Keywords

Training, Motivation, Discipline and Performance.

Introduction

The main problem in managing human capital is how to make employees in carrying out their duties able to produce optimal performance in line with the company expectation. Companies must realize that optimal performance cannot arise by itself, but must be encouraged and stimulated by an award (reward) which can be in the form of additional income or achievement allowances, promotions or training in an effort to get a promotion. Soegandhi et al. (2013) stated that organizations expect maximum individual performance to be able to achieve company excellence. Considering that individual or group performance ultimately affects the progress of the company as a whole. A company has a goal to achieve excellence, both to compete with other organizations and to maintain the sustainability of the organization. In line with ( Soegandhi et al., 2013; Horgan & Mühlau, 2006; Ivancevich et al. 2007; Jöreskog & Sörbom, 1996) states that individual performance is the substance of organizational performance. An essential aspect in the success of an organization is the manifestation of qualified and proficient employees equipped to uphold determination, to get satisfactory work outcomes. In fact, not all employees possess the capability, skills, and work spirit in line with the expectations of the organization. An employee who has the capability compliant with the expectations of the organization sometimes does not have high morale, causing the performance was not as expected ( Sumarsono, 2004; Umar, 2003; Laeeq et al., 2016; Mapparenta, 2011; Mas’Ud, 2004; Mozammel, 2019; Nawawi, 1983; Porwani, 2011; Saydanm, 2006; Schermerhorn et al., 1996; Shah et al., 2021).

PT Pegadaian (Persero) as one of the state-owned financial services companies in Indonesia is a very important financial institution. The Company currently has 1 head office, 12 regional offices, 540 Pawnshop Branches, 102 Sharia Pawnshop Branches, 2,913 Branch Service Units, and 560 Sharia Branch Service Units. With its coverage spread throughout Indonesia, it is not surprising that it has become a very important institution for business financing sources. Pegadaian is the main alternative for MSMEs to get access to funding other than banking. As the spearhead of the economy with nearly 60% of the total number of companies in Indonesia, the MSME business is expected to develop into a medium or large company. However, due to the difficulty of accessing funds to banks, mainly due to the limited fulfillment of the requirements for providing loan guarantees, MSMEs are often forced to use high-cost funding. This of course interferes with the sustainability of the MSME business which cannot compete with large companies that already have economies of scale in their business.

As the institution that has always been consistent and loyal to the economic empowerment of the small people in Indonesia and continues to accompany its customers to obtain a prosperous life, it is important for PT Pegadaian (Persero) to always increase its market share. However, in reality, based on the number of accounts from 2018 it showed a decline, even the target for adding accounts that had been lowered by management was not optimally achieved. Even turnover, service fees and outstanding loans in 2020 exceed the set target. The following are the performance plans and achievements that have been determined by the Company based on the KCA pawn product which is the main product of the pawnshop in the form of the amount of credit distributed to customers as a financing institution.

The Table 1 shows a downward trend in the number of customers. This means that competitors have captured the company's market share. The increasing number of competitors in the pawn business in recent years has caused the pawn business to no longer be the monopoly of PT Pegadaian (Persero). The existence of these competitors makes PT Pegadaian (Persero) to be able to improve services by developing technology that is supported by reliable human resources. to fulfill the company's vision.

Table 1 PAWN BUSINESS PERFORMANCE 2018 - 2020
KCA Pawn Products	Planning	Realisation	Target Achievement	Growth
2020
∑ Account	9,685,816	9,470,002	97.77%	-2.55%
Turnover (Million IDR )	120,231,603	129,987,851	108.11%	18.07%
Outstanding Loan (Million IDR)	36,426,691	38,737,357	106.34%	21.93%
Capital Lease Revenue (Million IDR)	8,849,197	9,496,234	107.31%	13.24%
2019
∑ Account	11,340,864	9,717,751	85.69%	2.32%
Turnover (Million IDR )	114,548,534	110,089,649	96.11%	4.58%
Outstanding Loan (Million IDR)	29,593,706	31,769,017	107.35%	11.93%
Capital Lease Revenue (Million IDR)	8,807,224	8,386,076	95.22%	7.51%
2018
∑ Account	11,779,702	9,505,283	80.70%	-3.20%
Turnover (Million IDR )	116,054,632	105,320,486	91%	0.40%
Outstanding Loan (Million IDR)	32,285,013	28,400,752	88.00%	1.20%
Capital Lease Revenue (Million IDR)	8,891,981	7,805,515	87.80%	2.20%

The decrease in the number of customers as mentioned above is estimated to come from the sub-optimal performance of employees. This can be seen from the internal data of PT Pegadaian (Persero) individual assessment Table 2 above, it is found that the achievement of employee performance has not been optimal, only reaching 84%. With the aim of to achieve the company's targets above, it is necessary to study the work motivation of employees in achieving the targets set. Motivation and performance are dual components that are mutually connected, beneficial and correlative each other. Member of staff performance will be condensed if they does not come up with the motivation to complete the job. On the other hand, if the employee has elevated motivation to finish their task, in general their performance will also high ( Harlie, 2010). The substance of Motivation is something that sources, circulates and encouragements human behavior to work intensely and enthusiastically to accomplish optimal results. Motivation becomes increasingly important when managers delegate their work with an expectation that it will be carried out well by their subordinates and integrated into the company goals. The company expects its qualified staff, and skilled, and also eager to do intensely in complete utmost performance (Hasibuan, 2008). Companies can increase motivation in employees through fulfilling the need for achievement the need for achievement, which can be in the form of giving positions and promotions ( Abboh et al., 2019; Abdullah, 2009a and 2009b; Dharma, 2004; Dessler, 2004). Hence, appointing and designing organization structure concerning their gender and leadership styles will raise on employees’ motivation ( Salleh et al., 2018). This is in accordance with the research of ( Cruz et al., 2009; Afful-Broni et al., 2012; Heldia, 2012; Hairina, 2020) which stated that motivation effect employee performance positively. However, this study is not in line with Tus's (2020) research which stated that motivation did not affect the academic performance of senior high school students, and Subari & Riady (2015) which stated that motivation did not directly influence performance nor indirectly via internal communication on employee’s performance.

Table 2 Individual Assessment Recapitulation
Component	Goal Achievement
1. Position knowledge	85%
2. Responsibility	89%
3. Discipline	86%
4. Cooperation	79%
5. Integration	90%
6. Quality and Quantity of work	88%
7. Initiative	75%
8. Work skills	80%
9. Service Orientation	87%
10. Desire to Achieve	80%

With 14,237 employees, it is a big task for management to improve employee performance. Since performance is the consequence results of an employee work achievement in quality and quantity in doing their responsibilities ( cMangkunegara & Prabu, 2005), training and education programs for employees have very important meaning. Effective training will affect employee’s performance ( Subari & Riady, 2015) stated that by combining various factors such as correct training techniques, careful preparation and planning, and commitment to the essence of training, companies can achieve enormous competitive benefits in a very tight market. In line with Rivai (2005) stated that education and training function to improve work behavior, which includes aspects of knowledge, skills, and personal attitudes required by their job duties. However, there is an important thing that is often overlooked by organizations, namely the implementation of the training at the workplace or transfer of training (Soleman, 2019). Many organizations suffer losses due to failure to implement adequate transfer of training within the organization. Therefore, it is important to ensure proper implementation of transfer training by ensuring good quality of supervisory support, perceived utility and higher education. Employee training can also improve communication within an organization, transfer skills and knowledge, improve work performance, encourage teamwork, increase employee morale and lead to job satisfaction ( Henry & Jennifer, (2011) Education and training provides an overview of the duties and responsibilities of a position, its relationship to other positions, the knowledge and skills required, and the working conditions in which the position is completed.

