Determination of Servant Leadership on Worker Performance Mediated by The Quality of Work -Life and Moderated by Work Engagement

Worker performance is volatile and does not just appear. Good worker performance is stimulated by various organizational and individual factors. The research objectives to analyze the impact of serv-ant leadership on worker performance mediated by the quality of work-life and moderated by work engagement. This research is an explanatory type. All staff of PT Alamjaya Wirasentosa as many as 134 people became the population and sample. Saturated sampling are used in determining the research sample. The analytical technique used causal step mediation test and the absolute difference value moderating test. The results prove that servant leadership has a significant positive impact on employee engagement mediated by the quality of work-life. Servant leadership has a significant negative impact on employee engagement moderated by work engagement. In this case, work engagement weakens the impact of servant leadership on worker performance. Further research needs to be done on a larger sample size by exploring variables that can strengthen the influence of servant leadership in improving employee performance to contribute to the organization’s competitive advantage.


INTRODUCTION
High-performing employees are important assets that determine the company's competitiveness.Hasibuan (2017) defines worker performance as a person's success in carrying out a job.Bakker & Schaufeli (2013) stated that worker performance indicators include vigor (spirit), dedication, and absorption (appreciation) in doing work.Worker performance in the company fluctuates from time to time.It is therefore important for companies to monitor and maintain a good level of worker performance.Based on Junita (2016), good worker performance does not just appear but is stimulated by various organizational practices implemented in the organization as well as aspects inherent in the employees themselves ( Junita, 2017).Individual and environmental factors are important in explaining worker performance (Bakker et al., 2013;Alessandri et al., 2014;Sittar, 2020).The research results of Devita & Musadad (2017), Sari et al., (2021) are well proven that worker performance is influenced by various factors, including company support and managerial ability of leaders by building a safe work system, harmonious industrial relations, facilities for the optimal growth of competence and work motivation.
One of the organizational factors that are an important concern of this research that affects worker performance is servant leadership.Servant leadership is a form of leadership that gives extra attention to subordinates in the form of mutual understanding, sharing, empathy, listening with a sense of togetherness and full attention (Spears, 2010;Greenleaf, 2011).Leaders can use their authority to move others.Leaders who can carry out their roles well can support employees to perform with good quality.Leaders play a critical role in supporting the quality of employee work and the success of the company.
Based on various variations of previous research findings, This research focuses on the determination of servant leadership on employee performance through the quality of work-life as a mediating variable.With the nature of the leader who protects and provides an example to his subordinates, it will increase the confidence of employees to work better.The quality of work-life means that the company's leaders become leaders and coaches.Openness and trust are very important, so the quality of work-life must change to be better in the future.Employees also need to be fully involved in the completion of their work.By implementing good quality of work-life, employees are healthier, more committed, and safer at work, and reduce organizational expenses (Horst et al., 2014).The existence of quality of work-life will maintain the desire of employees to continue to work together as a team and survive in the organization to improve worker performance (Pradana et al, 2013;Nurbiyati, 2014;Hedayati & Nabiee, 2017;Gunawan, 2018;Al-Hawasyi et al., 2019;Alberto et al., 2020;Mutholib, 2020;Setyaningrum & Pawar, 2020).
The individual factor that affects worker performance and becomes the focus of this research study is work engagement.Work engagement is the degree to which individuals identify with their work, actively participate in it, and consider their performance to be important for their self-worth (Liwun & Prabowo, 2015;Bakker & Leiter, 2010).The level of employee work engagement can be seen in the utilization of physical, cognitive, and emotional energy in completing work (Crawford et al., 2010).According to Abutayeh & Al-Qatawneh (2012), work engagement is the main attitude that refers to the employee's psychological identification of his job, where employees feel that work is representative of their life and many of their interests and life goals are related to their work.Therefore, employees who have high work engagement see their success at work as an indication of success in life as a whole.In contrast, employees with low work engagement feel that their lives are separated from work and other things that are much more important than their work.
Employee work engagement is diverse and fluctuates from time to time which can then lead to variations in worker performance levels (Beal et al., 2005;Sonnentag, 2011;Ferreira-Oliveira et al., 2017).Work engagement can have both positive and negative impacts on both the organization and employees (Listiau et al., 2017).The results of previous research prove that the effect of work engagement on worker performance provides different findings.Various researchers have proven that work engagement has a significant positive contribution in generating high worker performance (Kim et al., 2012;Jagannathan, 2014;Jackson, 2014;Azizah & Gustomo, 2015;Yongxing et al, 2017;Qodariah et al., 2019;Rana et al., 2019;Nas et al., 2020;Hendrik et al., 2021;Pitaloka & Putri, 2021, Satata, 2021).However, other researchers have proven the opposite (Kusumawati, 2017;Setyaningrum & Pawar, 2020;Baharsyah & Nugrohoseno, 2021).Work engagement has a weak correlation to worker performance between individuals and different genders (Sittar, 2020).The inconsistency of the findings of previous research shows the fact that the work engagement variable can play a role in strengthening or weakening worker performance.The findings of these studies become the basis for testing work engagement as a moderating variable in this research, to test whether work engagement participates in strengthening or weakening the impact of servant leadership and worker performance.
This research has interesting novelties in several ways.Integrated observation of organizational aspects (servant leadership) and individual aspects (work engagement) and their contribution to worker performance.In addition, this research also observes the impact of servant leadership on worker performance by including the mediator variables (the quality of work-life) and moderation (work engagement) in an integrated model.The research results are expected to provide an overview of the variables that indirectly involved in linking servant leadership and worker performance, in this case, the variable quality of work-life.Likewise, it will determine the variables that can strengthen/ weaken the impact of servant leadership and worker performance, that is, work engagement.This research gap will be resolved through this research.
The research was conducted at PT Alamjaya Wirasentosa which is one of the consumer goods distribution companies in Sumatra (Indonesia).PT Alamjaya Wirasentosa is a company headquartered in Tanjung Morawa Deli Serdang, North Sumatra, which is a product distribution network for the Indofood Group.The marketing area covers Aceh (NAD) & North Sumatra.To increase distribution optimally, worker performance needs to be improved to develop the company more rapidly.The phenomenon of the gap in worker performance at PT. Alamjaya Wirasentosa is seen in the achievement of sales targets in the last 3 years which are volatile (up and down).Various factors are thought to influence fluctuations in the achievement of employee sales targets.The conditions in the company environment can be perceived differently by employees, which will shape the individual attitudes and work behavior of employees.This research is limited to analyze the impact of servant leadership on worker performance mediated by the quality of work-life and moderated by work engagement.Based on the research and phenomenon gap.The findings of this research are expected to produce theoretical implications, especially to confirm organizational and individual factors that determine worker performance as well as practical implications for improving worker performance at a consumer goods distribution company, namely PT Alamjaya Wirasentosa.

