The impact of social support on job-related behaviors through the mediating role of job stress and the moderating role of locus of control: Empirical evidence from the Vietnamese banking industry

Abstract The aim of this research is to investigate the effect of social support on job-related behaviors, noting the mediating roles of job stress. Specifically, the study tests the prediction regarding the extent to which these different types of locus of control moderate the relationship between social support and job stress as well. The current study used self-report questionnaires from a sample of employees in the Vietnamese banking industry to test the hypothesis. Results from the partial least squares structural equation modeling (PLS-SEM) using the SmartPLS 3.0 program showed that social support positively impacted job-related behaviors (e.g., job satisfaction, job performance); these relationships were mediated partially through job stress. Besides, this research revealed that individuals with a higher internal locus of control were more likely to have lower levels of job stress and it also moderated the relationship between social support and job stress. This negative association was weaker for bankers who have a higher external locus of control. The main findings of this research demonstrated that social support and locus of control play an important role in the overall effectiveness of employees. It implied that it is crucial for organizations to set up a good work atmosphere that supports psychological well-being and have suitable policies to strengthen social support among supervisors and employees in an organization to reduce job stress and improve job-related behaviors.


PUBLIC INTEREST STATEMENT
Previous literature focused on job stress among employees in Vietnam while global banking sectors have instituted significant positive changes in this regard. However, research on these issues has received less attention with regards to the banking industry. Therefore, the current study makes an attempt to bridge this gap. It is perceived that human resource management has become fundamental in the sustainable development of the banking industry and needs to be studied. The findings of this study showed that social support positively influence job satisfaction and job performance which is partially mediated by job stress. In other words, an employee with adequate social support could experience reduced job-related stress that may lead to enhanced job satisfaction and performance. Additionally, the study also found that most employees with less education and experience, females, and married individuals tend to experience more job stress. demonstrated that social support and locus of control play an important role in the overall effectiveness of employees. It implied that it is crucial for organizations to set up a good work atmosphere that supports psychological well-being and have suitable policies to strengthen social support among supervisors and employees in an organization to reduce job stress and improve job-related behaviors.

Introduction
With the growing complexity of how work is executed and how organizations have functioned, in the past few decades, job stress and its related antecedences and outcomes have become a focal topic among organizational researchers and practitioners. Many employees experience periods of acute stress at some point in their working lives. However, job stress begins to have a stronger relationship with adverse outcomes when it becomes chronic and long-lasting (Darmody & Smyth, 2016;O. F. Malik et al., 2010;Yozgat et al., 2013). Recently, job stress has received a considerable amount of attention in research and practice. In other words, it continues to be a large issue for both employees and organizations (McCarthy et al., 2010;Moen et al., 2013). In the past, researchers and practitioners have considered job stress to be a less important variable compared to other organizational issues such as productivity (Finney et al., 2013;Hoboubi, Choobineh, et al., 2017;Kasl, 1996;Li et al., 2019). However, practitioners and researchers are observing that job stress directly relates to organizational factors such as turnover or loss of productivity that reflects as a cost for both their human capital and the bottom line. After decades of research in the area of job stress, it has become a well-established and widely examined topic at present (Nowrouzi et al., 2016).
In recent years, Vietnam's economic prospects have been facing a number of challenges while the difficulty in the banking sector still restrains domestic demand (SBV, 2019). With Circular No. 36/2014/TT-NHNN, the banks have divested the regulation level which has spurred waves of acquisitions and mergers in the banking industry. To overcome these challenges and establish sustainable development, one of the requirements identified for banks is improving business activities combined with managing human resources reasonably. The investment in people is an important and wise strategy to gain more profits in a competitive market at present. Some scholars (e.g., Vuong & Suntrayuth, 2020) expressed that in many countries if the development of the banking system is supported by the commensurate development of human resources, it will create sustainability for the banking sector in particular and the economy in general.
Nowadays, the merging and restructuring in Vietnamese banks has led to fluctuations in laborers because of strengthening the financial system. From 2010 to 2019, the number of employees has fluctuated significantly. The problem of the banking industry in Vietnam is that human resources are either deficient or weak and a brain-draining risk. In order to solve the above issues, it is essential for banks to retain their talented employees as a valuable resource by focusing on employees' job-related behaviors (e.g., job satisfaction, job performance). Jobrelated behaviors are an important key to decide the success of an organization (Hoboubi, Choobineh, et al., 2017). For the success of the banking sector, it is very crucial to handle issues related to human resources successfully to make the employees satisfied. When employees are happy and satisfied, they often enhance their performance and hold a good image of the bank (Bashir & Ramay, 2010). Therefore, employee's job-related behaviors need to be considered thoughtfully to improve the quality of human resources in banks.
While the banking industry has been growing faster and faster, job stress in this sector has become inevitable in the working process of banking staff. Job stress refers to any characteristics of the working environment posing a threat to an individual (Caplan et al., 1975; F. C. F. C. Yang et al., 2019). N. Malik (2011) indicated that the global banking sector has experienced rapid and significant changes due to globalization, harsh competition, pressure for maintaining profitability and large and maximized responsibilities. Hence, such a workplace seems to cause stress for bank employees. Besides, job stress has been found to have a strong relationship with certain organizational outcomes especially in the banking sectors requiring long-term, stable, and high-quality human resources. Some researchers stated that job stress has a strong negative relationship with job satisfaction (e.g., Darmody & Smyth, 2016), and job performance (e.g., Bashir & Ramay, 2010). In other words, high levels of job stress are associated with many negative organizational outcomes, underscoring the importance of developing effective methods to reduce job stress.
However, it is important to understand the factors that cause employees to suffer from stress syndrome to develop interventions, and it is also essential to provide effective solutions to prevent the problem. In this research, the authors focused on the role of social support, defined as "that subset of persons in the individual's social network upon whom he or she relies for socioemotional aid, instrumental aid, or both" (Thoits, 1982) at work as a key antecedent of job stress and a robust factor predicting the employees' job-related behaviors.
In addition, a few researchers have assumed that job stress depends on several individual personality characteristics, especially locus of control (Chen & Silverthorne, 2008). Locus of control is defined as an individual's beliefs about what determines whether or not they get a reward in life (Galvin et al., 2018). In the early definition, Rotter (1966) classified locus of control into the external and internal locus of control. Chen and Silverthorne (2008) proved that employees who have an external locus of control suffer a higher level of job stress. Furthermore, Karimi and Alipour (2011) conceptualized that locus of control moderated social support-job stress relationship and they provided just an integrated conceptual framework. However, there has been no other empirical study concentrating on this moderating effect of locus of control on the association of social support and job stress. Therefore, the moderating effect of locus of control on this relationship will be considered in order to manage human resources effectively.
In general, the Vietnamese banking industry has developed notably during the past few years and experienced many changes. At the same time, the industry is facing some issues and challenges i.e., poor quality labor, employee job stress, and job-related behaviors. Thus, the current study aims to examine the impact of social support on jobrelated behaviors (e.g., job satisfaction, job performance) through the mediating role of job stress. Additionally, this study attempts to find the moderating effect of locus of control on the relationship between social support and job stress in the Vietnam banking industry.