Performance of employee is also has en effect on discipline. Discipline is a starting point for all success in accomplishing the goals of an organization. Discipline does not automatically occure. In essence, discipline is to develop consciousness among employees to complete their assignment. Therefore, to form elevated discipline, it is necessary to form through formal and non-formal education, and be given motivation ( Harlie, 2010). Work discipline is a management feat to raise consciousness and enthusiasm of its members to fulfil the term of all organization regulations and social norms that apply willingly. The application of discipline in an organization aims to make employees in the company are willing to voluntarily obey every applicable order without any coercion. A decent discipline reflects someone appreciation of responsibility concerning their assignments consigned. This encourages work enthusiasm, morale and the realization of the company goals and its employees. Without the support of decent employee discipline, the company goals to accomplish optimal employee performance would face problematic. Discipline is the key success in achievement company goals (Hasibuan, 2008). Jalil et al. (2019) strengthen this statement according to the results of their research which concludes that training and work discipline had a positive impact on employee's performance. Utama & Sriathi (2021) also stated that work discipline would strengthen organizational culture in influencing employee task performance.

Based on the background stated above, this study purpose to prove empirically that the provision of training and motivation to employees plays a role in the formation of work discipline. This research is also to prove that good employee performance occurs because of providing training and motivation to employees which is strengthened by the formation of work discipline. The benefits of this study are expected to offer consideration for the Management to take appropriate steps in an effort to deal with the decline in the company's market share. To answer this research problem, the questionnaire was distributed to the respondents. The data processing technique was carried out using the SEM method and processed using AMOS.

Table 3 Operasional Variabel
No	Variable	Definition	Dimension	Indicator
1	Training	All forms of efforts used to improve employee performance. ( Gomes,2005),	- Quality	1. Increase the volume of target achievement
				2. Finding solutions to problems
				3. Content in the material relevant to the needs
			- Quantity	1. Training quantity
2	Motivation	Is a driving force that influences a person's actions and attitudes to be able to do something so that his desires are satisfied.	- Physiological needs	1. Fulfillment of basic needs such as food, clothing and shelter
				2. Income from salary, overtime pay, honorarium, allowances and incentives
				3. Guarantee of a person in position, position and responsibility
				4. Guarantee a sense of security in carrying out work and old age
			- Sense of security and safety at work	1. The need for affection such as attention and friendship (cooperation) from work groups or between groups
				2. Involved in increasing the realization with parties who need and
				3. The growth of a sense of togetherness including a sense of belonging in the organizationKedudukan dan promosi
			- Social Need for achievement (awards)	1. Develop the working capacity of employees properly;
			- Self-actualization	1. Show all abilities. Recognition of colleagues and leaders towards employees
3	Work discipline	Discipline is management's action to encourage awareness and willingness of its members to comply with all regulations that have been determined and social norms that apply voluntarily. (Handoko, 2005)	- Presence	1. Attendance at work time
				2. Employee compliance in working hours
			- Obedience	1. Service compliance with orders/instructions and superiors as well as complying with applicable rules and regulations
				2. Mental attitude that obeys rules, orders and instructions
			- Punctuality	1. Come and go home on time
				2. Punctual in completing a job
4	Employee Performance (Y)	Employee performance is the achievement obtained by a person in performing a task. Sutrisno (2009)	- Quality	1. Compliance with work standards
				2. Mastery of tasks
			- Quantity	1. Number of products produced
				2. Punctual in completing tasks
				3. The number of types of work that can be done

Literature Review

Performance

Raza et al. (2020) in the results of their research concluded that the role of employee willingness to perform contributes to increased levels of performance as well as the factors affecting employee effectiveness. Those employees who are more willing to perform, contribute more significantly to organizational success. Timpe (2007) states that performance is the level of achievement of a person or employee in an organization or company that can increase productivity. Meanwhile Mathis & Jackson (2006) explain that performance is what employees should act or don't. There are three principal factors that influence employee performance, namely the individual's ability to do the job, the level of effort devoted and the organizational support he receives ( Mathis & Jackson, 2006). Robbins (2003) defines performance as a function of the interaction between ability and motivation. If there is one factor that is not adequate, then the performance will be negatively affected. Mulyadi (2007) personnel performance is determined by three factors, namely talent and ability, perception of the role, effort. Where high performance is produced by personnel who have talent and ability and have a distinct role in the organization. Briefly, employee performance refers to a standards criteria or measurement of person's achievement set by the company/organization. Fareed et al. (2016) suggested the performance management system as a method that can be used to evaluate the performance of employees.

When the company wants to achieve high performance will strive to develop overall organizational performance. Some performance indicators according to experts are as follows: Tsu’s et al (1997) in Fuad Mas'ud (2004) formulate employee performance indicators as follows: 1) The quality of employee work, namely the level / level of good or bad work done by employees. 2) Professional standards, namely employee attitudes and work culture based on ideals and ethics to love, appreciate, improve and develop their skills in order to provide the best work results. 3) Employee work quantity, namely the amount/weight of work performed by employees in a certain period. 4) Employee creativity, namely the ability of employees to produce something new, either in the form of ideas or real works, either in the form of new works or combinations/modifications of existing works. As for the indicators of employee performance according to Parlinda & Mahyuddin (2005) are, (1) the ability to prepare work plans; (2) the ability to realize the work plan; (3) the ability to carry out orders/instructions from superiors; (4) the ability to provide services to the community; (5) ability in quality of work, including accuracy, neatness, speed, accuracy, and skills in performing tasks; and (6) the ability of employees to achieve work targets or desired work results.

Discipline

Discipline is a necessary capital in achieving the desired goals. In a disciplined environment, a company will be able to accomplish their programs and the goals that have been arranged. A good leader must strive so that his subordinates have good discipline and must also set an example in carrying out good discipline in an organization.

Siagian (2008) stated that discipline is a configuration of training in an effort to make progress and form the knowhow, manners and behavior of employees, in order to these employees do the assignment voluntarily and collectively among other employees and improve work performance. Sutrisno (2009) said that discipline is an attitude of someone’s willingness to follow the running norms of the regulations. Enhance, a decent employee discipline will accelerate the achievement of organization goals, while deteriorating discipline will become an obstacle and decelerate the achievement of goals. The placement of discipline in the life of an organization is intended, so that all employees in the organization are willing to voluntarily obey all applicable rules and regulations without coercion.

Discipline itself is divided into two types, namely preventive discipline and corrective discipline. Preventive discipline is an activity to support employees to act in accordance with existing regulations voluntarily. It is more aimed at preventing employees from making mistakes and violations. While corrective discipline is an action in the form of giving sanctions to employees or in other words more aimed at preventing employees from repeating mistakes that have been made. There are four perspectives in work discipline ( Handoko, 2005; Sihotang, 2007; Simamora, 2006; Siswanto, 2005; Sugiyono, 2009), namely (1) Retributive Discipline, which is attempting to penalize people who made a mistake. (2) Corrective Discipline, which is attempting to assist employees remedy their unsuitable behavior. (3) Individual rights perspective, namely attempting to guard the fundamental individuals rights during disciplinary actions. (4) Utilitarian Perspective, which emphases on applying of discipline when the consequences of disciplinary action exceeds adverse effects. There are indicators to appraise the level of employee discipline ( Handoko, 2005), namely attendance, obedience, punctuality, behavior. Utama & Sriathi (2021) and Jalil et al. (2019) concluded that work discipline had a positive impact on employee's performance, also would strengthen organizational culture in influencing employee task performance

Training

Gomes (2005) stated training is an attempt to enhance work performance in his particular duty. Ideally, training should be intended to accomplish the aims of the organization and also to complete the individual goals. Managers support training activities to make the employees more skillfull and even more productive, even though the time taken up while they are being trained. Supported by Saeed and team in 2021 that stated organization should enhance the use of organizational learning as a resource for better organizational performance and Fareed et al. (2016) that stated performance management system merupakan salah satu metode yang dapat dipergunakan mengevaluasi the performance of employees. Meanwhile, according to Wibowo (2007), training is a very important investment in human resources. Rachmawati (2008) explains that training is an environmental forum for employees, where they get lessons in matters relating to duties and responsibilities such as behavior, knowledge, abilities and skills and attitudes. Training is an effort and process to increase knowledge and skills that are used as an important investment asset in human resources.

