The Effect of Work Stress, Work Conflict and Work Ability and Their Impact on Employee Performance Through Motivation as an Intervening Variable

The purpose of this research is to be able to analyze the results of the effect of stress at work, work conflict and an increase in work performance through work motivation. The population used in this study were all employees of PT. Bank Rakyat Indonesia Tbk, Satellite Services Operation Department used as a sample. The analytical method will use an analytical tool in the form of a Structural Equation Model (SEM) program through AMOS software. The results of the research work stress, conflict at work and the ability to work will have a significant influence on work motivation. However, work stress, work conflict and ability to work will affect work performance, while motivation will not have a significant effect on performance. Motivation is useful for mediating work stress and work conflict on performance and motivation is not able to mediate work ability on


INTRODUCTION
In the business world, high-quality and competent human resources in a field have an important role and are needed to achieve the goals of every company. Human resources who have superior performance can be the main asset of an organization that will play an important role as a driver of productivity, innovation and creativity to meet organizational goals. The existence of dynamic changes in business will also have an impact on an organization, so this will require the organization to follow the existing changes. This change will then encourage human resources within the organization to be able to work better, so that the workload on these human resources will also increase (Wijaya, 2006).
Through various studies that have been carried out previously, changes in organizational environmental conditions, both from within and from outside, either directly or indirectly will affect employee stress levels and work conflicts and will be able to reduce employee performance levels, which in turn can lead to a decrease in employee performance. have an impact on reduced work motivation (Kelvin, 2021). The goals of an organization will be successfully achieved if there is a positive performance from its employees. Meanwhile, employees who do not have effective performance will provide obstacles to the organization in achieving its goals. One of the factors that affect the performance of employees is stress at work (Khairunnisa, 2021).
Job stress is part of a dynamic condition in which individuals face opportunities, constraints or demands regarding something related to what the employee wants, where the results are perceived as uncertain but important. Examples of job stress are pressure from superiors to achieve organizational goals, salaries given unfairly, and fear of accidents in the workplace that can cause injury and death. Apart from being caused by organizational factors, work stress can also be seen from external and individual factors of the employee in carrying out his work (Mukhtar, 2021).
In addition to work stress, another issue that is still related to the performance of employees is work conflict. Work conflict in an organization can make employees unhappy and make them not have awareness as part of the organization. The existence of this work conflict will reduce the performance of its employees. An example of work conflict is giving too much workload (Sikettang, 2021).
Another issue related to performance is the ability to work. The ability to work is related to knowledge, talents, interests and experience in completing tasks that are in accordance with the work they occupy. Employee work ability is also a skill possessed by employees in completing their work. If employees have the opportunity to develop their talents, and can use them appropriately, then this will affect the development of the company. Adequate employee capabilities will assist employees in carrying out their work, both now and in the future (Simanjorang, 2018).
To achieve its goals, performance has an important role for the organization. Performance is a result of work (output) or work performance, both in terms of quality or quantity that has been achieved by human resources in carrying out their duties in accordance with the responsibilities that have been given. Performance in accordance with work standards that have been set by the company, both in quality and quantity is an example of a form of good performance. Through this good employee performance, the productivity of a company will also increase rapidly. This increase in productivity is then expected by all organizations to achieve their goals (Jufrizen, 2018).
PT. Bank Rakyat Indonesia (Persero) Tbk, which is abbreviated as BRI, is one of the large banks that are in great demand by the Indonesian people and is owned by the government. BRI focuses on the micro, small and medium enterprise (MSME) segment and is a pioneer of microfinance in Indonesia. In 2016, BRIsat was launched by BRI, which made BRI the first bank to own and operate its own satellite in the world. The launch of this satellite is part of BRI's plan to strengthen the infrastructure supporting future digital services to be able to bring quality banking technology from the city center to remote rural areas.
The increase in the quality of BRI's management cannot be separated from the quality of BRI's human resources. Accuracy, accuracy and ability of human resources on employee performance are needed to be able to carry out all operational activities. Steps that can be taken to optimize employee performance are by dealing with work stress, work conflicts and employee motivation. Work stress and work conflict often become something that is vulnerable based on observations as a result of the many work demands and overtime hours that must be done. There needs to be an evaluation from the organization that is sustainable to maintain or even improve employee performance (Muzayyana, 2019).
For the past 4 years, BRI has evaluated the performance of its employees. Based on previous research, there are performance targets that are always different every year. Previous research has analyzed what factors affect performance, some of which are diverse and some are contradictory where there is a positive influence and a negative influence which then has a significant effect on performance variables (Amirudin, 2021).
The author wants to close the gap from the results of previous studies through the motivation variable as an intervening variable, so that this will result in new research and different from previous studies. This is because the author considers that motivation is important for an employee to fight and survive in an organization. Organizations will find it difficult to achieve their goals if employees are not enthusiastic in carrying out their work (Winarni, 2021).
Motivation is a skill in directing employees to work for the organization in order to produce the expected performance, so that this will lead to the achievement of the desires of employees and the organization simultaneously. The amount of performance produced by employees will be directly proportional to the motivation and encouragement of the leadership (Burhanuddin, 2021).