Social support
Social support refers to behaviors occurring during the interpersonal transactions of messages that make its recipients feel cared for, esteemed, and loved (Cobb, 1976). Social support is provided and received among social contacts of varying strengths of connection. It may serve to give advice, assist with problems, provide information, share personal concerns, and, when appropriate, comfort and encourage (Agneessens et al., 2006). Mack and Rhineberger-Dunn (2019) stated that social support is the product of interpersonal work relationships that has the potential to promote the well-being or coping abilities of the recipient. It also has been defined as the actions of others that are helpful (Deelstra et al., 2003). The social support is offered by organizational relationships from supervisors, peers, and subordinates. In this study, most of the participants are first-line employees. Therefore, the definition of social support in this study refers only to receiving support from co-workers and supervisors. Support from subordinates has been eliminated. Porter (1985)'s seminal work on market positioning and competitive advantage acknowledged the importance of effective human resource management in driving business deliverables. More specifically, Porter's study served to underscore the role of human resource management in creating and sustaining superior performance outcomes. Sung and Choi (2014) reported that more academic inquiry is necessary to understand which mechanisms affect worker performance. Due to the vastness of scholarly research on job performance, it was necessary to consolidate the review to a thematic assessment of performance in terms of ongoing research. Job performance has been studied in a number of contexts. In general, job performance refers to "the result of the fit between a person's behaviors and the demands of the job/task" (Brief & Weiss, 2002). Jacobs et al. (2013) defined job performance as the ability of an employee to do their job well. Therefore, the definition of job performance can be derived as the ability of an employee to reach their goal within a set amount of time within the context of a job or organization. Besides, Jamal (2007) stated that job performance is the work that an employee can perform with success within the boundaries of available resources. An important facet of job performance is task performance, which refers to behaviors of job performance that are beneficial to the organization (Kuyumcu & Dahling, 2014). Beavan and Ehrlich (2013) asserted that social support may have a positive effect in mitigating job demands and argued it is a necessary job resource at work. The construct of social support has its foundation in social exchange theory theories (Kossek et al., 2011). As social support is considered a basic human need (Cobb, 1976) the construct of social support within the organization, in fulfilling a basic need, leads to positive job-related behaviors (e.g., job satisfaction, job performance). Social support in the bank involves informative support (e.g., guidance in work or giving feedback), emotional support (e.g., acceptance, showing empathy), and material support (e.g., preparing resources and assistance that are related to the job). These supportive actions are expected to encourage positive and strong feelings of bankers' well-being and have a beneficial influence on their performance. Besides, getting support from co-workers and supervisors has a positive spillover effect on the bankers that results in high self-efficacy and more positive energy toward their work and workplace. In this case, job performance during an assignment would be enhanced. Moreover, previous scholars (e.g., Yuan-Duen et al., 2017) showed a positive association between social support and employees' job performance. Accordingly, the following hypothesis is proposed:

Job performance
H 1 : Social support will increase job performance of bankers.