Henry & Jennifer (2011) conclude that employee training and development is very important for a company in this competitive era, if the company is to survive and thrive. Therefore, training and development is an important issue in business strategy. According to Proctor and Thorton there are various benefits in training for employees and companies, as follows: increasing employee satisfaction, reducing waste, reducing absenteeism and employee turnover, improving work methods and systems, increasing income levels, reducing overtime costs, reducing costs. maintenance of machines, reduce the cost of maintaining machines, reduce employee complaints, reduce accidents, improve communication, increase employee multipurpose knowledge, improve employee morale and lead to better cooperation ( Manullang, 2004).

The general purpose of training ( Moekijat, 1993) is to improve skills in a attempt to complete the task more quickly and effectively, to form knowledge in a attempt to finish the task rationally, and also to build cooperatios among friends, colleagues and leaders. The benefits of training ( Rivai, 2005; Wardono, 2012) are divided into benefits for employees and benefits for the company. For employees, training is useful to help employees deal with stress, pressure, frustration, and conflict, increase job satisfaction and recognition, help eliminate fear of taking on new tasks, assist employees in making decisions and problem solving more effectively, and help encourage and achieve self-development and self-confidence. For companies, training is useful to help prepare and implement company policies, create a good climate for growth, help deal with conflicts so as to avoid work stress and pressure, improve HR morale, improve work knowledge and skills at all levels of the company, directing to increase profitability or attitudes, and positive towards profit orientation. Education and training indicators according to Hasibuan (2008) is the target of development, curriculum development, development facilities, participants, trainers and implementation. While the indicators of training according to Malthis (2006) are the ability, effort devoted and organizational support.

Previous study by Subari & Riady (2015) and Rivai (2005) about the essence of training to achieve enormous competitive benefits. Soleman, 2019 pressed in the essential of the implementationof the training at the workplace. Employee training can also improve communication within an organization, transfer skills and knowledge, improve work performance, encourage teamwork, increase employee morale and lead to job satisfaction ( Henry & Jennifer, 2011).

Motivation

The theories of motivation can be categorized into two kinds, namely the first is the theory of satisfaction (content theory) and the second theory is process theory. First theory is based on individual factors needs and satisfaction to act and behave in certain ways. This theory focuses on the factors why people reinforce, direct, support or stop their behavior. This theory tries to answer the question of what needs satisfy and encourage one's work enthusiasm. If the needs are increasingly met, then the spirit of work will be better (Hasibuan, 2008). These satisfaction theories include (1) Classical Motivation Theory by Fredrick Wislow Taylor. (2) Maslow's Need Hierarchy Theory. This theory is also called A Theory of Human Motivation, put forward by A. H. Maslow in 1943. (3) Herzberg's Two Factor Theory (4) Mc. Clelland's Achievement Motivation Theory. (5) Alderfer's Existence, Relatedness and Growth (ERG) Theory. (6) Motivation Human Relations Theory. (7) Claude's Theory of Motivation.

The second theory is process theory. According to Hasibuan (2008), the theory of process motivation is opposite to the theories of needs as described earlier, the theory of the process focuses its attention on how motivation occurs. In other words, process theory basically tries to answer the question of how to strengthen, direct, maintain, and stop individual behavior so that each individual works hard in accordance with the wishes of the manager. This shows that there is causal, namely cause and effect. This process motivation theory, known as Expectancy Theory, Equity Theory, Reinforcement Theory. This reinforcement theory is based on the causal relationship between behavior and compensation. So that the nature of dependence is related to the relationship between behavior and events that follow that behavior. The theory of reinforcement according to Hasibuan (2008) consists of two types, namely (1) positive reinforcement, namely increasing the frequency of behavior, occurs when positive reinforcement is applied conditionally. (2) Negative reinforcement, which is an increase in the frequency of behavior, occurs if negative reinforcement is conditionally removed. So the principle of reinforcement is always associated with increasing g frequency and response, when followed by a conditional stimulus.

Motivation has a purpose as in Hasibuan (2008): 1) Encouraging employee passion and morale, 2) Increasing employee morale and job satisfaction, 3) Increasing employee productivity, 4) Maintaining loyalty and stability of company employees, 5) Improving discipline and reducing employee absenteeism, 6) Streamlining employee procurement, 7) Creating a good working atmosphere and relationship, 8) Increasing employee creativity and participation, 9) Increasing employee welfare levels, 10) Enhancing employees' sense of responsibility towards their duties, 11) Improving the efficiency of the use of tools and raw materials. According to Sardiman (2006), there are three functions of motivation, namely (1) Encouraging humans to act, so as a mover or motor that releases energy. Motivation can provide direction and activities to be carried out; (2) Determine the direction of action, namely towards the goal to be achieved. Thus, motivation can provide direction and activities that must be carried out in accordance with the formulation of its objectives; (3) Completing actions, namely determining what actions must be done that are harmonious in order to achieve the goal, by setting aside actions that are not useful for that purpose.

According to Hamzah B. Uno the dimensions and indicators of work motivation can be grouped (1) Internal motivation in the form of responsibility in carrying out tasks, carrying out tasks with clear targets, having clear and challenging goals, providing feedback on the results of their work, have a sense of pleasure at work, always try to outperform others, prioritize achievement over what he does. (2) External motivation in the form of always trying to meet the needs of life and work needs, likes to get praise for what he does, works with the desire to get incentives, works with the hope of getting attention from friends and superiors. Hezberg describes six motivational factors that can be used as indicators, as follows: job satisfaction, achievements, opportunities for advancement, recognition from others, career development possibilities and responsibilities. So it can be argued that motivation is a driving force that influences a person's actions and attitudes to be able to do something so that his desires are satisfied.

Previous research by Hairina (2020); Salleh et al. (2018); Afful-Broni et al. (2012); Heldia (2012); Cruz et al. (2009), which stated that motivation effect employee performance. However, Tus's (2020) and Subari & Riady (2015) stated that motivation did not affect the performance.

Research Methods

The object of this research is the employees of PT Pegadaian (Persero) Indonesia. The research sample was conducted on employees of PT Pegadaian (Persero) Indonesia in Jakarta. The sampling technique used was random sampling with convenience sampling technique. This study will present how the level of influence of training and work motivation on discipline and its impact on the performance of employees of PT Pegadaian (Persero) head office. Determination of the number of samples as far as possible in accordance with the criteria for the ideal sample size for SEM modeling analysis, namely 100-200 and an absolute minimum of 50, or the minimum sample size is 5-10 x manifest variables or indicators of all latent variables. The sample that will be used in this study is determined by using the formula ( Santoso, 2001; Anyim et al., 2012; Arikunto, 2006) namely number of samples=number of indicators×5 to 10. Because in this study there are 26 indicators, the minimum number of samples that can be used is 130 respondents (Number of indicators×5). Furthermore, Hair et al. (2006) Ghozali (2008) found that the appropriate sample size for SEM is between 100 - 200 samples. With reference to Hair's opinion, the number of samples used in the study were 192 respondents from a total of 1306 employees.