B. METHOD
This study will use work stress, work conflict and work ability as independent variables (exogenous) which will then affect motivation as a form of intervening variable and employee performance as a form of endogenous dependent variable. The addition of the work ability variable that acts as an independent variable (exogenous) in this study is expected to strengthen the influence that comes from work stress, work conflict and motivation on employee performance. This addition was made because of the infrequent use of this variable in previous studies. The use of work ability as an independent variable in this study is because human resource planning is included in development planning in an organization. In order to facilitate the activities of an organization, it will also require good performance.

C.
RESULT AND DISCUSSION

Descriptive Analysis
Based on the questionnaire that has been given to the respondents, 39% of respondents answered agree about the work stress variable. The results of the distribution of answers regarding work stress also have the most agreeable answers at 39%, while the lowest results are disagreed with a value of 2.7%. As for the work conflict variable, there are 31.8% of respondents who answered agree. Meanwhile, the distribution of work conflict answers with the highest number of answers disagreed with 35.2% and the lowest answers strongly disagreed with 7.2%.
Analysis of the work ability variable is 38.3% of respondents who answered agree. The distribution of answers in work ability is 38.3% with the most agreeable answers, while the least agreeable answers are 3.1%. In the motivation variable, there are 36.9% of respondents who answered agree. The distribution of motivational answers is agreeing as the highest answer of 36.9% and strongly disagreeing with the lowest answer of 4.8%.
The last variable is employee performance, where almost half of the respondents answered agree, to be exact, 48% of respondents chose to agree. The distribution of answers regarding employee performance is agree with the highest result of 48%, while the lowest result is strongly disagree, where there are no respondents who answer strongly disagree or with a result of 0%.

Structural Equation Model (SEM) Test Analysis
SEM or Structural Equation Modeling is a combination of factor analysis and reflection analysis which aims to examine the relationship between variables in a research model. There are 2 main tests of the SEM model, namely the validity of the measurement model and the validity of the structural model. AMOS 22 software is used in this study to perform SEM testing (Sholihin & Ratmono, 2021).

Data Normality Test
The normality test has a purpose to test whether in a regression model there is a normal distribution or not in the independent variable and the dependent variable. When the data distribution is normal or close to normal, this indicates that the regression model is good. In the results of data processing from this study, there were no studies that exceeded the specified range, so the value of this study was considered to be normally distributed in a univariate manner (Ningsih & Dukalang, 2019).
Meanwhile, the normality test based on the AMOS output data showed that extreme values were not found, therefore, the data in this study can be stated to be normally distributed. Because there is no guarantee that the data to be analyzed shows no outliers at the univariate level, it is necessary to evaluate multivariate outliers. The mahalanobis distance was used to detect the presence or absence of outliers in multivariate behavior. The mahalanobis distance can be calculated and will show the distance of an observation from the average of all variables in a dimensional space for each observation (Triwidatin & Anwar, 2021).