Job satisfaction
The most commonly used definition of job satisfaction by Spector (1997) implicates how people feel about their jobs and different aspects of their jobs. Job satisfaction can refer to a pleasurable, positive attitude resulting from the appraisal of one's job or job experiences (Gustainiene & Endriulaitiene, 2009). It can be interpreted that the happier people feel in their job, the more satisfied they are. Job satisfaction is acquired when employees enjoy their work, or they get positive emotion after appraisal of his or her job. Moreover, Singh and Sinha (2013) suggested that job satisfaction can be understood as an overall attitude or it can apply to the parts in the job of an individual. It can be identified as an attitude, generally acquired over a period of time as an employee gains more and more information about the workplace.
Needs hierarchy theory of Maslow (1943) was one of the theories to test the main determinants of job satisfaction. Maslow's Hierarchy of Needs is the source of most of the motivation theories used in the area of job satisfaction. This theory is built from satisfaction on hierarchical ladders of human demand as motivation factors. This theory asserted clearly and obviously the importance of job-related satisfaction towards the activation of employees' motivation (Kian et al., 2015). According to Ozguner and Ozguner (2014), Maslow's needs hierarchy theory was used to examine the important contributors to job satisfaction. Maslow (1943) stated that social needs are the needs to communicate with other people to create social relationships; interact, affect, affiliate, share and collaborate with other members of the organization. The relationships with superiors, colleagues, and subordinates correspond to the satisfaction of social needs. Furthermore, many researchers have found that social support is strongly associated with job satisfaction (e.g., Yuh & Choi, 2017). They also showed that employees who had better social support were more likely to be satisfied with their jobs. Therefore, it is suggested that: H 2 : Social support will increase job satisfaction of bankers.
According to Vuong and Suntrayuth (2020), the relationship between employees and the organization is reciprocity. Thus, job satisfaction could be created by the exchange between the demands and the expectations of employees, and of an organization's expectations. The organization supplies the medium for employees to meet a series of needs and to realize their personal aspirations. Employees provide for the organization the resources in order to reach the organization's objectives. Mustapha (2013) argued that it is crucial to study employee's job satisfaction because it is often considered as an important factor in enhancing organizations' performance and competitiveness. The organizations having high job satisfaction of employees will increase the productivity of an employee and the organization's overall performance. Consequently, a banker experiencing a higher level of job satisfaction results in a higher level of job performance (e.g., Inuwa, 2016). Thus, the following hypothesis is suggested: H 3 : Job satisfaction will increase job performance of bankers.

Job stress
Job stress is a physiological state appearing when an employee is impacted and attacked intensely by anything at work. According to Beehr and Newman (1978), job stress is also a situation in which some characteristics of the work situation are thought to cause poor psychological or physical health or to create risk factors making poor health more likely. Darmody and Smyth (2016) defined job stress as the harmful emotional and physical responses that happen when the requirements of the work do not match the demands, resources, or abilities of an employee. Shirey (2004) suggested that social support is the actions of others that are helpful, and it can be tangible or intangible protection to shield people from the adverse effects of life stress. The relationship between social support and job stress can be clarified by the job demands-resources model (Demerouti et al, 2001). This model assumed that job stress has a tendency to happen when bankers lack sufficient resources to handle demands at work. From this viewpoint, social support (e.g., coworker, supervisor) could be considered significant job resources that help bankers cope with high demands at work (Charoensukmongkol et al., 2016). Additionally, when social support meets important human needs such as security, social approval, belonging, and affection, employees will be less vulnerable to the effects of role conflict, role ambiguity, and role overload. By reducing or avoiding the experience of job stress, its manifestations will be lessened or eliminated. The significant negative correlation between social support and job stress was found in previous studies (e.g., Shirey, 2004;Yu et al., 2014). They noted that social support grows to reduce job stress in the workplace. Therefore, it is suitable for this study and the following hypothesis is suggested: H 4 : Social support will reduce job stress of bankers.
On the other hand, Arasli and Tumer (2008) showed that there are many effects of job stress on the individual as well as on the organization. They can lead to losing adaptability and productivity of work, negative comments, and low job satisfaction. O. F. Malik et al. (2010) stated that in the environment of the workplace, job stress is the outlook of individual dysfunction attributable to the psychology affecting the tight, unfavorable, or unapproachable workplace position. The personenvironment (P-E) fit, and person-organization (P-O) fit theory have been applied to explain the negative relationship between job stress and job-related behaviors (e.g., job satisfaction, job performance) (L.-Q. L.-Q. Yang et al., 2008). When employees' abilities cannot fit the demands of the job, it means that they are suffering from stress. Subsequently, P-E fit and P-0 fit indicate that job stress decreases job satisfaction and job performance. Many researchers have also discovered the negative relationship between job stress and job-related behaviors (e.g., Darmody & Smyth, 2016;O. F. Malik et al., 2010;Yozgat et al., 2013). Therefore, the following hypotheses are indicated: H 5 : Job stress negatively influences job satisfaction of bankers.
H 6 : Job stress negatively influences job performance of bankers.

The mediating role of job stress
Social support could create a positive working environment. When employees perceive that they work in a supportive working atmosphere, they are likely to have less job stress. This lower level of job stress could lead to a result in creativity and remove boredom and mundanity, and even to higher job satisfaction and performance. Previous researches revealed the direct relationships between social support and job stress (e.g., Shirey, 2004;Yu et al., 2014), job stress and jobrelated behaviors (e.g., Darmody & Smyth, 2016;O. F. Malik et al., 2010;Yozgat et al., 2013), social support and employees' job-related behaviors (e.g., Yuan-Duen et al., 2017;Yuh & Choi, 2017). Linking the associations outlined above, the authors argue that job stress is likely to mediate the association between social support and job-related behaviors. Based on the above discussion, the following hypotheses are suggested: H 7 : Job stress mediates the relationship between social support and job satisfaction.
H 8 : Job stress mediates the relationship between social support and job performance.