Data collection technique. Primary data in the form of data collected directly by the author through a questionnaire consisting of question items distributed to employees of the head office of PT Pegadaian (Persero) Jakarta head office. Secondary data is in the form of data obtained from written reports as well as information about the state of the company that is the object of research, namely the head office of PT Pegadaian (Persero) Jakarta, including previous studies. This study uses a questionnaire model that is used using a multichotomous questionnaire (many answer choices) where the subject has five alternative responses using a Likert scale of 1 - 5. This scale is related to a person's attitude statement towards something. In this research questionnaire, the statement strongly disagrees (STS) to the statement strongly agrees (SS).

Operational definition. The operational variables in this study consist of independent variables, namely training (X1), and motivation (X2), then the intermediary variable is work discipline (Y) and the dependent variable is employee performance (Z). Independent variables or independent variables are variables that do not depend on other variables or can be referred to as variables that affect the dependent variable. The detailed explanation of the operational variables can be seen in the following matrix:

The variables in this study consist of the dependent variable of employee performance (Z), the intermediary variable of work discipline (Y), the independent variable which includes training (X1) and work motivation (X2). The research instrument of Employee Performance (Z), the intermediary variable of work discipline (Y), training (X1) and work motivation (X2) in the form of a questionnaire. Each respondent was asked to answer each question with a Likert scale of 5 (five) points, namely: number 1 (Strongly Disagree), number 2 (Disagree), number 3 (Medium/Neutral), number 4 (Agree) and number 5 ( Strongly agree).

Analysis Techniques and Hypothesis Testing

To test the research hypothesis by processing the sample data obtained, inferential statistical analysis was used. The Inferential Statistical Method used in the data analysis of this study was Structural Equation Modeling (SEM). The reason for using SEM is the consideration that the causal relationship formulated in this study uses a non-simple model that has multiple roles such as employee discipline. This form of causal relationship requires analysis that is able to explain simultaneously. Therefore, the method used in this study uses SEM. The use of SEM as an analytical tool is based on the reasons for the complexity of the model used, the limitations of multidimensional analysis tools that are often used in quantitative research, such as multiple regression, factor analysis, descriminant analysis and others. The weakness of this analysis tool is that it can only analyze one relationship at a time.

The data obtained and respondents used as research samples through questionnaires distributed. In accordance with the conceptual framework that has been developed based on the results of theoretical studies and hypotheses to be tested, the analytical technique used in this study is Structural Equation Modeling (SEM) using the AMOS 20 (Analysis of Structural Moment) program. SEM is a multivariate analysis technique that allows a series of analyzes of several latent variables to be carried out simultaneously so as to provide statistical efficiency (Hair et al., 2006). SEM has other advantages over other multivariate analysis techniques, because in the latent variable measurement errors are included in the model, namely the measurement error of the exogenous latent variable is given the symbol epsilon (e), while the measurement error of the endogenous latent variable is given the symbol zeta (z) ( Arbuckle, 1995)

SEM testing in this study includes: normality test, Confirmatory factor analysis (CFA), overall model fit test (overall model fit test), and individual test of significance (test of significance). Assumption test includes normality test and outlier test. Data normality test is important because SEM analysis can be done if the data is normally distributed. The normality test of the data used the univariate normality method by looking at the coefficients of the univariate skew index and the univariate kurtosis index (height-flat). The data meets the data normality requirements, if the coefficient of the univariate skew index and univariate kurtosis index is between 0 to ± 2.58. Detection of outliers uses Mahalanobis distance which shows how far the data is from a certain center point. A data is said to be an outlier if it has p1 and p2 numbers that are less than 0.05 ( Santoso, 2001). The test of the accuracy of the indicators or manifest variables in measuring the related latent variables is carried out through confirmatory factor analysis (CFA) by looking at the loading factor of each indicator.

The research hypotheses are:

H₁ Training affects work discipline

H ₂ Work motivation affects work discipline

H ₃ Training has a direct effect on employee performance

H ₄ Work motivation has a direct effect on employee performance

H₅ Work motivation and training have a direct and indirect effect on employee performance through work discipline variables

The model testing includes the overall model fit test and the individual test of significance (test of significance) of the estimated model parameter results. The overall model suitability test was carried out using the goodness of fit test (GFT), while the individual significance test was carried out using the t test statistic, the same as testing the ordinary classical regression coefficient, using a significance level (α) of 5%. That is, the value of the critical t statistic set is 1.96 (Joreskog & Sorbom, 1996).

The test criteria use several measures of model suitability (Goodness-of-Fit-Test) consisting of absolute fit measures, incremental fit measures, and parximonius fit measures. Based on the three types of Goodness-of-Fit-Test, a model is indicated to fit or fit the data if the model fits absolutely with the data, is relatively better than other models (baseline model or null model), and is relatively simple when compared to other models. alternative model (Bachruddin & Tobing, 2002:68). Absolute fit measures inform the model's ability to absolutely estimate the population covariance matrix based on the sample covariance matrix. The two most commonly used absolute fit measures are the Likelihood-Ratio Chi-Square Statistics (X2) statistic and the Root Means Square Error of Approximation (RMSEA) (Joreskog & Sorbom, 1996). The criteria for testing the goodness of fit overall structural model can be seen in the Table 4.

Table 4 Cut-Off Untuk Goodness of Fit Overall Structural Model
No.	Indeks	Cut-off
1	Chi-Square	Kecil
2	Probabilitas Chi-Square	≥ 0,05
3	CMIN/DF	≤ 2.00
4	RMSEA	≤ 0.08
5	TLI	≥ 0.90
6	CFI	≥ 0.90

The individual test of the significance of the structural model is carried out after the overall model that meets the goodness of fit criteria is obtained which aims to test the hypothesis of a causal relationship between constructs (Hair et al., 2006). The statistical test used is the t test which is based on the critical value. The calculated t value in the AMOS program is indicated by the critical ratio (CR). The significance of the relationship can be determined based on the CR value or the probability value (p) in the AMOS program. Based on the distribution table of the critical value t at the level of accuracy of 10% or 1.28, the level of accuracy of 5% is 1.65 and the level of accuracy of 1% is 2.33 (using two directions). The significance level in this study is 5%, so the relationship is said to be significant if the CR value >1.65 or P<0.05.

Research Results and Discussion

Research Result

This study raises the issue of the effect of training, and motivation on discipline and its impact on employee performance. Demographic characteristics of research respondents are presented in the following Table 5.

Table 5 Demographic Data of Research Respondents
Description		Total	Percentage (%)
Sex	Male	111	57.8
	Female	81	42.2
Status	Married	143	74.5
	Single	39	20.3
	Widower	3	1.6
	Widow	7	3.6
Age	<20 years	0	0
	>20-25 years	8	4.2
	>25-30 years	71	37.0
	>30-35 years	53	27.6
	>35-40 years	42	21.9
	>40 years	18	9.4
Recent Education	Senior High School	23	12.0
	Diploma	38	19.8
	S1	117	60.9
	S2	14	7.3
	S3	0	0
Work experience	<6 years	67	34.9
	>6-10 years	64	33.3
	>10-15 years	39	20.3
	>15 years	22	11.5

The majority of respondents who are permanent employees of the PT Pegadaian head office are men with an age range of >30 years with married status, this is because men at the age level >30 years, have become a household head who has responsibility so they try hard to produce good performance. Meanwhile, the majority of respondents' education level is undergraduate. Pawnshop services are services that require technical expertise, so they are easily produced by employees who have a higher education background. Employees who have a good educational background will be easy to receive and apply the training they get to improve performance.

Inferential Statistics - SEM Testing

A structural equation modeling (SEM) model can consist of a measurement model and a Structural model, and the purpose of SEM analysis is to test whether the model made is fit with the existing data. For testing using SEM analysis, estimation is carried out in stages, namely by performing confirmatory factor analysis (CFA) techniques and full structural equation modeling techniques ( Ferdinand, 2006; Flippo, 2003). CFA is to ascertain the validity and reliability of the questionnaire. A test was carried out to respondents who were excluded the analyzed respondents by spreading a preliminary questionnaire, namely 20 respondents from 150 contract employees (Figure 1).