Hypothesis Development The Impact of Servant Leadership on Worker Performance Mediated by The Quality of Work -Life
According to Nurbiyati (2014), efforts to condition quality of work-life are full of challenges, especially if it is not assisted by various parties.Manager support is very necessary and important to meet the needs of members and organizations.Quality of work life can meet the expectations and satisfaction of workers through experience in the organization.Its philosophy is that improving the quality of work life involves efforts from various parties at every level of the organization to protect human dignity.
Servant leadership uses a fundamental and long-term approach, which in the end will provide a complete change in the personal and professional lives of employees.The existence of a leader's privilege to serve and help the difficulties of subordinates will be more respected by employees.This mutual caring attitude triggers a conducive work climate which will ultimately improve worker performance (Astohar, 2012).The implementation of service leadership will have an impact on variations in worker performance (Tatilu, 2014;Rahayu, 2019).Based on Setyaningrum & Pa-war (2020) research, the quality of work-life mediates servant leadership and worker performance.The existence of quality of work-life will maintain the desire of workers to continue to work together as a team and survive in the organization to improve worker performance (Pradana et al, 2013;Nurbiyati, 2014;Hedayati & Nabiee, 2017;Gunawan, 2018;Al-Hawasyi et al., 2019;Alberto et al., 2020;Mutholib, 2020;Setyaningrum & Pawar, 2020).Based on these findings, the statement of hypothesis 1 research is: H1: The quality of work-life positively and significantly mediates servant leadership and worker performance.