Locus of control
According to the social learning theory of Rotter (1954), locus of control (LOC) comprises internal and external locus of control. Individuals with an internal locus of control believe that events happening in their lives are under their control. People with an internal locus of control could control their emotions and try by themselves to achieve success. Whereas, external locus of control individuals believe that events happen by chance, luck or under control of powerful others, etc. This means that individuals with an external locus of control usually blame their environment for failures.
Additionally, job stress can be impacted by individual characteristics, such as locus of control (e.g., Chen & Silverthorne, 2008). Chiu et al. (2005) suggested that locus of control impacts the coping styles that individuals choose in order to manage everyday problems and stress. Employees who have a high level of LOC, especially internal LOC, are more powerful to overcome stress at their workplace. Hence, internal LOC individuals usually feel less stress than external LOC. In other words, a banker with high LOC (internal LOC) could identify the outcomes of actions by his/her decision-making. Consequently, increasing locus of control could reduce job stress easily. Besides, many previous studies have identified the relationship between locus of control and job stress (e.g., Chen & Silverthorne, 2008). Therefore, the following hypothesis is suggested: H 9 : Internal locus of control will reduce job stress of bankers.
Locus of control is a useful variable to decrease job stress especially if an individual has high autonomy and ability to overcome conflicts while he/she is obtaining high social support (Chen & Silverthorne, 2008). Hence, the impact of social support could increase significantly to reduce job stress. Besides, they also argued that employees with a higher internal LOC have an efficient effect on social support to minimize job stress. "Employees sometimes receive social support to reduce their job stress, but they do not have an adequate locus of control; consequently, the role of social support on decreasing job stress will be negligible" (Karimi & Alipour, 2011). Therefore, the authors argued that LOC moderated the association between social support and job stress, but they didn't provide any empirical evidence to demonstrate the moderating effect of locus of control. According to the above discussions, the following hypothesis is proposed: H 10 : Locus of control moderates the influence of social support on job stress. The negative association between social support and job stress will be stronger for bankers who hold a higher internal locus of control.

Control variables
Some scholars stated that demographic characteristics (e.g., gender, marital status, age, income, experience, and educational level) might be associated with the degree of job stress (Ayyappan & SakthiVadivel, 2013;Balakrishnamurthy & Shankar, 2009;Lunau et al., 2015). Thus, they were incorporated into the analysis as control variables. Gender and marital status were measured as a dummy variable, where males and married were coded as 1 and female and single coded as 0. Educational level, income, experience, and age were measured on the ordinal scale. Based on the literature review, the original expectation of the relationship between gender and job stress was that there would exist a difference in the ratings of male versus female on the scales (Ayyappan & SakthiVadivel, 2013). Also, the authors state that the educational level will have an impact on how well the employee is able to proactively deal with work-related stress. The higher the educational experience, the more quickly an acceptable constructive thinking coping strategy will be developed and implemented. As this is proactive from the employee, employers will not need to expend as much effort on dealing with stressors for their highly educated professionals. Moreover, a high level of job stress is an apparent issue. Job stress also corresponds with their salary, bonus and other benefits. The Vietnamese bank staff's salary, bonus, and other benefits are accounted based on productivity which belongs to workload in the identified period. Besides that, when the productivity of employees increases and contributes to the profit in the business of banks, the working environment and remuneration policies are improved. Therefore, more workload results in more benefits (higher salary) that may lead to short term satisfaction. However, employees sometimes fail to handle larger workloads, simultaneously, workloads and high income create job stress. Therefore, the following hypothesis is postulated ( Figure 1):

Figure 1. The conceptual model.
H 11 : Employee characteristics (e.g., gender, marital status, age, income, experience, and education level) have a relationship with job stress of bankers.

Measurement scales
The measurement scales measuring five concepts appeared in previous academic journals (e.g., Bouckenooghe et al., 2013;Dysvik et al., 2014;Levenson, 1973;Mack & Rhineberger-Dunn, 2019;Paoline et al., 2015). In this research, the measurement scales of social support were adapted from the study of Mack and Rhineberger-Dunn (2019) consisting of eight items. Sample items included (Supervisor support: e.g., "My work supervisor really cares about my wellbeing"; Coworker support: e.g., "I am able to talk about work-related problems with my co-workers"). The six items of job stress were used from the measurement scale of Paoline et al. (2015). A sample item for this variable was "I am usually under a lot of pressure when I am at work". The job satisfaction variable is used with five items being adapted from the research of Paoline et al. (2015). A sample item for this variable was "Most days I am enthusiastic about my job". The job performance scale was applied with six items from the study of Bouckenooghe et al. (2013). A sample item for this variable was "I fulfill responsibilities specified in the job description". The locus of control scale is adapted by eight items from the internal control scale of Levenson (1973). A sample item for this variable was "My life is determined by my own actions". The measurement scale was used with an interval scale with a five-point Likert scale: Strongly disagree (1), disagree (2), neutral (3), agree (4) and strongly agree (5) in order to constitute the main content of the questionnaire. Besides, in the measurement scale of job stress, there are two adverse statements. They are "I am usually calm and at ease when I'm working" and "Most of the time when I'm at work I don't feel that I have much to worry about". These items were reverse coded.

Procedure and sample
Because the original questionnaire was in English, the participants were Vietnamese. Therefore, the back-translated procedure was used to prevent misunderstanding or unclear meaning. The original questionnaire was translated to the Vietnamese version by someone with a Bachelor's degree in English and a scholar in the human resources field who also understand English. Then, other people with Bachelor's degrees in English and another scholar in the human resource field back translated this Vietnamese version of the questionnaire to the new English version. This step was to ensure that the new English questionnaire does not have a different meaning from the initial questionnaire.
The draft questionnaire of this study is originally in English and adapted from different countries and different industries. Thus, before the main survey was conducted, a pilot study was carried out with qualitative research in order to make sure that Vietnamese banking employees can understand the content in the draft questionnaire clearly. Moreover, it also helps the measurement scale of this study increase the degree of accuracy and trust. Qualitative research was implemented by sending the questionnaire to 50 fulltime employees who work in five Vietnamese banks (e.g., Vietcombank, BIDV, Sacombank, Viettinbank, ACB) by a convenient method. After receiving and reading the questionnaires, interviewees were asked how they understood the measurement scales and what they thought the meanings of the measurement scales were. Then, all comments from the interviewees were collected. Based on the results of the pilot study, the items on the scale that were unclear or hard to understand by the Vietnamese participants were modified. Besides, Giao and Vương (2019) suggested that Cronbach's Alpha and corrected item-total correlation coefficients should be used to verify the reliability of the data and to ensure the internal consistency reliability for item scales. As shown in Table 1, the minimum value of Cronbach's Alpha coefficients was 0.73 which was above the acceptable threshold of 0.7 and the minimum value of the corrected item-total correlation coefficients of all scales was higher than the acceptable threshold of 0.3 (Giao, Vuong, Huan et al., 2020). Therefore, these scales were reliable enough to be used in the main survey.
After the final questionnaire was completed, it was distributed through social media via Google docs and e-mail and a hard copy of the survey was directly sent to bank employees. A survey with a convenience sample of about 600 fulltime employees in banks in Vietnam was conducted. Specifically, the researcher delivered 300 survey instruments to employees in the state-owned bank group and 300 instruments to the private bank group. Among 600 questionnaires distributed, 71 were not useable due to missing values. As a result, the final sample size was 529, including 237 state-owned banks and 292 private bank employees. Table 2 provides descriptive statistics on the characteristics of the samples.

