Role of motivation to learn in training transfer and job performance under peer and supervisor support in the Vietnamese public sector

Article history: Received 3 January 2020 Received in revised form 23 March 2020 Accepted 25 March 2020 This study examines the role of motivation to learn in both transferring skills and knowledge and improved job performance under the support of colleagues and supervisors. A structural equation modeling approach was used to analyze the responses to a survey of 399 employees in public sector organizations. The findings revealed that peer support had an insignificant direct effect on training transfer but had a significant indirect effect on training transfer throughout motivation to learn. However, supervisor support had both direct and indirect effects on training transfer through motivation to learn, which, in turn, improved job performance. The findings also indicated that both motivation to learn and skill transfer were significantly associated with job performance. The study provides theoretical contributions regarding the role of motivation to learn in training transfer and job performance. Managerial implications regarding stimulating job performance in public sector organizations are to generate incentive policies to inspire and motivate employees to learn new skills and knowledge and transfer them to improve organizational performance.


Introduction
*In this era of accelerated information and communication technology development, employees must constantly update their knowledge and skills. They must also improve their qualifications simply to meet the requirements of their current job and its increasing demands. Organizational leaders can encourage employees to learn and help create a dynamic workplace in which they are willing to share knowledge and experience, thereby encouraging the organization's sustainable development.
Practitioners and scholars recognize learned skills and knowledge transfer as crucial for improving organizational outcomes. Broad and Newstrom (1992) argued that the transfer of knowledge through training programs allows trainees effectively and continuously to apply such knowledge and accompanying skills, both on the job and in daily life. In other words, training transfer involves the maintenance, application, and synthetization of perceived knowledge, skills, and attitudes to enhance effective capacity (Baldwin and Ford, 1988;Blume et al., 2019). Therefore, learning and self-research promote the transfer of employee knowledge and skills to their work, effecting higher efficiency and improving service and product quality.
Scholars have identified a need to explore the factors that affect learning motivation and training transfer (Kozlowski et al., 2001). At the same time, managers believe it necessary to build a management system that promotes training transfer (Broad, 2003). Studies examining the correlation between learning motivation and training transfer in organizations have yielded different results, depending on whether a personality-oriented approach was used or the overall context of factors within the organization was addressed (Facteau et al., 1995;Tracey et al., 2001). However, researchers have reported that employees will be motivated to enhance job performance when they feel that they are receiving the support of their organization (Deconinck and Johnson, 2009).
In another study, Yamnill and McLean (2001) found that supervisors are an essential component of a work climate, and supervisor support is considered a moderator in the connection between training transfer and transfer design. Other studies have found that both a direct leader and peer support are essential in facilitating the application of learned skills and knowledge to a current job (Facteau et al., 1995;Holton et al., 1997;. In a recent study, Tran et al. (2018) demonstrated that a strong relationship between leaders and employees created a better quality workplace and improved the job performance of nurses in Vietnamese hospitals.
In general, the factors that stimulate motivation to learn and training transfer have been the subject of intensive research. These factors have been grouped under two broad categories: peer support and supervisor support. Driven by the need for greater clarity on the roles of these factors, this study concentrated on two variables, motivation to learn and training transfer, under the support of peers and supervisors, to estimate their influence on job performance in public sector organizations in Vietnam, where an understanding of the roles of supervisor and peer support for learning and training transfer is limited.
The results are anticipated to be useful to practitioners who design training courses and for the development of organizational policies, as training programs can be designed with an improved understanding of training transfer, motivation to learn, and their inter-relationship. The cause and effect relationship among the study constructs is also expected to interest scholars working in the field of human resource management.

