The Relationship Between Social Support And Social Withdrawal Of Left-Behind Adolescents: A Chain Mediating Model


 Background: Nowadays, as more and more Chinese farmers in rural area went to city for work, they left their kids at home. These kids were left-behind adolescents and they developed without their parental accompany. The family function of left-behind adolescents was deficient, which may result in their social withdrawal in social situations. Therefore, in this study, in order to improve left-behind adolescents’ psychological and behavior problems, we aimed to investigate their level of social withdrawal and its impact factors. Method: There were 339 left-behind adolescents and 289 non-left-behind adolescents recruited from a Chinese junior high school. Their social withdrawal, social support, relative deprivation, and resilience were measured through questionnaires. Results: The results showed that compared with non-left-behind adolescents, left-behind adolescents had lower social support and resilience, but their social withdrawal and relative deprivation were higher; besides, left-behind adolescents’ social support negatively predicted social withdrawal, while relative deprivation and resilience played a chain mediating role between them. Conclusion: This study found that compared with none-left-behind adolescents, left-behind adolescents had difficulty in social adaptation. However, there was a “context-process-outcome” model in which social support negatively predicted social withdrawal, while relative deprivation and resilience played a chain mediating role between them. In sum, this study provided suggestions to promote the mental health and social behavioral development of left-behind adolescents.


Introduction
In recent years, as more and more farmers found jobs in city and left their kids at home, left-behind adolescents have remarkably increased in China. Meanwhile, their psychological and behavioral problems are serious and need much attention. Left-behind adolescents refer to adolescents under the age of 18 who are left at home because of their parents or single parents working outside for a long time (Zhao et al., 2008). Adolescence is the key period of their academic and physical and mental development (Lopez-Zafra et al., 2019). However, compared with non-left-behind adolescents, left-behind adolescents lack of parental accompany and positive responses, thus it's hard to cultivate good parentchild relationship and form secure attachment (Ainsworth et al. 1978). Then, this insecure attachment may hinder adolescents' learning of social ability, resulting in poor psychology and social adaptation.
Besides, they are sensitive to others' words and behavior, and they may be also more likely to be rejected by peers, thus leading to external behavioral problems or internal psychological and emotional problems (Shamir-Essakow et al., 2005;van Brakel et al., 2006). Specially, social adaptation and communication problems of left-behind adolescents are particularly prominent, and it's vital to pay more attention to their mental health and alleviation of social anxiety and social avoidance (Li & Cai, 2012).

Social withdrawal
Social withdrawal is a comprehensive term, which refers to the individual who is willing to isolate himself/herself to familiar or unfamiliar others and consistently perform some solitary behavior (Rubin et al. 2009), such as shyness, long periods of solitude, avoidance of social contact. The model of developmental of social withdrawal believes that safe parent-child relationship is a crucial basis for development of children's social ability, which supports them to freely explore society and maintain good interpersonal relationships; besides, when parents response timely to children's needs, children will attain the feeling of control over the environment and self-e cacy in social skill learning (Ainsworth et al., 1978;Bowlby, 1969). However, for left-behind adolescents, insecure attachment may bring negative experience of interpersonal interaction (Ainsworth et al. 1978). Their limited social resources make it di cult to learn social ability e ciently, therefore resulting in social withdrawal (Hastings et al., 2019). Speci cally, adolescents who behave social withdrawal in social occasions are easy to be regarded as the outlier by their peers, then they may be ignored and rejected by them (Chen et al., 2006;Olweus & Breivik, 2014), even bullied by aggressive peers (Rubin et al., 2006). Thus, they will grow feelings of loneliness or depression (Coplan et al., 2007;Prior et al., 2000). Due to the prominent social withdrawal of left-behind adolescents, which has a negative impact on the healthy development of them, it is of great importance to explore left-behind adolescents' in uential mechanism of social withdrawal in depth, so as to provide some intervention measures to improve the current situation.