Partial least squares regression
The partial least squares structural equation modeling (PLS-SEM) was chosen to assess the studied hypotheses. "PLS-SEM is a statistical analysis technique for data exploration within the quantitative research discipline used to measure the observed variables collected from instruments to determine their influence on latent or unobserved variables" (Fornell & Larcker, 1981). Moreover, Vuong and Giao (2020) stated that SmartPLS 3.0 is one of the prominent software applications for partial least squares structural equation modeling. PLS is a soft modeling approach to SEM with no assumptions about data distribution because it has the capability to calculate p-values through a bootstrapping technique (Hair et al., 2014). It is well known for its capability of predictive accuracy. Because of all of the above strengths, the authors used PLS to analyze data in this study. Vuong and Suntrayuth (2020) suggested that the researchers should assess the reliability and validity of the multi-item variables before running PLS-SEM estimation for studied hypotheses testing. The results of scale accuracy (i.e., reliability and validity) of the studied constructs were presented in Table 3 and Table 4. To assess the reliability of the constructs, the thresholds of Cronbach's Alpha (0.7) and composite reliability (0.7) Hair et al., 2014) were applied; the data in Table 3  In addition, to assess the discriminant validity of the measurement model, the Fornell-Larcker criterion, and the Heterotrait-Monotrait ratio (HTMT) were used. Also, as can be seen from Table 3, the square root of the AVE of each variable was higher than the variable's highest correlations with the other variables. For example, the AVE value of job satisfaction was 0.628 and the square root of its AVE was 0.792. This value was both higher than the correlation values in its row (−0.533, 0.354, −0.436, and 0.419) and its column (0.602, 0.004, 0.371, −0.293, 0.470, and −0.052). Moreover, all HTMT values fell below the conservative maximum level of 0.681 (Table 4). Overall, both the reliability and validity of the measurement model were assured.

Assessment of structural model
Following the procedure to evaluate the structural model as proposed by Hair et al. (2014), the collinearity issues among each set of predictor variables were firstly checked; all VIF values (see Table 5) of less than 5.0 demonstrated that collinearity was unlikely to be a concern.
Next, the hypotheses testing results comprised of direct effects, mediating effects and moderating effects. They are presented below (Figure 3).
For hypothesis 1, the findings indicated that social support had a positive and significant relationship with job performance (beta coefficient = 0.124 and p-value = 0.007) ( Table 6). In other words, the more social support, the greater the possibility that bankers will have high levels of job performance, which was comparable with the outcome of the prior research of Yuan-Duen et al. (2017). Therefore, Hypothesis 1 was supported.
For hypothesis 2, the findings revealed that social support had a positive and significant association with job satisfaction (beta coefficient = 0.243 and p-value = 0.000) ( Table 6). In other words, the more social support, the greater the possibility that bankers will have high levels of job satisfaction, which corresponds with the prior research of Yuh and Choi (2017). This finding links to reality, employees in the banking industry interest in social support especially. When receiving significant social support, they have better relationships which support their work; they usually feel satisfied with their job. Thus, Hypothesis 2 was supported. For hypothesis 3, the findings confirmed that job satisfaction had a positive and significant relationship with job performance (beta coefficient = 0.458 and p-value = 0.000) ( Table 6). In other words, the more job satisfaction, the greater the possibility that bankers will have high levels of job performance, which was connected to the previous research of Inuwa (2016). Thus, Hypothesis 3 was supported.
For hypothesis 4, the findings demonstrated that social support had a negative and significant relationship with job stress (beta coefficient = −0.187 and p-value = 0.000) ( Table 6). When bankers have significant social support, they have less job stress, which was compatible with the previous research of Shirey (2004) and Yu et al. (2014). Besides, this result was accurate with the reality in the Vietnamese banking sector. If bankers do not have social support, they will face difficulties and pressure on the job. Consequently, if the policies and activities about increasing social support are not good, the employee will experience increased job stress. Thus, Hypothesis 4 was supported.
For hypothesis 5, the findings proved that job stress had a significant association with job satisfaction (p-value = 0.000) ( Table 6). This result indicated that job stress negatively affects job satisfaction. This result has strongly supported the fact that job stress was always a reliable predictor affecting job satisfaction with the beta coefficient = −0.435. In other words, the more job stress, the greater the possibility that bankers will have low levels of job satisfaction, which was similar to the previous research of Darmody and Smyth (2016) and O. F. Malik et al. (2010). Besides, this finding could be considered fit for the reality nowadays as employees of banks may want to  use all of their abilities for many different roles with most of the Vietnamese bank's culture. Again, the current solution of the banking industry in Vietnam is to improve the quality of labor. It consists of establishing suitable job progress for each position, criteria for training, and fostering and updating knowledge about basic business. Their aims are to meet the requirements of the competitive market and deal with the increasing pressure and complexity of Vietnam and the world. This situation has enabled banking staff to perceive their roles and position in the organization. Especially, high quality human resources have to be encouraged and motivated to improve the work environment. They want to undertake a different role with more frequency and intensity in some areas. Vietnamese banks want to perform well in all areas. So, when they are unable to perform their roles adequately, job stress will be high, which negatively impacts job satisfaction. Thus, Hypothesis 5 was supported.
For hypothesis 6, the findings indicated that job stress had a significant relationship with job performance (p-value = 0.001) ( Table 6). This result showed that job stress negatively affects job performance (beta coefficient = −0.172). In other words, the greater the job stress, the greater the possibility that bankers will have lower levels of job performance, which was shown in the previous research of Yozgat et al. (2013). One type of job stress is role ambiguity. It is easy to understand that role ambiguity is the ambiguity about aims, objectives, linking, and direction at work. When these things occur, bankers do not know what they want, whom they need to connect with in the workplace, what they need to do, and what their roles are. So, if the role ambiguity of an employee occurs, he or she may feel disappointed which leads to job stress and decreased job performance. Thus, Hypothesis 6 was supported.
For hypothesis 9, the findings showed that locus of control had a significant relationship with job stress (p-value = 0.002) ( Table 6). This result suggested that locus of control negatively affects job stress (beta coefficient = −0.137). In other words, bankers who have an internal locus of control are more likely to experience lower levels of job stress, which was  similar to the previous research of Chen and Silverthorne (2008). Therefore, hypothesis 9 was supported.