The Impact of Servant Leadership on Worker Performance Moderated by Work Engagement
Job engagement is a positive work-related state of motivation and energy as well as an employee's genuine desire to contribute to work and organizational success (Albrecht, 2010).Highly engaged workers will be more concerned and able to work together to contribute maximally to the interests of the company.Companies need employees who are tied to their work because highly engaged workers will show the best performance at work (Bakker & Leiter, 2010).
Employee work engagement is diverse and fluctuates from time to time which can then lead to variations in worker performance levels (Beal et al., 2005;Sonnentag, 2011;Ferreira-Oliveira et al., 2017).The results of previous research prove that the impact of work engagement on worker performance provides different findings.Various researchers have proven that work engagement has a positive impact on the high performance of workers (Kim et al., 2012;Jagannathan, 2014;Jackson, 2014;Azizah & Gustomo, 2015;Yongxing et al, 2017;Qodariah et al., 2019;Rana et al., 2019;Nas et al., 2020;Pitaloka & Putri, 2021, Satata, 2021;Hendrik et al., 2021;Manalu et al., 2021).However, other researchers have proven the opposite (Kusumawati, 2017;Setyaningrum & Pawar, 2020;Baharsyah & Nugrohoseno, 2021).Work engagement has a weak correlation to worker performance between individuals and different genders (Sittar, 2020).The inconsistency of the findings of previous research shows the fact that the work engagement variable can play a role in strengthening or weakening the impact of servant leadership on worker performance.Servant leadership has a responsibility to serve the interests of its employees so that they are more prosperous so that followers will do the opposite to be fully committed to working by the direction of the leader to achieve organizational goals.Based on these findings, the statements of hypothesis 2 research are: H2: Work engagement moderates positively and significantly servant leadership and worker performance.

METHOD
This research is a causal type and analyzes the impact between variables with a quantitative approach (Sugiyono, 2019).The research was conducted at PT Alamjaya Wirasentosa, as one of the consumer goods distribution companies in Sumatra (Indonesia) with a distribution network covering Aceh (NAD) & North Sumatra.The performance of PT Alamjaya Wirasentosa is largely determined by the performance of its employees.Various factors that are conditioned by the company such as Servant Leadership and Quality of WorkLife, as well as factors that exist within employees such as Work Engagement are thought to affect the level of worker performance.This is the important reason for researching at that place.
The population as well as the research sample are employees at PT Alamjaya Wirasentosa from all parts of the company with a total of 134 people.Determination of the number of samples is done by sampling technique nonprobability sampling that is saturation sampling.
The servant leadership variable is defined as the behavior of leaders who prioritize the needs of others, aspirations, and interests of others over their own (Sendjaya & Sarros, 2012).Servant leadership is measured by the indicator is.According to the indicators of love, empowerment, vision, humility, and trust (Dennis, 2014).Worker performance variables are defined as obligations and bonds between abilities and motivation (Robbins, 2016).Wayne (2018) defines uality of worklife as a management's concern about the impact of work on people in relation to their work, career, income and destiny at work.According to Walton (2014), the indicators used to measure the quality of work-life are growth and development, participation, an innovative reward system, and work environment.Work engagement is the composition of staff understanding of their duties, actively participating in their work, and viewing performance at work as more important for their goodness (Robbins & Judge, 2013).According to Liwun & Prabowo (2015), indicators for measuring work engagement include employee attitudes towards their work, work being self-identity, and attachment between self and work.
The data analysis technique used is the Barron & Kenny (1986) causal step mediation test to test hypothesis 1 and the absolute difference value moderation test (Frucot & Shearon, 1991) (Ghozali, 2017) so that it is stated that all of the variable questionnaire items are valid.The results of the variable reliability test found that the value of Cronbach's Alpha for the Servant Leadership variable was 0.849, the Quality of Work Life variable was 0.804, the Work Engagement variable was 0.761, and the Worker Performance variable was 0.812.These variables have Cronbach's Alpha values > 0.60 so they are declared reliable (Ghozali, 2017).

