Introduction

Mental health is associated with the positive personality and social development of adolescents, and it is an issue that the whole society pays more and more attention. In recent years, there are increasingly more adolescents experiencing various psychological problems that heavily impact their studies and daily life, where they are more likely to exhibit harmful and dangerous behaviors to themselves and others. Psychological problems of adolescents are closely related to their family environment and education levels. Studies have found that the family environment, including parenting methods, family education, relationships, etc., have a great impact on the mental health of adolescents (Li, 2020, b). Left-behind children, a group of underaged who are forced to stay in their registered birthplace by their parents and are unable to live with their parents, as both or either parent has to leave their hometown for work (Miao et al., 2021). This group of individuals are more likely to be affected by the negative environment, of which, the negative impact has been garnering more social attention.

By the end of 2020, there were 6.436 million left-behind children in rural China (the State Council Information Office of the People’s Republic of China, 2021). In 2016, China State Council (2016) issued the Opinions of the State Council on Strengthening the Work of Caring for the Left-behind Children in Rural Areas, which deemed the protection of the mental health of left-behind children as an important and urgent societal issue. In 2021, The National Mental Health Evaluation and Development Center of The Institute of Psychology, Chinese Academy of Sciences released a Survey report on Rural Children's Mental Health. The survey results show that left-behind children are at great risk of mental health. The detection rate of depression among left-behind children is 28.5%, and the detection rate of excessive anxiety among left-behind children is 27.7%, both higher than that of non-left-behind children (Huang, 2021). Despite increasing attention on the development of left-behind children, the continuous development of society and advancement of globalization has led to more Chinese opting to work abroad. This societal phenomenon brought along a rapidly expanding sub-population—overseas Chinese left-behind children.

Overseas Chinese left-behind children refer to the underage group aged 16 and under who are unable to live with their parents due to one or both parents reside abroad for a long time, the children stay in the birthplace. Overseas Chinese left-behind children may have a relatively high standard of living compared with other left-behind children, but their mental health status is negatively impacted, and may experience different struggles. Studies have found that children’s overall wellbeing is highly sensitive to changes in family structure (Fomby & Cherlin, 2007). When children are in an environment where their parents are not around for prolonged periods, they are able to sense family instability, so they are more likely to have psychological problems such as indifference and isolation (Akister et al., 2010; Senaratna, 2012). Similarly, left-behind experiences such as family dysfunction, lack of family support, lack of family education resources, and lack of psychological interactions with parents (Dong et al., 2020; Miao et al., 2021; Tan & Li, 2018), also have a negative effect on the personality and academic development of left-behind children (Jia, 2020), which is especially more prominent among preschoolers and junior high schoolers (Liu, 2019).

Overseas Chinese left-behind children make up a notable sub-population of the left-behind children population. Having their parents staying abroad, this group of children experience greater difficulty in navigating the distance and time between them and their parents, as well as possible social and cultural differences (Wang & Tong, 2019). Studies have found that transnational left-behind children are more likely to try suicide than non-transnational left-behind children (Knipe et al., 2019).

Transnational left-behind children are more likely to experience more symptoms of depression and anxiety (Suárez-Orozco et al., 2011). In addition, behaviors such as truancy, dropout and other academic problems are more likely to increase, and it is a challenge to resolve these psychological and behavioral problems (Adams, 2000). Some studies, however, have shown that despite the emotional pain that transnational left-behind children face, they experience a better quality of life as their parents’ overseas work greatly increases family income, and material support such as remittances and gift purchases (Ukwatta, 2010). In addition, with the development of science and technology, communication has become more convenient and diversified. Parents can now keep up with their children’s lives, academics, and emotional needs despite being abroad. Virtual companionship increases parent–child communication, and the transnational left-behind children eventually learn to empathize with their parents’ efforts and are able to regain a sense of family support, albeit virtually (Dreby, 2007; Madianou & Miller, 2011; Mazzucato & Cebotari, 2017). There are relatively few domestic research on overseas Chinese left-behind children, but some scholars have begun to pay attention to this special group. For example, Huang Chengyi, Liao Chuanjing, etc. have found that life events play an important role in the mental health of overseas Chinese left-behind children (Huang et al., 2016), Pan Yujin, etc. have found that the behavior, personality and family education resources of overseas Chinese left-behind children are comparably worse than those of domestic left-behind and non-left-behind children (Pan et al., 2010). In a general long-term perspective, especially with the continuous development of society, research on this sub-population is especially crucial.