Measurement Test
Another part of SEM modeling is the measurement model, in which this model consists of latent variables (constructs) and several manifest variables (indicators) to support the strengthening of these latent variables. This measurement model aims to determine the accuracy of these indicators to describe the existing latent variables. In general, Confirmatory Factor Analysis is used as a test of the validity of the measurement in general.

Model Alignment Test (Goodness of Fit Test)
Compared to an exploratory procedure, the SEM analysis procedure is confirmatory. This is because to determine if the variance and covariance plots in a data are consistent with the structural path, the SEM model alignment test will be used. However, the accepted model is only an affirmation model if other unobserved models can match the data. According to Ghozali & Fuad (2008), the following are some indices that are generally used as a reference for the model alignment test, namely: a. Chi square, where this analysis aims to test and develop whether a model is in accordance with the data. Chi square is quite sensitive by data samples that are too small or data that is too large. b. GFI (goodness fit index) is often used as a reference for assessing model fit as a model fit index. GFI is an index of model accuracy to explain the developed model. c. RMSEA (root mean square error of approximation), also called root mean square error. If the RMSEA value is less than or equal to 0.05, then it indicates a good model fit. d. CFI (Comparative Fit Index) is the comparison value of the model based on the ideal model. The requirement for CFI is to have a value above or close to 0.90. e. AGFI (Adjusted Goodness of Fit) is a fit index criterion, which is a development of CFI. f. NFI (Normal Fit Index) is a comparison of the proposed model with the null model. g. TLI (Tucker Lewis Index) is an incremental suitability index that compares the tested model using the baseline model.

Analysis of the Relationship of Each Indicator with Variables
Based on the results of the analysis that has been done, it can be concluded that the hypothesis about work stress has a positive and significant influence on work motivation at PT. BRI is acceptable. This means that if there is work stress among BRI employees, work motivation will also increase. Then the hypothesis of work conflict has a positive and significant effect on motivation among BRI employees can be accepted. This shows that an increase in work conflict at BRI will increase employee motivation as well. Furthermore, the hypothesis that work ability has a positive and significant effect on motivation among BRI employees can be accepted. This then shows that the work motivation of BRI employees will also increase along with the increase in work ability.
Furthermore, the hypothesis that work stress affects employee performance at BRI is acceptable. This means that increasing work stress at BRI will improve performance among BRI employees. Then the hypothesis that work conflict has a positive effect on the performance of BRI employees can be accepted. This shows that an increase in work conflict has an effect on an increase in performance among BRI employees. Then the hypothesis about work ability has a positive and significant effect on the performance of BRI employees can be accepted. This would indicate that performance would improve if there was an increase in employability among BRI employees.
However, the hypothesis that motivation has neither a positive nor significant effect on the performance of BRI employees is rejected. The rejection of this hypothesis indicates that motivation has a positive effect on performance, but it is not significant. Hypothesis testing 8,9 and 10 uses a mediation test (Sobel test), where this serves to determine whether there is an indirect effect of the independent variable on the dependent variable through the intervening variable.
The hypothesis that work stress through motivation has a positive and significant effect on the performance of BRI employees is acceptable. This means that if there is an increase in work stress of BRI employees, then there is an increase in performance through work motivation. Then the hypothesis that work conflict through motivation has a positive and significant effect on the performance of BRI employees can be accepted. This indicates that increasing work conflict can improve performance through motivation among BRI employees. Finally, the hypothesis that work ability has a positive and insignificant effect on performance through motivation is rejected. It can then be concluded that the ability to work through motivation has a positive and significant effect on BRI employees.

D. CONCLUSION
After testing the alignment of the model, it will be analyzed whether the indicator is an element that can explain each variable. Based on the results of calculations and data processing, it was found that there was a unidirectional relationship between work stress and motivation. Then it was also found that there was a direct relationship between work conflict and satisfaction. Then there is a direct relationship between work ability and motivation.
Meanwhile, with the employee performance equation, it was found that there was a unidirectional relationship between job stress and employee performance. Then it was also found that there was a unidirectional relationship between work conflict and employee performance. Then there is a unidirectional relationship between work ability and employee performance. Finally, there is a unidirectional relationship between motivation and employee performance, which means that motivation will increase, so employee performance will also increase.