Factors Affecting Employee Satisfaction at PT. Operational Division of Terminal-B TPK Belawan Pelindo-I Branch

. This study aims to determine whether compensation, work itself, promotion, work supervision, and co-workers partially and simultaneously have a positive and significant effect on employee job satisfaction at PT Pelindo-I branch TPK Belawan Terminal-B operational division. The data analysis technique used is the quantitative method with the help of the SPSS version 20.0 program. This research uses Confirmatory Factor Analysis and Multiple Linear Regression. The population in this study were all employees of the operational division at PT Pelindo-I branch TPK Belawan Terminal-B operational division, and a sample of 82 respondents. Primary data collection using a questionnaire. The results of the Confirmatory Factor Analysis test show that the components that most influence job satisfaction are compensation with eigenvalues of 2,644 and the work itself with eigenvalues of 1,274. The results of the study harmed job satisfaction and the work itself had a positive and significant effect on job satisfaction at PT Pelindo-I Branch TPK Belawan Terminal-B operational division. Compensation and the work itself simultaneously have a positive and significant effect on employee job satisfaction at PT Pelindo-I Branch TPK Belawan Terminal-B operational division.


INTRODUCTION
The aspect of human resources plays a very important and most dominant role in an organization. Human resources are an inseparable part of management in general, which is one of the determining factors for the company's success in achieving organizational goals. Employees in a company are not only positioned as factors of production but rather as company assets that must be managed and developed. Management in human resources is a field of management that specifically explores the relationship and role of human management in corporate organizations [1].Companies must be able to improve their quality to survive and develop their existence. To improve this quality, it is necessary to increase productivity both in quantity and quality as well as the ability to provide the best service that meets the desires and satisfaction of service users.Increased productivity by increasing the performance of employees of an organization, must be preceded by an increase in organizational commitment and job satisfaction of its human resources. Factors that encourage the workforce to continue to increase organizational commitment and job satisfaction of human resources need to be considered. Job satisfaction is a general attitude of an individual towards his work [2]. ISSN: 2722 -4015 http://ijstm.inarah.co.id To find out the extent to which employee job satisfaction at Belawan Container Terminal (CTB) Terminal-B Operational Division and its relationship to productivity, the authors conducted a preliminary research survey of 20 respondents.There are many factors that make an employee get job satisfaction, including compensation, the work itself, promotions, work supervision, and co-workers. Compensation is something that employees receive in exchange for their service contribution to the company. Based on a preliminary survey of 20 employees, the results show that there is a problem with the compensation received by employees because there are 61% of employees do not agree that employees receive payments as stated on the payslips.  individual or group of employees organizationally. Its purpose is to organize work assignments that meet organizational, technological, and behavioral needs.

Job Promotion
Promotion occurs when an employee is transferred from one job to another that is higher in pay, responsibility, and level [8]. Promotion is when a person is transferred from one job to another with greater responsibilities, a higher level of hierarchical position, and a higher income [11].

Work Supervision
Supervision is an activity carried out so that the work process is in accordance with the desired results [12]. Supervision in the company aims to find out whether the implementation of tasks/work is in accordance with the predetermined plan. Supervision is the process of observing the implementation of all organizational activities to be able to ensure that all work being carried out can be in accordance with a predetermined plan [11].Colleagues are everything that is done by the leadership and human resources department that will affect the relationship with employees, either directly or indirectly. Working relationships between employees need to be fostered so that employees can work together and help each other in achieving company goals. Employees who feel happy about their activities and duties and are friendly with others show a harmonious working atmosphere [8].

III. RESEARCH METHODS
The research approach used in this study is quantitative. Quantitative data analysis is used when researchers use survey research approaches, experimental research, quantitative research, and several other studies that contain data in the form of numbers [13]. Associative quantitative research aims to determine the degree of relationship and the pattern of influence between the independent variables (independent) on the dependent variable (dependent).This research was conducted at the office of PT. Pelindo-I TPK Belawan Branch has its address at GRHA Pelindo-I Jl. Harbor Ring No. 1 Belawan, Medan -North Sumatra from July to December 2021.
The population used in this study were employees of PT. Pelindo-I Branch TPK Belawan Terminal-B Operational Division, totaling 139 employees. The sampling technique in this study is a probability sampling technique using clusters. Cluster sampling is a technique used to determine a sample in groups. Samples taken at Terminal-B Operational Division amounted to 82 people.The type of data used in this study is primary data obtained through interviews and distributing questionnaires to respondents. Data collection techniques in this study were questionnaires, interviews, observations, and libraries. This study uses 5 (five) independent variables, namely: Compensation (X1), Work Itself (X2), Promotion (X3), Work Supervision (X4), Coworkers (X5), and 1 (one) dependent variable, namely Job Satisfaction (Y). ISSN: 2722 -4015 http://ijstm.inarah.co.id

Data Analysis Techniques
The data analysis technique used is: 1. Test data quality. The research results are said to be valid if there are similarities between the data collected and the data that occurs in the object under study.

Classic assumption test. A classical assumption test is a test of statistical
assumptions that must be met in linear regression analysis based on Ordinary Least Square (OLS). 3. CFA Factor Analysis (Confirmatory Factor Analysis). Factor analysis aims to find a way of summarizing the information contained in the original (initial) variable into a new set of dimensions or variate (factor). 4. Multiple linear regression. The data analysis model used to determine the magnitude of the influence of the independent variable on the dependent variable is the econometric model with the analysis technique using the ordinary least squares model. 5. Hypothesis testing. Hypothesis testing includes a simultaneous test (F test) to find out whether the effect of all independent variables simultaneously has a significant effect on the dependent variable at the confidence interval or the 5% hypothesis testing level with the F test using statistical formulas and partial test (ttest) for shows how far the influence of each independent variable partially on the coefficient of determination (KD). The coefficient of determination is used to find out how big the relationship between several variables is in a clearer sense [14].

