Luận văn determinants of employee engagement a case study of vnpt technology

1.2 Purpose of the research This research aims to examine the factors that influence engagement level at VNPT Technology, evaluate the impact of each factor on engagement, and propose s

INTRODUCTION

Research rationale

All organizations, regardless of size, whether local, national, or multinational, always consider their human resources to be a valuable asset The success of a company is largely determined by its human resources Effective human resource management is critical in today’s intensely competitive business environment, and an organization can differentiate itself with a group of creative, talented employees

According to the “Vietnam Best Places to Work” annual survey in

2019, turnover rates rose during the previous three years, putting a risk of exceeding 24% in 2019 Employees who worked for the company less than two years are the most likely to quit, particularly in management and directors, which is significantly higher than other positions In the field of ICT (Information and Communication Technologies), particularly in the software industry, over 60% of jobs lasted no longer than two years, and nearly 80% could last no longer than three years According to these statistics, employee engagement in the software industry in Vietnam appears to be a luxury

Despite the widespread and rising unemployment after two years of COVID-19 infection, mass resignations occurred during the first six months of 2022 As of April 2022, up to 260,000 out of 4 million LinkedIn members in Vietnam (6.5%) have changed their status to “looking for a new job”

Additionally, 90% of members who did not post the status “looking for a new job” were very open to feedback when receiving information about new employment opportunities Based on Anphabe survey results from December

2021, Thanh Nguyen, Managing Director of Inspiring Happiness, stated, “The level of engagement of working people in Vietnam is at an all-time low”

Anphabe reported that in 2022, 47%, 53%, and 66% of employees in the information technology, electronics, and telecommunications industries, respectively, desired to leave their organization and change careers

The organization performance is strongly impacted by employee engagement According to Aon Hewitt (2016), motivating employees is of utmost importance because they are an integral part of an organization’s effectiveness Engaged employees often feel at ease and enthralled by assigned tasks Therefore, they consistently demonstrate creativity at work, thereby assisting the organization in generating more creative ideas, innovating, making organizational breakthroughs, adapting to change, and initiating change proactively Teams that promote employee engagement see an 18-21% improvement in their productivity, according to a Gallup survey

Employee engagement will ultimately lead to higher customer satisfaction, a lower turnover rate, and the profitability of a firm Customers of companies with engaged workforces use their products and services more often and express greater levels of pleasure than those of businesses with disengaged workforces, according to a National Business Research Institute survey Employees that are engaged and dedicated in their work are less likely to leave Sometimes the best employees are not engaged, and the company might risk losing them

Boosting employee engagement fosters organizational commitment, leading to talent retention Gallup's 2021 study reveals a significant disparity in turnover rates, with low-engagement teams experiencing rates 18-43% higher than those with high engagement The substantial cost of replacing an employee, equivalent to half their annual salary, underscores the financial impact Notably, intellectual capital lost through employee departure further compounds these costs Gallup's 2022 global workplace report highlights the profitability advantage of engaged employees, indicating 23% higher profits for businesses investing in engagement initiatives.

3 those with disgruntled employees A highly engaged staff infuses every facet of the business with vigor and energy, which boosts earnings

The VNPT Group is powered by VNPT Technology in the domains of digital content, technology, electronics and telecommunications manufacturing, and information communication technology In 2022, over 8% of the employees, including experienced, knowledgeable, and highly skilled employees, including middle-level managers, quit their jobs or switched departments The author noticed that there are people who have a lot of energy both at work and in other activities, while other people do not There are employees who refused to participate in training courses like MBA, PMP There are some key engineers who quit the job after being trained at an expert level in their profession There are centers with very high turnover rates Are middle managers the source of this problem? Does the job at VNPT Technology allow employees to be independent in decision-making, flexible in time management, etc., without constant supervision from superiors?

So what factors will cause employees to disengage from the organization? Is it because of poor relations with colleagues, direct supervision, or a lack of faith in the company’s leadership? Does their perception of growth and advancement opportunities at work affect their engagement level? From this situation, the author chose the topic

“Determinants of Employee Engagement: A Case Study of VNPT Technology”.

Purpose of the research

This study analyzes the factors that affect employee engagement levels at VNPT Technology It assesses the impact of each factor and proposes strategies and management practices to enhance engagement within the organization and during job duties The research specifically focuses on examining the correlations among key variables, including leadership style, compensation, work-life balance, opportunities for growth, organizational culture, and employee well-being.

4 leader-member exchange (LMX), relations with colleagues, training and development opportunities, task autonomy, and employee engagement

Many research have been conducted on engagement in a range of service sectors, including hotels, banking, food service, medical, call centers etc., as well as in the manufacturing sector, but relatively few in the ICT sector Besides the factors affecting employee engagement that have been mentioned in previous studies, such as relationships with leaders, coworkers, and career growth, this study will examine the impact of task autonomy, which is part of motivational work characteristics, on employee engagement in the context of VNPT Technology

In this study, the variables and their interactions are examined with an aim of achieving the following results:

1 Find out how respondents feel about factors including the nature of the LMX relationship, the coworker or supervisor relationship, the level of task autonomy, the availability of training and development, and the degree of engagement

2 Determine how these factors relate to one another

3 Identify the level to which each of the above factors affect employee engagement.

Objects and scope of study

The purpose of this study is to determine the determinants influencing employee engagement at VNPT Technology The survey’s respondents are office division employees (R&D center, Technical and Service center, Business center, Solution and Product center, etc.)

The four key factors – Leader-member exchange, task autonomy, social support, and training and development – are selected based on observations of

5 the author within VNPT Technology These factors are explored to understand their direct and indirect impacts on overall employee engagement with the company as a whole, rather than isolated interactions with colleagues or managers The intent is to provide a holistic view of organizational engagement, thereby offering insights that can help VNPT Technology foster a more engaged and committed workforce at the organizational level

Regarding space, the research will be conducted at the VNPT Technology office, located at 124 Hoang Quoc Viet, Nghia Tan Ward, Cau Giay District

Regarding time, the research will be conducted between February and May of 2023 Questionnaires will be used to collect primary data, while secondary data will be based on actual VNPT Technology data from 2020 and

2022 and other sources from the internet.

Research methodology

To determine the effectiveness of the proposed model, the author employs quantitative methods to analyze the hypothesized relationships between various variables This approach allows for the exploration and examination of connections within the model as well as their influence on employee engagement, drawing on prior research and established theories.

Quantitative research is used to test or confirm theories and hypotheses based on numerical data, the measurement of variables, and the use of statistical procedures Using questionnaires from previous studies such as The Work Design Questionnaire (Morgeson & Humphrey, 2006) and The Drivers of Employee Engagement (Robinson, Perryman, and Hayday, 2004), measurements will be selected to fit the VNPT Technology situation A total of 354 questionnaires were distributed, with an expected response rate of about 60% SPSS 20 and AMOS 20 will be used to synthesize and analyze the survey results to create a picture of the level of employee engagement at VNPT Technology

Thesis structure

Chapter 1 introduces an overview of research problems Chapter 2, a review of the literature, contains theories about employee engagement, factors that influence employee engagement, and the current studies related to each factor Additionally, hypotheses are also evolved based on the review of the literature In Chapter 3, the research methodology is detailed, including the process of developing and evaluating constructs as well as the methods and tools utilized in the research Data analysis and findings will be presented in Chapter 4 The last chapter, Chapter 5, will be the recommendations and implications of the research.

LITERATURE REVIEW

Defining engagement

Along with its development over time, engagement has been defined from many different perspectives, often inconsistently, so that the concept has become ambiguous However, some say that these definitions sound a lot like concepts that are already well-known such as “organizational commitment” and “organizational citizenship behavior”

Regarding the concept of engagement, the academic community and the consultancy have two distinct schools of thought Within the academic community, it appears that Kahn (1990) was the first to give a definition of engagement, as he defines the term as the ability for people to fulfill work- related roles with varied degrees of their own physical, cognitive, and emotional identities After Kahn, many other researchers conceptualized it in different ways, mostly as a psychological and behavioral state From the viewpoint of employee engagement as a psychological state, Macey and Schneider (2008a) explored that it has several ideas that represent “some form of absorption, attachment, and/or enthusiasm” Engagement was described by Maslach, Leiter, and Schaufeli (2001, p 417) as a continuous, positive, affective-motivational level of fulfillment Schaufeli et al (2002) and Sonnentag (2003) reported that vigor, dedication, and absorption are the core dimensions of engagement When considering employee engagement as a behavioral outcome, Macey and Schneider (2008a) defined it as discretionary effort or a type of in-role or extra-role actions Within consulting and research institutes, there are also a lot of definitions of employee engagement, mostly

Employee engagement is a crucial concept that encompasses an employee's emotional and intellectual connection to their work, organization, and colleagues It goes beyond motivation and job satisfaction, requiring the employee's active contribution and an intrinsic desire to exceed expectations According to the Gallup Organization, employee engagement is associated with increased discretionary effort at work, while the Conference Board defines it as an employee's enhanced link to their work, leading to greater motivation Mercer's research emphasizes that employee engagement is a mindset characterized by a focus on the company's success and a willingness to go above and beyond job requirements, influenced by employee sentiment toward various aspects of the workplace.

