THE INFLUENCE OF TRANSFORMATIONAL LEADERSHIP STYLE AND COMPENSATION ON EMPLOYEE PERFORMANCE DIMEDIATION OF WORK SATISFACTION AT PT. TOR GANDA

This study aims to identify the influence of transformational leadership style and compensation on employee performance mediated by job satisfaction at PT. TOR GANDA. The number of samples in this study were 92 respondents who were determined using simple random sampling. The type of data used is associative research. Research connects causation or causation. which analyzes the independent variables, mediating variables and dependent variables that are affected. Data analysis used a simple linear regression equation with the SPSS 2023 tool. The results showed that Transformational Leadership had an effect on Job Satisfaction at PT. TOR GANDA, Compensation affects Job Satisfaction at PT. TOR GANDA, Transformational Leadership influences Employee Performance at PT. TOR GANDA, Compensation affects Employee Performance at PT. TOR GANDA, Job Satisfaction affects Employee Performance at PT. TOR GANDA, Transformational Leadership influences Employee Performance through Job Satisfaction at PT. TOR GANDA and Compensation affect Employee Performance through Job Satisfaction at PT. TOR GANDA.


INTRODUCTION
Performance is a concrete channel displayed by everyone in the resulting work performance. All performance or work results shown by good employees can be classified as performance. Performance comes from the root word "work" which translates the word performance which means work. Performance according to Mangkunegara (2019). is the result of work in quality and quantity achieved by an employee in carrying out his duties in accordance with the responsibilities given to him. Hasibuan (2019) suggests performance is a work result achieved by a person in carrying out the tasks assigned to him based on skills, experience and sincerity and time. Leadership is one of the factors in improving employee performance, because basically leaders are the backbone of organizational development to encourage and influence good morale for employees. Transformational leadership or transformational leadership is a leadership style that identifies the changes needed, develops a vision that will pave the way for changes to be made and implements the necessary plans for these changes to occur (Danim, 2017).
In addition to the leadership factor, compensation has a close relationship with employee performance. Compensation is anything that is received by employees as remuneration for their work that is balanced with the expectations of employees to meet the need for satisfaction with the achievements that have been achieved and aligned with the strategic objectives of the company's business ( Dewi et al, 2022). Compensation in the form of salary, incentives, bonuses, benefits, work and work environment is an important factor to influence the improvement of an employee's performance. Job satisfaction is an attitude of employees or employees towards work related to work situations, cooperation between employees, rewards received at work, and other matters related to physical and psychological factors (Hamali, 2017). Leadership and job satisfaction are aspects that are interrelated with one another and play an important role in determining the performance of an employee. With high leadership it will encourage employees to work in a disciplined manner which will ultimately lead to job satisfaction (Paendong et al, 2020). In addition, the compensation provided by the company can retain employees who are considered potential and qualified to continue working. In addition, compensation can also increase employee job satisfaction so that they are more active at work (Winata, 2019).
PT. TOR GANDA is a national private company founded in 1979 and is engaged in the plantation business and the oil palm plantation industry. PT. TOR GANDA is headquartered in Medan, with operational areas spread across Indonesia. Since its establishment, PT. TOR GANDAs have established sustainable relationships with relevant stakeholders, both locally, nationally and globally. The main business activities of PT. The TOR GANDA includes cultivating oil palm plants, harvesting fresh fruit bunches (FFB) and processing them into crude palm oil (CPO).
Performance indicators related to human resources on targets and realization in 2021 at PT TOR GANDA, can be seen as follows: In the PT TOR GANDA Key Performance Indicator report, which relates to human resources. The planned 2021 target is not equal to or exceeds the actual achievements in 2021. Successful achievements are in accordance with the KPI realization, namely monthly management reports, ideas/suggestions from investors (leaders), employee competencies and career planning systems. In addition to KPIs, researchers also conducted interviews with employees on leadership variables at PT TOR GANDA with indicators of respect from employees and understanding the needs of subordinates with employees in the field. In the results of the first pre-survey, namely "Is the leadership at PT TOR GANDA able to arouse respect from subordinates?", 23 people agreed (76.67%), 7 people said they did not agree (23.33%). The majority of respondents stated that they agreed that the leadership at PT TOR GANDA was able to inspire respect from subordinates. However, it should be noted that this survey was only conducted with a small number of respondents, and there are still respondents who disagree with these two question items so that the results cannot be generalized to all employees at PT TOR GANDA, and further research is needed. Leadership style has an influence on employee performance. If the leadership style for employees is further enhanced, then employee performance will also increase, and vice versa if the leadership given to employees is low, employee performance will decrease. One of the leadership styles is transformational leadership. To find out more about the problem gaps related to the two question items regarding compensation and presurvey with a sample of 30 respondents. Pre-survey results can be seen in the following table: On the question "Does the company pay salaries in accordance with the UMR set by the government?", 27 people agreed (90.00%), 3 people (10.00%) said they did not agree. The majority of respondents agreed that the company pays salaries in accordance with the UMR set by the government. This shows that most of the respondents believe that the company pays its employees fairly. In the next compensation question "Do the facilities provided by PT TOR DANDA make it easier for employees to work optimally?", 14 people agreed (46.67%), 16 people (53.33%) said they disagreed. More respondents answered disagree, that the facilities provided by PT TOR GANDA make it easier for employees to work optimally. This shows that a number of respondents felt that the facilities provided by the company were inadequate to support their performance. The