Overseas Chinese left-behind children are often found facing difficulties in life alone, due to the lack of immediate family support. Therefore, compared with ordinary adolescents, the impact of negative life events on overseas Chinese left-behind children may be more significant. Negative life events refer to adverse events that make individual life change negatively and cause negative consequences, thereby compelling the individual to make an adaptive response (Holmes & Rahe, 1967). Negative life events have been found to be important predictors of increasing mental health problems and decreasing life satisfaction among adolescents (Low et al., 2012; Ng et al., 2018). Some researchers found that negative life events experienced by individuals before the age of 18, such as family dysfunction, neglect, abuse, etc., can impair the development of adolescents’ cognitive function and the formation of personality characteristics, causing emotional problems and health-sabotaging behaviors (Wan & Tao, 2020; Gao et al., 2020). Negative life events, however, are not always detrimental to mental health. Some scholars have found that individuals with traumatic experiences, after experiencing the pain caused by negative events, may also find their meaning of life and experience psychological transformation, which will enhance their life satisfaction (Triplett et al., 2012). Although there is no definite consensus, there is a tellingly strong relationship between negative life events and mental health.

As overseas Chinese left-behind children face difficulties and problems in their daily lives, social support has become an important factor for them to overcome these difficulties. Social support refers to the external material or emotional support that an individual can get or perceive (Furmark et al., 2009), which can relieve psychological distress, improve social adaptability, and enhance mental health (Xiang et al., 2020). Perceived social support refers to the subjective feeling of emotional support (Kishimoto et al., 2016), which is the individual's experience of being supported, understood, respected and the availability of this support in society (Barrera, 1986; Xie, 2017). Studies have shown that the degree of perceived social support can affects the mental health development of college students continuously and dynamically (Zhang et al., 2018). An enhanced perception of social support can effectively reduce social anxiety and play a positive role in buffering the stress response generated under stress (Cohen & Wills, 1985; Wen et al., 2020), if adolescents can effectively perceive social support from family and school, it can not only reduce anxiety and depression symptoms, but also improve their subjective well-being (Hellfeldt et al., 2019). Perceived social support is, therefore, a strong variable in predicting the mental health (Shelton et al., 2017) status of overseas Chinese left-behind children.

Bandura has defined self-efficacy as an individual's ability to anticipate or perceive his own specific actions to achieve a specific goal, similar to confidence or belief in their own abilities (Bandura, 1977). It mediates the process of individual's thinking, choice, emotion, and motivation, which then regulates individual's psychological and behavioral activities. Schwarzer believes that people have a general sense of self-efficacy, that is, an overall self-confidence when individuals respond to challenges in different environments or accept new things (Schwarzer et al., 1997). Due to the long-term absence of parents in the growth process of overseas Chinese left-behind children, family then becomes an insignificant factor. In the face of various life and learning situations, the general self-efficacy of this group of individuals become indispensable. Studies have found that the social adaptability of left-behind junior high schoolers is affected by their self-efficacy (Xie et al., 2014). There is a positive correlation between self-efficacy and perceived social support. The stronger the perception of support, the higher the self-efficacy (Adler-Constantinesc et al., 2013; Ozsaker et al., 2015). The degree of social support perceived by left-behind high schoolers is a predictor of their general self-efficacy (Hu & Wang, 2009), and the self-efficacy and social support of freshmen with left-behind experience is a predictor of their mental health (Zhan et al., 2016). General self-efficacy is, therefore, likely to play an important role in regulating the mental health of overseas Chinese children with left-behind experience.

The theoretical model of psychological stress process model points out that the process of psychological stress involves three components: stressors (negative life events experienced by people, stimuli encountered, etc.), psychological mediators (social support, coping style, personality characteristics, etc.) and psychosomatic responses (various psychological, physical and behavioral responses) (Jiang, 2012). Psychological stress is a continuous process of multiple factors that affect the individual's mind and body due to stressors mediated by psychological mediators. Based on the psychological stress theory model, the current study focuses on the sub-population of overseas Chinese left-behind children. On the basis of investigating and analyzing the relevance of their negative life events, perceived social support, self-efficacy, and mental health status, this study will analyze the mediating effect of perceived social support between negative life events and mental health status of overseas Chinese left-behind children, as well as the moderating effect of general self-efficacy between the negative life events of overseas Chinese left-behind children and the perception of social support (Fig. 1). A questionnaire survey method was used to test the hypotheses, where the relationship between negative life events of overseas Chinese left-behind children and mental health status was explored, and the mediating effect of social support and the moderating effect of self-efficacy was analyzed.

