The impact of human resource management practices on service-oriented organizational citizenship behaviors: does positive psychological capital matter?

Abstract The aim of this study is to explore the relationship between perceived human resource management practices and service-oriented organizational citizenship behaviors through the mediating role of employee engagement. Besides, the moderating effect of positive psychological capital on the relationships will also be examined. The target population included 457 frontline employees who work at least one year at commercial banks in Vietnam. The inferential analysis to test the hypotheses was completed using SmartPLS 3.0 software for structural equation modeling with partial least squares (PLS-SEM). The result of this study indicated that a positive correlation between perceived HRM practices and service-oriented OCBs; employee engagement played a partial mediating role in this correlation. In addition, positive psychological capital heightened the positive relationship between perceived HRM practices and service-oriented OCBs of frontline employees. This research provided bank managers with an enhanced understanding of what the fundamental behaviors of their frontline employees are and how to develop and cultivate these employee contributions. It lets managers know how much the frontline employees take care of HRM practices at the workplace. From the point of view of the employees, managers may have the strategies to avoid the potential risk related to human resource management.


PUBLIC INTEREST STATEMENT
This study has some significant contributions to the literature. First, it shed light on the nature of the association between employees' perceptions of HRM practices and service-oriented OCBs. In other words, HRM practices played a vital role in the understanding of service-oriented OCBs. Besides, previous researches mostly emphasize on investigating the connection between HRM practices and service-oriented OCBs in terms of service sectors such as the hotel industry, local government, education, and transportation. However, it has yet to be verified in the banking industry, or only has little empirical evidence regarding this relationship. So, the findings of this study filled up the gap. Second, this research revealed employee engagement as a mediating mechanism that accounts for the connection between perceived HRM practices and serviceoriented OCBs. Finally, this finding also pointed to the importance of positive psychological capital which moderated this connection.