The definitions of employee engagement from the academic community focus more on engagement’s psychological condition, even though the behavioral state of engagement was also mentioned The relationship that exists between the organization and its employees is pointed out, but with very few solutions about what the organization should do in practice to enhance the experience of engagement In addition to the psychological state of engagement, consultancies and research institutions include the role of organizations in enabling employee engagement in their definitions It also indicates that the organization’s values and goals should be aligned with an individual’s or a team’s performance to help employees realize that their efforts are making a valuable and significant contribution Organizations ultimately put the theory of engagement into practice They offer a useful window into how engagement is perceived and applied in the

9 actual world There are some examples of employee engagement definitions by organizations across many public and private sectors

From Vodaphone’s perspective, employee engagement was a result that was assessed or seen as coming from people's willingness to give their all for something or someone in the company

Being engaged, according to Nokia Siemens Networks, is defined as having an emotional connection to the company, taking pride in it and being willing to speak up for it, understanding the business’s long-term objectives and values and how its employees fit into them, and being motivated and willing to put in extra effort to go above and beyond (Nokia Siemens Network, 2008)

The degree to which an employee feels a sense of commitment to the company they work for, trusts in its objectives, and embraces its values is what Barclays defines as employee engagement Barclays also suggests that you may get a good idea of someone’s level of engagement by simply asking them if they would recommend Barclays as a place to work (Barclay, 2008)

In general, there are some key differences in the definition of employee engagement from various sectors However, there are some aspects on which they all agree in the definition They agree that engagement is the feeling of attachment towards the organization for which they work, so that they invest their effort not only in the tasks but also in the organization as a whole They also propose that through discretionary effort, advocacy, pride, and organizational commitment, engagement is what employees give for the benefit of the organization.

Similar concepts

Some studies have equated the concepts of job satisfaction and engagement However, most consider engagement to be a broader concept

10 than satisfaction Gallup discovered through its surveys that employees are more engaged with the company when they are more satisfied in their job However, Gallup also found that job satisfaction is not closely related to performance and business outcomes, while engagement is According to Koscec (2003), whereas engagement is an active state for the employee, satisfaction is a passive state Even though their contribution to the company is negligible, Koscec thinks that some employees can be completely satisfied with the company as long as they get what they want in terms of pay, benefits, location of their workplace, paid time off, etc Such workers are content, regardless of how well or poorly the business performs From satisfaction studies, two things can be drawn:

1 Satisfaction is a minor outcome of employee engagement

2 Satisfaction differs from employee engagement in that, under work pressure, employees can show effort, but they don’t have the passion and vigor that tend to characterize engagement

In studies, engagement and commitment are frequently used interchangeably For example, according to the Corporate Executive Board, engagement is the degree to which workers devote themselves to a project or person inside their organization, the amount of effort they put in, and the length of time they stay as a result of that dedication It is believed that commitment is a psychological state of attachment or a unifying factor between a person and an organization Based on these definitions, commitment includes various forms of social behavior and organizational or job withdrawal From these findings, it can be said that commitment is one of a number of states of engagement and is encapsulated within the concept of engagement Besides shared characteristics, some distinct elements of

11 engagement, such as absorption and self-expression, are not found in organizational commitment

According to May et al., 2004, job involvement is the level of importance a job is to the individual and his identity

In some studies, involvement is sometimes confused with engagement Harter et al (2002) equated engagement with both job satisfaction and job involvement However, from May et al (2004) viewpoint, engagement and job involvement are not the same thing; job involvement is only a cognitive assessment of the job’s capacity to meet needs, while engagement requires the active use of feelings and actions in addition to cognitions Similarly, according to Maslach et al (2001), job involvement differs from engagement because it doesn’t seem to include the aspects of efficacy and energy As a result, engagement is seen to be more intricate and multifaceted, covering the idea of job involvement.

Previous research

Kahn's 1990 work drew inspiration from Hackman and Oldham's Job Characteristics Model, emphasizing the importance of job design in employee motivation Additionally, Goffman's research on social interactions and Maslach and Jackson's Burnout Syndrome Theory provided a framework for understanding the psychological impact of work on employees These influences shaped Kahn's exploration of employee disengagement, which he attributed to emotional exhaustion and diminished psychological capacity to contribute.

Kahn outlined three key elements that have a significant impact on how much an employee can connect to the culture, mission, and daily

12 responsibilities of their role These are psychological meaningfulness, psychological safety, and psychological availability

Kahn’s employee engagement theory promotes a more holistic approach Rather than concentrating on immediate motivational initiatives, it aims to create a more all-encompassing plan that takes into account every need of an employee The major goal of the Kahn employee engagement model is to increase employee engagement by:

- Establishing a psychologically safe workplace where employees can be themselves and talk openly

- Using a collaborative approach with employees rather than a hierarchical one

- Engaging employees, being open to their feedback, and allowing people to take part in the process of establishing their own roles and accomplishing company goals

May et al (2004) research contributes to Kahn’s theoretical framework by exploring the relations between elements, and propose a framework as below

Figure 2.1 Revised framework of engagement

From “The psychological conditions of meaningfulness, safety and availability and the engagement of the human spirit at work” (p 28), by May et al (2004)

May et al (2004) surveyed 213 insurance employees to establish a robust positive connection between psychological meaningfulness and psychological safety and engagement While psychological availability initially showed no significant relationship, it was found to affect engagement when considering the direct effect of resources.

In modifying their model, May et al found that the influence of psychological availability on engagement may be mitigated by an employee’s perception of the resources that are available

The Job Demands-Resources (JD-R) model of burnout (Demerouti, Nachreiner, and Schaufeli, 2001) proposes that working conditions can be

14 divided into two major groups, job resources and job demands, that have a differential relation to certain outcomes This study involved 374 employees from the northern region of Germany who were employed in one of 21 different occupations across three occupational fields: human services, manufacturing and chemical industry, and transportation The model suggests that there are two processes through which burnout can develop The first process involves extremely job demands, which cause continuous overtaxing and ultimately exhaustion (but not disengagement) In the second process, it becomes harder to meet job demands due to a lack of resources, which causes withdrawal behavior The long-term consequence of this is disengagement (but not exhaustion) from work The model predicts employees will get exhausted and disengaged from their job when there are high demands and inadequate resources This state represents burnout syndrome The research focuses on external factors of job resources such as job control, participation in decision-making, potential for qualification, support from colleagues and supervisors, etc., through which organizations can get employee engagement

Figure 2.2 The job demands-resources model of burnout

From “The Job Demands-Resources Model of Burnout” (p 502), by Demerouti,

Drawing on research from personality, transformational leadership (Bass, 1999), and the job characteristics theory (Hackman & Oldham, 1980), Macey and Schneider (2008) found that a number of antecedents should influence a person's desire to self-invest his own energies into achieving high work performance Consequently, Macey and Schneider (2008) stated that work engagement should be directly related to job characteristics, leadership, and personality traits, and as a result, indirectly related to performance

Figure 2.3 Framework for understanding components of employee engagement From “The Meaning of Employee Engagement” (p 6), by W.H Macey and

Phan Thanh Nha and Truong Thi Kim polled 284 employees at An My Fish Joint Stock company to investigate the elements influencing employee engagement with the organization The findings indicated that seven elements influence employee engagement: work compensation, leadership, colleague, training, working environment, nature of work and commendation, and bonuses and benefits The authors also discussed the management implications for the factors Work compensation, Work environment, and Leadership Work compensation is a crucial aspect in attracting and maintaining employees, so a fair and appropriate salary policy is required To develop a comfortable working environment and an emotional tie to the workplace, organizations must establish an identifiable corporate culture that assists employees in identifying the goals and direction of the company’s operations Leadership has significant impacts on employee engagement

Leaders must demonstrate outstanding, honesty, and consistency behaviors, as well as continuously learn, listen, respect employees’ ideas, and appropriately evaluate staff skills

A study by Truong Phuong Khanh found that transformational leadership, particularly its components of intellectual stimulation and individual consideration, positively influences work engagement in IT companies Leaders who foster innovation, support employees, and adapt to evolving practices encourage employees to acquire and share knowledge, fulfilling their self-actualization needs Moreover, by providing guidance, empowerment, and support, individual consideration promotes work efficiency, self-assertion, and overall work engagement among employees.