RESULT AND DISCUSSION
Research respondents can be described based on gender, age, work experience, education level, and family status.The majority of respondents were male (61.19%), aged 20-30 years old (50.75%), had worked at PT Alamjaya Wirasentosa for 11-15 years (31.34%),attained a graduate level of education (48.51%), and have a single status (77.61%).
Descriptively, employees at PT Alamjaya Wirasentosa perceive that company managers apply relatively good servant leadership (76.1%).The servant leadership variable is measured by the Dennis (2014) indicator, which includes love, empowerment, vision, humility, and trust.The mean value of the love indicator is 4.12; empowerment indicator of 4.15; vision indicator of 3.95; the humility indicator is 3.76 and the mean value for the trust indicator is 3.87.Thus, of the five indicators, empowerment is the important indicator in servant leadership variable.
In general, employees at PT Alamjaya Wirasentosa perceive that the company has conditioned the quality of work life at a relatively good level (66.4%).The variable quality of work life is measured by indicators of growth and development, participation, innovative reward system, and work environment (Walton, 2014).The mean value for growth and development indicators is 3.64; participation indicator is 3.60; the innovative reward system indicator is 3.76 and the mean value for the work environment indicator is 3.63.Of the four quality of work-life indicators, an innovative reward system is the most dominant indicator, which is perceived as the best compared to others.
From the aspect of work engagement, it is known descriptively that employees at PT Alamjaya Wirasentosa have a relatively high level of work engagement (61.9%).The work engagement variable is measured by 3 indicators, namely employee attitudes towards work, work becomes self-identity, and attachment between self and work (Liwun & Prabowo, 2015).The mean value of the employee attitude indicator towards their work is 3.78; for the work as self-identity indicator is 3.62 and the mean value for the employee's self-engagement indicator with work is 3.63.Thus the attitude of work engagement is the most important indicator in the work engagement variable.
For the description of worker performance variables, it is known that employees at PT Alamajaya Wirasentosa have relatively high performance (66.4%).Worker performance is measured by the dimensions of quality, quantity, knowledge and skills, timeliness, and communication (Bernadin & Russel, 2016).The mean value for the quality indicator is 4.18; work quantity indicator of 4.13; indicators of knowledge and skills of 4.09; the timeliness indicator is 4.13 and the mean value for the communication indicator is 4.09.Of the five dimensions, the quality dimension is the one that plays an important role in the worker performance variable.
The classical assumption test of the regression research model includes tests of normality, multicollinearity, and heteroscedasticity.The normality test results of the research data were based on the significance value of Asymp.(2-tailed) Kolmogorov-Smirnov Test is 0.077 for the 1 st hypothesis model and 0.220 for the 2 nd hypothesis model.Thus, the Asymp value.Sig.(2-tailed) KS-Test research model is greater than the significance level of 0.05 (a) (Ghozali, 2017) so it is stated that the assumption of data normality is fulfilled.
The results of the multicollinearity test of the first and second hypothesis models were carried out to assess whether there was a strong correlation between the independent variables of the study showing the Variance Inflation Factor (VIF) < 10 and Tolerance > 0.10 (Ghozali, 2017).The results of the multicollinearity test (Table 1) show the value of Variance Inflation Factor (VIF) < 10 and Tolerance > 0.10 of each variable in the hypotheses 1 and 2 models so that it is stated that in each research model there is no multicollinearity between independent variables.The heteroscedasticity test results with the Glejser test were carried out to determine the similarity or dissimilarity of the variance of the residuals between observations (Ghozali, 2017).The heteroscedasticity test results of hypothesis models 1 and 2 show the significant value of the Glejser test for each variable involved > a significance level of 0.05 so that it is declared free from heteroscedasticity (Ghozali, 2017).The heteroscedasticity test result is shown in Table 2.
The results of hypothesis testing 1 prove that servant leadership has a significant positive impact on worker performance mediated by the quality of work life (Table 3).Hypothesis 1 test refers to the causal step method proposed by Baron and Kenny (1986).Based on the significance test of the hypothesis 1 model, it was found that the significance value of the quality of worklife variable before and after being included in the research model, the value remained significant (sig.value < 0.050).Before the inclusion of the variable quality of work-life as a mediator, the servant leadership variable had a significant impact on worker performance (sig.value = 0.021).Furthermore, after the quality of worklife as a mediating variable was entered into the regression model, the servant leadership variable had no significant positive impact on worker performance with a significance value of 0.053 (sig.value > 0.050).
