 737 (35.9%)
whose mothers were away almost all the year. 737 (35.1%), and 43 (2.0%) whose mothers had passed
away.
Page 11/36
Table 1
Basic information of participants
Variable Number (%)/M(SD)
Synthesis High-level Parent-Child
Conict Group Low-level Parent-Child
Conict Group
Gender
Boys 905
(43.1%) 219 (38.6%) 290 (51.1%)
Girls 1,195
(56.9%) 348 (61.4%) 277 (48.9%)
Age 15.60
(1.78) 15.94 (1.81) 15.30 (1.83)
Residency
Urban 386
(18.4%) 111 (19.6%) 91 (16.0%)
Town 528
(25.1%) 151 (26.6%) 134 (23.6%)
Rural 1,186
(56.5%) 305 (53.8%) 342 (60.3%)
Living with my father or not
Living with my father lack of his
custody and care 78 (3.7%) 27 (4.8%) 16 (2.8%)
Within 3 months of my father's
absence annually 75 (3.6%) 36 (6.3%) 16 (2.8%)
Within 3–6 months of my
father's absence annually 415
(19.8%) 153 (27.0%) 89 (15.7%)
More than 6 months of my
father's absence annually 722
(34.4%) 168 (29.6%) 213 (37.6%)
Almost a whole year of my
father's absence 754
(35.9%) 172 (30.3%) 217 (38.3%)
My father has passed away. 56 (2.7%) 11 (1.9%) 16 (2.8%)
Living with my mother or not
Living with my mother lack of
his custody and care 135 (6.4%) 44 (7.8%) 23 (4.1%)
Within 3 months of my mother's
absence annually 103 (4.9%) 47 (8.3%) 19 (3.4%)
Page 12/36
Variable Number (%)/M(SD)
Synthesis High-level Parent-Child
Conict Group Low-level Parent-Child
Conict Group
Gender
Within 3–6 months of my
mother's absence annually 391
(18.6%) 156 (27.5%) 94 (16.6%)
More than 6 months of my
father's absence annually 691
(32.9%) 143 (25.2%) 200 (35.3%)
Almost a whole year of my
mother's absence 737
(35.1%) 169 (29.8%) 214 (37.7%)
My mother has passed away. 43 (2.0%) 8 (1.4%) 17 (3.0%)
The high level of parent-child conict group consisted of 567 individuals, 219 males (38.6%) and 348
females (61.4%), with a mean age of 15.94 years (age 12 to 20, SD = 1.81); the low level of parent-child
conict group consisted of 567 individuals, 290 males (51.1%) and 277 females (48.9%), with a mean
age of 15.30 years (age 12 to 20, SD = 1.83). Comparing the differences between the high level of parent-
child conict group and the low level of parent-child conict group on some demographic variables, the
descriptive statistics showed that the two groups were essentially similar on the variables of gender, age,
place of birth, and retention (see Table1). Through t-tests, there were no statistically signicant
differences (p > 0.05) in demographic variables other than age, township, and rural birthplace, suggesting
that in most respects the two groups were essentially equivalent.
The overall sample mean total score for the BSI was 33.29 (SD = 10.70), the MPAI mean total score was
48.67 (SD = 12.18), the RSCA mean total score was 87.43 (SD = 14.61), and the Parent-Child Conict
Scale mean total score was 9.66 (SD = 3.42). The mean and standard deviation of each dimension of the
scale, i.e., network nodes, are shown in Table2. The mean of the total score of the summary symptom
scale for the high level of parent-child conict group was 41.72 (SD = 10.53), the mean of the total score
of the index of mobile phone dependence scale was 52.60 (SD = 11.19), the mean of the total score of
the resilience scale for adolescents was 77.99 (SD = 12.29), and the mean of the total score of the
parent-child conict scale was 14.38 (SD = 2.54). The low level of parent-child conict group had a total
score mean of 26.58 (SD = 7.28) for the Brief Symptoms Scale, 42.37 (SD = 11.53) for the Mobile Phone
Dependence Index Scale, 97.57 (SD = 13.97) for the Adolescent resilience Scale, and 6.38 (SD = 0.49) for
the Parent-Child Conict Scale.
Page 13/36
Table 2
Mean, standard deviation, predictability, and centrality indicators for each node
Nodes Content Mean
(M) Standard
deviation
(SD)
Predictability Dissociation Bridge
Strength
BSI1 Somatisation 9.96 3.80 0.49 0.88 0.37
BSI2 Depression 11.65 4.14 0.64 1.10 0.48
BSI3 Anxiety 11.68 4.16 0.68 1.23 0.45
MPAI1 Inability to
control cravings 20.05 6.04 0.51 1.21 0.64
MPAI2 Feeling anxious
and lost 10.39 3.67 0.49 1.05 0.34
MPAI3 Withdrawal and
escape 8.95 2.95 0.28 0.68 0.12
MPAI4 Productivity loss 9.28 2.85 0.39 1.09 0.32
RSCA1 Goal planning 16.26 3.81 0.35 0.79 0.16
RSCA2 Affect control 18.03 4.27 0.42 0.90 0.49
RSCA3 Positive thinking 14.06 3.17 0.31 0.88 0.33
RSCA4 Family support 19.95 4.87 0.45 0.95 0.60
RSCA5 Help-seeking 19.13 4.83 0.23 0.57 0.24
3.3 Network structure
The network structure of the left-behind adolescents' mental health, mobile phone addiction, and
resilience is demonstrated in Fig.1a. There are 12 nodes in the network, and a total of 45 non-zero edges
actually exist, including 20 negative edges and 25 positive edges, accounting for 68.18% of the number
of possible connected edges. The proportion of the circle around a node that is lled represents the
predictability of that node, with a larger proportion of the lled portion indicating a higher predictability of
that node, with an average predictability of 0.44 (range 0.23 to 0.68, Table2).
The network module analysis displayed that the nodes of mental health, mobile phone addiction, and
resilience of the left-behind children clustered with each other to form three node communities (Fig.1b),
which was consistent with the three research variables and their dimensions. The communities for
mental health and mobile phone addiction were more strongly connected internally; whereas the
communities for resilience were weaker except for RSCA1 (goal planning) and RSCA3 (positive thinking)
which were deeply associated. The links between the three communities were also stronger, with the
dimensional nodes interacting with each other. The strongest connections were between BSI2
(depression) and RSCA4 (family support), BSI3 (anxiety) with RSCA2 (affect control) and MPAI2 (the
Page 14/36
feeling anxious and lost), and MPAI1 (inability to control cravings) with RSCA4 (family support) and
RSCA1 (goal planning) directly.
3.4 Indicators of centrality
The results of the centrality index of psychological status, mobile phone addiction and resilience network
of left-behind adolescents are shown in Fig.2, and the specic values are shown in Table2. Node BSI3
(anxiety, Strength = 1.23) has the highest strength centrality, and node MPAI1 (inability to control
cravings, Strength = 1.21) is the second highest. In terms of bridge strength centrality, nodes MPAI1
(inability to control carving, Bridge Strength = 0.64) and RSCA4 (family support, Bridge Strength = 0.60)
were signicantly stronger than the other nodes. The results of the variability test for the centrality index
also indicate that the high centrality nodes are stable and reliable.