Research model

While there have been significant studies on employee engagement, particularly in the hospitality industry, banking sector, healthcare, call center, etc., conducting research on engagement specifically in the Information and Communication Technology (ICT) sector remains necessary The ICT sector possesses unique characteristics and dynamics that can impact employee engagement differently compared to other industries In the ICT sector, task autonomy - the level of independence and control employees have over their job - is highly relevant because employees often require flexibility and freedom to solve complex problems The quality of leader-member exchange,

18 where leaders establish relationships with their team members, is essential for fostering engagement as it influences communication, trust, and support Moreover, social support from colleagues and direct managers plays a vital role in mitigating stress and promoting engagement in a sector known for its high-pressure work environment Lastly, the availability of training and development opportunities is crucial in the rapidly changing ICT landscape, as employees need to continually update their knowledge and skills to stay motivated Therefore, conducting research on employee engagement in the ICT sector with a focus on these specific factors can provide valuable insights to tailor engagement strategies, address sector-specific challenges, and enhance the well-being of employees and organizational outcomes

Newman et al (2011), in an effort to study the construct validity of employee engagement, clarified that there are two types of employee engagement: attitudinal engagement and behavioral engagement, and that attitudinal engagement will strongly affect the behavioral engagement Attitudinal engagement (or “state engagement”; Macey and Schneider, 2008) includes well-known content such as employee vigor, dedication, absorption, pride, and identification Attitudinal engagement, a central aspect of employee engagement, encompasses not only the enthusiasm and commitment that employees bring to their work but also the underlying emotional and cognitive dimensions that drive these attitudes Pride and identification are two important components of attitudinal engagement When workers recognize that their work, their team, or their organization adds value and makes a difference, they feel pride, which is a durable emotion that motivates them to share it with others Identification, which focuses on the way I see myself in connection to my company, has to do with how well one’s beliefs and values align with the culture, core values, and missions of the organization When employees identify with these values of their

Employees with organizational identification and pride exhibit a holistic understanding of engagement, encompassing both affective (emotional) and cognitive (reasoning) factors This engagement drives not only job satisfaction but also deep commitment to the organization's mission and goals Behavioral engagement, characterized by the shared variance of citizenship behavior, withdrawal from work, and job performance, represents the investment of time and effort into one's work It extends beyond basic task completion and includes shared values and expectations within the workplace.

Based on these points, the author proposes the following research model:

Leader-Member Exchange (LMX) theory posits that organizational members develop varying levels of positive relationships with their managers, based on trust, respect, and mutual obligation Due to limited time and resources, supervisors categorize subordinates into "in-groups" and "out-groups." In-group members form strong bonds with their leaders, receive preferred work assignments, and extend beyond their job descriptions as "trusted assistants." Out-group members, lacking these positive interactions, engage in routine tasks and have limited compatibility with the leader.

But as it has evolved, LMX has been known as a successful leadership process that gives everyone in the team the chance to build a partnership with their leader A prescriptive approach to leadership known as “leadership making” places emphasis on developing meaningful relationships with all followers rather than just a few It aims to give each follower a sense of belonging to the in-group, therefore avoiding the unfairness and unfavorable effects of being a member of an out-group Generally speaking, leadership decision-making encourages teamwork, with the goal of the leader creating productive relationships with each member of the team

Arun Aggarwal et al., utilizing a sample of 1.163 employees in the R&D division of companies operating in several Indian cities in the IT and pharmaceutical industries, confirmed a positive relationship between LMX

21 and engagement However, in the study about work engagement, Michael S Christian et al (2011) noted that the path weight of LMX was nearly zero (β

= 06) in relation to engagement, suggesting that when other factors are considered, the practical significance of LMX may be negligible This research will try to confirm and further clarify these results in the context of VNPT Technology

Maslow’s hierarchy of needs states that there are five levels in which human needs can be organized, with fundamental physiological needs at the bottom and needs for self-actualization at the top Self-actualization needs are the highest level of needs and are characterized by the desire to fulfill one’s full potential, achieve personal growth, and engage in meaningful activities

Autonomy refers to the degree of control an individual has over their work tasks, including the freedom to make decisions and set goals When employees have a high level of autonomy, they have the ability to use their skills and knowledge to solve problems and make decisions independently, which can lead to a sense of competence and mastery This sense of competence and mastery is consistent with the self-actualization needs of Maslow’s hierarchy, as individuals are able to use their full potential and engage in meaningful activities Similarly, according to Humphrey, Nahrgang, and Morgeson (2007), one definition of autonomy is the freedom a person has to pursue their own task or the level of individual freedom, independence, and discretion that the work allows the employee with regard to the selection of work scheduling, work method, and decision-making Autonomy, therefore, has three aspects, which relate to:

Employee autonomy is a crucial factor in fostering engagement It can be categorized into two types: real and perceived Real autonomy refers to the flexibility granted to an employee to organize their work independently On the other hand, perceived autonomy relates to the employee's belief in their ability to freely plan their own work Both actual and perceived autonomy play a significant role in employee engagement and satisfaction.

In the digital age, there have been many changes to the design of jobs Additionally, traditional job characteristics are also changing and entirely new tasks and jobs are emerging as well Large volumes of data can now be collected and analyzed more quickly and easily thanks to technology, which both simplifies certain tasks and complicates others This implied that a person’s job might consist of a lot more tasks and be more complex In a complex and unstable environment, a hug amount of information needs accessed and processed while the people ability to make decisions is constrained Delegating authority is necessary to avoid errors throughout the decision-making process The modern competitive world, on the other hand, necessitates that organizations be very creative and flexible in order to endure and develop This requires confronting challenges and performing creative tasks on the job With increased autonomy, it is believed that employees can effectively meet these demands Specifically, autonomy boosts employees’ inventiveness and responsiveness - that is, provided they possess the necessary knowledge and expertise and are passionate about what they do According to the research of Bakker, Tims, and Derks (2012), cited by Aleksandra Boskovic (2021), such a high degree of autonomy can have a powerful motivating effect that also helps to improve performance

During and after the COVID-19 pandemic, work-from-home or remote working became a trend This approach provides high autonomy and low control, so evaluating the benefits and drawbacks of autonomy in this context is crucial

Autonomy is supposed to increase job satisfaction among employees who work in the information and communication technologies (ICT) sector because it enables them to maximize their skills Since technology is fast growing and competition is severe, jobs are often complicated and need the use of a range of skills, continual data collecting and processing, as well as the fullest use of knowledge and freedom Therefore, employees with a higher degree of freedom will be able to achieve better performance Strict control or any mechanism that relies on giving commands and communicating top-down is said to be unable to generate successful ICT work, given that the majority of ICT work is performed by professionals with narrow specializations These professionals clearly know the job requirements and how to respond to them

Amanda Shantz et al (2015), in their research “Role of Employee Engagement” studied 671 workers for a UK-based consulting and construction company and showed that autonomy, with a standardized estimate of 0.19, was positively related to employee engagement Aleksandra Boskovic collected data from 158 employees work in the information technology (IT) industry in the Republic of Serbia and confirmed that two of the three engagement aspects, which are vigor (β = 887) and dedication (β 890), are positively impacted by autonomy in a statistically significant manner Another dimension, absorption, was removed from analysis because it didn’t meet the CFA’s criteria The result complemented previous research (Saks, 2006; Christian et al., 2011; Shantz et al., 2013) and suggested that if autonomy is increased, a higher employee engagement can be expected However, in the study about work engagement, Michael S Christian et al

(2011) noted that the path weight of autonomy was nearly zero (β = 04) in relation to engagement, suggesting that, when other considerations are taken into account, the practical relevance of autonomy may be small The present

24 research will try to validate and further clarify these findings in the context of VNPT Technology

Hypotheses

Based on the literature review, the following hypotheses are proposed between the constructs as illustrated in Figure 2.5

H1: (a) The quality of LMX relationship, (b) autonomy, (c) training and development opportunities, and (d) social support will positively affect attitudinal engagement

H2: (a) The quality of LMX relationship, (b) autonomy, (c) training and development opportunities, and (d) social support will positively affect behavioral engagement

H3: Attitudinal engagement will positively affect behavioral engagement

H4: Attitudinal engagement will mediate the relationship between (a) the quality of LMX relationship, (b) autonomy, (c) training and development opportunities, (d) social support and behavioral engagement

Overview of VNPT Technology

Vietnam Post and Telecommunications Group (VNPT) and two other owners, Pacifab and VNPost, established VNPT Technology on January 6 th ,

Established in 2011 as a joint stock company with a capital of VND 500 billion, VNPT Technology promptly acquired Siemens' and Alcatel Lucent's capital contributions in TELEQ and ANSV Joint Venture companies, incorporating both entities as subsidiaries.