Fig. 1
figure 1

Assumed model

Full size image

Methods

Participants

The target group of this study is the overseas Chinese left-behind children, namely the children aged 16 and under who are separated from their parents for a long time because at least one of their parents has been working abroad for half a year or more while the children live in their domicile in China. Taking this as the standard, we selected overseas Chinese left-behind children in grade 5–8 from two primary schools and three junior middle schools in Qingtian County and Ruian City of Wenzhou, Zhejiang Province in September 2020, which are two regions with many overseas migrant workers in China, as the samples of this study. We take the school as the unit and use the cluster random sampling method to carry out the questionnaire evaluation. A total of 600 questionnaires were distributed to 8 fifth grade classes, 9 sixth grade classes, 24 seventh grade classes and 24 eighth grade classes, and 585 questionnaires were recovered. After removing the questionnaires that did not meet the test criteria, obvious perfunctory and omission,568 valid questionnaires were obtained, with an effective rate of 97.09%. The age distribution of the sample ranges from 10 to 16 years, with an average age of 13.51 years and a standard deviation of 1.11. The general demographic information of the participants and their families is shown in Table 1.

Table 1 Demographic characteristics of the subjects and their families (n = 568)
Full size table

Design

We compiled the "Questionnaire on the psychological symptoms and impact of overseas Chinese left-behind children" (a combination of materials from “2.4 Materials” and demographic variables) and printed it into a sheet of front and back of A3 paper. It took about 30 min to complete the entire questionnaire. The paper questionnaires were distributed by teachers who are in charge of students’ mental health education in the school and trained by us to standardize their work. The students were briefed according to the following requirements: (1) Subjects were informed of the research purpose of the questionnaire. (2) Informed of the anonymity and confidentiality, where if subjects had objections and were unwilling to cooperate with the answering papers, they were not forced to do that, each subject was protected by research ethics. (3) The questionnaires were collected by the researcher after the students completed the self-evaluation on the spot. (4) After the invalid questionnaires were eliminated, the data were entered into the SPSS statistical software for analysis.

Materials

Adolescent self-rating life events checklist, ASLEC

ASLEC was compiled by Liu Xianchen and is suitable for assessing the frequency and intensity of adolescents' negative life events. The scale has a total of 27 items, including interpersonal relationship, learning stress, punishment, loss, health, and other six factors. Each item scores 0–5 points. The higher the score, the more negative the impact of the life event. Liu Xianchen et al. tested the reliability of ASLEC when compiling the questionnaire in 1997, and the results showed the internal consistency reliability of ASLEC. That is, Cronbach α was 0.8492, split-half reliability coefficient was 0.8809, and test–retest reliability was 0.6861(Liu et al., 1997). In recent years, many scholars have used ASLEC for research. For example, in the study of Zhang Shanshan et al., the Cronbach α of ASLEC was 0.85 (Zhang & Zhang, 2019). In the study of Zheng Yan et al., the Cronbach α of ASLEC was 0.932 (Zheng et al., 2020). In the study of He Anming et al., the Cronbach α of ASLEC in the pretest and posttest were 0.94 and 0.98, respectively (He et al., 2021). All this suggests ASLEC has a high degree of credibility. Our study conducted confirmatory factor analysis on the structure of ASLEC. The results: χ2/df = 4.647, RMSEA = 0.080, GFI = 0.817, AGFI = 0.768, NFI = 0.810, CFI = 0.844, TLI = 0.816, IFI = 0.845. The overall structural validity is good. The Cronbach α coefficient is 0.937.

Perceived social support scale, PSSS

PSSS was compiled by Zimet et al. and later revised into a localized version by Jiang Qianjin et al. The scale contains a total of 12 items, and is categorized into three dimensions: Family support, Friend support, and other support. The questions were tested using the 7 Point Likert Scale, with each item scores between 1–7 points. The higher the score, the higher the perceived level of social support (Zimet et al., 1988). A confirmatory factor analysis was conducted on the structure of the Perceived Social Support Scale. The results: χ2/df = 3.929, RMSEA = 0.072, GFI = 0.944, AGFI = 0.914, NFI = 0.928, CFI = 0.945, TLI = 0.929, IFI = 0.945, Overall, the validity of the questionnaire structure is good. In this study, the Cronbach α coefficient is 0.886.

