EFFECT OF COMPLEXITY OF WORK, CAREER DEVELOPMENT ORGANIZATION AND COMMUNICATION ON THE PERFORMANCE OF EMPLOYEES

This study aims to examine several factors that can influence employee performance including work complexity, career development, and organizational communication The sample used was 76 Bank “X” employees in Surakarta City. The sampling method uses the survey sampling method. The type of data used is primary data. The data analysis method uses multiple regression with the help of SPSS 23 The coefficient of determination test results shows that the complexity of work, career development and organizational communication can explain employee performance by 50.3% while the remaining 49.7% employee performance will be explained by other factors outside the three research variables. The results of the hypothesis test analysis show the conclusion that work complexity, career development, and organizational communication affect employee performance.


Introduction
Competition in the business world incuding the banking sector is experiencing very tight competition. The banking industry was faced with the era of the industrial revolution 4.0 which required banks to adjust their business to these conditions in order to remain able to complete with other banking companies. The emergence of increasingly advanced technologies and increasingly critical patterns of thinking of the public have an impact on the magnitude of the demand for banking companies to be able to bring up new innovations in order to attract customers and be able to show good performance in order to be able to maintain their customers. Innovation will not run without good performance and good performance cannot be achieved without the support of reliable human resource.
Banking companies as service companieswhere almost all of their business activities are managed and run by employees as their human resources. Therefore employee performance will provide a very significant influence on company performance. Good employee performance will have a positive impact on company performance (Lu et al., 2015;Victoria, 2019) ).
However, to improve the performance of good employees is not easy. The company needed an effort to improve the performance of employees. Several studies have shown several factors that affect employee performance including work complexity (Benish & Raza, 2017;Gomaa, 2005), career development (Kurniawan. et al., 2018;Nasution et al., 2018) and organizational communication (Rukmana et al., 2018;Titang, 2016).
The complexity of the work is a job that is not structured, confusing, difficult and consists

Literature Review 2.1.Work Complexity
The complexity of the work occurs because in carrying out one's work faced with high ambiguity and weak work structures (Restuningdiah & Indriantoro, 2000). As for Jamilah et al. (2007) views work complexity as unstructured, difficult and confusing work. Chung dan Monroe (2001) state that work complexity is influenced by the presence of inaccurate, relevant or incomplete information and the high ambiguity of expected results.

2.2.Career Development
Every employee who works in an institution or company would want their career to be able to develop better. Employee career development will indirectly increase the work motivation of these employees and increase their salaries and benefits and facilities. Therefore employee career development is what every employee wants. Marwansyah (2010) states that career development is an activity carried out by an individual to develop himself in order to plan his personal career in the future. Hadi (2013) states that career development is a place to improve status in his career in an organization. Bahri (2016) argues that career development is a joint effort between employees and HR managers to increase their existing potential so that they can be in a higher

Research Methods
This research uses descriptive research method that is to find out the value of the independent variable, either one variable or more (independent) without making comparisons, or connecting one variable to another variable (Sugiyono, 2016). The population in this study were 80 employees of the "X" Private Bank in the city of Surakarta. The sampling technique used is survey sampling. This technique was chosen because the total population is less than 100. According to Arikunto (2013) if the total population is less than 100 then the sample is taken from the entire population. Therefore the study sample was 80 "X" Private Bank employees in Surakarta City. The source of data in this study is primary data with data collection techniques using a questionnaire where the research instrument will be measured using 5 likert scale. Data analysis techniques were performed using multiple regression. Data analysis assistance is SPSS 23. Data analysis begins by testing data quality, namely testing the validity and reliability. Testing classic assumptions which include multicollinearity, autocorrelation, heterokedastisitas and normality is done as a condition of regression testing. Hypothesis testing will later be done by t test.

Results and Discussion
The questionnaire was distributed to all employees of the "X" Private Bank in the city of Surakarta. However, from the 80 questionnaires collected there were 4 questionnaires which were incomplete so that in the end only 76 questionnaires could be processed in the data analysis. Characteristics of research respondents including gender, respondent age, education and years of service will be presented in Based on table 1 regarding the characteristics of respondents it can be illustrated that the majority of employees involved in the study were women by 50 people (66%). Most respondents aged 26-35 years were 41 people (53.9%). Undergraduate education (S1) dominates respondents by 51 people (67%). The respondent's tenure of 1-10 years is 54 people (71%).

Data Quality Testing
Data quality testing will later use validity testing and reliability testing. The reliability test was conducted to find out that the questionnaire was an indicator of the research variables. In the results of the reliability analysis, the Cronbach's alpha value for the variable work complexity, career development, organizational communication, and employee performance is greater than 0.6 (table 2). The results of the analysis mean that the four variables are reliable for use in research. It also contains meaning in answering the questionnaire the respondents answered consistently from time to time. The results of testing the validity have been presented in table 3 which shows that the four research variables consisting of work complexity, career development, organizational communication, and employee performance provide the calculated r value for each item greater than 0.223 (r table). This shows that the instrument used to measure the four variables is valid or appropriate to measure the research variables.

Classical Assumption Testing
In this study, 4 types of testing will be carried out, which is testing classic assumptions. The four tests are heterokedasticity, multicollinearity, autocorrelation and normality. The results of heterokedasticity testing are generally shown in scatterplot images. The picture shows the data points spread up and down or around point 0 or it can be said that the points do not collect only on one side, namely the top or bottom. This shows that there is no heterocedasticity problem so that the regression model is good and can be used for research. The autocorrelation test results are indicated by the value of Durbin Watson (DW). Based on the analysis, the DW value is 1,950. This value is between the value of 1.7104 (du) and 2.22896 (4-du) which means that there is no autocorrelation found in the regression model.
In multicollinearity testing, the value of VIF and tolerance will be used to test whether there is a correlation between the independent variables in the research model. The results of the analysis clearly show that the tolerance value is greater than 0.10 and the VIF value is smaller than 10. This means that there is no multicollinearity in the regression model. Table 4 shows the results of multicollinearity testing as follows: Whereas in the Kolmogorov Smirnov normality test the significance value will be seen. The analysis showed that the significance level was 0.200, meaning that the data were normally distributed.

Hypothesis Testing
The results of the analysis with multiple regression give the adjusted R 2 value of 0.503. This implies that 50.3% of employee performance can be explained by the complexity of work, career development, organizational communication. But there are still 49.3% of the variables besides the complexity of work, career development, organizational communication which would affect employee performance. The ANOVA test showed an F value of 26.293 (0.000), meaning that this research model was good enough to be used to answer the proposed research problem.
The hypothesis in this study will be decided based on the results of the t test in the regression analysis. Table 5 will present the results of the multiple regression analysis to answer the proposed hypothesis.

Tabel 5. Regression Analysis Result
Dependent Variable: Employee Performance Source : Data Processed, 2019 The results of the above regression analysis can be stated in the following regression equation: Employee Performance = 0.330 Work Complexity + 0.363 Career Development + 0.297 Organizational Communication + e T test is done to make a decision on the hypothesis proposed. Table 5 illustrates the results of the t-test analysis which shows the positive influence of work complexity, career development and organizational communication on employee performance Each independent variable with a significance level of 5% has a greater t value of 1.96 (t table). This means that employee performance will be achieved if the company pays attention to the complexity of the work of its