According to Baron & Kenny (1986), if the effect of the independent variable on the dependent becomes insignificant after the mediation variable is included, then the mediating variable in the model acts as a full mediation.This means that in this research, the quality of work-life variable is only positioned as a mediating variable, and cannot be an inde-   3 also are supported by the results of the crosstabs test between the levels of servant leadership and worker performance as shown in Table 4.The data in Table 4 proves that in general respondents who perceive the level of servant leadership in companies in good condition will display a high level of performance (64.2%).Likewise, there are no employees who perceive the level of servant leadership as poor and underperforming (0.0%).This is in line with the descriptive output of the servant leadership perceived by employees at a relatively good level (76.1%) and empowerment is an important indicator in the servant leadership variable (mean value = 415).Employees at PT Alamjaya Wirasentosa also perceive that the company has conditioned the level of quality of work-life at a relatively good level (66.4%).An innovative reward system (mean value = 3.76) is the most dominant indicator, which is perceived as the best compared to others.For the description of worker performance variables, it is known that employees at PT Alamajaya Wirasentosa have relatively high performance (66.4%).The quality indicator (mean value = 4.18) is the important indicator in worker performance variable.
Similarly, the results of the crosstabs test between the levels of quality of work-life and worker performance as shown in Table 5.The descriptive data proves that in general respondents who perceive the quality of work-life in  companies in good condition will display a high level of performance (56.0 %).Likewise, there are no employees who perceive the quality of work-life as poor and have low performance (0.0%).The results of testing hypothesis 1 support the theoretical and empirical studies conducted by Nurbiyati (2014), Gunawan (2018), Dwijayanto & Priyono, (2019), Al-Hawasyi et al., (2019), Alberto et al., (2020), Mutholib (2020) which proves the same thing that the condition of quality of work-life in the organization has a significant effect on worker performance.The existence of quality of worklife will maintain the desire of employees to continue to work together as a team and survive in the organization to improve worker performance (Pradana et al, 2013;Nurbiyati, 2014;Hedayati & Nabiee, 2017;Gunawan, 2018;Al-Hawasyi et al., 2019;Alberto et al., 2020;Mutholib, 2020;Setyaningrum & Pawar, 2020).
The results of hypothesis testing 2 studies to prove whether servant leadership has a significant positive effect on worker performance moderated by work engagement are shown in Table 6.Hypothesis 2 testing refers to the absolute difference value method proposed by Frucot and Shearon (1991).The results of the significance test of hypothesis 2 found that the significance value of standardized servant leadership (ZX) was 0.121 > 0.050, the significance value of standardized work engagement (ZZ) was 0.000 < 0.050 the absolute value of the difference between servant leadership (X) and work engagement ( Z) (|ZX-ZZ|) of 0.048 < 0.05 with a negative beta coefficient value.Thus, it is concluded that servant leadership has a significant negative effect on worker performance moderated by work engagement because work engagement weakens the influence between servant leadership and worker performance.In other words, hypothesis 2 of this study was rejected.Source: SPSS processed data (2021) The results of the crosstabs test between indicators of work engagement and worker performance are shown in Table 7.The data in Table 7 proves that employee engagement with their work (9.7%) and employee attitudes about his job (7.5%) are the indicators of work engagement that contribute to low levels of worker performance.These indicators can specifically weaken servant leadership impact on worker performance.Indicators of employee engagement with their work are measured by seriousness, actively participating, and being fully involved in completing their work.While the indicators of workers' attitudes about their work are measured by the efforts of employees to appreciate their work, carry out their work as well as possible, and employees consider work important for their self-esteem.This is in accordance with the descriptive output of the work engagement variable that the indicators of employee attitudes (mean value = 3.78) and employee engagement with their work (mean value = 3.63) are 2 indicators that play an important role.If it is associated with the results of hypothesis 2 testing, these two indicators play an important role as moderating variables that weaken the influence between servant leadership and worker performance.
The existence of servant leadership and quality of work-life in the company which is relatively good is not enough to generate positive worker performance at PT Alamjaya Wirasento-sa when employee work engagement is not supportive.Work engagement can weaken servant leadership impact on worker performance.It seems that Person-Organization Fit and Person-Job Fit are important to fulfill (De Beer et al., 2016;Hendryadi et al., 2019).
Companies need employees who have a high attachment to their work because they will produce the best performance at work (Bakker & Leiter, 2010).Workers with high work engagement will really care about their work, be actively involved, master the areas of concern, and always try to put forth innovative ideas to complete work optimally because they feel that work is part of their life.For this reason, companies need to condition strategies to increase positive attitudes and high employee engagement with work, among others, by placing employees in jobs according to their competencies, providing opportunities for employees to master their fields of work, completing work according to their method, providing a sense of comfort at work, and opportunities to work actively participates in various activities in the company so that the results of their work contribute positively to the organization.
Thus the results of hypothesis 2 testing are not in accordance with the findings of Kim et al., (2012)