General self-efficacy scale, GSES

GSES was compiled by German psychologist Schwarzer et al. and is widely used to measure individual self-efficacy. Later, Chinese scholar Wang Caikang and others revised the scale and applied it in local studies (Wang et al., 2001). The scale has a total of 10 items, using the 4 Point Likert Scale, each item scores 1–4 points, the higher the score, the higher the general self-efficacy level (Schwarzer et al., 1997). A confirmatory factor analysis was conducted on the structure of the General Self-Efficacy Scale. The results: χ2/df = 2.504, RMSEA = 0.052, GFI = 0.972, AGFI = 0.951, NFI = 0.927, CFI = 0.954, TLI = 0.934, IFI = 0.955, Overall, the validity of the questionnaire structure is good. In this study, the Cronbach’ α coefficient of the scale is 0.769.

Mental health inventory of middle-school students, MMHI-60

MMHI-60 was compiled by Wang Jisheng. The scale is mainly used to measure the psychological symptoms of primary and middle school students to reflect their mental health status. The questionnaire has 60 items, including 10 dimensions: compulsiveness, paranoia, hostility, interpersonal sensitivity, depression, anxiety, learning stress, maladjustment, emotional instability, and psychological imbalance. The questionnaire uses the 5 Point Likert Scale. Each item scores 1–5 points. The higher the score, the worse the mental health status. The scores of each factor are the average scores of the items in the subscales, and the total scores make up the total average score (Wang et al., 1997). In recent years, many scholars have also used MMHI-60 for research. For example, in the study of Yao Qiang et al., the Cronbach α of MMHI-60 was 0.963 (Yao et al., 2017). In Yang Guoshun’s study, the Cronbach α of MMHI-60 was 0.964 (Yang, 2019). In the study of Yu Shiwei et al., the Cronbach α of MMHI-60 was 0.947 (Yu et al., 2021). All these evidences indicate that MMHI-60 has high reliability. This study conducted confirmatory factor analysis on the structure of MMHI-60. The results: χ2/df = 2.371, RMSEA = 0.049, GFI = 0.805, AGFI = 0.784, NFI = 0.759, CFI = 0.843, TLI = 0.832, IFI = 0.845. The overall structural validity is good. The Cronbach’ α coefficient of the scale is 0.962.

Data analysis

SPSS was used for descriptive statistics and correlation analysis to obtain the general demographic data of participants and their families, the average and standard deviation of each variable, and the correlation coefficient between variables. Amos was used for confirmatory factor analysis to test whether there was a common method deviation by adding a common method factor. Model 7 in Hayes macro program process was used to test whether the moderated mediation model assumed in this study is tenable. That is to test the mediating role of perceived social support between negative life events and mental health, and the moderating role of general self-efficacy between negative life events and perceived social support. To better present the interactive effects of negative life events and general self-efficacy, the general self-efficacy scores were divided into high and low groups according to the mean plus or minus one standard deviation.

Results

Common method deviation test

Data collected were from the self-reports of the same group of overseas Chinese left-behind children, which may cause common method bias (Han & Feng, 2020). Although we used anonymity and other methods in the questionnaire survey process to control possible common method bias, the common method bias test is still needed. Firstly, confirmatory factor analysis was carried out on the four factors of negative life events, general self-efficacy, perceived social support and mental health status of overseas Chinese left-behind children. The main fit indexes of the model were: χ2 / df = 2.282, RMSEA = 0.048, GFI = 0.664, NFI = 0.611, CFI = 0.736. Then, a latent method factor was added to the existing model to generate a new model, so that all measurement factors not only loaded on the corresponding personal construct dimension, but also on the newly added latent method factor. Finally, by comparing the fit index of the new model with the original model; Δχ2 /df = 0.023, ΔRMSEA = 0.001, ΔGFI = 0.005, ΔNFI = 0.005, ΔCFI = 0.005, it was found that the change of the fit index is less than 0.01, that is, after adding the common method factor, the fit of the original model did not change significantly. Results showed that there was no obvious common method bias in the data of this study.