Introduction
In recent years, Vietnam has been considered to be a dynamic economy in the ASEAN region and the world (Worldbank, 2022). Investors are attracted to Vietnam due to its largely untapped market with a population of over 97 million people (Worldpopulationreview, 2022), strategic geography position, stable political conditions, and high economic growth. Over the last five years, the changing market economy has brought Vietnam to average growth of six percent (6%) annually. Normally, the country aims to gain economic growth from seven to eight percent (7-8%) to bring Vietnam out of its position as a developing country.
One of the sectors in Vietnam exposed to a rapid change in the banking sector. It carries out the bridge for capital mobilization and distribution, satisfies the capital demand for the economy. Giao et al. (2020) stated that a healthy and stable banking and financial system will be the driving force for economic growth and development, stimulating industrialization, standardizing human life, and reducing poverty. The banking industry in Viet Nam has experienced many significant changes over the past two decades. Previously, the market share was predominantly owned by state-owned banks. However, nowadays restructuring, integration, internationalization trends, and technological progress, have contributed to changing the competition in the Vietnamese banking sector. Particularly by accessing the World Trade Organization (WTO) in 2007, the restriction in the financial and banking sector for foreign institutions gradually removed. The integration and liberation financial sector brings many opportunities but also many challenges for the Vietnamese banking system. The market entry of foreign banks not only increases the level of competition but also increases the effectiveness and safety of the banking industry. Domestic banks have opportunities to access the capital, technology, and management experience from international credit institutions. Although local banks have advantages of operating history and branch network, the weakness in management, technology, and service quality make the local banks cope with severe competition from foreign banks. Besides, the structural changes made more and more banks established in Viet Nam. As a consequence, the rapid development of the banking system in a short period has caused ineffectiveness in the banking sector. Thus, Decision No. 254 was approved by the State Bank of Vietnam (SBV) to restructure the credit institution system, and merger and acquisition (M&A) are one of the key solutions to improve the performance of Vietnamese's banking system.
Up to now, the credit institution system in Vietnam has been restructured and developed with various ownership types. As stated on the State bank of Vietnam's website current, there are a total of 93 banks in Vietnam include 9 wholly-owned foreign banks, 31 private commercial banks, 49 branches of foreign banks, and 4 joint-venture banks (SBV, 2022). Further, five of 100% of foreign banks have been allowed to set up a business in Vietnam. These banks are Standard Chartered, ANZ, HSBC, Shinhan, and Hong Leong (SBV, 2022). Consequently, the competition is increased not only commercial banks of Vietnam but also foreign banks regarding the operation, liquidity collection, and income generation. According to Vuong and Suntrayuth (2020), services of the banks are similar, easy availability and there is no significant difference. Thus, customers are in a strong bargaining position. As banks are pure service providing organizations improved service quality has been the crucial issue for finding and attracting new customers as well as look down on retaining satisfying the current customers. Long-term customers are not only a vital source of profit but also a free channel of advertisement for the business. Cultivating loyal customers can lead to increased sales and customer share, lower costs, higher prices as well as organizational performance. Kheng et al. (2010) stated that if commercial banks want to retain customer loyalty and succeed in the competitive market, they need to improve the unique quality of their services. The quality of services may become a strategic weapon that can help banks overcome rivals in order to capture the market. Sultana and Johari (2017) stated that banks provide services to customers at the branch level. In there, frontline employees have voice-to-voice and face-to-face contact with customers. Thus, frontline employees play a very important role in the performance of the bank as they do the most interacting with customers and receive customers' responses about the bank's products or services (e.g., foreign exchanging, loans, balance checking, account opening, credit cards) then contribute significantly to the customers' access to the services of the bank. The importance of frontline employees' interaction with customers is considered to be a psychological phenomenon that exerts a major impact on outcomes. Besides, frontline employees are often referred to as part-time marketers (Gummesson, 1987). They operate on the boundary of firms providing an instrumental link between the organization and its customers. Their influence on service effectiveness is disproportionate to their level in the organization's traditional hierarchy. In their role as part-time marketers and boundary-spanners, frontline employees can (1) enhance the image of the bank, (2) provide customer service, and (3) contribute to the development of the bank. In industries such as banking, insurance, and retailing, frontline employees are often the source of differentiation and competitive advantage for the firm (Bienstock et al., 2003). Thus, business is increasingly relying on the positive attitudes and service-oriented behaviors of these frontline employees for its success with customers.
According to Suan and Nasurdin (2014), "for ensuring high quality of service one of the recognized positive behavior of employees is service-oriented organizational citizenship behaviors (later denoted as service-oriented OCBs)". Service-oriented OCBs are defined as extra-role service performance that demonstrates discretionary behaviors to promote the effective function of service establishments (Bettencourt et al., 2001). It refers to behaviors of frontline employees that reflect loyalty (e.g., promoting the firm and its products and services), service delivery (e.g., responsible role performance in customer service), and participation in the governance and development of the service orientation of the firm (e.g., making service improvement suggestions). Sun et al. (2007) observed that the nature of services is intangible, with customer perceptions subject to indirect and sometimes irrelevant cues when assessing the service experience, note service-oriented OCBs' potential to enhance the customer's overall perception of the experience. The fact that production and consumption of service occur simultaneously also emphasizes the importance of service-oriented OCBs, as Chan et al. (2017) aptly note that "the customer experience is as important as, if not more important than, the consumer good" being delivered. Performance in a service setting can thus be defined as a function of the customer experience, so it is important to understand how service-oriented OCBs can be promoted in order to create the best customer experience possible. Moreover, some researchers (e.g., Philip M. Podsakoff et al., 2000;Sultana & Johari, 2017) stated that service-oriented OCBs improve the social relationship and social network, and lead to increase the organizational performance. They also argued that these behaviors can enhance the employees' interaction with colleagues, managers, customers, and display the extra-miles to draw customers' attention. According to Coyne and Ong (2007), as frontline employees exhibit organizational citizenship behavior during service delivery, it helps customers will be gratified, delighted and retain loyalty towards the bank. Consequently, it will decrease customer complaints, reduce the absenteeism and withdrawal behaviors of staff in the bank as well.
For promoting employees' service-oriented OCBs, numerous researches have begun to identify the antecedents of SO-OCBs from different perspectives (e.g., Alfes et al., 2013;Suan & Nasurdin, 2014;Sultana & Johari, 2017). A large number of empirical researches on the antecedents of service-oriented OCBs have focused on four major categories, including the firm-level variables, such as perceived organizational support (e.g., Bettencourt et al., 2001;Snape & Redman, 2010;Wang, 2009), human resource management practices (e.g., Cesário & Magalhães, 2017;Pohl et al., 2017), organizational justice (e.g., Aponno et al., 2017); employee-level variables, such as dispositions (e.g., Bettencourt et al., 2001), attitudes (e.g., Chen & Kao, 2012), personal abilities (e.g., Payne & Webber, 2006);leadership behaviors (e.g., Aponno et al., 2017); and task characteristics (e.g., Philip M. Podsakoff et al., 2000). Although there are many antecedents of serviceoriented OCBs identified, Sultana and Johari (2017) emphasized that human resource management (HRM) practices play the most important role in promoting employees' service-oriented OCBs. Snape and Redman (2010) also stated that "the function of HRM practices is to increase the outcome of the employee through offering intrinsic motivation as well as opportunity to accomplish through greater levels of perceived job effect and discretion". Sun et al. (2007) promoted that perspective as well, noting that HRM practices lead to higher levels of service-oriented OCBs, as reflected in employees' true willingness to be involved with the organization and its goals. Therefore, HRM practices have been seen to reinforce and promote service-oriented OCBs, firms should proactively invest in human capital because it is necessary to enhance employees' serviceoriented OCBs.
On the other hand, Alfes et al. (2013) recommended that further studies should be needed to discover the mechanisms through which HRM practices influence employees' organizational citizenship behaviors and organizational performance. Although previous studies (e.g., Aktar & Pangil, 2018;Pohl et al., 2017;Turek & Wojtczuk-Turek, 2015) have investigated job satisfaction, commitment, person organization-fit, and perceived organizational support as mediating variables in the relationship between HRM practices and employees' organizational citizenship behaviors, the results of these studies have been mixed. In this study, the author proposes an alternative mediator, which is known as employee engagement. Some researchers (Alagarsamy et al., 2020;Alfes et al., 2013;Kataria et al., 2012;Sugianingrat et al., 2019; recommended that employee engagement may be a core underlying mechanism in the workplace that explains a wide range of attitudinal and behavioral outcomes. Besides, it has become generally accepted as an important indicator of business performance (Aktar & Pangil, 2018). Bakker et al. (2011) posited that organizations need employees to be engaged, to invest themselves in the organization, committed to high performance, dedicated, and connected to their work. This level of engagement has resulted in employee engagement being viewed as a competitive advantage for firms because engaged workforces increase extrarole behaviors, productivity, profits, and retention (Sugianingrat et al., 2019). Although engagement is beneficial to organizations and the employees' psychological affective state, there is no published research to date has examined employee engagement as a mediating variable involving perceived HRM practices to employees' service-oriented OCBs and organizational performance in the banking sector. Thus, this study will contribute significantly to the present body of extant literature that focuses on the link between HRM practices and service-oriented OCBs through the mediating role of employee engagement in the Vietnamese banking industry.
Another gap is in this research is that how personal differences have an effect on the relationships between HR practices-employee engagement, and HRM practices-service-oriented OCBs. This study is concentrated on positive psychological capital as a personal resource. The author tries to fill these gaps in the current research by examining whether positive psychological capital moderates the HRM practices-employee engagement relationship and HRM practices-service-oriented OCBs. To researcher knowledge, no other study has investigated the moderating effect of positive psychological capital on these relationships in the banking sector.
In summary, given the vital role of HRM practices in promoting organizational performance, it is surprising that little attempt has been taken yet to investigate the role of HRM practices in improving organizational performance. Moreover, most of the researches about the relationship between HRM practices, employee engagement, and service-oriented OCBs was conducted in developed countries (Wang, 2009) and only focused on hotels and restaurants (Suan & Nasurdin, 2014). However, there were not many studies about these variables in the banking sector in developing countries such as Vietnam. Therefore, this research proposes and tests a framework that associates perceptions of HRM practices with service-oriented OCBs. Besides, the author proposes that employee engagement is a mediating mechanism that triggers this relationship. Finally, the author examines whether the relationships between HRM practices-employee engagement and HRM practices-service-oriented OCBs under consideration are moderated by positive psychological capital. Hussain and Rehman (2013) posited including HRM practices in an organization is significant to retaining employees of high value and critical to the success of the organization. HRM practices provide a satisfying relationship between human capital and business owners. When employees feel their professional interests align with HRM practices positive efforts will follow. In the business environment, the importance of HRM practices involves support for building trust in the business leaders from the employees, which equates to high-quality organizational performance (Cho & Poister, 2013). According to Kosiorek and Szczepańska (2016), the perception of HRMP varies depending upon the culture of an organization, structure, and human capital. The success of an organization presently reflects human resource management practices because it requires perceptual changes to govern activities, develop training, recruit employees, and determine performance parameters. Other researchers state recruitment and selection, training, performance appraisal, reward and recognition, career advancement, employee participation, and information sharing as the best human resource management practices as well as key areas of development for optimal human resource management practices (Vuong & Suntrayuth, 2020). Smith et al. (1983) defined organizational citizenship behaviors (OCBs) as actions that are not required on the job. The term service-oriented OCBs (SO-OCBs) was later established by Bettencourt et al. (2001) to describe "discretionary behaviors of contact employees in servicing customers that extend beyond formal role requirements". Bettencourt et al. (2001) settled a typology of serviceoriented OCBs that included three dimensions: loyalty SO-OCB, participation SO-OCB, and service delivery SO-OCB. First, loyalty SO-OCB is exhibited by "service providers acting as advocates for the organization, from developing customer relationships to promoting the organization's objectives and image, whether in favorable or adverse conditions". Consequently, it is crucial for employees to engage in loyal citizenship behaviors. Second, participation SO-OCB refers to "a voluntary effort aimed at improving service delivery, especially customer contact, from an individual level to coworkers and ultimately to the organization as a whole". Employees serve as a critical bridge between the internal operations and the external customers of the company. Employees can convey information based on customer comments to management and propose significant suggestions for improvement of services. Thus, participation SO-OCB plays a vital role in responding to the changing needs of its customers. Finally, service delivery SO-OCB mentions "individual initiative, with customer contact employees acting in a conscientious manner to enhance customers' experience with the organization". Employees should provide customers with reliable, responsive, and courteous service. These behaviors are one of the key factors that affect customer satisfaction. Therefore, it is important for employees to perform service delivery SO-OCB for the success of service organizations (Bettencourt et al., 2001). Some previous researchers (e.g., Alfes et al., 2013;Cesário & Magalhães, 2017;Pohl et al., 2017) also stated that HRM practices are a trigger to improve certain behaviors in employees, and may contribute to the emergence of service-oriented OCBs. The relationship between HRM practices and service-oriented OCBs is currently explained by the social exchange theory (SET). When employees receive benefits (good HRM practices) from their organization, employees will feel obligated to respond in kind and want to give something in return to their organization through positive attitudes and behavior. Moreover, Pohl et al. (2017) confirmed that HRM practices are a vital predictor of service-oriented OCBs. Thus, service-oriented organizations should promote HRM practices to inspire frontline employees' service-oriented OCBs for the organization's success. According to the above discussions, the first hypothesis is posited: H 1 : HRM practices have a positive impact on service-oriented OCBs in the banking industry. Kahn (1990) defined that employee engagement is staffs who are engaged in their work tend to invest their physical, cognitive, and emotional resources to perform their tasks well, while those who are not engaged in their role performance are not willing to devote themselves to their work emotionally, cognitively, and physically. In other words, an engaged employee invests the hand, head, & heart in his or her work within an organization. Schaufeli et al. (2002) conducted a study identifying three dimensions of employee engagement: vigor, dedication, and absorption, which are core dimensions of engagement. Vigor mentions "high levels of energy and mental resilience while working, the willingness to invest efforts in one's work, and persistence even in the face of difficulties". Dedication involves "a sense of significance, enthusiasm, inspiration, pride, and challenge". Absorption mentions "being fully concentrated and deeply engrossed in one's work, whereby time passes quickly and one has difficulties with detaching oneself from work".