After nearly 20 years of joint venture firms ANSV and TELEQ, VNPT Technology inherited accumulated resources and knowledge The company now owns and is developing a team of professionals that have grown since the beginnings of the digital telecommunications network VNPT Technology is solidifying its position as the leader in the field of research and development as well as the manufacture of electronic, telecommunications, and information technology equipment This is made possible by the infrastructure that has been amassed and the joint ventures that have continued to grow it

Important development milestones of VNPT Technology:

 1993: Alcatel Network Systems Vietnam (ANSV) was established This is a joint venture between Alcatel Group (France) and Vietnam Post and Telecommunications Corporation

 1995: Telecommunications Equipment Co., Ltd (TELEQ) was established This is a joint venture between VNPT Group and Siemens

 2011: established VNPT Technology JSC To become the parent company of TELEQ and ANSV, VNPT Technologies purchased the obligations and rights of Siemens and Alcatel in their joint venture, which are TELEQ and ANSV, respectively Vietnam Innovation Value-Added Services (VIVAS) was founded by VNPT Technology in

July 2011 and is focused on offering value-added services and digital content

 2016: researched, developed, manufactured, and distributed 5 million products to the market Being the first enterprise in Vietnam operating in the fields of information technology and electronic communications to receive the High-Tech Enterprise certification The company continued to invest in building a second factory in Hoa Lac Hi-Tech Industrial Park to be able to supply 26 million products per year to the market by 2030

 2018 to present: researched, developed, manufactured, and distributed over 8 million products to the market Established representative offices in Nepal, Bangladesh, Indonesia, and the Philippines

VNPT Technology’s main activities are:

 The company’s primary focus for generating value from knowledge, intellectual property, and high-tech product and services is research and development (R&D) of technological products and original design manufacturing In order to provide complete hardware/devices, platforms, research and develop core technologies in five areas: digital transformation, fixed and wireless broadband, IoT, and 5G technology The goal of VNPT Technologies is to become an Original Design Manufacturer by 2025

To meet domestic and international market demands, the company manufactures products developed through research and development using electronic manufacturing services (EMS) This expansion of EMS services allows the company to optimize its infrastructure and resources, enabling it to participate effectively in the global supply chain.

31 company’s yearly revenue is currently derived from 70% of technological items and equipment, a ratio that is expected to increase over the next years

 Integrating telecommunication network system, information technology: with over three decades of expertise integrating big systems, such telecom networks, VNPT Technology is committed to becoming a pioneer in integrating telecom network system and information technology Analyzing, creating a complete solution, creating, putting into practice, and imparting training are all parts of complete system integration operations

 ICT technical services and domestic and international commerce: offer technology devices, integrated solutions, technical services to traditional telco, B2B and B2C

 SME digital transformation: a recently defined field that will concentrate on R&D from 2020 to 2025, along with the National Strategy to 2030 Assisting small and medium-sized businesses in shifting their ways of doing business in order to use technological solutions and conduct operations in the digital age to create new products and add new values

Current status of VNPT Technology personnel

As the VNPT Group’s core firm in the field of manufacturing technology products, VNPT Technology has a high-quality workforce, with more than 90% of office staff holding university degrees or higher

Table 2.1 Headcount by level of education

Approximately 78% of employees are engineers, who concentrate on product research & development, technology solutions, and telecommunications projects for operators such as MobiFone and VNPT This personnel structure is relatively balanced and appropriate for a technology organization

Employees feel safe when contributing to the company because of good remuneration and increasing average income, which motivates them to increase labor productivity The annual income growth rate is a little bit slower than the productivity growth rate of each employee because the company is reinvesting in improving technology, expanding operations, developing new products, etc These investments can lead to long-term growth and sustainability, which can benefit employees in the future through more stable employment and potential future income growth

The turnover rate is around 8% and 10% This is a relatively low level in the ICT industry However, in recent years, this rate is on the rise.

Employee engagement is a psychological and motivational state that motivates workers to strive for difficult objectives and gives them motivation to achieve A leadership framework called leader-member exchange (LMX) uses three factors to assess the quality of the connection between an employer and an employee, which are trust, respect, and obligation LMX may have a positive effect on employee engagement because if employees frequently interact and get emotionally attached to their supportive leader, they can perceive psychological safety at work and enhance their level of engagement

To some extent, autonomy is reflected in self-actualization needs Employees’ rights to decide on the method and time as well as to make decisions when performing tasks will increase their perception of meaningfulness at work, which will lead to employee engagement Individuals vary their engagement based upon their perceptions of the benefits they receive Consequently, it is anticipated that workers will be more engaged in their work to the extent that they receive opportunities for training and development In terms of Social Exchange Theory (SET), when employees gain recognition from their company, they are obligated to reciprocate with a higher degree of engagement Similarly, social support, the degree to which a job offers chances for help and guidance from managers or colleagues, is a social attribute that is probably connected to engagement As a result, four (04) hypotheses describing the expected relationships between the latent variables of interest in this study have been proposed

METHODOLOGY

Research process

Research design

This study investigates the impact of various independent variables on employee engagement: Leader-Member Exchange (LMX), task autonomy, training and development opportunities, and social support Each variable is measured through validated, closed-ended questions presented in a questionnaire to staff and managers of ten (10) centers Employing these metrics, the study aims to uncover the relationships between these factors and employee engagement.

VNPT Technology through Google Forms The questionnaire items are translated from English to Vietnamese (forward translation) by the author The Vietnamese version is sent to two colleagues of the author, and they translate this version back into English (back translation) The final version of the scale is obtained after being carefully examined by the author and the two colleagues

Descriptive statistics and reliabilities of each of the variables and the latent factors are calculated, and a correlation matrix is used to identify the relationship between the constructs Since the current research utilizes previously well-known structures with sufficient theoretical and empirical basis to specify the model, Confirmatory Factor Analysis (CFA) will be used Hypotheses will be tested using Structural Equation Modeling Besides, a number of distinctions between groups will be found based on the demographic data received from participants.

Sample and data collection

After discussion with the managers of the 10 centers of VNPT Technology, full-time employees were invited to complete the online survey Among the 10 centers, the Technical and Service Center employs personnel in all three regions of Vietnam: the North, the Center, and the South The other centers have only northern (Hanoi) employees

An email was sent to the managers of each center, and with their help, we encouraged as much employee participation in the survey as possible The email and the survey form stressed the confidentiality of the survey process, aiming to ensure that this is just used for research objectives and that no individual information is collected in order to reduce response bias and maximize participation Besides, demographic data such as age, gender, education, length of service, etc were also collected for group differences

36 analysis purposes and to improve the participants perception of the anonymity of their responses

The survey was available from March 23 rd through April 4 th , 2023 During this time period, a questionnaire was distributed to 354 employees, including center staff and managers, of whom 224 provided valid responses The current total number of employees of VNPT Technology is 827, including office staff (354) and factory staff (473) The factors selected to study the impact on employee engagement come from the author’s observations when working with staff including engineers, sales, back office, etc These positions often require higher levels of education and professional training, which equip these employees with a deeper understanding of organizational processes, customer needs, and market trends By targeting these employees, the collected data will likely provide more insightful and relevant information for the research conclusions.

Measurement

The latent variables in this research includes independent and dependent variables listed below:

- Independent variables: Leader-member exchange (LMX), Autonomy, Training and development opportunities, Social support, Attitudinal engagement (*)

- Dependent variables: Attitudinal engagement (*) , Behavioral engagement

(*) There’s a multi-level dependence in which Attitudinal engagement is both an independent and dependent variable.