3.5 Accuracy and stability of the network
The results of the edge weight bootstrap procedure are shown in Fig.3, where the network estimation is
moderately accurate and there is a partial overlap between the 95% CI of the edge weights. The results
of the excluded cases bootstrap method are shown in Fig.4, with CS coecients of 0.75 for strength,
bridge strength, closeness, and edges, and 0.594 for betweenness, which are greater than 0.5, saying
that the network estimation has good stability. The result of bootstrapped difference tests is shown in
Fig.5, nodes and edges with strong centrality in the network are statistically more strongly different than
other nodes in the network, further indicating that the results of centrality analysis are stable and
generalisable.
3.6 Comparison of networks
Network Comparison Test (NCT) was performed on the Parent-Child Conict High-Level Group and Low-
Level Group. The results show that both networks contain 12 nodes, and the result of the parent-child
conict high-level group contains 47 edges, while the parent-child conict low-level group contains 42
edges, and the visualisation of the network is shown in Fig.6. By centrality analysis, in the parent-child
conict high-level group, the core nodes and core bridge nodes are BSI3 (anxiety) and MPAI1 (inability to
control cravings). In the low-level group of parent-child conict, the core nodes were BSI3 (anxiety),
MPAI2 (the feeling anxious and lost), and the core bridge nodes were MPAI1 (inability to control
cravings), and RSCA4 (family support).
The results of the network invariance test showed a signicant difference in structure between the high
and low-level groups of parent-child conict (M = 0.265, p < 0.001), and the results of the global strength
invariance test did not nd a signicant difference in the global strength of the network (high-level group:
5.526 vs. low-level group: 4.952; S = 0.566, p = 0.162). Tests of centrality invariance revealed signicant
differences in both intensity centrality (p < 0.001, cohen's d = 0.456) and bridge intensity centrality (p <
0.001, cohen's d = 0.828). A total of 6 node centrality of BSI1 (somatization), RSCA5 (interpersonal
assistance), and BSI2 (depression) were signicantly different (p < 0.05), accounting for 50% of the total.
Page 15/36
The results of the borderline weight invariance test showed that a total of 15 borderlines such as RSCA4
(family support) differed signicantly (p < 0.05) from BSI2 (depression), MPAI1 (inability to control
cravings), and RSCA2 (affect control), and BSI1 (somatization) differed signicantly (p < 0.05) from BSI2
(depression), and MPAI4 (productivity loss), which accounted for about 28% of the total number of
borderlines.
4. Discussion
In this study, we used network analysis to explore in depth the associations between the dimensions of
resilience and mental health and mobile phone addiction among Chinese left-behind adolescents, and to
further compare the core dimensions and network structure differences in the networks of resilience and
mental health and mobile phone addiction among left-behind adolescents with different levels of parent-
child conict. The results of the study found that: (1) there exists a structurally stable network
relationship between resilience, psychological health, and mobile phone addiction among left-behind
adolescents; (2) BSI3 (anxiety) is the most central node in the network model, followed by MPAI1
(inability to control cravings); (3) MPAI1 (inability to control cravings) and RSCA4 (family support) are the
most central bridge nodes connecting resilience, psychological health and mobile phone addiction in the
study sample; (4) there were signicant differences in the network structure between the high and low-
level groups of parent-child conict, specically no signicant differences in the global strength of the
network, and signicant differences in both strength centrality and bridge strength centrality.
4.1 Network structure and its core dimensions of resilience, mental health, and mobile phone addiction
of left-behind adolescents,
4.1.1 Network structure
The study showed that there were three relatively independent clusters in the networks of resilience and
mental health and mobile phone addiction among left-behind adolescents. The mental health and mobile
phone addiction communities are more closely connected internally, in line with psychopathology
network theory [86], i.e. some symptoms are more closely connected to each other than others, and the
clusters form manifestations of mental disorders. Connections within the resilience community are
looser overall, with only a strong positive correlation shown between goal planning and positive thinking,
with the ve dimensions representing different dimensions of the individual, the environment, and so on,
which work together to contribute to the overall resilience of the individual. Looking at the network as a
whole, mental health and mobile phone addiction in general have strong negative associations with
resilience, while a positive correlation was shown between mental health and mobile phone addiction.
Enhancing resilience may help to reduce the risk of mobile phone addiction and promote mental health
among left-behind adolescents.
4.1.2 Network core dimensions
Page 16/36
This study found that inability to control carving occupies a crucial position in the network of left-behind
adolescents' resilience and mental health, and mobile phone addiction. is both a core node and a core
bridge node, which has a profound impact on the overall network structure. Inability to control carving is
manifested in the individual's diculty in self-regulation, investing too much time in mobile phone use
without being able to manage it effectively [67]. The results of this study are similar to those of existing
network analysis studies [87, 88], and are also consistent with the Interaction of Person-Affect-Cognition-
Executionmodel (I-PACE) proposed by Brand [89], diminished control over decision-making can be
transferred to behavioural addictions and specic Internet-use disorders. Network visualisation results
showed a stronger direct relationship between inability to control carving and goal planning and family
support in resilience, possibly because the inability to control carving creates an attention bias toward
the automation of addictions [90] that affects addicts' attention allocation and cognitive resource use.
According to the social displacement hypothesis, addiction to smartphones can neglect face-to-face
interactions with friends and family members [91] and lack of real-world social support [92] and left-
behind adolescents are already more deprived of parental companionship, leading to lower levels of
family support and interpersonal assistance. For mental health, uncontrollability was only directly related
to depression. This may be due to the fact that both mobile phone addiction and depression are related
to the dopamine system in the brain[93, 94], and both show similar symptoms such as loss of interest,
social withdrawal, and mood swings [95]. In contrast, adolescent somatization and anxiety are
inuenced by a wide range of factors, including genetics, environment, past experiences, psychological
states, and psychological changes [96–98], and so were only indirectly affected by runaway sex in this
network. Notably, the results of the present study showed that inability to control carving in mobile
phone addiction was positively associated with positive perceptions of resilience, in contrast to existing
research where hopeful attitudes may reduce adolescents' dependence on smartphones [99] The results
are not consistent with the results of Left-behind adolescents may need to take on family responsibilities
earlier because of the unique nature of their home environment, an experience that may allow them to
hone their independence and coping strategies [100] and thus be able to adapt to adversity more quickly
in certain situations, and to make self-determination, self-planning and problem-solving with a more
optimistic and positive attitude. This also suggests that left-behind adolescents may realise the
seriousness of the problem after experiencing uncontrolled mobile phone addiction, and may engage in
self-reection and seek adjustment, in which their positive thinking may be enhanced.