CONCLUSION AND RECOMMENDATION
The research findings conclude that the quality of work-life mediates positively and significantly servant leadership and worker performance at PT. Alamjaya Wirasentosa Servant.Indirectly, through the quality of work-life variable, servant leadership has a greater effect on worker performance than directly.The quality of worklife acts as a full mediating variable in servant leadership and worker performance relationship.Leaders who apply leadership characters of compassion, empowerment, vision, and humility will be able to increase the quality of work-life at PT Alamjaya Wirasentosa and subsequently be able to produce high worker performance.Furthermore, work engagement moderate negatively and significantly servant leadership and worker performance.The work engagement variable weakens the impact servant leadership on worker performance.Specifically, it means that the servant leadership style that is applied well in the company will be able to reduce worker performance if the employee does not have a high attachment to the company, and vice ver-sa.Indicators of employee engagement and attitudes towards work are identified as indicators that specifically weaken the servant leadership impact on worker performance.
Therefore, the theoretical implications for examining the influence of organizational and individual factors that lead to worker performance in an integrated model are important to present a broad picture of worker performance determinants in the organization.Practical implications of the research are expected to be meaningful for PT.Alamjaya Wirasentosa.The opportunity to actively participate in various activities in the company so that the results of their work contribute positively to the organization.In addition to conditioning servant leadership and quality of work-life, companies need to ensure a match between people and organizations (Person-Organization Fit) and people and their work (Person-Job Fit) to foster better employee work engagement.Further research needs to be done on a larger sample size by exploring variables that can strengthen the impact of servant leadership in improving worker performance and organizational competitive advantage.Other strategic organizational and individual factors that influence employee and organizational performance need to be explored more broadly through various further studies.
 Correspondence AddressInstitutional address : Jl.Sembada XI No. 4 Medan 20131, Indonesia Email: audia.junita@yahoo.com to test hypothesis 2. Hypothesis testing is carried out by first fulfilling the required assumptions in multiple regression tests.The results of the validity test of 10 Servant Leadership variable questionnaire items, 8 Quality of Work-Life variable questionnaire items, 6 Work Engagement variable questionnaire items, and 10 Worker Performance variable items have a value of r count > r table

Table 1 .
Multicollinearity Test Results

Table 2
Heteroscedasticity Test Results

Table 3
Hypothesis 1 Test Results

Table 4
Servant Leadership and Worker Performance Cross Tabulation Results Source: SPSS processed data(2021)

Table 5
Cross Tabulation Results of The Quality of Work-Life and Worker Performance Wanta & Augustine (2021) of Work Engagement Indicators and Worker Performance Similarly, in accordance with the findings ofListiau et al., (2017),Sittar (2020),Wanta & Augustine (2021)that work engagement can have a negative impact on worker performance.The work engagement variable weakens the servant leadership impact on worker performance.Specifically, it means that the servant leadership style that is applied well in the company will be able to reduce worker performance if the employee does not have a high attachment to the company, and vice versa.Servant leadership will contribute positively to worker performance if employee work engagement is reinforcing.