The demographics of interest in this research includes: gender, age, level of education, and job tenure

In order to measure each of the constructs in this study, the author used well-tested, valid, and reliable measurements that have been extensively

37 utilized by researchers These instruments include the LMX-7 (Graen, Uhl- Bien, 1995) and the Work Design Questionnaire (Morgeson, Humphrey,

2006) to measure task autonomy and social support Training and development opportunities and employee engagement are measured using items derived from Robinson D et al.’s (2007) study about employee engagement

The LMX-7 includes seven questions outlining the characteristics of the relationship between leaders and subordinates, including the quality of the relationship, the ability to understand job problems and needs, the ability to realize potential, and the willingness to support each other Five replies are provided for each item in the format of a nominal frequency scale such as (1) Rarely (2) Occasionally (3) Sometimes (4) Fairly Often (5) Very Often The quality of exchanges between the employee and the supervisor increases with higher scores

There’s controversy about the dimension of LMX The first to express concern regarding whether LMX is multidimensional or unidimensional was Dienesch and Liden (1986, p 624) They discovered that LMX is multidimensional in terms of perceived contribution, loyalty, and affect in response to this question After Dienesch and Liden research, LMX dimensions were tested by some researchers such as Duran, Howell, Maslyn in 1993 and 1994 The majority of the findings are homogeneity on a single dimension (Cronbach’s alpha ~ 80% - 90%), whereas multidimensionality is found when factors are forced in confirmatory factor analysis but not exploratory factor analysis Respect, trust, and obligation are the three (03) dimensions that Graen and Uhl-Bien (1995) theorized make up LMX-7 Based on prior studies as well as their own testing, they concluded that although the LMX-7 are multiple dimensions, yet are strongly correlated that

38 a single LMX measure may be used to assess them all Therefore, in this research, the author utilizes LMX as a unidimensional construct

The following latent variables will be measured using a five-point scale, with ratings defined as follows: 1 = strongly disagree, 2 = disagree, 3 neutral, 4 = agree, and 5 = strongly agree

The autonomy measurement originally designed by Hackman and Oldham (1974), and then expanded further by Morgeson and Humphrey

(2006), was included in this research Originally, autonomy was understood to be the degree of personal freedom and independence that a person has when performing their job; it is then enlarged to include the extent to which a job grants independence, freedom, and flexibility to plan work, make decisions, and select the most effective method to finish tasks

The nine (09) items of autonomy used in this research were developed by Morgeson and Humphrey (2006) by taking or adapting from previous research by Campion and McClelland (1991), and Karasek et al (1998) In Morgeson and Humphrey’s (2006) research, they tested five different models, and the 21-factor model in which they separated autonomy into three factors (work scheduling, work methods, and decision-making autonomy) fitted their data the best However, Brady et al (1990) in their studies found indications that prefer autonomy as a single scale Christian and Slaughter (2011) also viewed autonomy as a single dimension The present research utilizes the autonomy construct as a single structure

The eight-item measure of training and career development derived from Robinson D et al.’s (2007) study about employee engagement was used None of these items have a reverse score In Robinson D et al.’s (2007) study, a Cronbach’s alpha of 0.923 was achieved

Social support in the workplace encompasses the availability of guidance and assistance from managers, colleagues, and friends As a key indicator of job quality, social support is measured through a five-item scale that assesses managerial support, peer support, and opportunities for workplace friendship This measure was developed by Morgeson and Humphrey (2006), adapting previous research by Karasek et al (1998) and Sims, Szilagyi, and Keller (1976).

The twelve-item measure of employee engagement derived from Robinson D et al.’s (2007) study about employee engagement was used Combined with the research by Macey and Schneider (2008), and Newman et al (2011), these items are grouped into two constructs: the first to the seventh items showing advocacy, pride, positive emotion, etc., are treated as observed variables of the attitudinal engagement construct, while the others are grouped into the behavioral engagement construct None of these items have a reverse score.

Data analysis

SPSS version 20 is used to perform descriptive statistics for each observed variable and to conduct ANOVA analysis to check if there’s a difference between groups based on demographics

AMOS version 20 will be utilized to check the measurement model fit through Confirmatory Factor Analysis (CFA) and test the proposed hypotheses through Structural Equation Modeling (SEM) According to Fabrigar et al (1999), CFA is likely to be the appropriate method because it allows for focused testing of particular data hypotheses when there is enough theoretical and empirical support for a researcher to specify the model

There are multiple measures that can be used to evaluate the overall fit of the model Adopting Hu and Bentler’s (1995) criteria for evaluating model fit, apart from the relative chi-square statistic (χ 2 /df or CMIN/df, which is considered good if the value is less than 3), this research uses three more indices: the comparative fit index (CFI), the root mean square error of approximation (RMSEA), and the standardized root mean square residual (SRMR) For the CFI, a value above 95 indicates a close fit, above 90 suggests an acceptable fit, and above 80 is sometimes permissible With the RMSEA, a value below 06 indicates a good fit, and below 08 is an adequate fit model The value of 08 or less for SRMR indicates a reasonably good fit between the actual data and the hypothesized model

Cronbach’s alpha is the most frequently used to report reliability coefficient However, it is controversy that Cronbach’s alpha is typically inadequate Cronbach’s alpha assumes a form of measurement invariance called “tau-equivalence” Tau-equivalence means that all indicators of a latent construct have the same factor loading (i.e., indicator weights) on the relative latent construct In other words, it assumes that all indicators are equally related to the latent construct and contribute equally to its measurement However, in empirical studies, the observed variables (or indicators) demonstrate unequal factor loadings Therefore, in this research, construct reliability, or composite reliability (CR), a more appropriate reliability measurement, was used Hair et al (2010) noted that a CR value of 70 or higher indicates good reliability

Validity tests consist of convergence and discriminant validity tests Convergent validity refers to the degree to which different measures of the same construct are positively related or converge, indicating that they are measuring the same underlying concept Discriminant validity, on the other hand, is the extent to which measures of different constructs are not strongly

41 related to each other, indicating that they are distinct and measure different concepts

There are some ways to test convergent validity Researchers proposed that the magnitude of the standardized factor loading could be a criterion For example, Hair et al (2010) proposed that it should be at least 50 and, ideally, at least 70 In addition to standardized factor loading, it was proposed that convergent validity is confirmed when a latent construct accounts for at least half of the variability observed in its associated indicator The Average Variance Extracted (AVE) with the value of 50 or more is used to establish an acceptable level of convergent validity

Similar to testing convergent validity, there are some ways to test discriminant validity For example, some researchers proposed a fixed threshold of 75 (Voorhees et al., 2016), 85 (Kenny, 2016), or 90 (John and Benet Martínez, 2000) to compare the correlation between two constructs If the correlation between two constructs is less than 75, 85, or 90, the acceptable level of discriminant validity is reached This study used the AVE value as a criterion for examining discriminant validity Two criteria are applied to test discriminant validity: the AVE higher than the maximum shared variance (AVE > MSV) and the square root of the AVE higher than correlations between two constructs

The result of the analysis will be explained in detail through figures, tables, and narratives in Chapter 4

Chapter 3 briefly presents the research process, research design, sampling, and information collection methods for this quantitative research A total of 29 observed variables from 04 latent constructs that affect employee engagement and 12 observed variables that measure engagement level were collected and analyzed

RESEARCH FINDINGS

Summary of the survey result

A total of 354 questionnaires were sent to the employees of 10 centers of VNPT Technology The number of valid responses is 224, which represents a response rate of 63.3%, which is a good rate

The gender distribution of the respondents (127 males:97 females) reflects a significant imbalance (56.7% male:43.3% female), mirroring the gender disparity within VNPT Technology, an ICT industry workforce known for its male dominance.