The mediating role of job stress
For hypothesis 7, the findings revealed that job stress mediated the relationship between social support and job satisfaction. The outcomes in Table 6 showed that the p-values for the direct path SS→JST and JST→JS were 0.000; 0.000, respectively, which were less than 0.05. So, they were statistically significant. Besides, Table 7 showed that the p-value of the indirect influence (SS→JST→JS) was 0.000, which was statistically significant as well (p < 0.05). Thus, the mediating role of job stress existed (Giao & Vương, 2019). Moreover, the p-value for the direct effect of social support on job satisfaction was 0.000. Consequently, this mediation was partial. The total impact of social support on job satisfaction through the mediating role of job stress was 0.324 (Table 7). This finding discovered that bankers who had social support had less job stress, which leads to high job satisfaction. Therefore, hypothesis 7 was supported.
For hypothesis 8, the findings showed that job stress mediated the relationship between social support and job performance. The outcomes in Table 6 showed that the p-values for the direct path SS→JST and JST→JP were 0.000; 0.001, respectively, which were less than 0.05. So, they were statistically significant. Besides, Table 7 indicated that the p-value of the indirect influence (SS→JST→JP) was 0.000, which was statistically significant as well (p < 0.05). Thus, the mediating role of job stress existed (Giao & Vương, 2019). Moreover, the p-value for the direct effect of social support on job performance was 0.007 (p < 0.05). Consequently, this mediation was partial. The total impact of social support on job performance through the mediating role of job stress was 0.156 (Table 7). This finding revealed that bankers with significant social support have less job stress, which leads to high job performance. Therefore, hypothesis 8 was supported.

The moderating role of locus of control
Hypothesis 10 predicted that locus of control would moderate the association between social support and job stress. The research justified that the moderating influence of the interaction between social support and internal locus of control with job stress was negative and statistically significant (beta coefficient = −0.130 and p-value = 0.000) ( Table 6). This discovery indicated that internal locus of control negatively moderated the relationship between social support and job stress. Besides, social support negatively affected job stress (Hypothesis 4). Therefore, the negative relationship between social support and job stress was stronger for bankers who have a high internal locus of control ( Figure 4). Therefore, Hypothesis 10 was supported.

Control variables
Hypotheses 11a, 11e and 11 f indicated that gender, educational level, and experience had a negative and significant association with job stress (beta coefficient = −0.158; −0.323; −0.242 and p-value = 0.003; 0.000; 0.000, respectively) ( Table 6). These results indicated that female, less educated and less experienced bankers tended to have high levels of job stress, which were comparable with the previous research of Ayyappan and SakthiVadivel (2013), Balakrishnamurthy andShankar (2009), andLunau et al. (2015). Thus, Hypotheses 11a, 11e, and 11 f were supported. Hypothesis 11b showed that marital status had a positive and significant relationship with job stress (beta coefficient = 0.180 and p-value = 0.003) ( Table 6). This result revealed that married bankers tended to have high levels of job stress, which was similar to the previous research of Ayyappan and SakthiVadivel (2013). Thus, Hypothesis 11b was supported.
Hypothesis 11 c and 11d displayed that age and income did not have a significant connection with job stress (p-value = 0.400; 0.762, respectively) ( Table 6). This result indicated that age and income positively affect job stress (beta coefficient = 0.026; 0.010, respectively), which means that higher income and older bankers tended to show a higher level of job stress. However, these connections were not statistically significant (p-value > 0.05), which means that there is a high potential that these connections may occur purely by chance. Thus, hypotheses 11 c and 11d were rejected.  recommended that the model's predictive power should be measured by the coefficient of determination (R 2 ), which ranges from 0 to 1. R 2 is the amount of variance independent constructs in the structural model explained by the independent variables connected to it. "The higher the R 2 coefficient, the better the construct is explained by the latent constructs in the structural model. The high R 2 coefficient also reveals that the values of the variables can be accurately predicted by the PLS path model" (Giao, Vuong, Huan et al., 2020). The R 2 value for the job performance was 0.404 (Figure2), which meant that 40.4% of the variation in job performance was explained by all independent variables such as social support, job stress, and job satisfaction. Besides, the R 2 value for the job stress was 0.390, which showed that 39% of the variation in job stress was explained by social support, locus of control, gender, marital status, educational level, and experience. Moreover, Giao, Vuong, Huan et al. (2020) suggested that "R 2 values and the effect for endogenous latent variables in behavioral sciences can be assessed as 0.26 (large effect), 0.13 (moderate effect), and 0.02 (weak effect)". Since R 2 values for the job performance and job stress were larger than 0.26, the model of this research demonstrated good model-data fit.