THE INFLUENCE OF TRANSFORMATIONAL LEADERSHIP STYLE AND COMPENSATION ON EMPLOYEE PERFORMANCE DIMEDIATION OF WORK SATISFACTION AT PT. TOR GANDA Venna Natala Lingga, Prihatin Lumbanraja, Meilita Tryana Sembiring
results of the two pre-surveys are important for further analysis, because these results do not represent all the employees studied at PT TOR GANDA.
The results of the pre-survey show that it is still necessary to analyze the employee's performance on the work carried out. A decrease in employee performance can have a negative impact on the organization, because poor performance can interfere with productivity and efficiency, and can affect the company's reputation.

IMPLEMENTATION METHOD Types of research
This type of research is associative research. Associative research is research that analyzes the relationship between two or more variables Sugiyono (2019). Research connects causation or causation. which analyzes the independent variables, mediating variables and dependent variables that are affected.
A causal relationship arising from the independent variables, namely (X1) Transformational Leadership, (X2) Compensation, mediating variable (Z) Job Satisfaction, and the dependent variable (Y) Employee Performance at PT. TOR GANDA.

Location and Time of Research
This research was conducted at PT. TOR GANDA Kandir Jl. Abdullah Lubis No. 26 Kel. Babura, Kec. Medan Baru, Medan City. This research was conducted for 3 (three) months, from January 2023 to March 2023.

Population and Research Sample
The population in this study were all employees of PT. TOR GANDA Kandir, totaling 118 people. The sample is part of the population that is considered representative of that population, (Sinulingga, 2019). The sampling technique used in this research is probability sampling. Sampling by simple random sampling, taking samples from members of the population in a random way without regard to strata or levels in the members of the population in this study. Rounding off the calculation of the slovin formula in this study amounted to 92 samples.

Data collection technique
Data collection techniques in research carried out by means of: a. Interview.
Interviews are data collection techniques that are carried out through face-to-face and direct question and answer between researchers and informants, namely employees at PT. TOR GANDA b. Questionnaire Questionnaires or questionnaires are data collection using a list of questions/questions that have been prepared in advance and given to PT employee respondents. TOR GANDA c. Documentation Study.
Collect and study data or information needed through documents at the company.