HRM practices and employee engagement
Aktar and Pangil (2018) stated that HRM practices are one of the most dominant concepts claiming to have robust impacts on employees' attitudes, behavior, and performance in academic literature. HRM practices are the process or the basic activities of managing employees who utilize other resources to do their tasks successfully and accomplish the organizational goals (Suan & Nasurdin, 2014). Proper HRM practices could create healthy psychological well-being for employees that may motivate them to achieve organizational objectives. Besides, HRM practices could influence individual employees' knowledge and skills, their opportunities to express their talent, and their willingness to exert effort in their work (Alfes et al., 2013). It's likely that as employees perceive effective HRM practices, they would exhibit a positive attitude at work such as being happy, energetic, and dedicated. Some empirical studies (e.g., Alfes et al., 2013;Huang et al., 2018) confirmed that HRM practices had a positive association with employee engagement in service organizations. Therefore, the second hypothesis is proposed.  Kataria et al. (2012) stated that employee engagement is an active psychological state of employees' which is often connected and described in congruence with OCBs. Uddin and Akther (2016) found that employee engagement leads to proactive behavior. Employees who are engaged in their tasks become more involved in discretionary behaviors, and they could increase the occurrence of serviceoriented OCBs. Sugianingrat et al. (2019) also believed that as engaged employees are more dedicated, absorbed, and highly involved in their tasks, they will be more likely to go the extra mile beyond the formal job requirements. Moreover, some empirical evidence (e.g., Sugianingrat et al., 2019;Uddin & Akther, 2016) indicated that employee engagement could affect OCBs. Since the general idea of service-oriented OCBs derives from OCBs, such empirical evidence helps to explain the role of employee engagement in increasing service-oriented OCBs. However, despite the widespread knowledge on the concept of OCBs and the empirical studies, there has been no research on the relationship between EE and service-oriented OCBs in the banking industry done before. Therefore, on the basis of previous empirical evidence, the author developed the following hypothesis:

Employee engagement and service-oriented OCBs
H 3 : Employee engagement has a positive impact on service-oriented OCBs in the banking industry.