In addition, similar to previous network analysis studies [101, 102], anxiety is also one of the most central
nodes of the network model in this study. Anxiety is the brain's response to danger, stimuli, and is a state
that an organism will actively try to avoid [103]. In this network, anxiety mainly affects withdrawal or
escape in mobile phone addiction. the over-comfort pathway in the pathway theory of problematic
mobile phone use proposed by Billieux [104] states that individuals with increased anxiety contribute to
their mobile phone dependence addiction out of needs such as comforting reassurance. Meanwhile,
according to the Attention Gate Model (AGM), anxiety overestimates the time interval of negative stimuli
[105, 106] by paying attention to them [107]. When addicts are in withdrawal, their attention is more likely
to be focused on the time they are waiting to use their mobile phones, leading to distorted perceptions of
Page 17/36
time and making withdrawal more dicult for addicts by making it feel more dicult. Emotional control
and family support in the resilience of left-behind adolescents are directly and negatively affected by
anxiety. Relevant studies have shown that chronic anxiety leads to a more sensitive response to external
stimuli [108] which leads to diculties in regulating negative emotions [109]. Due to parental absence
and limited resources, left-behind adolescents have diculties in obtaining immediate emotional
support and proper guidance, and lack effective strategies for emotional expression and control. Anxiety
may affect individuals' perception and utilisation of family support [110], or due to communication
barriers [111], the inability to effectively access needed support from family members.
Family support is another one of the core bridge nodes in the network structure of this study, connecting
depression, anxiety, somatization in left-behind adolescents' mental health and uncontrollability and
ineffectiveness in mobile phone addiction. It refers to the tolerant, respectful and supportive attitudes of
family members [68], an important external factor in resilience [112]. Increased family support can
enhance family members' psychological well-being [113, 114]. In the present study, the family support of
the left-behind adolescents also had direct effects on their depression, anxiety and somatization to
varying degrees. Among them, the effect on depression was the most signicant, and lack of parental
care leading to depression was one of the most prominent problems among left-behind children [115].
The lack of direct parent-child interactions and emotional communication, and the sense of stress
caused by the impairment of parental care resources or unmet needs of left-behind adolescents have
both immediate and delayed negative predictive effects on depression [116]. The effect of family support
on anxiety was also more signicant, consistent with previous research ndings [117]. The relatively
weak effect on somatization, on the other hand, may be related to the indirect nature of somatization
symptoms and the multiple inuencing factors [96] of interference. Existing research suggests that
family support signicantly and negatively predicts adolescent mobile phone addiction [118], but in the
network structure of this study, family support was negatively associated with inability to control carving
and showed positive results with inecacy. Inecacy refers to excessive mobile phone use resulting in
lower academic or work productivity [67] Ineffectiveness Left-behind adolescents usually grow up with
their grandparents, who are more lenient than their parents based on their love for their grandchildren,
and do not know whether to stop, encourage, or ignore adolescents' problematic behaviours due to their
lack of knowledge and backwardness [119]. Parents who work outside the home tend to feel indebted to
their children and indulge them completely, and lack proper guidance and supervision of their children's
learning in education management. Such intergenerational education is often prone to spoiling, and
although a certain degree of family support is provided for children, improper discipline in life also leads
to problematic behaviours of excessive mobile phone use and affects learning eciency.
4.2 Differences in parent-child conict levels between resilience, mental health, and mobile phone
addiction networks in adolescents who are left behind
In this study, it was found that the differences in network invariance test at different levels of parent-child
conict were signicant, and the differences in the global strength of the networks were not signicant.
It indicated that the two networks had similar levels of connectivity overall and maintained some stability
Page 18/36
at different levels of parent-child conict. Families with high levels of parent-child conict were more
connected within the mental health and mobile phone addiction communities, and distant within the
resilience community, and even showed a signicant negative internal correlation. This result reveals
that parent-child conict may undermine left-behind adolescents' resilience [56] to further disintegrate
mental health and lead to mobile phone addiction problems. In addition, the network structure of the
high-level parent-child conict group had stronger direct associations among the three associations, the
correlations among the nodes were more disordered, and changes in individual dimensions were more
likely to spread across different associations. This suggests that intense parent-child conict may lead
to a rapid spread of risks or negative effects and cause individuals to show a high degree of exibility
and diversity in their psychological adaptations [120] that can cope with different situations through
multiple psychological mechanisms. In contrast, in the group with low levels of parent-child conict,
inter-community connections were relatively looser, intro-community associations within resilience were
stronger, and the direct link between mental health and mobile phone addiction communities was
signicantly reduced, suggesting that resilience effectively buffers the interplay between mental health
and mobile phone addiction problems [29, 45], playing a better protective role which played a better
protective role.
Somatization was the node with the most signicant central difference in intensity, and was signicantly
more associated with depression, positive thinking, and inecacy in left-behind adolescents with high
levels of parent-child conict than in the group with low levels of parent-child conict. Somatization is a
unique response to psychosocial stresses [66] and parent-adolescent conict can exacerbate
adolescent somatization symptoms [121].. Parent-child conict is one of the stressors for adolescents in
puberty [122], prolonged exposure to high-pressure and stressful environments increases the risk of
mental health problems in adolescents [123] and is more likely to lead to multiple co-morbidities of
psychological problems [124]. In the present study, somatization showed a stronger negative correlation
with inecacy in the high levels of parent-child conict group, which may be due to the physical
discomfort of somatization [125] and intense parent-child conict [126] are the more dominant causes
of academic ineciency among left-behind adolescents. It is also possible that the separation anxiety of
parents of left-behind adolescents (Scharf & Goldner, 2018) and severe family conict [127] lead to
stricter parental psychological and behavioural control, which prevents excessive mobile phone use from
affecting the adolescents' eciency.
Positive thinking was the node with the most signicant difference in bridge strength centrality. In the
network structure of high parent-child conict, positive thinking increased the direct positive correlation
with lack of self-control, avoidance, and ineciency in the mobile phone addiction community. This
implies that more severe smartphone addiction problems such as lack of self-control, escapism, and
ineciency are intertwined with more optimistic attitudes, which is inconsistent with existing research
[129]. In addition to the previously mentioned ability of left-behind adolescents to adapt more quickly to
adversity, make positive self-decisions and plans, and potentially reect deeply and seek change, it is
also possible that negative cognition allows them to feel the realities of the situation such as family
conict more acutely, rather than being fully immersed in the world of mobile phones, which reduces
Page 19/36
performance in areas such as loss of control [100]. However, at the same time, according to the "loss of
compensation" hypothesis [130], their addiction to mobile phones may be more of an emotional
attachment, and their inner dependence on mobile phones will be strongly manifested during withdrawal.
In addition, The connecting line between inability to control carving and family support is borderline with
the most signicant difference in weights and is strongly negatively correlated in the low-level group.
Family members in families with low parent-child conict tend to use positive coping strategies [131],
and family support may be more accessible and effective. In contrast, if the parent-child conict is high,
although family support may be present, its effectiveness may be diminished by the conict, and
adolescents may seek other forms of social support to cope with the conict, thus attenuating the direct
effect of family support on mobile phone addiction loss of control, and instead having interpersonal
assistance directly associated with loss of control.