Social support and social withdrawal
The model of developmental of social withdrawal believes that insecure parent-child relationship and negative interpersonal interaction experience are the important factors to cause individual social withdrawal (Ainsworth et al., 1978;Bowlby, 1969). Therefore, it's particularly important to support leftbehind adolescents and provide them interpersonal resources. As one of the interpersonal resources, social support means the support and concern given by others (Brugha, 1990 More speci cally, the positive effect of social support can work in two aspects. On the one hand, socialcognitive process theory holds that positive social interaction can promote individual cognitive adaptation (Tedeschi & Calhoun, 2004;Kong et al., 2018). It also holds that social support is a kind of supportive social environment which provide individual potential social resources Tedeschi & Calhoun, 2004;Xu & Yuan, 2014), and prevent negative environment attribution and emotion caused by the lacking of environment resources (Lazarus & Folkman, 1984). Thus, individual could deal with various problems effectively and enhance the skills of coping setback (Yuan et al., 2018). On the other hand, social support is a feeling of being concerned, understood, accepted, respected and cherished (Kane et al., 2012;Reis & Shaver, 1988). While a positive interpersonal experience for individuals can improve their self-evaluation and control over environment Feeney & Collins, 2015 , develop social ability, promote social approach motivation, and actively construct new interpersonal relationships (Reis et al., 2010). Many studies showed that instrumental support can negatively predict social withdrawal (Nonaka & Sakai, 2021), and social support can relieve negative mental states or behavioral reactions, such as depression, sleep disorders and social withdrawal. (Ginter et al., 1994;  Due to the lack of family support, left-behind adolescents have less psychological and material resources than their peers, and they are likely to think that they do not gain what they deserve, which results in a sense of relative deprivation. Relative deprivation means that compared to the referenced target, the individual or the ingroup is in a relatively disadvantaged position, and this kind of judgement invokes the feelings of anger, resentment and entitlement (Smith & Pettigrew, 2015). Relative deprivation theory believes the entitlement and feeling of deserving are the core of relative deprivation. In the process of social comparison, individuals realize that there is a gap between their expectation value and reality; when they experience that their basic rights are deprived, they will have a strong sense of injustice . This further leads to depression, loneliness, social anxiety or other psychological problems Nevertheless, social-cognitive process theory suggests that social support can accelerate individual cognitive adaptation process (Tedeschi & Calhoun, 2004;Kong et al., 2018). When the support system from left-behind adolescents' families is lack, individual can receive enough social support from others, thus reducing the sense of injustice and relative deprivation. Relative research also veri ed that the higher social support the college students had, the weaker the relative deprivation they would feel (Zhang & Tao, 2013). Further, individual would feel less hostility and fear to environment or others with reduction of relative deprivation, and then the social avoidance motivation and social withdrawal behavior would also decrease (Devos et al., 2003;Greitemeyer & Sagioglou, 2019;Osborne et al., 2012). Therefore, this study hypothesized that relative deprivation played a mediating role between social support and social withdrawal.
1.4 The mediating role of resilience between social support and social withdrawal Resilience refers to the factor that promotes individual's effective coping and adaptation under the pressure, frustration or adversity (Masten et al., 2011). It is an important ability to recover from adversity and maintain good mental functioning under pressure (Laird et al., 2019;Theron & Theron, 2010). Besides, it can effectively predict the social behavior of disadvantaged children (Rahat & Ilhan, 2016), and is also a protective factor for their behavior problems (Breda, 2017;Theron & Theron, 2010). More importantly, it can decrease individual social withdrawal . As an important psychological protective factor, the development of resilience is closely related to the growing environment of children, in which a good family circumstance helps its development (Lopez & Snyder, 2009). In the process of interaction between children and their parents, they will form an overall understanding of themselves and make judgments about their own abilities and values, which will also . According to "contextprocess-outcome" model proposed by Roeser (1996), individual characteristics can play a mediate role between school environment factors and their behavior outcomes. Therefore, some researchers have found that resilience, an individual characteristic, can mediate the relationship between teacher-student relationship and related behavioral outcomes of social withdrawal, while teacher-student relationship is regarded as a factor of school environment . Therefore, this study also assumed that resilience could mediate the relationship between social support and social withdrawal.
In addition, based on the triadic theory of learning and the "context-process-outcome" model, individual cognition, behavior and environmental factors interact with each other (Bandura, 1978;Roeser et al., 1996). . Although there has no study directly investigating the relationship between relative deprivation and resilience, some researchers believed that people with high relative deprivation would experience the unfair environment, be disappointed with the current environment, and believe future is hard to improve (de la Sablonnie`re et al., 2015). This may also affect the resilience of individuals facing pressure and di culties which as a function of psychological adaptation, and then change the behavior, namely social withdrawal. Therefore, this study also hypothesized that relative deprivation and resilience played a chain mediating role between social support and social withdrawal.
The present research aimed to explore left-behind adolescents' and non-left-behind adolescents' states of social support, social withdrawal, relative deprivation and resilience, and explore the left-behind adolescents' in uential mechanism of social withdrawal. Based on the triadic theory of learning and the "context-process-outcome" model, we proposed a chain mediation model (see in Fig. 1). The social support system of left-behind adolescents which was regarded as environmental background would in uence their interpretation of the environment. We assumed that there was a signi cant negative correlation between social support and social withdrawal. In addition, relative deprivation and resilience mediated the relationship between social support and social withdrawal.