The mediating role of employee engagement
The preponderance of research suggests OCBs are positive actions, which result in advantageous outcomes for organizations and employees. OCBs created positive outcomes for the employee in the form of good employee performance evaluations (Sugianingrat et al., 2019). However, a true analysis of the literature yielded findings of potential negative outcomes that OCBs could have on an employee or the organization (Bergeron, 2007). He argued that individuals who engage in behaviors that are not defined by their job role are not concentrating their efforts on actual task performance. This behavior would take attention and resources away from the specific job that an employee is hired to do. If an individual devotes their time to helping others do their jobs, their own work could suffer. Smith et al. (1983) stated employees may have to decrease their work productivity to assist others with their own. Besides, the traditional definition of OCBs defines the construct as engaging in behaviors outside of one's prescribed role or going above and beyond the call of duty to assist others or the organization (e.g., Alagarsamy et al., 2020;Rai & Maheshwari, 2021). Despite an organization's reliance on these behaviors, employees will still be measured on the tasks that are core to their position. Hence, an employee should engage in their tasks before they try to exhibit service-oriented OCBs. Based on aforementioned discussions, in order to avoid these negative outcomes, this study added to the body of literature on OCBs by suggesting that employee engagement would mediate the relationship between HRM practices and service-oriented OCBs to yield beneficial outcomes.
Moreover, previous researches showed that the relationships between HRM practices and employee engagement (e.g., Alfes et al., 2013;Huang et al., 2018); HRM practices and serviceoriented OCBs (e.g., Alfes et al., 2013;Cesário & Magalhães, 2017;Pohl et al., 2017), and employee engagement and service-oriented OCBs (e.g., Sugianingrat et al., 2019;Uddin & Akther, 2016) existed. Based on linking the relationships outlined above, the author contends that there is a possibility that employee engagement mediates the associations between perceived HRM practices and service-oriented OCBs. Thus, it is hypothesized that: H 4 : Employee engagement mediates the relationship between HRM practices and service-oriented OCBs in the banking industry.

The moderating role of positive psychological capital
Positive psychological capital (PsyCap) is "an individual's positive psychological state of development" (Luthans et al., 2015). PsyCap is specifically defined as having the confidence to succeed in any difficult tasks by giving the required effort, being optimistic about today's success and the future, staying hopeful in order to succeed, and when facing difficulty and able to bounce back (Luthans et al., 2015). PsyCap consists of self-efficacy, resilience, optimism, and hope that is grounded in theory and research and each of them has positive impacts on attitudes, behaviors, and performance. In particular, Self-efficacy involves "have the confidence to take on and put in the necessary effort to succeed at challenging tasks"; Resilience mentions that "when beset by problems and adversity, sustaining and bouncing back even beyond to attain success"; Optimism implies to "make a positive attribution about succeeding now and in the future"; hope refers to "persevere towards goals and, when necessary, redirecting paths to goals to succeed" (Luthans et al., 2015). Yildiz (2019) stated that positive psychological capital is a composite construct of employees' positive virtues and strengths. Employees who reflect positive states in their organizations can enhance positive outputs better than others. Thus, employees with high positive psychological capital are more engaged with their task, because they are more hopeful, confident, optimistic, and resilient. Luthans et al. (2015) also argued that when employees having high positive psychological capital may "perceive what they are now by understanding their potentials and strengths; these employees can also foresee what they are capable of becoming". So, positive psychological capital could encourage employees to elicit extra-role behavior. Additionally, when organizations provide effective HRM practices, this would foster employees to meet their goals. In turn, employees would become more dedicated to their tasks and exhibited extra-role behavior, because they derive fulfillment from it (Xanthopoulou et al., 2007). Thus, the availability of effective HRM practices could lead to employee engagement (Alfes et al., 2013;Huang et al., 2018) and serviceoriented OCBs (Alfes et al., 2013;Cesário & Magalhães, 2017;Pohl et al., 2017). Although previous studies emphasize the relationship between "HRM practices and employee engagement" and "HRM practices and service-oriented OCB", the probability that the relationship can vary with a moderating variable has rarely been considered. Some scholars (e.g., Aybas & Acar, 2017; Shaheen et al., 2016) recommended that high levels of HRM practices could exert high levels of employee engagement and service-oriented OCB because of positive psychological capital factor. This factor may buffer the relationship between "HRM practices and employee engagement" and "HRM practices and service-oriented OCB". They also argued that a moderating variable might play an essential role in buffering effect to modify these relationships. Moreover, Aybas and Acar (2017) found that positive psychological capital moderated the association between HRM practices and employee engagement. Also, Shaheen et al. (2016) demonstrated the moderating role of positive psychological capital in the relationship between organizational support and organizational citizenship behaviors. Based on prior researches, in this study, the author proposes that positive psychological capital could heighten the relationships between "HRM practices and employee engagement" and "HRM practices and service-oriented OCB". Therefore, the following hypotheses can be detailed (Figure 1):  Figure 1. The conceptual model.   Bello et al. (2018) suggested that demographic variables could affect on OCBs. Nimran (2011) identified the relationship between individual characteristics and extra-role behaviors (i.e., OCBs). Besides, Sharma et al. (2017) also found that demographic factors impacted on employee engagement. However, in contrast, some scholars initiated that there was no relationship between demographic variables and organizational citizenship behaviors (e.g., Philip M. Podsakoff et al., 2000;Uzonwanne, 2014), and employee engagement (George & Venkatapathy, 2018). Based on the aforementioned discussion, the hypotheses are proposed: H 7 : There is an association between control variables (e.g., gender, marital status, age, income, experience, and education level) and service-oriented OCBs in the industry.