Peer support in motivation to learn and training transfer
Transfer of training can be defined as "the degree to which trainees apply to their jobs the knowledge, skills, behaviors, and attitudes they gained in training" (Holton et al., 1997). In addition, training outcomes are defined as "the amount of original learning that occurs during the training program and the retention of that material after the program is completed" (Baldwin and Ford, 1988). The work environment can influence the transfer of learning to job performance through its transfer of training climate; trainees realize more benefits from a climate with organizational supports than from psychological cues. The results of a study of social support for training in four organizations in the United States suggested that pretraining motivation stimulated perceived training transfer, and perceived transfer had a positive relationship with peer support but was negatively related to supervisor support. Individuals who attended training because it was an organizational requirement were less motivated to learn. The findings also implied that managers would like to perceive better transfer of training skills when they believed that employees were willing participants (Facteau et al., 1995).
In another study on learning transfer in the United States,  reported that transfer systems vary in organizational characteristics, cultural differences, and learning approaches. Moreover, their results illustrated that supervisors do not pay attention to training activities as an expectation. Specifically, learning transfer systems are considerably distinctive across private and public sector organizations. Employees in private sector organizations perceive that they will have a chance to exercise their learning to improve job performance and create valued outcomes. By contrast, employees in public sector organizations are more likely to perceive that their supervisors will resist their application of new approaches learned in training.
Several studies have identified possible predictors for training transfer. For example, in a study on training transfer at an organization in the United States, Chiaburu and Marinova (2005) found that peer support has a significant influence on both training transfer and pretraining motivation, whereas supervisor support is not associated with either factor. Another study of a manufacturer in India illustrated that the impact of peer support on training transfer is greater than that of supervisor support (Chauhan et al., 2016).
Peer support at work is seen to play an essential role in motivation to learn and to improve job performance through learning outcomes. Bates et al. (2000), for example, reported on the important role of peer support in creating an opportunity for learning and stimulating employees to share knowledge learned from work experience in chemical manufacturers. Moreover, peer support encourages trainees to use their learned knowledge on the job (Hawley and Barnard, 2005). Nijman et al. (2006) proposed that peer support may help and motivate trainees to maximize transfer. With peer support, trainees have to participate in training activities and take responsibility for their learning (Broad and Newstrom, 1992). Others at work may observe how trainees apply skills acquired in training and assist training transfer (Gilpin-Jackson and Bushe, 2007). Finally, peers stimulate training transfer through a networking and information sharing mechanism (Martin, 2010). From this literature, the following hypotheses were developed: H1: Peer support will positively impact motivation to learn H2: Peer support will positively impact training transfer

Motivation to learn and training transfer under the supervisor support
Supervisors work closely with subordinates and understand their training demands. Therefore, supervisors encourage subordinates to participate in professional training and support them as they apply newly learned skills as well as to identify scenarios where the acquired skills can be implemented. Supervisors also provide feedback on improvement. As a result, supervisor support stimulates the positive transfer of training (Elangovan and Karakowsky, 1999). Particularly, supervisors are supposed to contribute throughout the entire training process, including by providing guidance and recommendations in selecting a training program, suggesting potential projects for using learned knowledge, and assigning duties to develop learned skills (Chiaburu and Tekleab, 2005). Baldwin and Ford (1988) regard supervisor support as a multi-dimensional construct that can include the practice of new skills, and apply of learned knowledge (Chiaburu and Marinova, 2005); toleration changes (Lancaster et al., 2013); and discussion with subordinates approaches to apply new learning (Lim and Johnson, 2002). Supervisors, then, play an essential role in creating a working environment wherein employees can share ideas for designing course content based on job demand and feedback from others, use new learning, and promote knowledge transfer after training. Therefore, supervisor support becomes essential to steering employee motivation, behavior, and attitude (Hutchins, 2009;Nijman et al., 2006). In addition, supervisor support is expected to have a crucial direct or indirect influence on training transfer through motivating trainees and interacting dynamic components in the transfer climate (Cromwell and Kolb, 2004). Transfer outcomes can lead to reducing production costs due to applying new ways to decrease in scrap rates electronics manufacturing companies after training (Xiao, 1996). Chauhan et al. (2017) showed supervisor support has a significant impact on training transfer in the Indian manufacturing sector. Most previous studies have found a positive relationship between supervisor support and transfer of training (Cromwell and Kolb, 2004;Facteau et al., 1995;Holton and Baldwin, 2003). In a study at a petrochemical manufacturing facility, Holton et al. (1997) reported that supervisor support has a stronger impact on the transfer of acquired learning than peer support does. In addition, McCracken et al. (2012) showed managers in the public sector play an important role in creating an organizational training climate that encourages employees to participate in training programs and transfer learning to their workplace. Specifically, in a recent study of the Agency Supervisor model, Westman et al. (2019) indicated the more supervisory support, the higher quality treatment perform. However, other empirical studies are inconsistent with this perspective (Klink et al., 2001).
These studies lead to the following hypotheses: H3: Supervisor support has an influence on motivation to learn. H4: Supervisor support has an influence on training transfer.