IV. RESEARCH DISCUSSION ANALYSIS A General Description of the Company
The port that is given the task of carrying out port activities in order to serve the public interest is realized in the form of a company that is a State-Owned Enterprise (BUMN). One of the SOEs assigned to carry out port activities is PT. Pelabuhan Indonesia (Pelindo)-I (Persero) which is domiciled in Belawan, Medan City, North Sumatra. PT. Pelindo-I (Persero) has several branches, one of which is Belawan Port. Furthermore, Belawan Port has several business units, one of which is the Belawan Domestic Container Terminal (TPKDB).

Validity Test
Through the results of the SPSS version 20 output, it is known that the validity value is in the Corrected item-total Correlation column, which means the correlation value between the scores of each item and the total score in the tabulation of respondents' answers.
The results of the validity test of 10 (ten) questions on the compensation variable can be declared valid (valid) because all coefficient values are greater than 0.30. (Table 1) The results of the validity test of 6 questions on the work variable itself can be declared valid (valid) because all coefficient values are greater than 0.30 (Table 2). The results of the validity test of 18 (eighteen) questions on the promotion variable can be declared valid (valid) because all coefficient values are greater than 0.30 (Table 3). The results of the validity test of 6 (six) questions on the work supervision variable can be declared valid (valid) because all coefficient values are greater than 0.30 (Table 4). The results of the validity test of 8 (eight) questions on the coworker variable can be declared valid (valid) because all coefficient values are greater than 0.30 (Table  5). The results of the validity test of 10 (ten) questions on the job satisfaction variable can be declared valid (valid) because all coefficient values are greater than 0.30 (Table 6).

Reliability Test
The questionnaire is said to be reliable if Cronbach's alpha > 0.60 and unreliable if it is equal to or below 0.60.  Figure 3, for the results of testing the normality of the data using the PP Plot image, it can be seen that the data points spread around the diagonal line so that the data is normally distributed. From the two images above, it can be concluded that after the data normality test was carried out, the data for the employee performance variable was normally distributed. To further confirm whether the data along the diagonal line is normally distributed or not, the Kolmogorov Smirnov test (1 Sample KS) was carried out by looking at the residual data and whether the distribution was normal or not. If the value of Asym.sig (2-tailed) > significant level (α = 0.05) then the residual data is normally distributed.  a. Test distribution is Normal. b. Calculated from data. In table 8 it can be seen that the results of the data processing, the significance value of Kolmogorov Smirnov are 0.637, so it can be concluded that the data is normally distributed, where the significance value is greater than 0.05 (p = 0.637> 0.05).

Multicollinearity Test
The multicollinearity test from the results of the questionnaire that has been distributed to respondents can be seen in the following table. Based on table 9 above, it can be seen that the Variance Inflation Factor (VIF) number is less than 10, including compensation 2,374 < 10, the work itself 4,721 < 10, promotion 1,383 <10, work supervision 3,093 <10 and coworkers 5,424 < 10 and the value of compensation tolerance is 0.421 > 0.10, the work itself is 0.212 > 0.10, promotion is 0.723 > 0.10, work supervision is 0.323 > 0.10 and co-workers are 0.184 > 0.10 so that they are free from multicollinearity. ISSN: 2722 -4015 http://ijstm.inarah.co.id

Heteroscedasticity Test
The results of the heteroscedasticity test can be seen where the scatterplot image shows that the resulting points spread randomly and do not form a certain pattern or trend line.The figure 4 also shows that the distribution of the data is around the zero point. From the results of this test, it shows that this regression model is free from heteroscedasticity problems, in other words: the variables to be tested in this study are homoscedasticity. Table 10 shows the results of the total variance explained, it is known that only 2 variables are components that affect job satisfaction. Eigenvalues show the relative importance of each factor in calculating the variance of the 5 variables analyzed.  Table 11 shows that 2 factors are the most optimal number, so it can be seen in the component matrix table. The process of determining which variable will enter which factor is carried out by comparing the magnitude of the correlation in each row. Based on the results of the component matrix values, it is known that the 5 factors, that are feasible to influence job satisfaction are compensation and the work itself. So that the OLS equation model in this study is formulated:

Fig 4. Heteroscedasticity Test Scatterplot Factor Analysis
In the analysis of a study, after knowing which factors will represent a dependent, the next analysis is carried out with multiple regression.

Multiple Linear Regression
Multiple linear regression analysis was conducted to determine the effect of the independent variable, namely compensation on job satisfaction. a. If everything in the independent variables is considered non-existent, then job satisfaction (Y) is 28.055 units. b. If there is an increase in compensation by 1 unit, then job satisfaction (Y) will increase by -0.121 units. c. If there is an increase in the work itself by 1 unit, then job satisfaction (Y) will increase by 0.516 units.

Hypothesis Testing
Partial Test (t) shows how far the independent variables individually explain the variation. This test is carried out using a level of significance of 5%.  Based on table 14 above, it can be seen that the number R Square is 0.099 which can be called the coefficient of determination which in this case means 9.9% job satisfaction can be obtained and explained by compensation and the work itself. While the remaining 100% -9.9% = 90.1% is explained by other factors or variables outside the model, such as opportunities for advancement, performance, physical and nonphysical work environment, and others.
b. The resulting R-value is 0.315. This R-value shows a not close relationship between compensation and the work itself on job satisfaction. The closer to number one, the model issued by the regression will be better.

V. CONCLUSION
Based on the results of the research described above. Then it can be concluded that: 1.