Discussion
This research proved that both social support and job stress play an important role in improving job-related behaviors (e.g., job satisfaction and job performance) and human resources management. On the basis of social exchange theory, this study investigated the effect of social support on job stress and job-related behaviors. The results are consistent with the past empirical findings of   -Duen et al. (2017), Yuh and Choi (2017), Shirey (2004), and Yu et al. (2014). In terms of P-E fit and P-O fit theory, the research demonstrated that lower job stress may significantly increase job satisfaction and job performance. Specifically, employees who suffer from a higher level of job stress obtain a lower level of job satisfaction and job performance. This finding was also comparable with the theory and prior studies of Darmody andSmyth (2016), (O. F. Malik et al., 2010), and Yozgat et al. (2013). Besides, in terms of social learning theory, the research explored LOC as a moderator of social support and job stress relationship. The negative impact of social support on job stress is stronger on employees with internal LOC than external LOC. In addition, this study also explored that LOC has a direct negative relationship with job stress. In other words, employees who have internal LOC obtain lower job stress. In general, the current study is an attempt to clarify the link between LOC, which is a new concept that has received little attention, and a social support-job stress relationship. In addition, the measurement scale has been applied in different industries and different countries, therefore, this study modified the scale to be suitable for Vietnam. The survey, which had been created in developed countries, had to be modified to be suitable for Vietnamese participants.

Conclusions
In conclusion, the literature on job stress has been conducted over the past several years and employers are now in need of its practical implications. This is because bank employees have become more and more stressed since the global banking sector has experienced rapid and significant changes. However, there has been little research on these issues with regards to the banking industry and the current study attempts to bridge this gap. From the background about the banking industry context in Vietnam and the lack of relevant information related to the banking industry in previous studies, it is perceived that human resource management has become fundamental in the sustainable development of the banking industry and needs to be focused on. This study was implemented to research the impact of social support on jobrelated behaviors (e.g., job satisfaction, job performance) through the mediating role of job stress and the moderating effect of locus of control. Besides, this research also considered the impact of demographic characteristics on job stress in the Vietnamese banking industry.
To achieve the above purposes, this study was conducted with references to prior research about the related topics. From the literature review, a hypotheses testing model was built to create the preconditions for this research. From the development of the questionnaire, the data collection method was experienced through two main phases. Firstly, a pilot study was used through qualitative research in order to ensure that the questionnaire was suitable for Vietnamese participants. Secondly, the main survey was carried out by quantitative analysis. There were 600 questionnaires sent to participants and 529 questionnaires were used as data for this study. Based on the findings, this study extended prior research with a better understanding of social support, job stress, job-related behaviors, and locus of control as well as the relationships of these concepts in the banking industry in a developing country. The findings of testing discovered correlations between a system of constructs. Specifically, social support positively influences job satisfaction and performance. Moreover, these relationships could be partially mediated by job stress. In other words, bankers who have significant social support could reduce their stress at the workplace and the decreased stress can lead to enhanced job satisfaction and performance. In addition, this study suggested a moderating mechanism of a locus of control on the relationship between social support and job stress. This negative connection was stronger for bankers who hold a higher internal locus of control. Finally, the findings showed that some demographic characteristics (e.g., gender, marital status, educational level, and experience) had a significant association with job stress. Particularly, married, female bank employees with low education and little experience tended to have high levels of job stress. After obtaining the results, this study discussed their connection with results form previous research and the context of the Vietnamese banking industry.