Nearly seventy-four percent (73.7%) of the respondents are under 35 years old, with about fifty-nine percent (58.9%) between the ages of 25 and

34 The other group consists of 26.3% of the respondents older than 35 years of age It demonstrates that the company’s workforce is of an appropriate age to work in the ICT industry in Vietnam

Thirty-six percent of respondents have less than five years of experience with VNPT Technology This is nearly equivalent to the percentage of respondents who have worked for the company between 5 and

10 years (41%) The remainder have employment tenures longer than ten years

A great number of respondents achieved an undergraduate or graduate degree Among the 224 respondents, 77% of them earned at least a bachelor’s degree, and 17% of them have a master’s or doctoral degree

Figure 4.3 Level of education distribution

Descriptive statistics

The mean and standard deviation of the latent variables are provided in Table 4.1, which shows that the scales display good variability with few minimum or maximum values

Factor N of items Mean SD

Between group differences

SPSS version 20 is used to conduct an ANOVA so that the means of each factor can be compared across various demographic groups If statistically significant differences are discovered, a confidence interval of 95% is calculated to identify the groups between which the differences are found The following demographic variables are considered: age, gender, education level, and job tenure

The result of the test of homogeneity of variance shows that the sig values of all constructs are greater than 05, which means there is no difference in variance between gender groups Therefore, the F-test results in the ANOVA table are used Since all the sig values in the ANOVA table are greater than 05, it can be concluded that there is no statistically significant gender difference relative to all the factors (see Table 4.2 and Table 4.3)

Table 4.2 Test of homogeneity of variance for gender

Factor Levene Statistic df1 df2 Sig

Table 4.3 ANOVA: between group differences X gender

Relationship with Gender Sum of

The sig values of the Levene test for LMX, SS, and ATTI are greater than 0.05, whereas they are less than 0.05 for the other three constructs (Table 4.4) This means there is an equal variance between groups for LMX, SS, and ATTI The F-test results from the ANOVA table (Table 4.5) are used to test differences in LMX (F = 5.63, sig = 00) on age, whereas the F-test results from the Welch table are used to test differences in BEHA (F = 4.83, sig .01) on age The results of the Bonferroni test indicate that employees over

48 the age of 35 perceive their relationships with organizational leaders to be superior to those of their younger counterparts Similarly, the results of the Tamhane test show that age groups over 35 believe they are more actively involved with the company than age groups under 25 (

Table 4.4 Test of homogeneity of variance for age Factor Levene Statistic df1 df2 Sig

Table 4.5 ANOVA: between group differences X age Relationship with Age Sum of

Table 4.6 Multiple comparisons between groups of age

1: less than 25 years old; 2: from 25 to 35 years old; 3: more than 35 years old

The results of the Levene test indicate that all sig values are greater than 05, suggesting that there are no significant differences in the variance between groups Therefore, the F-test in the ANOVA table is used to examine differences in mean scores between groups based on level of education Statistically significant differences are observed in LMX, TD, AUT, SS, and BEHA, indicating that these dimensions vary across different educational levels.

The aforementioned constructs have different mean scores, most notably between the “College” group and the other two groups In all five constructs, the “College” group differs from the “Graduate” group, whereas LMX, SS, and BEHA have additional differences between the “College” and the “Undergraduate” group According to the data, the group with the lowest level of education perceives its relationship with the company’s leadership and coworkers to be inferior to that of the other groups In addition, they do not exhibit the same level of involvement in their jobs and organizations as the other groups Due to the considerably smaller number of respondents in the “College” group (N = 14) compared to the “Undergraduate” group (N 51

172), it is not possible to confirm that there is a statistically significant difference between these two groups when LMX, SS, and BEHA are considered The data demonstrates that the mean scores for all five constructs increase with increasing levels of education

Table 4.7 Test of homogeneity of variance for level of education

Factor Levene Statistic df1 df2 Sig

Table 4.8 ANOVA: between group differences X level of education

Level of Education Sum of

Table 4.9 Multiple comparisons between groups of level of education

1: Graduate; 2: Undergraduate; 3: College/Technical school

The result of the Test of Homogeneity of Variance (Table 4.10) reveals that the sig values of all constructs are greater than 05, which means there’s no difference in variance between groups Therefore, the F-test results (Table 4.11) are used Since all the sig values in the ANOVA table are greater than 05, it is possible to draw conclusion that there is no statistically significant difference in job tenure relative to all the factors

Table 4.10 Test of homogeneity of variance for job tenure

Factor Levene Statistic df1 df2 Sig

Table 4.11 ANOVA: between group differences X job tenure

Structural equation modeling

As mentioned in Chapter 3, the present research uses well-tested, valid, and reliable measures that have been extensively utilized by researchers The latent variables in this model are as follows: Leader-Member Exchange (LMX), Task Autonomy (AUT), Training and Development (TD), Social Support (SS), Attitudinal Engagement (ATTI), and Behavioral Engagement (BEHA)

All of the observed variables are also included in the model These consist of Graen and Uhl-Bien’s LMX-7 items (LMX1 – LMX7), nine items of task autonomy measures (AUT1 – AUT9), and five items of social support measures (SS1 – SS5) from the Work Design Questionnaire (Morgeson and Humphrey, 2006); and Robinson D et al.’s (2007) eight items of training and development (TD1 – TD8) The twelve items of engagement derived from Robinson D et al.’s

(2007) study, adapted from Macey and Schneider’s (2008) and Newman et al.’s

(2011) research, are distributed into attitudinal engagement (ATTI1 – ATTI7) and behavioral engagement (BEHA1 – BEHA5)

In this model, 102 distinct parameters are estimated, and the critical ratio (CR) for each regression weight shows that all the paths are statistically significant (p < 001)

The measurement model of the relationship between variables in a standardized solution indicates an acceptable fit with CFI = 877, χ 2 /df 2.113, SRMR = 072, and RMSEA = 071

By removing some observed variables from the full model, a slightly modified model with fewer parameters to estimate was created to improve the model fit This model may help analyze and interpret the data better

The multiple R-squared correlations (R 2 ) indicate the proportion of variability in the observed variables elucidated by the corresponding underlying variables Each observed variable’s R 2 and standardized regression weight (factor loading) were carefully checked, and these factors had a significant impact on which variables were removed from the model The remaining observed variables also need to build a balanced structure for the latent variable Therefore, the items removed from the full model consist of

01 item of leader-member exchange, training and development, attitudinal, and behavioral engagement; 02 items of social support, and 03 items of task

57 autonomy Table 4.12 lists the removed observed variables, while Table 4.13 outlines the retained observed variables of each construct

Table 4.12 Observed variables removed from the model Table 4.12 Observed variables removed from the model

Table 4.13 Observed variables retained in the modified mode

The model’s parameter estimates were once again generated using these revisions The result is illustrated in Figure 4.4 with the standardized estimates In the revised model, 83 distinct parameters were estimated With CMIN/df = 1.530, CFI = 952, RMSEA = 049, and SRMR = 052, this data fits the model well

Figure 4.4 Revised measurement model: standardized estimates

Despite prior empirical validations, translating the scales into Vietnamese before the survey may have introduced deviations Therefore, it is essential to examine the reliability, convergence, and discriminant validity of the scales before utilizing them in analyses or hypothesis testing to ensure accurate and meaningful interpretation of the results.

The values of CR, AVE, and MSV were calculated by utilizing an Excel macro developed by Assoc Prof James Eric Gaskin, Ph.D (Marriott School of Business, Brigham Young University) The results of these tests are summarized in Table 4.14 The CR values of all the latent constructs exceed 80, indicating good reliability of the scales The AVE values also confirmed that a good level of convergent validity was established, with the lowest value of 524 Moreover, every standardized factor loading within the model exhibited statistical significance, surpassing 70, with the exception of 3 out of

The measurement model demonstrated good internal consistency and reliability, with high loadings for all observed variables (LMX2 = 666, ATTI7 = 683, and BEHA5 = 595) Discriminant validity was also established, as evidenced by AVE values for each latent construct exceeding their corresponding MSV values Additionally, the square roots of AVE for all constructs (values in bold) were greater than the correlation coefficients between any two constructs, further supporting discriminant validity.

Table 4.14 The result of the reliability and validity test

CR AVE MSV ATTI TD AUT SS LMX BEHA ATTI 906 618 518 786

In order to test the relationships between constructs, bivariate correlations are calculated In order to calculate the correlations, the mean scores of each construct are calculated The constructs in this analysis are composite variables of all the survey items of the respective construct The result in Error! Reference source not found shows that except TD has negative impact on BEHA, AUT doesn’t have statistically significant impact on ATTI, other independent variables have statistically significant, positive impact on dependent variables (ATTI, BEHA)

The structural equation modeling was completed using AMOS 20 All observed variables were revised, and their corresponding latent constructs were placed in the model The model contains the measurement model and the structural model The measurement model includes the measurement equations for each observed variable For example, LMX2 is regressed on the LMX factor plus measurement error e23 The structural model contains (i) the regression of ATTI and BEHA on the LMX, AUT, TD, and SS factors (one regression equation with each of ATTI and BEHA as dependent variables), as well as the regression of BEHA on the ATTI factor, displayed by single- headed arrows, and (ii) the estimated covariances between exogenous factors such as LMX, AUT, TD, and SS, displayed by double-headed arrows The maximum likelihood method is utilized to estimate the parameters The structural model is shown in Figure 4.5

Figure 4.5 Structural equation model: standardized estimates

The standardized solution is presented to facilitate the comparison of path coefficients Notably, path coefficients between observed variables and their corresponding latent variables exhibit statistical significance Specifically, LMX possesses statistically significant path coefficients with each dimension of attitudinal engagement (γ).