The results of the model suitability test on the confirmatory factor analysis are presented in the following Table 6:

Table 6 Feasibilitv Test Results of the Confirmatorv Factor. of Performance Variables
Goodness-of-Fit Index	Performance Variable rDf -6			Discipline Variable Df - 10
	Cut-off Value	Result	Model Evaluation	Cut-off Value	Result	Model Evaluation
1'-Chi-sQuare	<12.59159	11,996	Good	<l&.30704	12,014	Good
Significance Probabilitv	0.05	0.062	Good	0.05	0.284	Good
RMSEA	<0.08	0.204	Mar•inal	<0.08	0.092	Good
GFI	>0.90	0.897	Good	<2.00	1.201	Good
CMIN/DF	<2.00	1.999	Good	>0.95	0.907	Good
TL!	>0.95	0.742	Good	>0.95	0.907	Good
CF!	0.95	0.897	Good
Goodness-of-Fit Index	Training Variable ffif -27			Motivation Variable Df -36
	Cut-off Value	Result	Model Evaluation	Cut-off Value	Result	Model Evaluation
1'-Chi-sQuare	<31.41043	31.66&	Good	<50.99846	50.729	Good
Significance Probabilitv	0.05	0.245	Good	0.05	0.053	Good
RMSEA	<0.08	0.0&5	Good	<0.08	0.131	Mar•inal
CMIN/DF	<2.00	1.173	Good	<2.00	1.409	Good
TL!	>0.95	0.942	Good	>0.95	0.&49	Mareinal
CF!	>0.95	0.965	Good	>0.95	0.901	Mareinal

From some of these feasibility tests, the model feasible if at least one of the model's met feasibility test methods ( Widarjono, 2010). Based on the results (Table 6), the feasibility index is within the range of expected values or is within good criteria so that all of these models can be accepted.

The validity test with the CFA test or construct validity test (indicator) that measure whether the constructs (indicators) are not reflect the latent variable. The results meet the criteria (the value of Critical Ratio (CR)>1.96 with Probability (P)<0.05. This means that these indicators are significantly the dimensions of the formed latent factors (Table 7).

Table 7 Regression Weight Factor Confirmatory Variable Analysis Regression Weights: (Group no. 1- Default Mode)
Performance	Estimate	SE	CR	P	Label
K2 ß Performance	10.376	4.506	2.303	0.021	par_1
K3 ß Performance	17.232	2.686	6.416	0.001	par_2
K4 ß Performance	14.077	10	4.527	0.001	par_3
K5 ß Performance	16.83	3.468	4.853	0.001	par_4
K1 ß Performance	1
Discipline	Estimate	SE	CR	P	Label
D1 ß Discipline	1
D2 ß Discipline	0.403	0.149	2.7	0.007	par_1
D5 ß Discipline	0.137	0.220	3.623	0.001	par_2
D6 ß Discipline	0.428	0.113	3.788	0.001	par_3
D3 ß Discipline	0.645	0.183	3.521	0.001	par_4
D4 ß Discipline	1.543	0.536	2.877	0.004	par_5
Training	Estimate	SE	CR	P	Label
P6 ß Training	2.286	1.411	2.92	0.001	par_1
P5 ß Training	2.18	1.299	2.978	0.001	par_2
P7 ß Training	2.596	1.562	1.962	0.001	par_3
P8 ß Training	1.274	0.867	1.969	0.001	par_4
P9 ß Training	1.465	1.03	1.922	0.001	par_5
P3 ß Training	1.646	1.006	1.936	0.001	par_6
P4 ß Training	1.269	0.857	1.981	0.001	par_7
P1 ß Training	1
P2 ß Training	2.597	1.569	1.961	0.001	par_8
Motivation	Estimate	SE	CR	P	Label
Motivation ß z11	9.891	2	6	0.001	par_1
M6 ß Motivation	0.432	0.118	3.652	0.001	par_2
M7 ß Motivation	0.503	0.154	3.257	0.001	par_3
M8 ß Motivation	0.598	0.144	4.159	0.001	par_4
M9 ß Motivation	0.545	0.161	3	0.001	par_5
M10 ß Motivation	0.490	0.140	3	0.001	par_6
M1 ß Motivation	l.000
M2 ß Motivation	0.485	0.145	3	0.001	par_7
M3 ß Motivation	0.462	0.141	3.267	0.001	par_8
M4 ß Motivation	0.548	0.168	3..261	0.001	par_9
M5 ß Motivation	0.404	0.11	3.122	0.001	par_10

Validity Test using Convergent Validity Test, which is testing the construct (indicator) whether it has a high proportion of variance or not. Meets the criteria if “Loading Factor” or “Standardized Loading Estimate” >0.5. The test results presented above, each indicator variable in the latent variable has a loading factor value greater than 0.50 which indicates a close relationship, the direction of the relationship is positive. ( Santoso, 2018). Except for the indicator variable K1, which is 0.205<0.5 and the indicator variable D5, which is 0.309<0.5, it shows a weak relationship. After the measurement model is analyzed through confirmatory factor analysis and it is seen that each indicator can be used to define a latent construct, the test can be continued with a structural model test.

Descriptive Statistics

This analysis aims to obtain an overview of respondents' answers to the variables studied by measuring the average value of respondents' answers which describe respondents' perceptions of the questions asked. The scores obtained will be divided into 3 categories, namely 1.00-1.99 in the low category, 2.00-3.99 in the medium category, and 4.00-5.00 in the high category. Based on the range of values, the researcher determined the respondent's perception of the variables used in this study and processed it with SPSS 22.

Employee Performance Variable (Y2). The index calculation shows that the employee performance variable has a perception index of 4.32, so it can be concluded that the tendency of the performance of PT Pegadaian's head office employees is in the very good category. Work Discipline Variables. From the calculation of the index number, it is shown that the job satisfaction variable has a perception index of 4.54 so it can be concluded that it is in the very good category. This shows that the respondent's work discipline towards the work in the workplace can be said to be optimal. Training Variable (X1). From the calculation of the index number, it shows that the training style variable has a perception index of 4.42 so it can be concluded that the training variable is in the very good category. This shows that the training at the head office of PT Pegadaian has been very good. Work Motivation Variable. From the calculation of the index number, it shows that the organizational culture variable has a perception index of 4.48 so it can be concluded that the work motivation variable is in the very good category. This shows that the level of discipline at the head office of PT Pegadaian is very good.

Structural Equation Model

This research uses Structural Equation Model (SEM) analysis. The software used for this research is IBM AMOS 20. The theoretical model that has been described in the path diagram is analyzed based on the data obtained.