Demographic variables
H 8 : There is an association between control variables (e.g., gender, marital status, age, income, experience, and education level) and employee engagement in the banking industry.

Procedure and sample size
The target population of this study only focuses on all frontline employees who work for 31 private commercial banks in Vietnam. A questionnaire will be distributed to the respondents via a paper version by using convenience sampling. The authors contacted these banks to collect data of the person in charge-human resource manager.
All the constructs in the model were measured with multiple items, which were developed by other researchers (e.g., Cesario, 2015;Delery & Doty, 1996;Gillani et al., 2018;Karatepe et al., 2007;Vuong & Suntrayuth, 2020;Wilches-Alzate, 2009) and also based on the results of a qualitative study to adequately capture the domain of the constructs. All the item of scales was transferred to the questionnaire. After that, the questionnaire in these surveys was translated and back-translated into Vietnamese in line with cross-cultural translation procedures. Committee approaches were used for translation. Committee or parallel translation involved several translators who made independent translations of the same questionnaire. At a reconciliation (consensus, revision) meeting, translators compared the translations, reconciled discrepancies, and agreed on a final version that taped the best of the independent translations or appeared in the course of discussion. In this study, the author used the committee of two translators.
After completing the translation of the questionnaire, the author conducted a pretest by a focus group with ten people included five managers in the banking management field and five frontline employees of the banking sector. Participants of this study would not be interviewed in the main study in the next step. The detailed questionnaire was shown to the respondents for checking their understanding. During the focus group, the author also wanted to find out if the chosen measurement scale was suitable for conducting the research in Viet Nam. The comments related to the translation of the questionnaire were mainly suggestions concerning words and phrases that make the translations better and less sensitive so that respondents could give a rating that best fit with their reality. Therefore, all the comments from the focus group were gathered with the aim to modify the measurement scale. After the revision finished, the author had the final conceptual model and draft questionnaire.
Concerning the quantitative research method, there are two phases in this step: quantitative pilot survey and main survey. The authors used the final draft questionnaire completed in the above qualitative research to carry out face-to-face interviews with the subjects of the study. The quantitative pilot survey aimed to test how the respondents understood not only the content of questionnaires but also the interviewees' attitudes toward the questionnaire. The author interviewed 50 people who frontline employees and managers in Vietnamese banks in Ho Chi Minh City. Based on the data collected, the evaluation of the questionnaire was processed (e.g., did the subjects of the study understand the questions correctly? Were they willing to provide information?). The reliability of the measurement scales was preliminarily tested with collected data using Cronbach's Alpha (Table 1). Hair et al. (2017) stated that the scales were found to be reliable due to the minimum Cronbach's Alpha of each factor equals 0.70. Table 1 showed the value of Cronbach's Alpha coefficients of all constructs were fluctuating from 0.790 to 0.924 (e.g., Loyalty SO-OCB, employee engagement). In general, all observed variables of the above scales had a Cronbach's Alpha coefficient was more than 0.7. Therefore, these scales were reliable to be used for collecting data in the main survey.
After the revisions of the pilot survey were completed, 500 questionnaires were distributed directly to frontline employees in Vietnamese banks. From the 500 questionnaires, 493 responses were collected (98.6% return), but there were only 457 respondents were eligible for the data analysis process. 36 respondents were excluded because they did not satisfy the inclusion criteria (e.g., they did not complete all the items in the survey). Table 2 summarized the output from the response data using IBM SPSS

Measurement instrument
The questionnaire included two parts: part one consists of measurement scales using a five-point Likerttype scale to ask respondents to choose one item that ranged from (1) = "strong disagree" to (5) = "strongly agree"; part two was some general demographic information of the respondent. All the constructs in this study were measured with multiple items. Human resource management practices scale was measured by seven dimensions with 33 items of prior authors (e.g., Cesario, 2015;Delery & Doty, 1996;Gillani et al., 2018;Karatepe et al., 2007;Vuong & Suntrayuth, 2020;Wilches-Alzate, 2009). Sample items included (Recruitment and selection: e.g., "The recruitment and selection processes in this organization are impartial"); (Training: e.g., "Employees in this organization receive continued training to provide good service"); (Performance appraisal: e.g., "Performance appraisals are administered in an equitable manner among all employees"); (Reward and recognition: e.g., "My superiors generally appreciate the way I do my job"); (Career advancement: e.g., "Individuals in this job have clear career paths within the organization"); (Employee participation: e.g., "Employees in this job are allowed to make many decisions"; (Information sharing: e.g., "Information I receive through the internal communication channels helps me to perform better my job").
Employee engagement scale was measured by three dimensions with seventeen items of Schaufeli et al. (2002). Sample items included (Vigor: e.g., "When I get up in the morning, I feel like going to work"); (Dedication: e.g., "I am enthusiastic about my job"); (Absorption: e.g., "Time flies when I am working").
Service-oriented organizational citizenship behaviors were measured by three dimensions with sixteen items of Bettencourt et al. (2001). Sample items included (Loyalty: e.g., "Encourages friends and family to use firm's products and services"); (Participation: e.g., "Encourages co-workers to contribute ideas and suggestions for service improvement"); (Service delivery: e.g., "Follows customer service guidelines with extreme care").
Positive psychological capital was measured by eight items of Luthans et al. (2015). Sample items included (e.g., "I am confident in representing my work area in meetings with management").

Partial least squares regression
SmartPLS 3.0 software was employed to run the hypothesis testing for the research. The influences of perceived HRM practices on employee engagement and service-oriented OCBs, and the moderating role of positive psychological capital simultaneously, and not in isolation. So, to simulate reality, this study was addressed by applying the structural equation modeling, based on the method of partial least squares (PLS), which bring all the effects of variables together into one model . Moreover, Hair et al. (2017) also stated that PLS-SEM can operate effectively with nominal, ordinal, and interval level variables, and does not assume that the variables are normally distributed. Some scholars (e.g., Chin et al., 2012;Jordan & Troth, 2019) highlighted that self-reported data could lead to common-method bias. To address this problem, this paper followed the guidelines of P. M. Podsakoff et al. (2003) and employed Harman's single-factor test. They described that "if a single factor explains less than 50% of the covariance, it indicates that common-method bias is not affecting results". The results of single exploratory factor analysis showed that the single factor explained below 50% of the total variance, which established that the model is free from common method bias.