4.3 Research limitations and future research perspectives
This study used a network analysis model to examine the relationship between psychological variables,
providing a multidimensional understanding of left-behind adolescents' resilience in relation to mental
health, mobile phone addiction, and parent-child conict from a cross-sectional perspective for
psychological research, as well as expanding and deepening the theory of resilience. This study reveals
that teachers and clinical interveners can maximise psychosocial interventions by focusing on the high
centrality dimensions such as anxiety, inability to control carving, and family support when confronting
left-behind adolescents' mental health and mobile phone addiction issues [88] The However, this study
still has some limitations. First, this study only used cross-sectional data to construct a partial
correlation network, and was unable to infer causal relationships. Although the important role of core
nodes can be armed based on the network model centrality feature [88], it should be veried by
longitudinal or experimental design in the future. Second, the average predictability of the nodes of the
network analysis model in this study was not high, indicating that the networks could not predict each
other well internally and were inuenced by factors outside the network (e.g., environmental, biological
factors, other psychological variables) [62]. Future research can collect more data with more
representativeness and accuracy, identify and control for external variables that may affect the
predictability of the model, and also incorporate multimodal indicators to construct the network based
on relevant theories. Third, mobile phone addiction measured through self-report may be affected by
social expectation bias. Some experiments have found that self-reported mobile phone use does not
match the actual situation [132], future research should be cautious in interpreting estimated
smartphone use with more objective metrics or a combination of personal interviews and guardian
observations for evaluation.
5 Conclusion
(1) The network structure of left-behind adolescents' resilience, mental health, and mobile phone
addiction is stable, in which anxiety and inability to control cravings are core nodes, and controlling
Page 20/36
inability and family support are core bridge nodes. Practitioners should focus on the high centrality
dimension for effective intervention for left-behind adolescents.
(2) There are signicant differences in the network structure of resilience, psychological health, and
mobile phone addiction among left-behind adolescents at different levels of parent-child conict. In
families with higher levels of parent-child conict, the network structure is more complex, and the
resilience of left-behind adolescents is undermined, with risks and negative effects spreading faster;
while in the lower counterpart, resilience has a protective effect.
Declarations
Ethics approval and consent to participate
This study was conducted in strict accordance with the Declaration of Helsinki and received ethical
approval from the Institutional Committee of Law School, Southwest University of Science and
Technology in Mianyang, China (No. LL23001). Informed consent was signed by each adult participant,
or their parent(s) or legal guardian(s) on behalf of adolescent participants.
Consent for publication
Not applicable.
Availability of data and materials
The data supporting the ndings of this study are available from the corresponding author, upon
reasonable request, immediately following publication and no end date. We can share individual
participant data that underlie the results reported in this article, after deidentication (text, tables, gures
and appendices).
Competing interests
The authors declare that they have no competing interests.
Funding
This work was supported by the General Project of Sichuan Philosophy and Social Science Planning
Fund of Sichuan Province [Project No. SCIJ23ND226], Steering Committee for Teaching Psychology in
Higher Education, Ministry of Education [Project No. 20232010], and Institute of Psychology, Chinese
Academy of Sciences [Project No.GJ202003].
Authors' contributions
All authors contributed to the study conception and design. XY curated and analyzed the questionnaire
data, visualized the results, interpreted the results of the network analysis model, and was the main
Page 21/36
contributor to writing the rst draft of the manuscript. YM curated the data, conducted formal analyses,
reviewed and edited the manuscript. XX investigated, curated the data and conducted a formal analysis.
RH translated, reviewed and edited the manuscript. YW investigated and formally analyzed the data. MT
investigated and formally analyzed the data. GD investigated and formally analyzed the data. XT
investigated and formally analyzed the data. HFproposed the conceptualization and methodology and
performed the result validation. XZ proposed the conceptualization and methodology, performed
supervision, reviewed and edited the manuscript. GZ performed supervision, provided related resources,
reviewed and edited the manuscript. BW proposed the conceptualization and methodology, provided
funding for the project, provided related resources, performed supervision, and was the project
administrator. All authors read and approved the nal manuscript.
Acknowledgements
The authors would like to acknowledge the participants in the study.
Clinical trial number
not applicable.
References
1. State Council of the People’s Republic of China. State council on strengthening rural left-behind
children advice on care and protection work [Internet]. 2016 [cited 2024 Apr 25]. Available from:
https://www.gov.cn/zhengce/content/2016-02/14/content_5041066.htm
2. National Bureau of Statistics of China, UNICEF China, UNFPA China. What the 2020 Census Can Tell
Us About Children in China: Facts and Figures [Internet]. 2023 [cited 2024 Apr 24]. Available from:
https://www.unicef.cn/en/reports/population-status-children-china-2020-census
3. Yao YS, Kang YW, Jin YL, Chen Y, Gong WZ, Zheng L, An Z, Tao FB, Hao JH. Analysis on physical and
mental health and related inuential factors among those “left behind” adolescents in Anhui
province. Zhonghua Liu Xing Bing Xue Za Zhi. 2012 Jul;33(7):681–4.
4. Wang F, Lin L, Xu M, Li L, Lu J, Zhou X. Mental Health among Left-Behind Children in Rural China in
Relation to Parent-Child Communication. Int J Environ Res Public Health. 2019 May 2;16(10):1855.
5. Zhang X, Dai Z, Antwi CO, Ren J. A Cross-Temporal Meta-Analysis of Changes in Left-Behind
Children’s Mental Health in China. Children-Basel. 2022 Apr;9(4):464.
. Cai J, Wang Y, Wang F, Lu J, Li L, Zhou X. The Association of Parent-Child Communication With
Internet Addiction in Left-Behind Children in China: A Cross-Sectional Study. Int J Public Health.
2021 Sep 10;66:630700.
7. Zhou M, Bian B, Zhu W, Huang L. The Impact of Parental Migration on Multidimensional Health of
Children in Rural China: The Moderating Effect of Mobile Phone Addiction. Children-Basel. 2023
Jan;10(1):44.
Page 22/36
. Wang X. A Longitudinal Analysis of Mobile Phone Dependence in Chinese Adolescents: The Risk
and Promotive Factors of Mobile Phone Dependence Trajectories. Advances in Psychology. 2021
Jan 1;11:9–19.
9. JING J, GAO C, Niu G. The effect of internet use on empathy. Advances in Psychological Science.
2017 Jan 1;25:652.
10. Tang CSK, Koh YYW. Online social networking addiction among college students in Singapore:
Comorbidity with behavioral addiction and affective disorder. Asian J Psychiatr. 2017 Feb;25:175–8.
11. Chai X, Lin D. School transition during adolescence: Turning crisis into opportunity. Advances in
Psychological Science. 2021 Jan 1;29:864.
12. Ge Y, Se J, Zhang J. Research on relationship among internet-addiction, personality traits and mental
health of urban left-behind children. Glob J Health Sci. 2014 Dec;7(4):60–9.
13. de Sola J, Fonseca F, Rubio G. Cell-Phone Addiction: A Review. Frontiers in Psychiatry. 2016 Oct
24;7.
14. CNNIC. The 44th China Statistical Report on Internet Development [Internet]. 2019 [cited 2024 Aug
20]. Available from: https://www.cac.gov.cn/2019-08/30/c_1124938750.htm
15. Zhen R, Li L, Ding Y, Hong W, Liu RD. How does mobile phone dependency impair academic
engagement among Chinese left-behind children? Children and Youth Services Review. 2020 Sep
1;116:105169.