Subjects
A total of 675 students from the rst and second grade of a junior high school in Jiangsu province were recruited to conduct a questionnaire, and 628 valid questionnaires were collected with a questionnaire recovery rate of 95.41%. Among them, 276 were male students (41.90%) and 368 were female students. There were 339 left-behind adolescents (54%), 289 non-left-behind adolescents. There were 312 (49.70%) in their rst year and 316 in their second year of junior high school.

Social support
The adapted social support rating Scale by Xiong and Ye (2013) was adopted to measure social support of left-behind adolescents. The scale is based on the Social support rating scale developed by Xiao (1999). There are 11 items in this scale, including objective support (3 items), subjective support (5 items) and utilization degree of social support (3 items). The Cronbach's α of the scale was 0.69, which was within the acceptable range.

Social withdrawal
The social withdrawal scale developed by Tian (2014) was adopted. There were 16 items in the scale, including three dimensions of avoiding unfamiliar environment, outlier and avoiding speaking in public. This scale used likert-5-point scoring. In this study, the Cronbach's α was 0.85.

Relative deprivation
The relative deprivation scale compiled by Ma (2012) contained four questions, and the likert-6 point scoring method was adopted. The higher the score, the higher the individual's relative deprivation had. Cronbach's α in this study was 0.63, which was acceptable.

Resilience
The adolescent resilience scale was developed by Hu and Gan (2008). There are 27 items in the scale. Among these, there were 12 reverse items, including two dimensions of personal power and support power. Personal power includes three factors: goal focus (5 items), positive cognition (4 items) and emotional control (6 items). And support power dimension includes family support (6 items) and interpersonal assistance (6 items). Likert-5 point scoring was used in the scale, and Cronbach's α of the scale in this study was 0.86.

Common-method variance
we assessed results for common-method variance because of all variables measured via questionnaire.
According to the Harman single factor test (Zhou & Long, 2004), fourteen factors were extracted, and variance contribution rates were 58.23%. The variance contribution rate of the rst factor was 16.57%, <40%, so common variance caused by this questionnaire could be eliminated.
3.2 Differences in social support, social withdrawal, relative deprivation and resilience between leftbehind adolescents and non-left-behind adolescents An independent sample t-test was conducted on the social support, social withdrawal, relative deprivation and resilience of left-behind adolescents and non-left-behind adolescents, as shown in Table 1. The results showed that there were signi cant differences in those ve variables between left behind leftbehind adolescents and non-left-behind adolescents. In short, compared with non-left-behind adolescents, left-behind adolescents had a lower level of social support and resilience, while they had a higher level of social withdrawal and relative deprivation.

correlation analysis among variables
The correlation of social support, relative deprivation, resilience and social withdrawal of left-behind adolescents are calculated. The results showed that there were signi cant correlations between all variables (See in Table 2). Among them, there was a signi cant positive correlation relationship between social support and resilience, a signi cant negative correlation relationship between social support and relative deprivation, social withdrawal. There was a positive correlation relationship between relative deprivation and social withdrawal, and a signi cant negative correlation relationship between resilience and relative deprivation, social withdrawal. Note: * p <. 05, ** p <. 01, *** p <. 001, the same below.