Motivation to learn, training transfer, and job performance
Human resource management practices serve to generate value for organizations by enhancing inrole and extra-role efficacy if employees are perceived to benefit from organizational support. Support policies can promote the well-being of employees by rewarding their contributions and enhancing emotional commitment (Tremblay et al., 2010). In an era of fast-paced technological change and innovation, to maintain business performance, employees must have the ability to overcome new challenges. Training then becomes an essential strategy for organizations to assist employees in acquiring the knowledge and skills required to confront these challenges.
According to Noe and Schmitt (1986), motivation to learn is described as a specific desire on the portion of a trainee to discover the content of a training program, acquire new skills, and apply these skills to job tasks. In general, organizations must evaluate incentive policies for motivating learning and assisting employees in updating skills and knowledge (Chuang et al., 2005).
When employees perceive meaningful benefits from organizational supports and are satisfied with the rewards for their valuable contributions, they are more likely to achieve better performance and shoulder a continuing commitment (O'Driscoll and Randall, 1999). When organizations pay attention to creating a good working environment, positive work attitudes and employee behavior are enhanced, and employees are encouraged to devote more time and effort to learning new skills and knowledge and absorbing new attitudes during the training process. As a result, employees are better able to apply and transfer newly learned knowledge, skills, and attitudes. Hence, the more organizational support on the motivation to learn, the higher the expectation that employees will apply the learned skills, knowledge, and attitudes gained in training to the workplace (Zumrah and Boyle, 2015).
Empirical studies have found that motivation to learn drives significant results, such as by encouraging employees to attend the training program (Noe and Wilk, 1993); stimulating employees to devote greater effort to learning and gaining benefits from the training (Chuang et al., 2005;Colquitt et al., 2000); and transferring acquired knowledge and skills to applications in the workplace (Cheng and Ho, 2001); considered as the most crucial predictor of training effectiveness (Kodwani and Prashar, 2019). In practice, employees only have more motivation to learn and upgrade their education and skills when they perceive that participating in a training program can improve job performance or promote their career. Higher levels of pretraining motivation are associated with a larger return on training (Quinones, 1995). Foxon (1993) confirmed that transfer intention is the most essential phase in the transfer process for anticipating the capacity to establish the level of transfer. Cheng and Ho (2001) reported that trainees often embrace a trial and error method to control training transfer, which can take more time and resources without delivering a desirable result. In other words, if they do not understand training content, they are often stumbled by training transfer outcomes. Studies have also provided strong evidence of the positive relationship between training transfer and motivation to learn (Chiaburu and Marinova, 2005;Colquitt et al., 2000). However, Chiaburu and Tekleab (2005) found no evidence to support this relationship after analyzing survey data from an organization in the United States. Whereas, in a recent study, Kim et al. (2019) reported that learning motivation had the largest influence on intention to transfer knowledge from a professional training program in secondary schools in the United States. The relationship between motivation to learn and training transfer, then, is not entirely clear. Hence, this study proposes the following hypothesis: H5: Motivation to learn has an influence on training transfer.
Previous studies have explored training motivation theory, attempting to identify antecedents and relationships of training outcomes. For instance, Tracey et al. (2001) highlighted the importance of training preparation by examining a model of pretraining self-efficacy and motivation under the mediating effects of training reactions, knowledge acquisition, and other determinants. In addition, training motivation enhances the effects of cognitive ability, which leads to incremental variance in training outcomes (Colquitt et al., 2000).
In the era of globalization and technology development, every leader must deal with recruiting and keeping talent in the organization. To do so, organizations often support a professional training program and encourage trainees to use training outcomes to improve job performance. As a result, they develop the individual ability and competitive advantage to maintain market share in a dynamic business environment (Michaels et al., 2001). Indeed, the main objective of training programs is to transfer skills and knowledge from training to improve current employee task performance. However, several studies have reported that only about 10 percent of training outcomes actually transfer to job performance (Georgenson, 1982;Holton and Baldwin, 2003;Kupritz, 2002). Therefore, training efforts in private and public organizations have concentrated on individual outcomes such as "self-esteem, self-efficacy, performance evaluations, merit increases, and organizational performance outcomes," (Lim and Morris, 2006), instead of evaluating learning outcomes (Ford and Weissbein, 1997).
Meanwhile, Yamnill and McLean (2001) described expectancy, equity, and goalsetting theories as supporting theories for motivation to transfer. Kontoghiorghes (2002) categorized motivation to transfer as two types: Motivation to learn (Tracey et al., 2001) and motivation to transfer (Ruona et al., 2002). According to this categorization, motivation to learn is a trainee's intrinsic or extrinsic passion for attaining a high level of learning, and motivation to transfer is the trainee's passion for using the learned skills and knowledge in current tasks (Facteau et al., 1995).
Empirical studies have illustrated that motivation to learn can impact the extent to which workers are willing to attend in training (Tharenou, 2001), devoting effort to get benefit from training (Chuang et al., 2005;Colquitt et al., 2000), to transfer updated knowledge and skills to the workplace (Baldwin and Ford, 1988;Cheng and Ho, 2001). Clark et al. (1993) proposed that employees would have a motivation to learn if they perceived training outcomes leading to improving either job performance or career development. On the other hand, Mathieu et al. (1993) found that no direct linkage between motivation and job performance. Hence, from this review, the next hypotheses are proposed: H6: Motivation to learn has a direct influence on job performance. H7: Training transfer has a direct influence on job performance.
The literature review demonstrates a further need for a theoretical framework, empirical study, and practical implications regarding the antecedents of training transfer and motivation to learn in the public sector. The prior studies highlight the varying effects of training transfer initiatives in both private and public organizations but have not provided a sufficient base for a comprehensive model that relates to both antecedents and consequence of motivation to learn and to train transfer in the public sector. Moreover, the literature review also indicates that antecedents of training transfer such as supervisor and peer support have differential effects on the basis of which we may derive that the findings of previous studies are not consistent. Scholars claim that these inconsistent results may be owing to divergent contexts such as cultures, sectors, countries. Fig. 1 illustrates an integrated theoretical framework intended as a starting point in the Vietnamese public sector.