= 320, p < 001) and behavioral engagement ( = 259, p = 002) The path coefficient between social support and attitudinal engagement is statistically

65 significant ( = 217, p = 017), while the same does not exist between social support and behavioral engagement ( = 171, p = 280) The coefficient representing the influence of task autonomy on attitudinal engagement, however, is not significant ( = 114, p = 227) Even though a small negative linear relationship is found to exist between training and development and behavioral engagement, the path coefficients between training and development and both attitudinal engagement ( = 219, p < 001) and behavioral engagement ( = - 128, p = 015) are still statistically significant In addition, the path coefficient between attitudinal engagement and behavioral engagement is also statistically significant ( = 397, p < 001) Table 4.16, Table 4.17, and Table 4.18 summarizes this information together with estimates of path coefficients, standard errors, p-value, standardized coefficients, residuals, and multiple R- squares identified by the relationship between observed and corresponding latent variables

Table 4.16 Summary of relationships between latent and observed variables Observed variable

Std error p-value Standardized coefficient Residual R 2 Paths from LMX to:

Paths from Task Autonomy to:

Paths from Training and Development to:

Paths from Social support to:

Paths from Attitudinal engagement to:

Paths from Behavioral engagement to:

Table 4.17 summarizes the direct effects of independent latent variables on the two dependent variables

Table 4.17 Direct relationships among latent variables

Latent variable Path coefficient Std error Standardized coefficient p- value Direct effect of LMX on:

Direct effect of Task Autonomy on:

Direct effect of Training and

Direct effect of Social support on:

Direct effect of Attitudinal engagement on:

The strength of the linear relationships between independent and dependent variables is demonstrated in Table 4.18 through the multiple R-squared (R 2 ) of the dependent variables.

Table 4.18 Relationships between latent variables Dependent variable Independent variable Residual R 2

Attitudinal engagement LMX, AUT, TD, SS 193 506 Behavioral engagement LMX, AUT, TD, SS, ATTI 096 676

The findings relative to four hypotheses are outlined below

H 1 : (a) The quality of LMX relationship, (b) task autonomy, (c) training and development opportunities, and (d) social support will positively affect attitudinal engagement

A statistically significant path coefficient was found to exist between attitudinal engagement and each of LMX ( = 320, p < 001), training and development ( = 219, p < 001), and social support ( = 217, p = 017) A positive and significant relationship between attitudinal engagement and each of LMX, training and development, and social support based on correlation estimation (Error! Reference source not found.) with respective correlation coefficients of r = 609, r = 618, and r = 584 was also found to support hypotheses H1a, H1c, and H1d However, the path coefficient between task autonomy and attitudinal engagement was not statistically significant ( 114, p = 227) This means H1b is not supported Consequently, H1 is partially supported

H 2 : (a) The quality of LMX relationship, (b) task autonomy, (c) training and development opportunities, and (d) social support will positively affect behavioral engagement

A significant relationship exists between behavioral engagement and each of LMX (r = 623), task autonomy (r = 573), and social support (r 641, Error! Reference source not found.) In addition, the path coefficient

69 between behavioral engagement and each of LMX ( = 259, p = 002), and task autonomy ( = 312, p < 001) provides support to hypotheses H2a and

H2b Nevertheless, even though there is a significant relationship between behavioral engagement and training and development (r = 476, Error! Reference source not found.), training and development has a negative effect on behavioral engagement ( = -.128, p = 015) The analysis result indicates that the relationship between social support and behavioral engagement is not statistically significant ( = 171, p = 28) H2c and H2d are not supported Therefore, hypothesis H2 is partially supported

H 3 : Attitudinal engagement will positively affect behavioral engagement

H3 is supported A statistically significant path coefficient ( = 397, p < 001) between attitudinal engagement and behavioral engagement supports this hypothesis In addition, the correlation analysis also supports this hypothesis (r = 72, Error! Reference source not found.)

H 4 : Attitudinal engagement will mediate the relationship between (a) the quality of LMX relationship, (b) task autonomy, (c) training and development opportunities, (d) social support and behavioral engagement

This hypothesis assesses the mediating role of attitudinal engagement in the relationship between independent variables (LMX, task autonomy, training, and social support) and behavioral engagement Bootstrapping, a statistical technique for population parameter estimation, is employed for mediation analysis This technique involves random resampling with replacement, generating multiple samples from a dataset to approximate potential scores within the population.

70 dataset being viewed as the population In this study, the number of samples was 2000

The mediation analysis summary is presented in Table 4.19

Table 4.19 Indirect effects mediated by attitudinal engagement

Relationship Direct effect Indirect effect Conclusion

SS  ATTI  BEHA 0.187 0.065 0.094 0.018 Full Mediation

TD  ATTI  BEHA -0.191 0.044 0.13 0.001 Partial Mediation LMX  ATTI  BEHA 0.265 0.006 0.13 0.004 Partial Mediation AUT  ATTI  BEHA 0.269 0.001 0.039 0.163 No Mediation

SS = Social Support, TD = Training and Development, LMX = Leader- Member Exchange, AUT = Task Autonomy, ATTI = Attitudinal engagement, BEHA = Behavioral engagement

Finding interpretation

The quality of the relationship between a leader and their team members (LMX) positively influences employee engagement, both in terms of attitudes and behaviors For the entire sample, LMX had a statistically significant linear relationship with engagement ( = 320, p < 001 for attitudinal engagement and  = 259, p = 002 for behavioral engagement) Furthermore, ANOVA analysis revealed that perceived LMX quality increased with age and seniority, particularly among employees over 35 and those in managerial or directorial roles These findings suggest that employees with more time in the workforce may have developed stronger interpersonal, emotional intelligence, and communication skills, fostering greater trust and mutual respect within their relationships with leaders.

72 understanding, which can contribute to positive and effective interactions with their leaders, enhancing the quality of LMX The fact that staff perceived the quality of the relationship with their direct managers as lower than what managers perceived suggests that mid-level managers probably received the right amount of attention from high-level managers, while mid-level managers need to spend more effort to improve the relationship with their staff

TD exhibits a positive relationship with attitudinal engagement (ATTI; γ = 219, p < 001) due to its ability to enhance skills and foster emotional connections In contrast, TD and behavioral engagement (BEHA) are negatively related (γ = -.128, p < 015) This is attributed to a misalignment between employees' training expectations and actual outcomes, leading to decreased job performance and behavioral engagement.

73 engagement Employees might feel frustrated if they perceive the training as ineffective or irrelevant to their day-to-day work Another possible reason for the negative coefficient could be challenges in implementing the knowledge and skills acquired from training into practical work situations If employees encounter barriers or lack opportunities to apply what they have learned, it can hinder their ability to demonstrate behavioral engagement or even cause them to leave the organization for opportunities that are more suitable for their competence This is reasonable, especially for the ICT field, as the technology life cycle is getting shorter and the emergence of new technologies requires employees to constantly update their knowledge and skills Being trained but having little chance of practical application as well as having little opportunity to demonstrate one’s ability may lead to a discrepancy between the positive impact of TD on ATTI and the lack of a corresponding positive impact on BEHA

Task autonomy empowers employees, particularly in the ICT sector, by granting them control over their work methods, time management, and decision-making This autonomy enhances employees' perceived meaningfulness of their work, allowing them to fully utilize their abilities Elevated meaningfulness and autonomy lead to stronger organizational alignment, emotional attachment, and a sense of pride and belonging, ultimately fostering employee engagement and organizational commitment.