Assessment of the measurement model
A valid and reliable measurement model measured through objective means was necessary to demonstrate quality in an operational definition of the concept. Validity is concerned with a measure of goodness of fit of the operational definition and the concept (Hair et al., 2017). Reliability measures the stability and consistency of the measures. Hair et al. (2017) stated that Cronbach's Alpha value of each latent variable should be more than 0.7 to indicate the internal consistency of the items. In this study, the Cronbach's Alpha coefficient for each construct was greater than 0.7 (Table 3). For example, the Cronbach's Alpha coefficient for recruitment and selection = 0.810; training = 0.866; performance appraisal = 0.877; reward and recognition = 0.837; career advancement = 0.830, employee participation = 0.855, information sharing = 0.813; vigor = 0.851; dedication = 0.788; absorption = 0.841; loyalty SO-OCB = 0.877; participation SO-OCB = 0.889; service delivery SO-OCB = 0.872, and positive psychological capital = 0.879. According to Vuong and Giao (2020) also suggests using composite reliability as a suitable replacement for Cronbach's alpha because it is too conservative in PLS-SEM. Composite Reliability (CR) scores were reported for each of the variables. According to Hair et al. (2017), the composite reliability estimates the extent to which a set of latent construct indicators share in their measurement of a construct and that composite reliability should be equal to or greater than 0.7. In this model, the reliability score for each relationship was greater than 0.7 (Table 3) In addition, Vuong and Giao (2020) suggested that the reliability coefficient (rho_A) is the most important reliability measure for PLS-SEM and they also stated that the rho_A value should be greater than 0.7. In this model, the rho_A value for each construct was greater than 0.7 (Table 3). For, rho_A value for recruitment and selection = 0.820; training = 0.870; performance appraisal = 0.880; reward and recognition = 0.854; career advancement = 0.834, employee participation = 0.856, information sharing = 0.827; vigor = 0.864; dedication = 0.815; absorption = 0.865; loyalty SO-OCB = 0.879; participation SO-OCB = 0.890; service delivery SO-OCB = 0.873, and positive psychological capital = 0.896. Therefore, the authors confirm that all of variables in this study achieved a high reliability.
In PLS approach, as a part of convergent validity test, the items outer loading must be above the thumb rule of 0.5 (Hair et al., 2017) to indicate a convergent validity for constructs. In Table 3 above, all outer loading values were acceptable and above minimum 0.5. For example, the minimum indicator of outer loading values for perceived HRM practices = 0.505. Besides, Fornell and Larcker (1981) suggested the examination of the average variance extracted (AVE) to determine convergent validity. According to Hair et al. (2017), the prevailing measure of convergent validity is an AVE above 0.5 threshold is highly recommended. In this study, Table 3 presented AVE values were than 0.5 for all constructs tested. For example, AVE for recruitment and selection = 0.574; training = 0.651; performance appraisal = 0.670; reward and recognition = 0.612; career advancement = 0.663, employee participation = 0.697, information sharing = 0.581; vigor = 0.581; dedication = 0.548; absorption = 0.567; loyalty SO-OCB = 0.672; participation SO-OCB = 0.693; service delivery SO-OCB = 0.609, and positive psychological capital = 0.544. Thus, convergent validity was considered acceptable. The square root of AVE for each construct was computed to test for discriminant validity. Fornell and Larcker (1981) proposed that "the square root of AVE should be higher than its correlations with all other factors in the model". The research confirmed discriminant validity because the square root of AVE scores (in bold) is higher than the off-diagonal correlations in the corresponding rows and columns. For example, the variable "HRM practices" has an AVE of 0.511 and the square root of the AVE for "HRM practices" in Table 3 above (Fornell-Larcker Criterion) is 0.715. As stipulated 0.715 is both greater than the correlation values in its row (0.508 and 0.646) and its column (0.663; Table 4).
The measurement model was assessed the collinearity among the indicators using the variance inflation factor (VIF). A VIF value below 5 indicates the absence of collinearity (Hair et al., 2017). As seen in Table 5, the maximum of Outer VIF Values of constructs was 4.009 (e.g., positive psychological capital). It indicated that all of the VIF values were below 5. Therefore, no collinearity existed between the indicator variables and the latent variables.