RESULTS AND DISCUSSION Validity and Reliability Test a. Validity test
The validity test was carried out to ensure that the measuring instruments used in the study could measure the variables studied accurately and appropriately. In this study, the validity test was carried out at PT. TOR GANDA involving 30 respondents. Validity means that the data obtained through a questionnaire can answer the research objectives. To assess the validity, the number of n (samples/respondents) is used in the degrees of freedom r-table (df = nk), where k is the number of items in the instrument. Referring to the validity output and it is known that the number of samples or n is 30, with a significance of 5% (0.005) in the distribution of r product moment tables, an r table of 0.361 is found which can be seen in the table image of the r product moment values. In the context of person validity testing, the reference used reads r tables, the value of n does not need to be reduced by 2. In the distribution of product moment r tables, it is found that r tables of 0.361 validity are carried out on the Transformational Leadership variable (X1) which can be seen in the correlation column, it is known that all r count values are greater than r tables and the significance level is less than 0.005 (5% ) which means that all indicators in the questionnaire are declared valid. For more details, it can be seen in table 5 as follows: In the distribution of product moment r tables, it is found that r tables of 0.361 validity are carried out on the Compensation variable (X2) which can be seen in the correlation column, it is known that all r count values are greater than r tables and the significance level is less than 0.005 (5%) which means that all indicators on the questionnaire are declared valid. For more details, it can be seen in table 6 as follows: In the distribution of product moment r tables, it is found that r tables of 0.361 validity are carried out on the Job Satisfaction variable (Z) which can be seen in the correlation column, it is known that all r count values are greater than r tables and the significance level is less than 0.005 (5% ) which means that all indicators in the questionnaire are declared valid. For more details, it can be seen in table 7 as follows: In the distribution of product moment r tables, it is found that r tables of 0.361 validity are carried out on the Employee Performance variable (Y) which can be seen in the correlation column, it is known that all r count values are greater than r tables and the significance level is less than 0.005 (5% ) which means that all indicators in the questionnaire are declared valid. For more details, it can be seen in table 8 as follows:

b. Reliability Test
According to Ghozali (2017), a variable is said to be reliable or reliable if its Cronbach Alpha is greater than 0.60. A higher Cronbach Alpha value indicates a higher level of reliability, which means that the questions in the instrument have good consistency and can be relied upon. Reliability test data on the instruments tested can be seen in table 9 below: In the table, the Cronbach Alpha value for Transformational Leadership (X1) is 0.699, Compensation (X2) is 0.632, Job Satisfaction is 0.656, and Employee Performance is 0.719. Cronbach's Alpha value greater than 0.6 means the data is reliable. It can be concluded that all statement instruments from variables are reliable so that the data can be used in further analysis.

Structural Equation Modeling (SEM) Analysis
Data The path diagram is an illustration of the research conceptual framework. The path diagram shows the constructs of exogenous or endogenous variables with the manifest variables of each of these latent variables. Before the data will be analyzed, the variables in the study will be described by a structural model. Structural models are models regarding the structure of relationships that form or explain causality between factors by looking at their probability values and their tstatistics.

a. Measurement Model (Outer Model)
The outer model measurement model is implemented in testing the construct validity and instrument reliability. Validity test to determine the ability of research instruments to measure what should be measured. The reliability test is used to measure the consistency of a measuring instrument in measuring a concept or it can also be used to measure the consistency of respondents in answering question items in a questionnaire or research instrument.

b. Structural Model (Inner Model)
The inner model analysis aims to assess the significance of the path coefficients or t-values for each path between variables. The inner model is a structural model that is used to predict causal relationships (causation relationships) between latent variables or variables that cannot be measured directly. The value of the path coefficient or inner model indicates the magnitude of the influence of the exogenous latent variable on the endogenous latent variable. The value of the standardized path coefficient is shown by the latent variable relationship arrow. If the value of the standardized path coefficient ≥ 0.1, the influence of exogenous variables on endogenous variables is significant (the greater the value, the greater the effect). The path coefficient score is indicated by the T-statistic value, where the value must be above 1.96 for the two-tailed hypothesis and above 1.64 for the one-tailed hypothesis for hypothesis testing at α = 5% in the study This.

Hypothesis test
Hypothesis testing is based on the path coefficient and the total effect of the research variables. Testing the significance of these effects is done by the bootstrapping method. If this value is statistically significant, then the research hypothesis is accepted: a. Testing the main effect (independent effect on the dependent) >> must be significant. b. Testing the effect of the independent variable on the mediating variable >> must be significant. c. Calculate the Variance Accounted For (VAF) value. The VAF value is calculated by the equation: Testing the mediation effect is carried out using a bootstrap approach with the following stages: a. If the VAF value is > 80%, the mediating variable is full mediation. b. If 20% ≤ VAF ≤ 80% then the mediating variable is partial mediation. c. If VAF <20%, the mediating variable is not a mediator.

Coefficient of Determination (R2)
The coefficient of determination analyzes the ability of the independent variable model to explain the variance of the dependent variable data. The criteria used in assessing the coefficient of determination are 0.75; 0.5; 0.25, each of which represents a model assessment that can explain