1. Yang LL, Guo C, Li GY, Gan KP, Luo JH. Mobile phone addiction and mental health: the roles of sleep
quality and perceived social support. Front Psychol. 2023;14:1265400.
17. Cimadevilla R, Jenaro C, Flores N. Impact on Psychological Health of Internet and Mobile Phone
Abuse in a Spanish Sample of Secondary Students. Rev Argent Clin Psicol. 2019 Nov;28(4):339–47.
1. Park SY, Yang S, Shin CS, Jang H, Park SY. Long-Term Symptoms of Mobile Phone Use on Mobile
Phone Addiction and Depression Among Korean Adolescents. Int J Environ Res Public Health. 2019
Oct;16(19):3584.
19. Desouky DES, Abu-Zaid H. Mobile phone use pattern and addiction in relation to depression and
anxiety. East Mediterr Health J. 2020;26(6):692–9.
20. Li Y, Li G, Liu L, Wu H. Correlations between mobile phone addiction and anxiety, depression,
impulsivity, and poor sleep quality among college students: A systematic review and meta-analysis.
J Behav Addict. 2020 Sep;9(3):551–71.
21. Kang Y, Liu S, Yang L, Xu B, Lin L, Xie L, Zhang W, Zhang J, Zhang B. Testing the Bidirectional
Associations of Mobile Phone Addiction Behaviors With Mental Distress, Sleep Disturbances, and
Sleep Patterns: A One-Year Prospective Study Among Chinese College Students. Front Psychiatry.
2020 Jul 17;11:634.
22. Nahidi M, Ahmadi M, Fayyazi Bordbar MR, Morovatdar N, Khadem-Rezayian M, Abdolalizadeh A. The
relationship between mobile phone addiction and depression, anxiety, and sleep quality in medical
students. Int Clin Psychopharmacol. 2024 Mar;39(2):70–81.
Page 23/36
23. Davis RA. A cognitive-behavioral model of pathological Internet use. Computers in Human Behavior.
2001 Mar 1;17(2):187–95.
24. Kardefelt-Winther D. A conceptual and methodological critique of internet addiction research:
Towards a model of compensatory internet use. Computers in Human Behavior. 2014 Feb 1;31:351–
4.
25. Hammen C. Stress and Depression. Annual Review of Clinical Psychology. 2005 Apr 27;1(Volume 1,
2005):293–319.
2. Shuan S. Development of the Smartphone Addiction Scale for College Students. Chinese mental
health journal. 2014;
27. Shuan S. Development of the Smartphone Addiction Scale for College Students. Chinese mental
health journal [Internet]. 2014 [cited 2024 Apr 29]; Available from:
https://www.semanticscholar.org/paper/Development-of-the-Smartphone-Addiction-Scale-for-
Shuan/e25c7a6ce49e96646ab94de5e1382bbe5e174474
2. Vella SLC, Pai NB. A Theoretical Review of Psychological Resilience: Dening Resilience and
Resilience Research over the Decades. Archives of Medicine and Health Sciences. 2019
Dec;7(2):233.
29. Shang R, Pang H, Jiang J, Ji Y, Liu Q, Zhang M, Yang R, Li S, Li Y, Liu Q. Internet addiction and
depressive and anxious symptoms among Chinese rural left-behind adolescents: Mediating roles of
resilience and friendship quality. Child Care Health Dev. 2024 Jan;50(1).
30. Fan X. Unpacking the Association between Family Functionality and Psychological Distress among
Chinese Left-Behind Children: The Mediating Role of Social Support and Internet Addiction. Int J
Environ Res Public Health. 2022 Oct;19(20):13327.
31. Mesman E, Vreeker A, Hillegers M. Resilience and mental health in children and adolescents: an
update of the recent literature and future directions. Curr Opin Psychiatr. 2021 Nov;34(6):586–92.
32. Kim, Eun Joo. Effects on mobile phone functional use of ego resilience, peer attachment and mobile
phone-related characteristics in male and female middle school students - focused on uses of SNS
& messenger, music and internet in era of convergence-. Journal of Digital Convergence. 2016 Aug
28;14(8):383–91.
33. Ma A, Yang Y, Guo S, Li X, Zhang S, Chang H. Adolescent resilience and mobile phone addiction in
Henan Province of China: Impacts of chain mediating, coping style. PLoS One. 2022 Dec
27;17(12):e0278182.
34. Zhang LQ, Gao HN. Effects of sports on school adaptability, resilience and cell phone addiction
tendency of high school students. World J Psychiatr. 2023 Aug 19;13(8):563–72.
35. Xie G, Wu Q, Guo X, Zhang J, Yin D. Psychological resilience buffers the association between cell
phone addiction and sleep quality among college students in Jiangsu Province, China. Front
Psychiatry. 2023 Feb 8;14:1105840.
3. Hao Z, Jin L, Huang J, Lyu R, Cui Q. Academic Burnout and Problematic Smartphone Use During the
COVID-19 Pandemic: The Effects of Anxiety and Resilience. Front Psychiatry. 2021 Oct
Page 24/36
20;12:725740.
37. Li S, Cui G, Yin Y, Tang K, Chen L, Liu X. Prospective Association Between Problematic Mobile Phone
Use and Eating Disorder Symptoms and the Mediating Effect of Resilience in Chinese College
Students: A 1-Year Longitudinal Study. Front Public Health. 2022 Apr 27;10:857246.
3. Choliz M. Mobile-phone addiction in adolescence: The Test of Mobile Phone Dependence (TMD).
Prog Health Sci. 2012 Jan 1;2:33–44.
39. Connor KM, Davidson JRT. Development of a new resilience scale: the Connor-Davidson Resilience
Scale (CD-RISC). Depress Anxiety. 2003;18(2):76–82.
40. Luthar SS, Cicchetti D, Becker B. The construct of resilience: a critical evaluation and guidelines for
future work. Child Dev. 2000;71(3):543–62.
41. Qiu C, Qi Y, Yin Y. Multiple Intermediary Model Test of Adolescent Physical Exercise and Internet
Addiction. Int J Environ Res Public Health. 2023;20(5).
42. Carbonell X, Chamarro A, Oberst U, Rodrigo B, Prades M. Problematic Use of the Internet and
Smartphones in University Students: 2006-2017. Int J Environ Res Public Health. 2018
Mar;15(3):475.
43. Lissak G. Adverse physiological and psychological effects of screen time on children and
adolescents: Literature review and case study. Environ Res. 2018 Jul;164:149–57.
44. Mak KK, Jeong J, Lee HK, Lee K. Mediating Effect of Internet Addiction on the Association between
Resilience and Depression among Korean University Students: A Structural Equation Modeling
Approach. Psychiatry Investig. 2018 Oct;15(10):962–9.
45. Hu B, Wu Q, Xie Y, Guo L, Yin D. Cell phone addiction and sleep disturbance among medical students
in Jiangsu Province, China: the mediating role of psychological resilience and the moderating role of
gender. Front Psychiatry. 2024 May 15;15:1405139.