Research design and measurement
A questionnaire was designed for a paper-based survey using a five-point Likert scale anchored by 1="strongly disagree" and 5="strongly agree" The questionnaire included two parts: Demographic information of respondents and a structured questionnaire related to five constructs, with a total of 29 items derived either from previous studies or developed through group discussion. First, supervisor support was measured using five items adapted with minor modifications from Yarnall (1998). Second, peer support was assessed using seven items, in which five were self-developed following group discussion, and two were adapted with minor modifications from Xiao (1996). Third, motivation to learn was measured through four items adapted from Lepine et al. (2004). Fourth, training transfer was assessed using seven items, in which five were adapted from Facteau et al. (1995) and the other items adapted from Zumrah and Boyle (2015). Finally, job performance was measured through six items adapted from Rupp and Cropanzano (2002 All questions were translated from English into Vietnamese. We then conducted a group discussion with five managers of human resource departments to ensure that the content was consistent with public organizations in the Vietnamese context. The managers made recommendations for adjusting the Vietnamese version so as to produce a more appropriate questionnaire that would allow respondents to easily understand the items. They also suggested adding five more items to measure peer support. We conducted a back translate approach to complete the English questionnaire. The 29 items are denoted and described in the appendix. Determining sample size relies on the statistical estimating accuracy and number of variables; larger sample sizes allow for more reliable results. Hair et al. (2009) suggested that sample sizes ranging from 200 to 400 are critical for multiple regression and path analysis. In addition, Green (1991) proposed a rule of thumb for deciding sample size. Specifically, the sample size should be equal to "50+(8 * number of measurement items)." The number of items was 29 in this study; hence, a total of 282 observations was considered appropriate for testing the hypotheses.