The chain mediated effect test of relative deprivation and psychological elasticity between social support and social withdrawal
The PROCESS 3.3 provided by Hayes (2013) was used for analysis. Model 6 was selected, with grade and gender of adolescents as covariant variables, social support as the independent variable, social withdrawal as the dependent variable, and relative deprivation and resilience as the chain intermediary variables. The chain mediating effect of the region was tested using the bootstrap method, repeating sampling 5000 times, and setting a 95% con dence interval was set. The results of regression were shown in Table 3, the path coe cients were shown in Fig. 2, and Table 4

Discussion
This study explored the left-behind adolescents and non-left-behind adolescents' social support, relative deprivation, resilience and social withdrawal, and clari ed the left-behind adolescents' in uential mechanism of social withdrawal of, thus giving theoretical and practical support for reducing the psychological and behavioral problems of left-behind adolescents.

Psychological state of left-behind adolescents and non-left-behind adolescents
This study showed that compared to non-left-behind adolescents, left-behind adolescents had lower levels of social support and resilience, and higher levels of social withdrawal and relative deprivation. Speci cally, the family support of left-behind adolescents was relatively weak compared to non-leftbehind adolescents, and the lack of psychological and social resources given by parents in the family, as part of the social support system, leaded to less social support, which was similar to the result of past study (Liu et al., 2007), and the lack of parent-child relationship affected the establishment of their social support system. At the same time, compared to their peers, left-behind adolescents would realize that they were in a relatively disadvantaged position and that they did not receive social resources comparable to their peers, thus they would develop a stronger sense of relative deprivation (Smith & Pettigrew, 2015). Past researches on relative deprivation also found that children from poor single-parent families had signi cantly higher levels of relative deprivation than children from non-poor single-parent families (Callan et al., 2017), suggesting that sound family functioning was important for children's social cognitive development. Furthermore, since the development of resilience was inextricably linked to the child's upbringing (Lopez & Snyder, 2009), de cit in family functioning of left-behind adolescents also leads to a hindrance in their development of resilience, and thus (they)/ left-behind adolescents had a lower degree of resilience compared to non-left-behind adolescents. Finally, for adolescents who lacked secure attachment, their parents' inability to respond sensitively to their thoughts and behaviors leaded to deviations in their beliefs, attitudes, and values (Rubin, 1993), and they become fearful of new environments, resulting in inhibitory behaviors. In addition, due to the lack of secure parent-child attachment, it was di cult for them to develop good social ability and actively construct relationships with others, and negative interpersonal experiences in the family also leaded them to develop lower selfe cacy, which in turn resulted in social withdrawal (Ainsworth et al., 1978;Bowlby, 1969). This demonstrated the important role of good parent-child relationships in the development of children's social ability and social adjustment.

Effect of social support on social withdrawal
This study found that social support of left-behind adolescents negatively predicted social withdrawal, indicating that the higher the level of social support, the lower the level of social withdrawal. Therefore, we should pay more attention to compensate for the absence of their parents by making full use of other social relationships in their social support system, thus providing them with a safe environment for social ability development.