Analysis of the structural model
The hypotheses in this study and represented in the structural model were analyzed with the SmartPLS 3.0 software package. The structural model was designed in the software tool and the raw data loaded. The resulting model in the SmartPLS tool is shown in Figure 2. Table 5 below showed a mixture of path coefficients (t-statistics) that are both higher and lower than 1.96. Those that are higher than 1.96 indicate that they exceed the minimum requirement and therefore, significant. The t-statistic is below the minimum requirement indicating that they are not significant. It is also noteworthy that all the t-statistic below the minimum also have p-values greater than 0.05.
H 1 : Perceived HRM practices have a positive impact on service-oriented OCBs in the banking industry. H 1 was significantly supported by the path HMRP → SO-OCB with a path coefficient of 0.328, and a p-value of 0.000 (p < 0.05). This result corresponded with many studies which suggested that HRM practices are one of the preliminary factors promoting frontline employees elicit service-oriented OCBs (e.g., Alfes et al., 2013;Cesário & Magalhães, 2017;Pohl et al., 2017). Frontline employees who perceived that HR practices in their organization are helpful and positive could be encouraged to promote service-oriented OCBs. HRM practices could be viewed as signaling intent for long-term investment in employees that obligates them to respond with discretionary role behavior.
H 2 : Perceived HRM practices have a positive impact on employee engagement in the banking industry. H 2 was significantly supported by the path HMRP → EE with a path coefficient of 0.694, and a p-value of 0.000 (p < 0.05). This result was similar to the prior researches of Alfes et al. (2013) and Huang et al. (2018). When organizations provide resources through HRM practices to their employees, the employees will be obliged and perform their work at a higher level of engagement.
H 3 : Employee engagement has a positive impact on service-oriented OCBs in the banking industry. H 3 was significantly supported by the path EE → SO-OCB with a path coefficient of 0.471, and a p-value of 0.000 (p < 0.05). This hypothesis was linked to the previous researches of Sugianingrat et al. (2019), and Uddin and Akther (2016). Frontline employees who are engaged in their tasks become more involved in discretionary behaviors, and they could increase the occurrence of service-oriented OCBs.
Hypothesis 4 predicted that employee engagement mediates the relationship between human resource management practices and service-oriented OCBs. This supposition was tested through several steps. First, the individual path coefficients of HRMP → EE and EE → SO-OCB were collected using the SmartPLS 3.0 bootstrapping path analysis algorithm. Table 6 showed that coefficients of HRMP → SO-OCB; HRMP → EE, and EE → SO-OCB were 0.694, and 0.471 respectively. Besides, the p-value for these paths was 0.000. Thus, the direct relationship between these constructs existed. Second, according to Table 7, the indirect path coefficient of HRMP→EE→ SO-OCB was 0.327 with its p-value of 0.000. It revealed that the indirect relationship between HRMP and SO-OCB through EE has also happened. Lowry and Gaskin (2014) suggested that when the direct and indirect relationships between A and B through C have occurred simultaneously, the researcher can conclude that C was a mediator of A and B. In other words, employee engagement mediated the relationship between human resource management practices and service-oriented OCBs. Moreover, the direct effect of HRMP on SO-OCB was still significant. Thus, employee engagement was a partial mediation on the relationship between human resource management practices and service-oriented OCBs. H 4 was supported.
Through the PLS-SEM model, the outcome of the study showed service-oriented OCBs were associated with the interaction of perceived HRM practices and positive psychological capital (Moderating effect 1 = HRMR*PPS) with a coefficient = 0.060 and p-value = 0.039 as presented in Table 8 and Figure 3. Moreover, there was a positive relationship between HRMP and SO-OCB. Thus, positive psychological capital positively moderates the relationship between HRMP and SO-OCB. In other words, positive psychological capital heightens the positive effect of perceived HRM practices on service-oriented OCBs. Therefore, hypothesis 5 was supported.
To better understand the interaction term, a median split was conducted for the positive psychological capital variable. There was a significant difference in service-oriented OCB between those with low positive psychological capital scores and those with high positive psychological capital scores. In other words, frontline employees with high "positive psychological capital" scores will exhibit higher service-oriented OCB when they get high human resource management practices than those with low "positive psychological capital" (Figure 4).
On the other hand, the result indicated that employee engagement was not associated with the interaction of human resource management practices and positive psychological capital (Moderating effect 2 = HRMR*PPS) with a coefficient = 0.034 and p-value = 0.261 as presented in Table 7. Thus, hypothesis 7 was not supported. In other words, there was no significant difference in employee engagement between frontline employees with low "positive psychological capital" scores and those with high "positive psychological capital" scores when they get high HRM practices.
H 7 : The results showed that gender, age, income, experience, and education didn't impact serviceoriented OCBs in the banking industry because the p-values of these variables were more than 0.05 (p-value = 0.887, 0.965, 0.534, 0.543, and 0.559, respectively; Table 3). However, marital status impacted service-oriented OCBs in the banking industry. This relationship was supported by the path Marital status → SO-OCB with a path coefficient of 0.065, and a p-value of 0.048 (p < 0.05; Table 3). In other words, married frontline employees had stronger service-oriented OCBs than single frontline employees.
H 8 : The findings indicated that age, income, and experience didn't influence employee engagement in the banking industry because the p-values of these variables were more than 0.05 (p-value = 0.130, 0.183, and 0.942, respectively; Table 3). However, gender, marital status, and educational level impacted employee engagement in the banking industry. These relationships were supported by the p-values of this variables were less than 0.05 (p-value = 0.018, 0.023, and 0.004, respectively; Table 3). According to Hofstede (2001), people with strong masculinity tend to place greater value on tasks, performance, and purposefulness, whereas people with strong femininity tend to emphasize people, helping others. Therefore, male frontline employees show stronger job engagement than female frontline employees. Besides, married frontline employees engaged in jobs stronger than single frontline employees. Similarly, frontline employees with high education levels tend to engage in their job stronger than those with low education levels. Hair et al. (2017) suggested that "the primary criterion for the assessment of a PLS path model is the coefficient of determination (R 2 )". R 2 ranges are always between 0 and 1, where higher R 2 values indicate a better fit of the data to the model. Typically, R 2 values are ranked, where 0.1 indicates a small effect size, 0.3 indicates a medium effect size, and 0.5 indicates a large effect size (Hair et al., 2017). With partial least squares, however, the ranks are lower, where 0.02 indicates a small effect size, 0.15 indicates a medium effect size, and 0.35 indicates a large effect size . In this study, R 2 values of employee engagement and service-oriented OCBs were 0.440 and 0.528 respectively. This means the exogenous latent variable, perceived HRM practices, explains 44.0% of the variance of the endogenous variable employee engagement, and 52.8% of the variance of the endogenous variable service-oriented OCBs. Thus, they showed that perceived HRM practices had a large effect size on employee engagement and service-oriented OCBs.