Data collection
We conducted a pilot test of face-to-face interviews with 35 public servants in July 2019. The result of the pilot test enabled us to verify and adjust the final Vietnamese questionnaire for easy understanding. Then, after finishing minor modification of the Vietnamese version, we conducted data collection. We use a stratified sampling method to conduct the questionnaire survey.
The people's committee of Tien Giang province, Vietnam, is charged with control of 32 key public organizations. These include 18 departments, 11 districts, and three non-business units. With the valuable support of managers in the human resource management offices, the authors distributed 10 questionnaires to each department, 20 questionnaires to each district, and 15 questionnaires to each non-business unit. A total of 445 paper-based surveys were therefore delivered to the target respondents, all working at key public organizations in Tien Giang. The authors collected 430 completed questionnaires, of which 399 were valid for use in data analysis. Three months, from July to September 2019, were required to conduct the data collection.
The demographic information described a relative gender balance. There were 185 female respondents or 46.36% of the total, and the remaining 214 respondents (53.63%) were male. The majority of respondents were public servants, taking up 78.45% (N=313), whereas the remaining 21.55% (N=86) were managers or the equivalent. In all, 89.47% (N=357) of the respondents held a bachelor's or master's degree. The majority of respondents ranged from 26 to 45 years old, accounting for 79.45% (N=317). In terms of work experience, 42.10% (N=168) had worked in public organizations for between 1 and 9 years, and the remaining 57.90% (N=231) had 10 or more years' experience in such organizations.

Reliability validity and construct validity
To assess reliability, the authors performed exploratory factor analysis (EFA) and calculated the Cronbach's alpha of each variable by using SPSS software version 22. According to Kline (1998), a Cronbach's alpha indicator of 0.9 is considered excellent, 0.8 very good, 0.6 to 0.7 adequate, and below 0.5 unreliable. The indicators varied from 0.859 to 0.908, indicating good internal reliability of the construct measures. In addition, the corrected item-total correlation of all items was greater than 0.3, satisfying the reliability requirements of the measurement criteria (George and Mallery, 2003).
The EFA was conducted to explore potential underlying factors and identify whether a set of constructs consistently loaded on the same factor was based on strong correlations. This test is implemented to "reduce the number of variables like the measurement indicators for the path analysis of the overall model" (Lee, 2009). Hair et al. (2009) suggested that factor loading can be accepted if it is higher than 0.5. Excepting two items (TT1 and TT2), all loading indicators were greater than 0.5. TT1 and TT2 were therefore dropped as inadequately loaded. Moreover, KMO and Barlett's test indicator was equal to 0.919, greater than the threshold of 0.5, with a significance level of less than 0.001. The total extracted variance of 58.37% (>50%) proved that the five factors were explained by 58.37% of the data variability. Table 1 presents the descriptive statistics and results of the EFA. Notes: SD: Standard deviation. 5-point Likert-type scale, with one representing "strongly disagree" and 5 "strongly agree." Loading items were extracted from maximum likelihood with rotation. TT1 and TT2 were eliminated because of loading results equal to 0.420 and 0.476, respectively, lower than the threshold of 0.5