4. Ma A, Yang Y, Guo S, Li X, Zhang S, Chang H. The Impact of Adolescent Resilience on Mobile Phone
Addiction During COVID-19 Normalization and Flooding in China: A Chain Mediating. Front Psychol.
2022;13:865306.
47. Bronfenbrenner U. The Ecology of Human Development: Experiments by Nature and Design
[Internet]. Harvard University Press; 1979 [cited 2024 Jul 1]. Available from:
https://www.jstor.org/stable/j.ctv26071r6
4. Paikoff RL, Brooks-Gunn J. Do parent-child relationships change during puberty? Psychol Bull. 1991
Jul;110(1):47–66.
49. Niu G, Yao L, Wu L, Tian Y, Xu L, Sun X. Parental phubbing and adolescent problematic mobile phone
use: The role of parent-child relationship and self-control. Children and Youth Services Review. 2020
Sep 1;116:105247.
50. ZHANG Y. A Review of Studies on the Inuence of Family Environment on Adolescent Cell Phone
Dependence. Advances in Social Sciences. 2023 Jan 1;12:1305–9.
Page 25/36
51. Gao Q, Sun R, Fu E, Jia G, Xiang Y. Parent-child relationship and smartphone use disorder among
Chinese adolescents: The mediating role of quality of life and the moderating role of educational
level. Addict Behav. 2020 Feb;101:106065.
52. Qu Y, Li X, Ni B, He X, Zhang K, Wu G. Identifying the role of parent-child conict and intimacy in
Chinese adolescents’ psychological distress during school reopening in COVID-19 pandemic. Dev
Psychol. 2021 Oct;57(10):1735–47.
53. Li C, Jiang S, Fan X, Zhang Q. Exploring the impact of marital relationship on the mental health of
children: Does parent-child relationship matter? J Health Psychol. 2020;25(10–11):1669–80.
54. Oh YH. Parent-Child Conict, Forgiveness, and Mental Health of College Students. .
2004;18(3):59–77.
55. Bouteyre E, Duval P, Pietri M. Children’s Physical Proximity to Interparental Conict: Resilient Process
and Retrospective Perceptions of Parent-Child Relationships. Violence Against Women. 2024;30(3–
4):854–72.
5. Tian L, Liu L, Shan N. Parent-Child Relationships and Resilience Among Chinese Adolescents: The
Mediating Role of Self-Esteem. Front Psychol. 2018;9:1030.
57. Belsky J, Pluess M. Beyond diathesis stress: differential susceptibility to environmental inuences.
Psychol Bull. 2009 Nov;135(6):885–908.
5. Hair JF, Hult GTM, Ringle CM, Sarstedt M, Danks NP, Ray S. An Introduction to Structural Equation
Modeling. In: Hair Jr. JF, Hult GTM, Ringle CM, Sarstedt M, Danks NP, Ray S, editors. Partial Least
Squares Structural Equation Modeling (PLS-SEM) Using R: A Workbook [Internet]. Cham: Springer
International Publishing; 2021 [cited 2024 Aug 1]. p. 1–29. Available from:
https://doi.org/10.1007/978-3-030-80519-7_1
59. Rey L, Pena M, Neto F. Editorial: Protective Resources for Psychological Well-Being of Adolescents.
Front Psychol. 2020;11:720.
0. Liu Y, Ge T, Jiang Q. Changing family relationships and mental health of Chinese adolescents: the
role of living arrangements. Public Health. 2020 Sep;186:110–5.
1. Fergus S, Zimmerman MA. Adolescent resilience: a framework for understanding healthy
development in the face of risk. Annu Rev Public Health. 2005;26:399–419.
2. Cai Y, Dong S, Yuan S, Hu CP. Network analysis and its applications in psychology. APS2. 2022 Jul
13;28(1):178–90.
3. Borsboom D. Psychometric perspectives on diagnostic systems. Journal of Clinical Psychology.
2008;64(9):1089–108.
4. Borsboom D. A network theory of mental disorders. World Psychiatry. 2017 Feb;16(1):5–13.
5. Cramer AOJ, Waldorp LJ, van der Maas HLJ, Borsboom D. Comorbidity: a network perspective.
Behav Brain Sci. 2010 Jun;33(2–3):137–50; discussion 150-193.
. Derogatis LR. BSI 18, Brief Symptom Inventory 18: Administration, scoring and procedures manual.
NCS Pearson, Incorporated; 2001.
Page 26/36
7. Leung L. Leisure boredom, sensation seeking, self-esteem, and addiction. Mediated Interpersonal
Communication. 2008;359.
. Yue-Qin H, Yi-Qun G. Development and Psychometric Validity of the Resilience Scale for Chinese
Adolescents. Acta Psychologica Sinica. 2008 Aug 30;40(08):902.
9. Nelissen S. The Child Effect in Media Use: Investigating Family Dynamics Concerning Media
Behavior in Parent-Child Dyads. 2018;
70. Epskamp S, Borsboom D, Fried EI. Estimating psychological networks and their accuracy: A tutorial
paper. Behav Res. 2018 Feb 1;50(1):195–212.
71. Kelley TL. The selection of upper and lower groups for the validation of test items. Journal of
Educational Psychology. 1939;30(1):17–24.
72. van Borkulo C, van Bork R, Boschloo L, Kossakowski J, Tio P, Schoevers R, Borsboom D, Waldorp L.
Comparing Network Structures on Three Aspects: A Permutation Test. Psychological Methods. 2022
Apr 11;28.
73. Epskamp S, Cramer AOJ, Waldorp LJ, Schmittmann VD, Borsboom D. qgraph: Network
Visualizations of Relationships in Psychometric Data. Journal of Statistical Software. 2012 May
24;48:1–18.
74. Tibshirani R. Regression Shrinkage and Selection Via the Lasso. Journal of the Royal Statistical
Society: Series B (Methodological). 1996 Jan;58(1):267–88.
75. Chen J, Chen Z. Extended Bayesian information criteria for model selection with large model
spaces. Biometrika. 2008 Sep 1;95(3):759–71.
7. Haslbeck JMB, Waldorp LJ. mgm: Estimating Time-Varying Mixed Graphical Models in High-
Dimensional Data. Journal of Statistical Software. 2020 Apr 27;93:1–46.
77. Liu X, Wang H, Zhu Z, Zhang L, Cao J, Zhang L, Yang H, Wen H, Hu Y, Chen C, Lu H. Exploring bridge
symptoms in HIV-positive people with comorbid depressive and anxiety disorders. BMC Psychiatry.
2022 Jul 5;22(1):448.
7. Katzgraber HG. Spin glasses and algorithm benchmarks: A one-dimensional view. arXiv.org. 2007
Nov 9;
79. Fruchterman TMJ, Reingold EM. Graph drawing by force-directed placement. Software: Practice and
Experience. 1991;21(11):1129–64.