Descriptive statistics

Kolmogorov–smirnov test for the four variables in this study was performed. The results showed that the data of the four variables in this study were not obey normal distribution (p < 0.05), so we showed the median and interquartile range (IQR) in the Table2. Correlation analysis among various variables showed (Table 2) that there was no significant correlation between negative life events and general self-efficacy; There were significant positive correlations between general self-efficacy and perceived social support, and between negative life events and mental health status. There were significant negative correlations between negative life events and perceived social support, between general self-efficacy and mental health status, and between perceived social support and mental health status.

Table 2 Median(IQR) and correlation coefficients between variables
Full size table

Testing of the moderated mediation model

According to the moderated mediation model test procedure proposed by Wen & Ye (2014), if the following three conditions are met at the same time, it indicates that the influence of negative life events of overseas Chinese left-behind children on mental health status is a moderating mediating effect. (1) In Eq. 1, the negative life events have a significant effect on mental health status; (2) In Eq. 2, the negative life events have a significant main effect on perceived social support, and a significant interactive effect on general self-efficacy; (3) In Eq. 3, perceived social support has a significant effect on mental health status.

Using the SPSS macro program PROCESS compiled by Hayes, regression analysis was performed after controlling the gender, grade, etc. The results are shown in Table 3: Negative life events have a significant and positive association with mental health status (β = 0.46, p < 0.001); negative life events have a significant and negative association with perceived social support (β = -0.13, p < 0.05); General self-efficacy has a significant and positive association with perceived social support (β = 0.75, p < 0.001), and negative life events × general self-efficacy has a significant and negative association with perceived social support (β = -0.31, p < 0.01); Perceived social support have a significant and negative association with mental health status (β = -0.10, p < 0.001). The above results satisfy three conditions of moderated mediation model. Deviation-corrected percentile Bootstrap method was also tested. The average of the general self-efficacy score and plus or minus one standard deviation were taken as three levels of general self-efficacy: high, medium, and low. The mediating effect and 95% Bootstrap confidence interval of perceived social support between negative life events and mental health status are shown in Table 4. In summary, the moderated mediation model proposed in this study has been confirmed (Fig. 2), that is, the negative life events of overseas Chinese left-behind children are not only directly associate with the mental health status, but also associate with the mental health status through the mediating effect of perceived social support, and the first half of this mediating effect is regulated by the level of general self-efficacy.

Table 3 Test of the mediating effect of negative life events on mental health status
Full size table
Table 4 The mediating effect of perceived social support between negative life events and mental health status under different levels of self-efficacy
Full size table
Fig. 2
figure 2

The moderated mediation model

Full size image

To better present the interactive effects of negative life events and general self-efficacy, a simple effect analysis chart was drawn (Fig. 3). The simple slope test shows that at a high general self-efficacy level, negative life events have a significant negative association with the perception of social support; at a low general self-efficacy level, negative life events have no association with the perceived social support (Table 4).

Fig. 3
figure 3

The adjustment of general self-efficacy to negative life events and the perception of social support

Full size image

Negative life events variance test

The variance analysis was used to test the difference of the average score of the six dimensions of negative life events, and the results showed that the sensitivity of overseas Chinese left-behind children to different types of negative events was significantly different (F = 17.415, p < 0.001). The simple effect test is further conducted, as shown in Table 5. The results show that: the sensitivity of overseas Chinese left-behind children to interpersonal relationship, learning pressure and punishment is significantly higher than the loss, health adaptation and other negative events. The sensitivity of overseas Chinese left-behind children to loss negative events is significantly higher than health adaptation and other negative events. However, there is no significant difference in sensitivity to interpersonal relationship, learning pressure and punishment, and there is no significant difference in sensitivity to health adaptation and other negative events.

Table 5 Anova results for average scores of different types of negative life events
Full size table

Discussion

Negative life events are directly associated to the mental health of overseas Chinese left-behind children

This study found that negative life events of overseas Chinese left-behind children ha a significant negative association with psychological symptoms, that is, the greater affected by negative life events, the worse overseas Chinese left-behind children’s mental health status is. Some scholars have found that the higher the level of negative life events of left-behind children, the more negative coping styles they adopt, the fewer positive coping styles they adopt, and the worse their mental health is (Wang et al., 2013). The results of this study concur with that. Also, it is consistent with the previous studies by Nancy and other scholars, that is, negative life events have indeed caused negative changes in the lives of overseas Chinese left-behind children, and have a negative impact, which has led to a decline in the mental health of overseas Chinese left-behind children (Low et al., 2012; Ng et al., 2018; Wan & Tao, 2020; Gao et al., 2020). The results of this study, however, do not support the findings of Triplett et al., that negative life events contributed to the psychological transformation of overseas Chinese left-behind children, making their mental health more resilient (Triplett et al., 2012). This may indicate that although negative life events may have a positive possibility of “frustration builds strong” on the mental health of adolescents, they are more likely to have a negative association, especially for individuals such as overseas Chinese left-behind children, whose weak family support highlights the negative association between negative life events and their mental health.