Analisis Full Model Structural Equation Modeling (SEM) Analysis

After the measurement model is analyzed through confirmatory factor analysis and it is seen that each indicator can be used to define a latent construct, then the testing process is continued with the Structural Model testing. The results of the full SEM model Structural Model test in the image below (Figure 2):

The results of the analysis of data processing in the Table 8 above show that there are two marginal indices, namely GFI and AGFI. AMOS provides two types of information that can be used to detect model specification errors, namely the standardized residual value and the modification index ( Ghozali, 2008). Based on the Table 8, although GFI and AGFI have not reached a value of 0.90, the results of the other indexes can inform that the modified model formed can represent all the constructs used in the study, so it can be said that the full SEM model analysis has met the goodness of fit criteria that have been set. The probability value in this analysis shows a value above the significance limit of 0.062 (p>0.05). This value indicates that there is no difference between the predicted covariance matrix and the estimated covariance matrix. This conclusion was put forward after comparisons were made to similar studies which resulted in the GFI and AGFI indexes not meeting the requirements and reinforced by the theory presented by Ghozali (2008)

Table 8 Standardized Regression Weight - Variable Confumatory Factor Analysis
Variable	Estimate	Variable	Estimate
K2 ß Performance	0.879	D1 ß Discipline	0.998
K3 ß Performance	0.990	D2 ß Discipline	0.817
K4 ß Performance	0.989	D5 ß Discipline	0.309
K5 ß Performance	0.995	D6 ß Discipline	0.894
K1 ß Performance	0.205	D3 ß Discipline	0.880
		D4 ß Discipline	0.920
Variable	Estimate	Variable	Estimate
P6 ß Training	0.996	Motivasi z11	1.000
P5 ß Training	0.996	M6 ß Motivation	0.993
P7 ß Training	0.995	M7 ß Motivation	0.987
P8 ß Training	0.976	M8 ß Motivation	0.999
P9 ß Training	0.976	M9 ß Motivation	0.989
P3 ß Training	0.995	M10 ß Motivation	0.991
P4 ß Training	0.986	M1 ß Motivation	0.990
P1 ß Training	0.946	M2 ß Motivation	0.988
P2 ß Training	0.991	M3 ß Motivation	0.987
		M4 ß Motivation	0.987
		M5 ß Motivation	0.984

Table 9 Hasil Pengujian Kelayakan Full Model SEM
Goodness-of-Fit Index	Cut-off Value	Result	Model Evaluation
χ2 -Chi-square (DF= 113)	< 450.81	438.80	Good
Significance Probabilitv	≥ 0.05	0.062	Good
RMSEA	≤ 0.08	0.050	Good
GFI	≥0.90	0.869	Marginal
AGFI	≥ 0.90	0.822	Marginal
CMIN/DF	≤ 2.00	1.260	Good
TLI	≥ 0.95	0.971	Good
CFI	≥ 0.95	0.976	Good

Structural Model Evaluation

Sample Size

To get a good model, the deviation problem will be tested against the SEM assumptions first. Based on the opinion of Wijaya (2009) and Santoso (2001), the number of samples that must be met for analysis using SEM is that the sample ranges from 100-200 or at least five times the number of indicators. And the opinion of Hair et al. (2006) in Ghozali (2008) which says that multivariate analysis with SEM will be effective on the number of samples between 140-400. So in this study using 30 observed variables or indicators so that the minimum sample is 30×5=150 respondents. Thus the number of samples collected as many as 192 respondents in this study was able to meet the criteria.

Data Normality Evaluation

Evaluation of normality is carried out using the criteria for the critical ratio skewness value of ± 2.58 at a significance level of 0.01. The data can be concluded to have a normal distribution if the critical ratio skewness value is below the absolute value of 2.58 ( Haryono & Wardoyo, 2012). The results of the normality test of data by AMOS 20 (attachment 5.10) are shown in Table 5.19. The results of the data normality test show that there is no C.R. for skewness and kurtosis which are outside the range of ± 2.58. Thus, the research data used have met the data normality requirements, or it can be said that the research data has been normally distributed. Thus the assumption of normal data in this case can be fulfilled.

Evaluation of Multicollinearity and Singularity

Wether multicollinearity and singularity in a combination of variables is exist, can be seen at the determinants of the covariance matrix. A determinant that is really small or close to zero indicates the existence of multicollinearity or singularity. Thus the data cannot be used for the current analysis. From the output produced by AMOS 20 for research data, the determinant of sample covariance matrix is 5.568. Because it is far from zero, therefore there is no multicollinearity or singularity in the data of this study. Thus the research data can be used in the analysis.

Outlier Evaluation

Outlier evaluation is the observation conditions of data that have unique characteristics, due to very different from other observations and appear in extreme forms (Hair et al., 2006; Ghozali, 2008). Outlier detection is done to see multivariate outliers by looking at the mahalanobis distance value. The cut off value generally used to detect the presence or absence of outlier data is the value of p1 and p2 must be>0.05. If there is a mahalanobis distance value, it means that there is a multivariate outlier problem ( Ferdinand, 2000). The criteria used are based on the chi-square value at 40 degrees of freedom (according to the number of indicator variables) at a significance level of p < 0.001. The value of mahalanobis distance 2 (40, 0.001)=73.40196. This means that all cases that have a mahalanobis distance greater than 73.40196 are multivariate outliers ( Haryono & Wardoyo, 2012). Therefore, it can be concluded that in this study there is no multivariate outlier. Because there is no multivariate outlier, it means that the data is suitable for use.

Construct Reliability Test and Extract Variance

Basically, the reliability test shows the extent to which a measuring instrument can provide relatively the same results when repeated measurements are made on the same subject. There are two ways that can be used, namely composite (construct) reliability and extravted variance. The cut off value of construct reliability is at least 0.70 while the cut off value of the extravted variance ( Ghozali, 2008). The calculation results of construct reliability and variance extracted from each latent variable are based on the results of AMOS 20 calculations. Meanwhile, the processing of reliability and variance extract values uses Microsoft Excel. The results of the reliability and variance extract calculations show that all reliability values are above 0.7. This means that the measurement of this SEM model has met the reliability requirements of the gauge. The variance extract value is also above 0.5. This means that the measurement of this SEM model has met the requirements of good factor extraction. The test results show that all construct reliability and variance extracted values are above the required values. This means that the measurement of the SEM model has met the requirements.

Discriminant Validity

Discriminant Validity measures the extent to which a construct is completely different from another. A high value of Discriminant Validity provides evidence that a construct is unique and able to capture the phenomenon being measured. The way to test it is by comparing the value of the square root of the Average Variance Extracted (AVE) with the correlation between constructs ( Haryono & Wardoyo, 2012). The square root value of the variance extracted construct is training of 0.7307, motivation of 0.7674, discipline of 0.7970 and employee performance of 0.7436

Hypothesis Testing

After assessing the suitability of the model through goodness of fit analysis and the assumptions in SEM, then hypothesis testing will be carried out. The testing of the five hypotheses proposed in this study was carried out by analyzing the Critical Ratio (CR) value and the Probability (P) value of the data processing results compared to the required statistical limits, which were above 1.967 for the CR value and below 0.05 for the P value (probability). ). If the results of the data processing show a value that meets these requirements, the proposed research hypothesis can be accepted. In detail, the research hypothesis testing will be discussed in stages in accordance with the proposed hypothesis. Based on the data obtained, the results of the AMOS 20 analysis as a hypothesis testing step are as follows (Table 10):

Table 10 Hypothesis Testing
	Estimate	S.E.	C.R.	P	Label
Discipline ßTraining	0.364	0.105	3.454	0.001
DisciplineßMotivation	0.243	0.11	2.218	0.027
PerformanceßTraining	0.301	0.136	2.216	0.027
PerformanceßMotivation	0.48	0.142	3.389	0.001
PerformanceßDiscipline	0.479	0.181	2.645	0.008

Testing Hypothesis 1: The Effect of Training on Work Discipline

The estimated parameter of the relationship between training variables and discipline is 0.364. The test shows the results that meet the requirements, namely producing the value of CR=3.454>1.967 with probability=0.001<0.05. Thus, hypothesis 1 is accepted and it can be concluded that training has a significant effect on employee work discipline.

Hypothesis Testing 2: The Effect of Work Motivation on Work Discipline

The estimated parameter of the relationship between work motivation and discipline is 0.243. The test shows the results that meet the requirements, namely producing the value of CR=2.218>1.967 with probability=0.027<0.05. Thus, hypothesis 2 is accepted and it can be concluded that work motivation has a significant effect on work discipline.

Hypothesis Testing 3: The Effect of Training on Performance

The estimated parameter of the relationship between training variables and employee performance is 0.301. The test shows the results that meet the requirements, namely producing the value of CR=2.216>1.967 with probability=0.027<0.05. Thus hypothesis 3 is accepted and it can be concluded that employee training has a significant effect on employee performance.