Model fit, convergent validity, and discriminant validity
To examine the measurement model fit, confirmatory factor analysis was conducted. The ꭕ 2 was checked to identify the overall fit of the structural model, but this indicator is sensitive to sample size and model complexity (Chen, 2008). Hence, we combined other indicators to determine model fit. The combination rules included ꭕ2/df (threshold between 1 and 3), comparative fit index (CFI>0.90), standardized root mean squared residual (SRMS<0.08), and root mean squared error of approximation (RMSEA>0.06; Hu and Bentler, 1999). The results suggested goodness of fit, demonstrating a good model fit (ꭕ 2 /df=2.734,CFI=0.915,SRMS=0.047,and RMSEA=0.066). Therefore, the model fit indices were in accordance with the acceptance criteria.
To evaluate the convergent validity of a measurement scale, two indicators-average variance extracted (AVE) and composite reliability (CR)-are evaluated. The AVE values must at least 0.5, and that of the CR higher than the 0.6 cut-off point to ensure convergent validity (Fornell and Larcker, 1981). In addition, based on the simulation results of Fornell and Larcker (1981), an AVE value of from 0.4 to 0.5 is acceptable when the value of the CR is higher than 0.6 in all the measurement scales. Malhotra et al. (2010) also argued that AVE is too strict, and reliability can be established through CR alone.
As can be seen in Table 2, all estimates, and the AVEs of all constructs, were greater than 0.5, and the composite reliability values of all constructs were good (>0.8), except for the AVE of PEER, which was equal to 0.475, though its CR was equal to 0.863. The results fully met the above criteria. The convergent validity of all constructs was confirmed, and reliability measures for all model constructs and items met the criteria.
The correlation matrix between the two constructs was used to examine discriminant validity. When the AVE of any measurement item was higher than the square of the largest correlation estimates of that factor with the maximum shared variance (MSV), discriminant validity was asserted (Fornell and Larcker, 1981). When the values of the average shared variance (ASV) and MSV were lower than their respective AVE values, discriminant validity conquered (Hair et al., 2009). All indicators are reported in Table 2. They assured the cut-off criteria for test convergent and discriminant validity.  Table 3 presents the testing results for the structural equation modeling in Fig. 2. The overall goodness of fit statistics satisfied the criteria, and the estimated coefficients were standardized. The finding supports Hypothesis 1, that is, peer support has a significant impact on motivation to learn (0.366, p<0.001). Meanwhile, peer support had no direct effect on training transfer (i.e., Hypothesis 2 was not accepted) but had a significant indirect impact on training transfer due to motivation to learn (0.366×0.573=0.210; p<0.001; following the calculation rule of Bollen (1989). Furthermore, the findings illustrate that motivation to learn positively mediates supervisor support (H3 was accepted with the estimate=0.197, p<0.01), and training transfer (H5 was accepted with the estimate=0.573, p<0.001). The testing result of H4 was accepted, implying that supervisor support is positively associated with training transfer (0.224, p<0.001). Moreover, motivation to learn also had both an indirect effect on job performance (0.264, p<0.001) and a significant direct effect on job performance (H6 was accepted with the estimate=0.389, p<0.001). Therefore, it generated total effects on job performance (the estimate=0.653, p<0.001). This is explained by the AMOS software version 22 report: When the motivation to learn is increased by one standard deviation, job performance increases by 0.653 standard deviations (Kline, 1998). In addition, Hypothesis 7 accepts, since the estimated coefficient illustrated a positive and statistically significant association between training transfer and job performance (0.461, p<0.001).

Discussions
These results are relevant to the intensity and key role of motivation to learn and training transfer under the support of peers and supervisors for the improvement of job performance in public sector organizations. They have contributed to clarifying the study questions. Previous literature did not explore causal linkages among motivation to learn, training transfer, and job performance under the effects of peer and supervisor support. The findings demonstrate the mediating roles of training transfer and motivation to learn in the linkage between peer and supervisor support. The study also investigated the total effect of motivation to learn on job performance.
First, a key contribution is that motivation to learn plays an important role in encouraging training transfer and improving job performance. The data analysis uncovered a significant and positive connection between motivation to learn and both training transfer and job performance and that between training transfer and job performance. The findings suggest that managers should encourage public servants to transfer skills and knowledge in the workplace and, in turn, improve job performance. They should launch incentive policies for motivating employees to learn and encourage public servants to update skills and knowledge. The findings are consistent with previous studies (Blume et al., 2019;Chuang et al., 2005;Zumrah and Boyle, 2015).
Second, the significant total effect of motivation to learn on job performance via training transfer is an essential finding, one that had not previously been empirically tested for, particularly in public sector organizations. This study extends recent research that found training readiness and motivation to learn were mediators in the connection between supervisor support and intention to transfer; these results also indicate that the supervisor has both significant and positive direct and indirect effects on training transfer (Kim et al., 2019).

Supervisor support
Motivation to learn Peer support Third, the findings provide evidence to support the essential roles of motivation to learn and training transfer in job performance that is consistent with two types of theories for motivation to transfer: motivation to learn and motivation to transfer (Kontoghiorghes, 2002;Ruona et al., 2002;Tracey et al., 2001). The results are also consistent with findings such as that supervisor support enhances positive transfer of training outcomes (Elangovan and Karakowsky, 1999); supervisor support is an important instrument for stimulating employee motivation, behavior, and attitudes (Hutchins, 2009;Nijman et al., 2006). A prominent concern is how to encourage employees' passion for applying learned skills and knowledge to their current tasks (Cromwell and Kolb, 2004;Facteau et al., 1995;Holton and Baldwin, 2003).
Fourth, peer support has no direct effect on training transfer but has a significant and positive one throughout motivation to learn. This result is inconsistent with the previous finding that the impact of peer support on training transfer is greater than that of supervisor support (Chauhan et al., 2016). It also implies that peer support motivates employees to participate in training activities and take responsibility for their learning (Broad and Newstrom, 1992). To promote the transfer of training of employees in the context of public sector organizations, supervisors should create an organizational climate where employees' transfer of skills and knowledge in the workplace is greatly inspired and can be exercised quickly (Kim et al., 2019).