The chain mediating effect of relative deprivation and resilience between social support and social withdrawal
Results of this study showed that social support directly predicted social withdrawal, and in addition, relative deprivation and resilience could play a serial mediating role between social support and social withdrawal. Among them, social support had a negative correlation with relative deprivation, relative deprivation had a negative correlation with resilience, and resilience also had a negative correlation with social withdrawal.
Based on the triadic theory of learning and the "context-process-outcome" model (Bandura, 1978;Roeser, 1996), social support is the environment/context for left-behind adolescents and affects their cognitive evaluation of their environment and themselves, which in turn affects their psychological adaptive functioning and reduces social withdrawal behavior. Speci cally, social support can, to some extent, compensate for their missing family support system, bring positive interpersonal experiences to individuals, and enhance their self-evaluation and control over their environment (Feeney & Collins, 2015).
According to social-cognitive process theory, positive social interactions promote individuals' cognitive adaptation, and a supportive social environment also provides social resources to left-behind adolescents to prevent them from developing unfair environmental perceptions due to a shortage of family resources ). In addition, support from family plays an important role in adolescents' development, and even for left-behind adolescents, who have less direct parental presence, they still need the care and support of their families. Therefore, the families of left-behind adolescents also need to give more understanding and care, pay attention to the psychological changes of children, and provide timely responses.
Second, this study found that relative deprivation had a negative effect on social support. Relative deprivation is a cognitive evaluation that accompanies the shortage of social and psychological resources for left-behind adolescents and needs to be given more attention and more resources and care from the social level. In addition, since the core of relative deprivation is social comparison and the perception of unfair environment (Smith & Pettigrew, 2015), interventions on adolescents' value system and self-perception are also needed. As the standard of living in current society increases, the gap between the rich and the poor becomes more pronounced, and the vulnerability of adolescents left behind in this situation becomes more highlighted. In addition to giving more care to them in terms of policies, it is also necessary to intervene in the value system of children, guiding them to reduce the pursuit of material things and focus more on the enrichment of the spiritual world, so that they can learn in an equal and harmonious classroom atmosphere. After that, we can also cultivate self-con dence of left-behind adolescents and help them develop their specialties, at the same time make efforts to reduce the discrimination and devaluation of left-behind adolescents by non-left-behind adolescents.
Finally, this study found that resilience was also an important protective factor for social withdrawal.
Resilience is an individual's ability to recover from setbacks and is an important psychological factor that allows individuals to face and effectively solve problems with a positive mindset despite stress and distress (Laird et al., 2019; Theron & Theron, 2010). To address this, life skills training can be provided to enable them to solve various problems independently, so that these problems do not lead to their psychological stress and affect their normal study and life. In addition, to help the left-behind adolescents strengthen their resilience, they can be given lectures on relevant topics and group counseling on selfpleasure, positive coping with setbacks and stress, and positive mindset guidance. In addition, counseling on interpersonal communication can also be provided directly to middle school students to improve their shyness and avoidance when interacting with others, and to train their social ability so that they can interact with their peers in a harmonious manner.

Limitations and Future Research Directions
This study found differences in interpersonal, cognitive, and psychological characteristics and behaviors between left-behind and non-left-behind adolescents, discovered a chain mediating model affecting social withdrawal of left-behind adolescents, and found that relative deprivation and resilience played a chain mediating role between social support and social withdrawal from an interpersonal perspective. However, there are certain limits in this study. First, this study regarded social withdrawal as a holistic concept and did not distinguish between active withdrawal and quiet withdrawal, but actually the psychological nature of these two is different, and whether different withdrawal behaviors have similar psychological mechanisms needs to be further tested in the future; Second, although the present study found a negative correlation between relative deprivation and resilience, and although it showed that relative deprivation as a cognitive and emotional integrative psychological state affects individuals' psychological adaptive functioning, the core reasons for this pathway of in uence still deserve more detailed analysis. In addition, in this study, due to the convenience of sampling, the subjects were selected from the same school, which may affect the external validity of the study results, and future studies may consider the in uence of the subjects' geography on the study to test this model extensively.
Finally, as a disadvantaged group, there are still few studies on left-behind adolescents, but due to the lack of family functions, left-behind adolescents may have a lack of social adaptation ability in many aspects, which seriously affects their healthy development, so more attention should be paid to this group in the future, and timely intervention and care should be provided to help them grow up healthily.

Conclusion
(1) Compared to non-left-behind adolescents, left-behind adolescents have a lower degree of social support, a higher degree of social withdrawal, a higher degree of relative deprivation, and a lower degree of resilience.
(2) There is a negative correlation between social support and social withdrawal among left-behind adolescents, while relative deprivation and resilience play a chain mediating role between social support and social withdrawal.

Declarations Ethical Approval and Consent to participate
All experiments were carried out in accordance with the Ethics Committee of the Center for Studies of Social Psychology at Central China Normal University (CSSP-2019026), and all participants enrolled in the study voluntarily, with informed consent prior to its initiation.
Written informed consent was obtained from all participants.

Consent for publication
Written informed consent for publication was obtained from all participants.

Availability of data and materials
The datasets analyzed during the current study are available in the supplementary. The materials during the current study are available from corresponding author at: Bin Zuo (zuobinccnu123@163.com) Competing Interests

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. data.xls