Managerial implications
This study has supplied for managers in Vietnamese banks a clear viewpoint of the social support, employees' job stress, job-related behaviors and locus of control. As mentioned in the research background, human resources has been an area of interest after the merging and restructuring in the Vietnamese banking industry. In particular, job stress and job-related behaviors (e.g., job satisfaction and performance) have been focused on in order to keep talented people and allow the organization to experience sustainable growth. Based on the results achieved to reduce job stress and improve job-related behaviors in the workplace, there are many issues to solve. Increasing social support is a necessary attribute to decrease job stress. For developing job satisfaction and performance of employees in Vietnamese banks, the managers need to restrict job stress associated with promoting social support. Therefore, the study can give some concrete and suitable recommendations to help the Vietnamese banks in developing human resource management as follows: Firstly, social support was a crucial component significantly influencing both job stress and jobrelated behaviors. Therefore, the most important obligation of managers in banks is intensifying social support of the organization to increase job-related behaviors and contemporarily decrease job stress. Therefore, managers should concentrate on organizing social activities for employees, i.e., team building, sports, art, and other community activities. In the context of Vietnam, managers in the banking sector also promote the role of trade unions and youth groups. By doing so, the relationships between employees at all levels can improve, which may lead to strong solidarity. Accordingly, it creates social support in work, every staff in the organization feels comfortable with each other, colleagues become good friends, and the communication is easier and more commodious. As a result, it helps to reduce employees' job stress and improve job-related behaviors.
Secondly, job stress also had an impact on job satisfaction. When employees do not understand their role carefully, they can be more stressed and less satisfied with their job. Managers need to eliminate the role of the ambiguity of employees. The job description is an important thing to prevent role ambiguity. With a clear and specific job description (e.g., work conditions, competency standard, difficulty, task functions, and position in the organization), employees are more focused on the goals and objectives of their job, their authority, responsibility, and linking to other tasks. Human resource managers are responsible to analyze these contents for each position every month, every quarter, and every year. These points enable employees to deal with all aspects of their work without conflicting with the development of the entire system. In addition, our country's economy is in a deep and large integration period; therefore, the elements of openness and transparency in finance, selection, and planning of human resources in Vietnamese banks are not evaluated highly. Therefore, it needs to be emphasized in order to create confidence and employee loyalty which can limit the job stress of each employee.
Thirdly, job stress influences job performance significantly and negatively when it increases, overwhelming the ability of the employee. At Vietnamese banks where job performance is in a slowdown, managers need to be concerned about reducing the job stress of their employees. They should seek out ways to keep job stress at a certain level. This level is enough for employees to show all of their abilities, but not exceeding their competences. In order to do this, managers need to demarcate regulations clearly about functions and the responsibility of each position. Again, job description plays an important role. Managers should intensify concern on the job description of each position. This focus can create an opportunity for staff performing their abilities in more than one role, but those roles should be concrete, clear, and suitable to their abilities. These roles should not overlap in terms of responsibility, especially working in a team. All people in an organization should build targets consistently. By doing so, managers can avoid a mismatch between the abilities of employees and requirements of the job, employees can receive compatible requests from different people without experiencing job stress. Moreover, managers should attempt to carefully recruit more qualified people who fit the performance standards. It should be very objective, impartial, not bound by the relationship in recruitment and assignment of tasks, especially in the Vietnamese context. Every manager should enhance risk management skills; purification, enhance their professional qualifications, professional ethics qualities, ability to work and dedication to the staff. Besides that, if banks are fiscally sound, they can invest in more technology to reduce the workload for employees.
Fourthly, locus of control impacted significantly on job stress and it also moderated the relationship between social support and job stress. An employer may take this variable into account to hire the right person. From the results, current research suggests that bank managers should conduct a survey to identify whether their staff is an internal or external locus of control rather than strictly selecting those candidates with a pre-existing internal locus of control, for example, this study suggested that there may be some sort of individual differences that may cause employees to respond to similar situations differently. If that is the case, human resources departments may wish to screen candidates based on individual character differences which could be identified in later research. It further suggests that at least some percentage of their current employees can be trained to feel a stronger sense of personal control and to have a more internal locus of control. This could involve training designed to shift employees' locus of control to be internal.
Finally, this finding indicated that demographic characteristics would need to be considered within the everyday work in order to minimize job stress. The individuals will have an impact on the employee's ability to embrace an overall coping strategy of constructive thinking style. This study showed that gender had an impact on job stress. A banker's gender does have an effect on how well he/she proactively, conscientiously and with a positive mindset copes with job stressors. This positive attitude may even reach unrealistic and optimistic levels. The gender groups will hold different levels of private endorsement of superstitions as well as a tendency to see stressors and stress in terms of black and white. Management will have to adapt their coping strategies accordingly. Besides, married employees tended to have high job stress. The authors stated that information about marital status would evoke cultural assumptions in evaluators about gender division of work in families, where married mothers and fathers would be expected to prioritize caregiving and breadwinning. Thus, in order to reduce job stress, banks should also have policies to support work-life balance issues relative to married employees.
Additionally, a banker's experience has a correlation to the results of job stress. The original expectation of the relationship between experience and job stress had been that as the respondent experienced fewer experience segments, he/she would score higher on the job stress. This would be due to the increased range of emotions, experiences, and personal triumphs and failures the respondent had experienced. As a worker progresses through the various stages (experience segments) of life, the manner in which job created stress and stressors will impact his/her performance with a minimum of stress will change. Management must, therefore, be aware of the effect of experience and adjust its coping strategies accordingly. If the industry under consideration is one with high employee turnover (e.g., a cadre of fresh-faced college graduates who stay for only short periods of time), management, in this case, will also have to align its stress management techniques to the appropriate experience group. This may mean more hands-on supervision or additional mentoring or an overall adjustment of job performance expectations consistent with the experience group of the employees.
From the literature review, it was also expected that the higher the education level of the employee, the lower would be the ratings on job stress. It was implicitly anticipated that the highly educated employee would more likely embrace constructive thinking as an effective stress coping strategy. For management, this significant relationship tends to indicate that stress coping strategies need to be amended for educational level prerequisites of particular positions. There would be a need for coping strategies to be revised when new job-related stress or stressors are experienced by employees with higher levels of education that have resulted in promotion to higher positions in the organization.

Limitations and further research
This research has some limitations. Firstly, this study assumed that employees in the same bank type have the same level of job stress and job-related behaviors regardless of the level of their position (clerk, officer, supervisor, manager). Therefore, the level of the position should be examined in future research. Results will make it possible to find more effective solutions for different position levels. Secondly, although the banking sector was divided into only two categories: state-owned and private banks, this study didn't compare the level of job stress on these two categories of the bank. There have been more and more private banks established in recent years, which makes the banking sector competitive. Therefore, private banks usually set a higher target and higher pressure on their employees. This can be another reason why private bank employees experience a higher level of stress than state-owned banks. Future research should compare two types of banks and provide empirical evidence about this difference. In addition, the next limitation is that this study researches the impact of just social support on job stress and job-related behaviors. Therefore, further research can add other new factors such as empowerment, commitment, work-life balance, job enrichment, and work treatment. Finally, the opinions of participants can be biased due the convenient sampling method. From this limitation, further research may use other sampling methods instead of a convenient method.