Managerial implications
The findings address the crucial role of supervisor support in stimulating employee motivation to learn. Supervisor support not only directly impacts training transfer and motivation to learn but also indirectly affects training transfer. When supervisors can inspire employees to participate in professional training programs, employees are more likely to be ready for training and motivated to learn. Supervisor support also stimulates employees to apply new skills and knowledge to their current tasks, when the organization can establish a working environment where employees are motivated and feel valued. By doing so, public sector organizations not only keep employees productive but also prevent employees from leaving when the private sector offers opportunities for career advancement.
Public organizations should also create a work environment that recognizes the support and effort of peers and supervisors. They are more likely to accept greater responsibility for stimulating employee accountability, which will positively affect employee attitude and performance. Specifically, the results indicate a significant and positive relationship between job performance with both motivations to learn and training transfer. This suggests that the more employees are motivated to learn, the better their performance and training transfer. Thus, leaders of public sector organizations should ensure that incentive policies inspire and motivate employees to learn new skills and knowledge and transfer them to improve organizational performance. To do so, leaders should pay careful attention to designing professional training programs that meet job demands so that public servants are willing to share skills, knowledge, and experience in the workplace to improve job performance. It also implies that the more successful training transfer-related improvements in job performance will ultimately lead to a more effective training cost.
This study contributes to the extant research in two ways. First, the paper examines the process through which motivation to learn and to train transfer interactions to affect job performance under the support of peers and supervisors in public sector organizations in Vietnam. Second, the significant result of the contribution path is highlighted throughout a peer support-motivation to learntraining transfer-job performance. In other words, the study confirms the significance of motivation to learn in stimulating training transfer and improving job performance in the public sector. Therefore, leaders in the public sector should create a dynamic working climate to provide support and motivate employees to transfer learned skills, knowledge, and experience at the workplace. Overall, the findings suggest that public sector organizations should establish motivation to transfer and ultimately ensure training transfer throughout generating a working environment enriched with peer support.
Finally, the findings provide valuable evidence and implications for human resource management in public organizations with regard to the importance of enhancing interpersonal interactions and motivation to learn, training transfer, and job performance. Adequate attention should be paid to prioritizing solid interactions among public servants. Therefore, public organizations should generate incentive policies that recognize contribution in the workplace and attend to employee well-being and career development. Furthermore, the findings enable us to suggest that managers in the public sector should concentrate on motivation to transfer throughout initiating distinct intrinsic and extrinsic benefits of training programs to employees. By doing so, employees are provided with applications of learned skills and knowledge at the workplace. Organizational leaders can stimulate the perception of employees towards the training benefits and build motivation to transfer.

Conclusion
This paper investigated the role of supervisor and peer support in motivating public servants to learn and transfer knowledge and experience to improve job performance in the public sector in Vietnam. The findings provide recommendations for organizational leaders to reform policies as well as nurture a working climate in which employees are motivated to learn, share experience, and propose novel ideas. In this understanding, public organization leaders should design professional training courses and host counseling seminars for public servants so that they are willing to devote more time and effort to improving job performance. Finally, public organizations should create a work climate that fosters employee pursuit of lifelong learning.

Limitations and suggestions for further study
One limitation of the study is that data were collected at a single point in time and in one province of Vietnam with a cross-sectional design methodology. Chiaburu and Lindsay (2008) mentioned that the collection of data at one point of time in training research makes it difficult to drive inferences of causality (Dysvik and Kuvaas, 2008), restricting the explanatory power of the results (Gegenfurtner et al., 2009). Thus, implementing a longitudinal study in this area would assist in validating the current findings.
Another limitation is that data were collected from public sector organizations. Therefore, future studies could replicate the research model used here in other organizations, such as private enterprises and multinational companies. Conducting a study across different businesses will allow a fuller understanding of the role of the proposed constructs in motivation to learn and transfer skills and knowledge at the workplace.