Theoretical implication
The present study made three important contributions to the literature. First, it shed light on the nature of the association between employees' perceptions of HRM practices and service-oriented OCBs. In other words, HRM practices played a vital role in the understanding of service-oriented OCBs. Besides, previous researches mostly emphasis on investigating the connection among HRM practices and service-oriented OCBs in terms of service sectors such as the hotel industry (e.g., Suan & Nasurdin, 2014), local government, education, transportation (e.g., Alfes et al., 2013;Cesário & Magalhães, 2017;Husin et al., 2012). However, it has yet to be verified in the banking industry, or only has little empirical evidence regarding this relationship. So, the findings of this study filled up the gap.
Second, this study revealed employee engagement as a mediating mechanism that accounts for the connection between perceived HRM practices and service-oriented OCBs. These findings supported the social exchange theory (Blau, 1964) which refers to the social exchange between employers and employees. When HRM practices are positively perceived, frontline employees are more likely to be grateful to their organization, then increase their engagement with jobs and, hence exhibit service-oriented OCBs. By proposing a new mediator in the association between perceived HRM practices and service-oriented OCBs, this finding extended prior studies, which have mainly focused on job satisfaction, employee commitment, person organization-fit, and perceived organizational support (e.g., Aktar & Pangil, 2018;Pohl et al., 2017;Turek & Wojtczuk-Turek, 2015). Besides, the author states that employee engagement covers a more holistic attitude as it is consisting of physical, cognitive, and emotional elements, is a better construct to mediate the connection between perceived HRM practices and service-oriented OCBs. So, this paper encourages future research to integrate employee engagement as a mediator through which perceived HRM practices impact outcome variables.
Finally, this research demonstrated that the association between perceived HRM practices and service-oriented OCBs might not be a straightforward relationship as is supposed in most prior researches to date. This finding also pointed to the importance of positive psychological capital which moderated this connection. Frontline employees' positive psychological capital had a place in respecting how their perception in HRM practices came out in the organization and it also did matter to some extent with respect to service-oriented OCBs. In other words, frontline employees both high in perceived HRM practices and high in positive psychological capital could enact the highest propensity of service-oriented OCBs compared to those who were low in positive psychological capital. In this respect, psychological capital heightened the causal relationship between HRM practices and service-oriented OCBs.

Practical implication
The findings in this study had practical implications that could provide additional insight into HRM practices in the banking sector in Vietnam. They are important guidelines for banks to apply HRM practices as a useful tool for increasing work engagement and promoting employees' service-oriented OCBs. In particular, the results of this study showed that HRM practices included seven dimensions such as recruitment and selection, training, performance appraisal, reward and recognition, career advancement, employee participation, and information sharing. Therefore, banking managers should use these determinants as an incentive to promote their employees' service-oriented OCBs. Some ideas are recommended to improve the current HRM practices in the Vietnamese banking industry as the following: First, bank managers should keep open communications with employees and pay attention to the bankers' requests or spend time listening to their staff. They should also discuss with bankers about general strategies and related decisions, the feedbacks of bankers will be recognized to improve bank strategies. Besides, employees should be encouraged to contribute ideas related to important decisions of the department or bank. Second, the goals of the bank should state clearly. When an employee recognizes the organization's purpose and understands content clearly, they will have the right direction in their job. An employee should know the priorities task that they ran the job. All of these helps to raise the sense of initiative and engaging in their job. Thus, banks should create a clear organizational purpose and mission and articulate well the organizational goals of all members. Third, banks need to create and enforce a fair recruitment and selection system that includes both internal and external resources such as the opportunity to be selected equally each person. So, banks could attract competent people from various resources. Fourth, creating and maintaining helpful information sharing by establishing procedures, policies, office facility, information, work units, planning and control systems, and so on, that facilitate the organizational efforts. Good information sharing helps employees make them identify with the organization or feel a vital part of the organization, hence their serviceoriented OCBs would be increased. Finally, bank organizations should establish clear and appropriate procedures to guide and control the performance appraisal, reward, and recognition process. The procedures should be ensured fair for all employees at levels. Once perceived HRM practices are positive, employees could be willing to work hard and engage to work longterm. Employees like going to work and concentrate on their job enthusiastically. They feel happy when working intensely. When employees engage physically, cognitively, and emotionally in their job, they have a strong tendency to show service-oriented OCBs.
Additionally, the relationship between perceived HRM practices and employees' service-oriented OCBs is stronger with a higher level of positive psychological capital for employees. In more detail, people with a high level of positive psychological capital tend to have higher levels of career planning and decision making confidence than people with diffidence. On the other hand, people with diffidence tend to think of themselves as victims, are in a constant state of helplessness, blame themselves for problems, unable to control the situations around them, and see negative events as permanent and insoluble. Hence, the role of HRM practices and employees' serviceoriented OCBs is stressed. Bank managers equip their employees with the necessary knowledge and skills, abilities, and motivation to succeed. Thus, bank managers would help their employees taking charge and being in control of their own destiny.
Furthermore, in order to increase the level of service-oriented OCBs, banks should take into account the characteristics of an employee (marital status and education level). They had significant influences on the impact of control factors on employees' service-oriented OCBs and employee engagement. The findings of this study confirmed that employee engagement and service-oriented OCBs were influenced by marital status. Married employees tend to elicit more service-oriented OCBs than unmarried employees. Additionally, employee engagement was also affected by gender and educational level. Male employees were found to be more engaged in their jobs as compared to their female employees. Besides, this study also indicated that employees with high educational levels were more engaged than employees with less educational levels. The results of this research could help bank managers devise appropriate recruitment and selection processes so that service-oriented OCBs and employee engagement can be influenced in a better way. Moreover, it also helps managers to have a better understanding of employees' demographic characteristics which are actually one of the necessary factors to reinforce positive volunteer personal among bankers. Psychometric testing can be employed to assess personality factors that have emerged as a significant predictor of serviceoriented OCBs and employee engagement.

Limitation and further research
There were also some limitations of this study. First, the survey to collect data used for this research was only conducted by a convenience sampling method with a quite small sample size (457 frontline employees), therefore it may not reflex truly the whole Vietnamese banking system. Further researches should conduct a survey by a probability sampling method with a large sample size for the whole country, it will increase the reliability of the results. Second, there are many determinants that affect employees' service-oriented OCBs, but in this study, only HRM practices were analyzed, so further researches could extend other factors that may explore and contribute more valuable knowledge to theory and practice. Third, this research was only done on core frontline employees. However, employing temporary staff, or short-term contract staff or collaborators is a developing trend nowadays. Because of their advantages such as cost savings, less responsibility in labor laws. And usually having effectiveness with the job does not arise frequently, or work with specific sales targets must be achieved. So, understanding the needs and aspirations of this employee group will help bankers easy to manage better and avoid disturbance in business.