0. Jones P. networktools: Tools for Identifying Important Nodes in Networks [Internet]. 2024. Available
from: https://cran.r-project.org/web/packages/networktools/index.html
1. Hallquist MN, Wright AGC, Molenaar PCM. Problems with Centrality Measures in Psychopathology
Symptom Networks: Why Network Psychometrics Cannot Escape Psychometric Theory. Multivariate
Behav Res. 2021;56(2):199–223.
2. Rodrigues FA. Network Centrality: An Introduction. In: Macau EEN, editor. A Mathematical Modeling
Approach from Nonlinear Dynamics to Complex Systems [Internet]. Cham: Springer International
Page 27/36
Publishing; 2019 [cited 2024 Aug 24]. p. 177–96. Available from: https://doi.org/10.1007/978-3-319-
78512-7_10
3. Jones PJ, Ma R, McNally RJ. Bridge Centrality: A Network Approach to Understanding Comorbidity.
Multivariate Behavioral Research. 2021 Mar 4;56(2):353–67.
4. Rodrigues FA. Network centrality: an introduction [Internet]. arXiv.org. 2019 [cited 2024 Mar 26].
Available from: https://arxiv.org/abs/1901.07901v1
5. Podsakoff PM, Organ DW. Self-Reports in Organizational Research: Problems and Prospects.
Journal of Management. 1986 Dec 1;12(4):531–44.
. Borsboom D. A network theory of mental disorders. World Psychiatry. 2017 Jan 26;
7. Shen X, Zhou X, Liao HP, Mcdonnell D, Wang JL. Uncovering the symptom relationship between
anxiety, depression, and internet addiction among left-behind children: A large-scale purposive
sampling network analysis. J Psychiatr Res. 2024 Mar;171:43–51.
. Huang S, Lai X, Xue Y, Zhang C, Wang Y. A network analysis of problematic smartphone use
symptoms in a student sample. J Behav Addict. 2020 Dec;9(4):1032–43.
9. Brand M, Young KS, Laier C, Wöling K, Potenza MN. Integrating psychological and neurobiological
considerations regarding the development and maintenance of specic Internet-use disorders: An
Interaction of Person-Affect-Cognition-Execution (I-PACE) model. Neuroscience & Biobehavioral
Reviews. 2016 Dec 1;71:252–66.
90. O’Neill A, Bachi B, Bhattacharyya S. Attentional bias towards cannabis cues in cannabis users: A
systematic review and meta-analysis. Drug Alcohol Depend. 2020 Jan 1;206:107719.
91. Verduyn P, Schulte-Strathaus JCC, Kross E, Hülsheger UR. When do smartphones displace face-to-
face interactions and what to do about it? Computers in Human Behavior. 2021 Jan 1;114:106550.
92. Yong-zh J. College Students Rely on Mobile Internet Making Impact on Alienation:the Role of Society
Supporting Systems. Psychological development and education [Internet]. 2014 [cited 2024 Jul 15];
Available from: https://www.semanticscholar.org
93. Corominas M, Roncero C, Bruguera E, Casas M. The dopaminergic system and addictions. Rev
Neurologia. 2007 Jan 1;44(1):23–31.
94. Ebert D, Lammers CH. Das zentrale dopaminerge System und die Depression. Nervenarzt. 1997
Jul;68(7):545–55.
95. Ding X, Jin X, Tang YY, Yang Z. Associations between mobile phone addiction and depressive
symptoms in college students: A conditional process model. Ann Med-Psychol. 2024
Mar;182(3):258–65.
9. Erkolahti R, Sandberg S, Ebeling H. Somatisointi ja somatoformiset hairiot lapsilla ja nuorilla.
Duodecim. 2011;127(18):1904–10.
97. Kang KI, Kang CM. Factors Inuencing Adolescent Generalized Anxiety Disorder A Zero-Inated
Negative Binomial Regression Model. J Psychosoc Nurs Ment Health Serv. 2024 Jun;62(6):46–55.
Page 28/36
9. Wang F, Ma X, Zhao L, Li T, Fu Y, Zhu W. The Inuence of Genetic and Environmental Factors on
Anxiety among Chinese Adolescents: A Twin Study. J Genet Psychol. 2024 Feb 17;
99. Xiao L, Yao M, Liu H. Perceived Social Mobility and Smartphone Dependence in University Students:
The Roles of Hope and Family Socioeconomic Status. Psychol Res Behav Manag. 2024;17:1805–17.
100. Liu W, Wang Y, Xia L, Wang W, Li Y, Liang Y. Left-Behind Children’s Positive and Negative Social
Adjustment: A qualitative Study in China. Behav Sci (Basel). 2023 Apr;13(4).
101. Tang Q, Zou X, Gui J, Wang S, Liu X, Liu G, Tao Y. Effects of childhood trauma on the symptom-level
relation between depression, anxiety, stress, and problematic smartphone use: A network analysis. J
Affect Disord. 2024 Aug 1;358:1–11.
102. Tullett-Prado D, Doley JRR, Zarate D, Gomez R, Stavropoulos V. Conceptualising social media
addiction: a longitudinal network analysis of social media addiction symptoms and their
relationships with psychological distress in a community sample of adults. BMC Psychiatry. 2023
Jul 13;23(1):509.
103. Beesdo K, Knappe S, Pine DS. Anxiety and anxiety disorders in children and adolescents:
developmental issues and implications for DSM-V. Psychiatr Clin North Am. 2009 Sep;32(3):483–
524.
104. Billieux J, Maurage P, Lopez-Fernandez O, Kuss DJ, Griths MD. Can Disordered Mobile Phone Use
Be Considered a Behavioral Addiction? An Update on Current Evidence and a Comprehensive Model
for Future Research. Curr Addict Rep. 2015 Jun 1;2(2):156–62.
105. Zakay D, Block RA. Temporal Cognition. Curr Dir Psychol Sci. 1997 Feb 1;6(1):12–6.
10. LIU J, LI H. How state anxiety inuences time perception: Moderated mediating effect of cognitive
appraisal and attentional bias. Acta Psychologica Sinica. 2019;51(7):747–58.
107. Van Bockstaele B, Verschuere B, Tibboel H, De Houwer J, Crombez G, Koster EHW. A review of
current evidence for the causal impact of attentional bias on fear and anxiety. Psychol Bull. 2014
May;140(3):682–721.
10. Muris P, Schmidt H, Merckelbach H, Schouten E. Anxiety sensitivity in adolescents: factor structure
and relationships to trait anxiety and symptoms of anxiety disorders and depression. Behav Res
Ther. 2001 Jan;39(1):89–100.
109. Hurrell KE, Hudson JL, Schniering CA. Parental reactions to children’s negative emotions:
relationships with emotion regulation in children with an anxiety disorder. J Anxiety Disord. 2015
Jan;29:72–82.
110. Geng C. The Relationship between Self-Construction and Social Support and Anxiety of College
Students. Advances in Psychology. 2020 Jan 1;10:1647–55.
111. Halls G, Cooper PJ, Creswell C. Social communication decits: Specic associations with Social
Anxiety Disorder. J Affect Disord. 2015 Feb 1;172:38–42.