The mediating effect of perceived social support

After examining the mediating effect of overseas Chinese left-behind children’s perceived social support, it was found that perceived social support mediates the association between negative life events and mental health. Of which is manifested as negative events by weakening the perceived social support of overseas Chinese left-behind children, and indirectly reducing their mental health status. This result just supports the theory of Psychological Stress Theory, that is, under the influence of negative life events and other stressors, individuals show some psychological, physical, or behavioral reactions through the role of mediating factors such as social support, coping, cognition and personality characteristics. Previous studies have found that excessive negative life events can reduce people's positive experience of life (Zimmer-Gembeck & Skinner, 2010), such as perceived positive social support. Failure to perceive social support effectively can lead to negative emotions such as anxiety and hopelessness (Li et al., 2020a, b). Other studies have found that individuals affected by punishment-type negative life events are less likely to have mature coping strategies (Cai & Jin, 2014), where they can not perceive social support and will be unlikely to seek help. Individuals who have left-behind experience and struggles to perceive social support will be more likely to have negative emotions, which can cause loneliness in early adulthood (Liu et al., 2020). The results of this study are consistent with previous studies. The greater the intensity of negative life events of overseas Chinese left-behind children, the less they can perceive social support, thus, the worse their mental health is.

The moderating effect of general self-efficacy

After clarifying the direct and indirect relations of negative life events on mental health status, we propose that this is a moderated mediation model. We tested the moderating effect of general self-efficacy and found that general self-efficacy moderates the path of negative life events for perceiving social support. Specifically, the negative association between negative life events and the perceived social support of overseas Chinese left-behind children is still significant at high levels of self-efficacy but becomes insignificant at low self-efficacy levels. It may be attributed to the locus of control. Previous studies have found that locus of control directly affects college students' level of perceived social support, and internal controllers tend to perceive the behavior of others as supportive. Therefore, the perceived level of social support is relatively high (Ye, 2005). There is also a significant positive correlation between internal control and general self-efficacy, and internal control significantly and positively predicts general self-efficacy (Mo et al., 2010). We can, therefore, speculate that individuals with high self-efficacy may have stronger internal control characteristics, which enable them to perceive social support more effectively than individuals with low self-efficacy when facing the negative effects of negative events. Based on this speculation, we can incorporate the factor of locus of control into the follow-up further research.

Education advice

According to the results of this study, there is a negative relationship between negative life events and the mental health of overseas Chinese left-behind children. Meanwhile, among the six negative life events, the sensitivity of overseas Chinese left-behind children to interpersonal relationship, learning pressure and punishment is the highest, followed by loss. This is consistent with some previous research findings. For example, some scholars have found that teenagers’ mental health problems are related to their interpersonal relationships in school(Li et al., 2020a, b). The decline of mental health level is also related to high learning pressure(Hosseinkhani et al., 2020). Parents who use corporal punishment are more likely to develop mental health disorders in children(Durivage et al., 2015). Mental health problems of adolescents also increase after bereavement(Stikkelbroek et al., 2016). This result suggests that parents should avoid long-term separation from their children, otherwise children will feel the loss of parental love. At the same time, parents and educators should pay more attention to the social and learning status of overseas Chinese left-behind children, reduce punitive educational measures, and adopt more educational means of encouragement and guidance.

Conclusion

Perceived social support plays a partial mediating role between negative life events and mental health of overseas Chinese left-behind children. And this mediating role is regulated by the level of general self-efficacy. Specifically, when the general self-efficacy of overseas Chinese left-behind children is at a medium and high level, the greater the negative impact of negative life events, the weaker the ability to perceive social support from family, friends and other aspects, resulting in the worse their mental health. While the general self-efficacy of overseas Chinese left-behind children is at a low level, the relation between negative life events and perceived social support is not appear.