Hypothesis Testing 4: The Effect of Work Motivation on Performance

The estimated parameter of the relationship between work motivation and employee performance is 0.480. The test shows the results that meet the requirements, namely producing the value of CR = 3.389> 1.967 with probability = 0.001 <0.05. Thus, hypothesis 4 is accepted and it can be concluded that work motivation has a significant effect on employee performance.

Hypothesis Testing 5: The Effect of Discipline on Performance

The estimated parameter of the relationship between work discipline variables and employee performance is 0.479. The test shows the results that meet the requirements, namely producing the value of CR = 2.645 > 1.967 with probability=0.008<0.05. Thus, hypothesis 5 is accepted and it can be concluded that work discipline has a significant effect on employee performance.

Analysis of Direct, Indirect And Total Effects

The strength of the influence between constructs is analyzed both direct, indirect, and total effects. The direct effect is all coefficient lines with one end arrow. Indirect effects are effects that arise through an intervening variable. The total effect is the effect of various relationships. Based on the results of calculations with AMOS 20 Standardized Direct Effects, the values of direct and indirect effects are obtained as follows (Figure 3):

Direct Influence in Research is as Follows

From the information presented in the Table 11 below, it is known that the influence of training and motivation on employee performance through the mediating variable of discipline each has a greater total effect than the direct effect, meaning that discipline has an important role in mediating the effect of training and motivation on employee performance. From the results of the study, it was also found that motivation had a greater effect than training on employee performance.

Table 11 Testing the Effect of Intervening Variables
Variable	Effect
	Direct	Indirect	Total
Training	0.271	0.508×0.309 = 0.157	0.428
Motivation	0.381	0.299×0.309 = 0.092	0.473

Discussion

Based on the findings, this study concluded that training has a direct and significant effect on discipline cannot be rejected. Thus, the level of employee discipline can be improved by providing employee training. Employees who are disciplined and obey the rules, obey all applicable norms and regulations in the organization or agency will be able to increase efficiency, effectiveness and productivity. Meanwhile, organizations or agencies that have undisciplined employees will find it very difficult to implement their programs to increase productivity. Therefore, training is needed to improve employee discipline. This study is in line with previous studies by Soleman (2019), Subari & Riady (2015) using SEM proceed by LISREL, Ongori et al. (2011) and Rivai (2005).

Motivation has a significant effect on employee work discipline, thus the level of employee discipline can be explained by motivation. Motivation can foster discipline that is directed to improve and shape employee attitudes and behavior voluntarily. This study support previous study by Utama & Sriathi (2021) using path analysis concluded work discipline influenced task performance.

Discipline has a significant effect on employee performance. Thus the high and low work performance of employees can be explained by the work discipline factor. Discipline must be enforced in a company organization. Without the support of good employee discipline, the company is difficult to realize its goal of achieving optimal employee performance. This study support previous study by Jalil et al. (2019)

Training has a significant effect on employee performance. Based on the research, it was found that training had a direct effect on employee performance or indirectly through disciplinary mediation variables. In mediating the effect of training on employee performance. So it can also be concluded that if you want to improve employee performance through training, it will be more effective if it is accompanied by increasing employee discipline than without the discipline factor. Utama & Sriathi (2021) and Jalil et al. (2019) concluded that training and work discipline had positive impact on employee's performance, also would strengthen organizational culture in influencing employee task performance

Based on the research, it is found that motivation have directly effect on employee performance and also indirectly through the disciplinary mediating variable. From the research, it was found that the influence of motivation on employee performance through the mediating variable of discipline has a greater total effect than the direct effect, meaning that motivation has an important role in mediating the effect of training on employee performance. So it can be concluded that if you want to improve employee performance by using the variable of increasing work motivation, it will be more effective if it is accompanied by an increase in work discipline. From the results of the study also obtained that motivation has a greater effect than training on employee performance.

Conclusion

Based on the findings and discussion above, it is concluded that training and motivation have a significant effect on employee work discipline. Thus, the level of employee discipline can be improved by providing employee training. Training and motivation have a direct and significant effect on acceptable discipline. Discipline has a significant effect on employee performance. Thus, the high and low work performance of employees can be explained by the work discipline factor. “Discipline, training has a direct and significant effect on performance” is acceptable.

Based on the research, it was found that training had a direct effect on employee performance and indirectly through the disciplinary mediating variable. From the research, it was found that the effect of training on employee performance through the mediating variable of discipline has a greater total effect than the direct effect. This means that discipline has an important role in mediating the effect of training on employee performance. So it can also be concluded that if you want to improve employee performance through training, it will be more effective if it is accompanied by increasing employee discipline than without the discipline factor.

Motivation has a significant effect on employee performance. Based on the research, it is found that motivation has a direct effect on employee performance or indirectly through the disciplinary mediating variable. From the research, it was found that the influence of motivation on employee performance through the mediating variable of discipline has a greater total effect than the direct effect, meaning that motivation has an important role in mediating the effect of training on employee performance. So it can be concluded that if you want to improve employee performance by using the variable of increasing work motivation, it will be more effective if it is accompanied by an increase in work discipline. From the results of the study also obtained that motivation has a greater effect than training on employee performance. Based on the findings of this study, it can be concluded that the hypothesis: “The Influence of Training and Motivation on Discipline that has an Impact on Performance at PT Pegadaian (Persero) head office employees” can be accepted.

Research Implication

By paying attention to the average value of respondents' answers from each variable which on average is very good, can be implied that employee performance is the main factor that determines the success of the company which is influenced by the existence of training, motivation and work discipline. For this reason, to be a concern for the management of PT Pegadaian (Persero), it is necessary to strive to encourage the maintenance of the current conditions, to survive, and even further improvement. So that the vision and mission organization can actually be realized. Implementing training, motivation and work discipline requires serious commitment and effort from the management so that this system can provide progress to the company and the trust of the community who use pawnshops.

For future study, to reduce the limitations of this study we suggest stratifying the research object, especially on education or gender to sharpen the results. The addition of other variables such as compensation will certainly provide more in-depth results in seeing the purpose of the study. The level of discipline and employee performance will certainly change over time, so it is necessary to conduct regular research to determine the level of these changes.

References

Abboh, U.A., Majid, A.H.A., & Fareed, H. (2019). Examining the effect of HR practices on job performance in Nigerian polytechnics. International Journal of Economics, Commerce and Management, 7(9), 98-115.

Abdullah, D.K.K. (2009a). The effect of training in order to improve skills and work motivation on lecturer performance. Management Application Journal, 7(1), 72-79.

Abdullah, H. (2009b). Major challenges to the effective management of human resource training and development activities. Journal of International Social Research, 2(8), 13-25

Afful-Broni, A. (2012). Relationship between motivation and job performance at the University of Mines and Technology, Tarkwa, Ghana: Leadership Lessons. Creative Education, 3(3), 309.

Anyim, C.F., Chidi, O.C., & Badejo, A.E. (2012). Motivation and employees' performance in the public and private sectors in Nigeria. International Journal of Business Administration, 3(1), 31.

Arbuckle, J.L. (1995). Amos™ 7.0 user’s guide. Amos Development Corporation.

Arikunto, S. (2006). Research procedures, a practical approach. PT Rineka Cipta, Jakarta.

Cruz, N.M., Pérez, V.M., & Cantero, C.T. (2009). The influence of employee motivation on knowledge transfer. Journal of Knowledge Management

Dessler, G. (2004). Human resource management ninth edition. Jakarta: PT. Gramedia Group Index.

Dharma, A. (2004). Supervision management, revised edition. PT Raja Grafindo.

Fareed, M., Mohd Isa, M.F., Ahmad, A., & Laeeq, H. (2016). Performance management system (PMS) as a predictor of employee performance of private bank managers in (Pakistan). The Social Sciences, 11(20), 5035-5042.