112. Mandleco BL. An Organizational Framework for Conceptualizing Resilience in Children. Journal of
Child and Adolescent Psychiatric Nursing. 2000;13(3):99–112.
Page 29/36
113. An J, Zhu X, Shi Z, An J. A serial mediating effect of perceived family support on psychological well-
being. BMC Public Health. 2024 Apr 2;24(1):940.
114. Yang C, Gao H, Li Y, Wang E, Wang N, Wang Q. Analyzing the role of family support, coping strategies
and social support in improving the mental health of students: Evidence from post COVID-19. Front
Psychol. 2022 Dec 23;13:1064898.
115. Chang B, Wei Y, Fang J. Lack of parental care increases depression of rural left-behind children in
China: a moderated mediating effects*. Curr Psychol. 2024 Jul 1;43(25):21830–9.
11. FAN X, FANG X, HUANG Y, CHEN F, YU S. The inuence mechanism of parental care on depression
among left-behind rural children in China: A longitudinal study. Acta Psychologica Sinica.
50(9):1029–40.
117. Wu X, Tang L, Gong J. Correlation analysis of mental toughness, family social support, and anxiety
of nursing staff. Am J Transl Res. 2024;16(6):2563–70.
11. Sun R, Gao Q, Xiang Y, Chen T, Liu T, Chen Q. Parent–child relationships and mobile phone addiction
tendency among Chinese adolescents: The mediating role of psychological needs satisfaction and
the moderating role of peer relationships. Children and Youth Services Review. 2020 Sep
1;116:105113.
119. Wang CD, Hayslip B, Sun Q, Zhu W. Grandparents as the Primary Care Providers for Their
Grandchildren: A Cross-Cultural Comparison of Chinese and U.S. Samples. Int J Aging Hum Dev.
2019 Dec 1;89(4):331–55.
120. Moyer DN, Sandoz EK. The Role of Psychological Flexibility in the Relationship Between Parent and
Adolescent Distress. J Child Fam Stud. 2015 May 1;24(5):1406–18.
121. Bursch B, Lester P, Jiang L, Rotheram-Borus MJ, Weiss R. Psychosocial predictors of somatic
symptoms in adolescents of parents with HIV: a six-year longitudinal study. AIDS Care. 2008
Jul;20(6):667–76.
122. Anniko MK, Boersma K, Tillfors M. Sources of stress and worry in the development of stress-related
mental health problems: A longitudinal investigation from early- to mid-adolescence. Anxiety, Stress,
& Coping. 2019 Mar 4;32(2):155–67.
123. Slavich G. Psychoneuroimmunology of Stress and Mental Health. 2018 Jun 7;
124. von Klitzing K, White LO, Otto Y, Fuchs S, Egger HL, Klein AM. Depressive comorbidity in preschool
anxiety disorder. J Child Psychol Psychiatry. 2014 Oct;55(10):1107–16.
125. Ayoub MA. Ergonomic deciencies: I. Pain at work. J Occup Med. 1990 Jan;32(1):52–7.
12. Dotterer AM, Hoffman L, Crouter AC, McHale SM. A Longitudinal Examination of the Bi-Directional
Links between Academic Achievement and Parent-Adolescent Conict. J Fam Issues. 2008 Jun
1;29(6):762–79.
127. Steeger CM, Gondoli DM. Mother–adolescent conict as a mediator between adolescent problem
behaviors and maternal psychological control. Developmental Psychology. 2013;49(4):804–14.
Page 30/36
12. Scharf M, Goldner L. “If you really love me, you will do/be…”: Parental psychological control and its
implications for children’s adjustment. Developmental Review. 2018 Sep 1;49:16–30.
129. Guan J, Ma W, Liu C. Fear of missing out and problematic smartphone use among Chinese college
students: The roles of positive and negative metacognitions about smartphone use and optimism.
PLoS One. 2023 Nov 28;18(11):e0294505.
130. Gao W, Chen Z. A Study on Psychopathology and Psychotherapy of Internet Addiction. Advances in
Psychological Science. 2006 Jul 15;14(4):596.
131. Camisasca E, Miragoli S, Di Blasio P, Grych J. Children’s Coping Strategies to Inter-Parental Conict:
The Moderating Role of Attachment. J Child Fam Stud. 2017 Apr 1;26(4):1099–111.
132. Andrews S, Ellis DA, Shaw H, Piwek L. Beyond Self-Report: Tools to Compare Estimated and Real-
World Smartphone Use. PLOS ONE. 2015 Oct 28;10(10):e0139004.
Figures
Figure 1
Network structure diagram (a) and cluster diagram (b) of mobile phone addiction, mental health and
resilience among left-behind adolescents
Note: Nodes are psychological variable dimensions, and connecting lines are partial correlations
between node dimensions. The thicker the line, the higher the correlation, and the colour of the line
indicates the direction of the correlation (green for positive correlation, red for negative).
Page 31/36
Figure 2
Centrality index of each node of the network
Note: BSI1, somatization; BSI2, depression; BSI3, anxiety; MPAI1, inability to control carving; MPAI2,
feeling anxious and lost; MPAI3, withdrawal or escape; MPAI4, productivity losse; RSCA1, goal planning;
RSCA2, affect control; RSCA3, positive thinking; RSCA4, family support; RSCA5, help-seeking.
Page 32/36
Figure 3
Bootstrap condence intervals for edge weights in network
Note: Red dots indicate sample values, black dots indicate values for each edge weight, and grey areas
indicate 95 % condence intervals.
Page 33/36
Figure 4
Stability test diagram for the case-dropping subset
Note: Lines represent the average relationship between the original sample centrality and the
subsamples. Regional color blocks indicate the range between the rst quartile and the third quartile.
Page 34/36
Figure 5
Bootstrapped difference tests result of the network strength (a), bridge strength (b) and edges (c)
Note: Black boxes indicate signicant differences between two nodes (α= 0.05). BSI1, somatization;
BSI2, depression; BSI3, anxiety; MPAI1, inability to control carving; MPAI2, feeling anxious and lost;
Page 35/36
MPAI3, withdrawal or escape; MPAI4, productivity losse; RSCA1, goal planning; RSCA2, affect control;
RSCA3, positive thinking; RSCA4, family support; RSCA5, help-seeking.
Figure 6
Comparison of network structure between high- and low-level groups of parent-child conict
Note: Nodes are dimensions of psychological variables, and lines are partial correlations between the
dimensions of nodes. Thicker lines indicate higher correlation, and line colors indicate the direction of
correlation (green for positive correlation, red for negative correlation).
Page 36/36
Figure 7
Comparison of network centrality indicators between high- and low-level groups of parent-child conict
Note: The centrality graph depicts the strength centrality (z-score) and the bridge strength centrality (z-
score) of each node in the network, with higher scores representing greater inuence of nodes in the
network. BSI1, somatization; BSI2, depression; BSI3, anxiety; MPAI1, inability to control carving; MPAI2,
feeling anxious and lost; MPAI3, withdrawal or escape; MPAI4, productivity losse; RSCA1, goal planning;
RSCA2, affect control; RSCA3, positive thinking; RSCA4, family support; RSCA5, help-seeking.