However, in this study, the relationship between AUT and BEHA is statistically significant ( = 312, p < 001), while that between AUT and ATTI is not ( = 114, p = 227) The work characteristics in the ICT sector can allow employees a high degree of autonomy to adapt in a rapidly changing environment to make decisions and test new ideas The direct link between AUT and BEHA may therefore be clearer and more observable in the

74 context of technology companies, resulting in a statistically significant relationship with BEHA Moreover, attitudinal engagement is a multifaceted construct that may be influenced by various factors beyond task autonomy alone, such as organizational culture, leadership, belief In the context of VNPT Technology, where innovation and creativity are highly valued, employees may prioritize factors other than task autonomy when forming their attitudes towards work

The analysis results show that SS has a direct effect on ATTI with a moderate relationship ( = 217, p = 017) and no direct effect on BEHA ( .171, p = 280) The relationship between SS and BEHA is fully mediated by ATTI ( = 094, p = 018)

Social support from direct managers or coworkers can positively impact employees’ attitudes, beliefs, and perceptions, leading to higher levels of attitudinal engagement Subsequently, when employees have higher levels of ATTI, they are more likely to exhibit higher levels of behavioral engagement

The full mediation suggests that the influence of SS on BEHA operates through psychological processes that are primarily captured by ATTI Social support can help meet employees’ psychological needs, arouse their sense of belonging, and stimulate their motivation for learning and development, all of which contribute to positive attitudes and emotional well-being These positive attitudes act as a foundation for subsequent behavioral engagement, as employees who are highly attitudinally engaged are more likely to exhibit proactive, dedicated, and persistent behaviors

SPSS version 20 and AMOS version 20 were utilized to analyze the data One-way ANOVA was used to check between group differences based on demographic variables to assess if a substantial difference in the mean

75 value of each latent variable between groups is found Significant differences were found in the mean scores of the survey items based on age and level of education, while the other two variables, job tenure and gender, did not show any significant difference between groups CFA and SEM were used to evaluate the fit of the data to the research model as well as the relationships between latent constructs SEM is also used to test the hypotheses mentioned in Chapter 2 The results show that the collected data is relatively fit with the proposed research model The results of the hypothesis testing are also illustrated in Figure 4.6

Note: p-value are in parentheses, s = support, ns = not support

Figure 4.6 Result of the hypotheses testing

RECOMMENDATION AND IMPLICATION

Chapter 5 presents recommendations and implications to maintain and improve the degree of employee engagement

In order to improve the role of LMX in employee engagement, mid-level managers, particularly direct managers, need to put more effort into:

- Synthesizing ideas and communicating effectively to convey these ideas to the target audience as well as inspiring others to take action

- Properly assessing the strengths and weaknesses of subordinates in terms of knowledge, skills, relationships, awareness, and potential

- Respecting and harmoniously coordinating the diversity in employees’ personalities and qualifications

- Coaching subordinates, leading with emotional intelligence, attract and retain talents

To achieve this, companies need to offer workshops and seminars on active listening, empathy, effective communication, and conflict resolution for managers By holding periodic meetings, summaries, sharing organizational updates, and setting forth clear expectations regarding roles and responsibilities, senior management can help employees understand the significance of the work they are doing as well as their responsibilities and personal interests in completing tasks This will increase inspiration and enthusiasm among the workforce

The result of the study confirms that there is a significant effect of autonomy on engagement Here are some recommendations to leverage task autonomy and its direct impact on engagement:

- Encourage employees to participate in decision-making processes related to their work Involving employees in decisions that directly

77 impact their tasks and responsibilities can increase their sense of empowerment and engagement

- Offer flexible work arrangements, for example, remote work opportunities or flexible hours, to provide employees with additional autonomy in how they manage their work-life balance Trusting employees to manage their time effectively can boost their motivation and commitment to achieving their goals

- While promoting autonomy, create opportunities for collaborative projects Collaboration allows employees to benefit from diverse perspectives and expertise while still having the freedom to contribute their own ideas and approaches This is a form of job enrichment and helps utilize the resources of each department

- Understand that different employees have varying preferences for autonomy Some may thrive with higher levels of autonomy, while others might prefer more guidance and structure

However, it’s worth noting that, while promoting autonomy, it’s crucial to provide employees with clear guidelines and boundaries, as well as based on individual expertise, to ensure they understand the parameters within which they can make decisions This helps prevent potential issues and ensures overall organizational performance In addition, it is essential to offer constructive feedback on their efforts and outcomes to help them understand how autonomy is contributing to their overall performance

The analysis results also show that TD has a moderate impact on employee engagement Here are some recommendations to improve the impact of TD on attitudinal engagement as well as resolve the negative relationship between

- Ensure that the training programs offered are relevant, practical, and directly applicable to employees’ job roles Even a well-planned

78 training program can fail if employees do not see the link between the program’s value and their professional goals Training that equips employees with the skills and knowledge they need to perform their tasks more effectively can lead to increased confidence and motivation, ultimately positively impacting their behavioral engagement

- Identify and address any organizational or operational barriers that may prevent employees from applying their training Ensure that challenging tasks are given to employees and that the necessary resources, tools, and support are available for them to implement their new skills effectively

- Ensure that leaders are supportive of employees’ training and development endeavors as well as regularly evaluate the impact of training programs on employees’ performance

- Encourage employees to participate in personalized training programs via online platforms such as Coursera and Udemy, so that they can manage their time flexibly

A good relationship between the employees and their coworkers and supervisors will positively impact employee engagement It is recommended to:

Creating a positive and supportive work environment fosters a sense of belonging and encourages employees to seek and provide social support Team-building activities, collaboration, and social events provide platforms for informal interactions and relationship-building among colleagues By fostering an environment where employees feel comfortable connecting and supporting one another, both personally and professionally, workplaces strengthen team cohesion and enhance overall well-being.

- Provide training for managers and supervisors on the importance of social support in the workplace and how to effectively provide support to their team members This can include active listening, empathy, and understanding the diverse needs of employees

- Develop official policies that emphasize the importance of social support within the organization For example, remote work opportunities or flexible hours that enable employees to manage their personal and professional responsibilities effectively can help employees better manage their time and reduce stress, leading to improved social support and engagement

This study explores the components of employee engagement within VNPT Technology using Macey and Schneider's (2008) and Newman et al.'s (2011) framework Results support Newman's categorization of employee engagement into attitudinal and behavioral elements, with attitudinal engagement being an antecedent to behavioral engagement Additionally, factors like leader-member relationships, training opportunities, task autonomy, and coworker and supervisor support emerged as distinct antecedents of employee engagement.

The additional contribution of this study is examining the mediating role of attitudinal engagement between other antecedents and behavioral engagement Although different studies in Vietnam investigated the antecedents and outcomes of employee engagement, few examined the mediating relationship The relationship between factors seems to be complex and may vary from one time to another time within individuals Emphasizing factors with stronger direct effects may lead to more immediate and tangible improvements in behavior engagement However, considering the mediated effects through attitudinal engagement can help address underlying attitudes and perceptions, which could have longer-term impacts on behavioral

80 engagement In addition, the identification of competitive mediation, where attitudinal engagement plays a mediating role while the direct effect is in the opposite direction, also has practical implications for organizations

The purpose of this research is to identify the determinants of employee engagement at VNPT Technology The research successfully covers concerns introduced at the beginning when exploring the multidimensional nature of employee engagement, comprising attitudinal engagement (ATTI) and behavioral engagement (BEHA), and its associations with key antecedents in the workplace The factors investigated included the quality of leader-member relationships (LMX), task autonomy (AUT), opportunities for training and development (TD), and support from coworkers and supervisors (SS) The research findings shed light on the complex interplay of these factors and their implications for organizational strategies to enhance employee engagement

The results demonstrated that LMX had the strongest positive effect on attitudinal engagement (ATTI), highlighting the significance of fostering supportive and positive relationships between leaders and employees to promote employees’ emotional connection to the organization On the other hand, task autonomy (AUT) emerged as the primary driver of behavioral engagement (BEHA), emphasizing the importance of providing employees with the freedom and empowerment to make decisions, take ownership of their tasks, and display their creativity and initiative Training and development (TD) exhibited a positive influence on attitudinal engagement (ATTI), emphasizing the value of investing in employees’ growth and professional development to foster a sense of fulfillment and satisfaction in their roles However, it was also found that TD had a negative direct effect on behavioral engagement (BEHA), indicating potential challenges related to skill application or support for implementation Moreover, the study revealed

81 that support from coworkers and supervisors (SS) had a moderate, direct, and statistically significant impact on attitudinal engagement (ATTI) only Furthermore, the relationship between SS and BEHA was fully mediated by ATTI, suggesting that the influence of coworker and supervisor support on behavioral engagement is channeled through employees’ attitudes towards their work