Ferdinand, A. (2005). Structural equation modeling in management research (3rd Edn.). Diponegoro University Publishing Agency, Semarang.

Flippo, E.B. (2003). Personnel management, (Moh. Mas' ud, Translator) Erlangga.

Ghozali, I. (2008). Structural equation models: Concepts and applications with the AMOS 16.0 program. Diponegoro University Publishing Agency.

Gomes,  F.C.  (2005). Human resource management.  Andi Offset, Yogyakarta:

Hair, J.F., Anderson, R.E., Thatam, R.L. dan Black, W.C. (2006). Multivariate data analysis with reading, (4th Edn.). Englewood Cliffs, Prentice Hall, NJ.

Hairina, D. (2020). The relationship of emotional intelligence and work motivation through work discipline on the performance of state vocational high school teachers in banjarmasin city. Journal of K6 Education and Management, 3(3), 327-341.

Handoko, T.H. (2005). Manajemen sumber daya manusia. BPFE, Yogyakarta, 52.

Harlie, M. (2016). The influence of work discipline, motivation and career development on the performance of civil servants at the tabalong regency government in tanjung, south kalimantan. Journal of Management and Accounting , 11 (2).

Haryono, S., & Wardoyo, P. (2012). Structural equation modeling for management research using AMOS 18.00. Publisher PT Intermedia Personalia Utama, Bekasi.

Hasibuan, M.S. (2008). Human resource management: Jakarta, Indonesia: PT. Earth Literature.

Heldia, P. (2012). The influence of discipline, motivation and work environment on employees in Central Tapanuli district. Universitas Terbuka.

Henry, O., & Jennifer, C.N. (2011). Training and development practices in an organisation: An intervention to enhance organisational effectiveness. International Journal of Engineering and Management Sciences, 2(4), 187-198

Horgan, J., & Mühlau, P. (2006). Human resource systems and employee performance in Ireland and the Netherlands: A test of the complementarity hypothesis. The International Journal of Human Resource Management, 17(3), 414-439.

Ivancevich, J.M., Konopaske, R., & Matteson, M.T. (2006). Organizational behavior and management (7th Edn.). Erlangga, Jakarta.

Jalil, R., Perizade, B., & Widiyanti, M. (2019). effect of training and work discipline on employee performance PT PLN (Persero) power plants control unit keramasan. International Journal of Management and Humanities, 3(11), 37-42.

Jöreskog, K.G., & Sörbom, D. (1996). LISREL 8: User's reference guide. Scientific Software International.

Laeeq, H., Shahzad, A., Ramalu, S. S., & Fareed, M. (2016). Determinants of organizational performance: a case of Punjab Police. Sukkur IBA Journal of Management and Business, 3(2), 50-60.

Mangkunegara, A.P., & Prabu, A. (2005). Evaluation of human resource performance. Refika Aditama, Bandung.

Manullang . (2004). Personnel management. Ghalia Indonesia, Jakarta.

Mapparenta. (2011). Effect of education, training and work discipline on work productivity of regional secretariat employees, East Halmahera Regency, North Maluku Province.

Mas’Ud, F. (2004). Organizational diagnostic survey of concepts and applications. Diponegoro University Publishing Agency.

Mathis, R.L., & Jackson, J.H. (2006). Human resource management, translation (10thEdn.). Jakarta, Salemba Empat.

Moekijat. (1993). Evaluation and training in order to increase productivity (company). Mundur Maju, Bandung.

Mozammel, D. (2019). Understanding post-training factors and job performance relationship: review of literature for transfer of training enthusiasts. Annals of Contemporary Developments in Management & HR (ACDMHR),1(3), 33-38.

Mulyadi.  (2007). Management planning and control system. Salemba Empat, Jakarta.

Nawawi, H.H. (1983). Administration and organization of guidance and counseling. Indonesian Gholia.

Parlinda, V., & Mahyuddin. (2005). The influence of leadership, motivation, training, and work environment on employee performance at a regional water company in surakarta city. Scientific journals. Muhammadiyah University of Surabaya

Porwani, S. (2011). Education and training relations. Scientific Journal of Business Administration Darussalam Polytechnic Palembang Polytechnic, 3(3), 1-9.

Raza, A., Wan Mohd Noor, W.S., & Fareed, M. (2020). Mediating the role of employee willingness to perform between career choice and employee effectiveness (case study in public sector universities of Pakistan). International Journal of Innovation, Creativity and Change, 11(1), 33-389.

Rivai, V. (2005). Leadership and organizational behavior (First Printing). PT Raja Grafindo Persada, Jakarta.

Robbins. S.P. (2003). Organizational behavior: Concepts, controversy and applications(1st Edn.). PT. Prehalindo Persada, Jakarta.

Salleh, S.S.M.M., Zubair, M.F., & Hamzah, A.W. (2018). Gender differences in leadership styles and its impact on employees’ motivation. International Journal of Management and Business Research, 8(2), 86-97.

Santoso,  S.  (2001). SPSS version 10.01, processing statistical data professionally. PT Elex Media Komputindo, Jakarta.

Santoso, S. (2018). Basic concepts and SEM applications with AMOS 24. Elex Media Komputindo.

Saydanm, G. (2006). Human resource management. Jambatan, Jakarta.

Schermerhorn, R.J., Hunt, G.J., & Osborn, N.R. (1996). Organizational behavior (6thEdn.). John Willey & Son, NY.

Shah, S.A., Shah, M.H., Matloob, S., & Abbas, S.I. (2021). Managing organizational performance of pharmaceutical smes in indonesia: examining the role of entrepreneurial orientation and organizational learning in a mediated model. Annals of Contemporary Developments in Management & HR (ACDMHR), 3(1), 24-32.

Siagian, S.P. (2008). Human resource management. Raja Grafindo Persada, Jakarta.

Sihotang. (2007). Human resource management (1st Edn.). Pradnya Paramita, Jakarta.

Simamora, H. (2006). Human resource management (2nd Edn.), YKPN College of Economics, Yogyakarta..

Siswanto, H.B. (2005). Introduction to management, cet. 1, PT. Bumi Aksara, Jakarta.

Soegandhi, V.M. (2013). The effect of job satisfaction and job loyalty on organizational citizenship behavior in employees of PT. Surya Timur Sakti Jatim. Agora, 1(1), 808-819.

Subari, S., & Riady, H. (2015). Influence of training, competence and motivation on employee performance, moderated by internal communications. American Journal of Business and Management, 4(3), 133-145.

Sugiyono. (2009). Business research methodology. Alfabeta, Bandung.

Sumarsono, H.M. (2004). Human resource research methods.

Sutrisno, E. (2009). Human resource management first edition. Jakarta: Kencana Prenada Media Group.

Timpe, D. (2007). The science and art of business management (leading people) series.PT. Elex Media Komputindo, Jakarta.

Tus, J. (2020). Academic stress, academic motivation, and its relationship on the academic performance of the senior high school students. Asian Journal of Multidisciplinary Studies, 8(11).

Umar, H. (2003). Organizational behavior research methods. Gramedia Main Library.

Utama, K.M., & Sriathi, A.A. (2021). The effect of organizational culture on employee performance in mediation by work discipline at PT. American International Journal of Business Management, 4(6), 52-57.

Wardono, M.N.S. (2012). The effect of education, training and motivation on employee performance at the semen district office, kediri regency. Journal of Management Science, Revitalization, 1(2), 115-124.

Wibowo. (2007). Work management. PT Raja Grafindo Persada, Jakarta.

Widarjono, A. (2010). Applied multivariate statistical analysis. UPP STIM YKPN, Yogyakarta.

Wijaya, T. (2009). Analysis of structural equation modeling using AMOS. Atma Jaya University Yogyakarta, Yogyakarta.