570 THE INFLUENCE OF WORK BEHAVIOR AND COMPETENCE ON PERFORMANCE THROUGH JOB SATISFACTION IN PERMANENT EDUCATION WORKERS NON PNS FACULTY OF ENGINEERING BRAWIJAYA UNIVERSITY MALANG Volume: 4 Number: 2 Page: 570 - 584 Ratna Sri HARTATIK1, Ernani HADIYATI2, Gunadi GUNADI3 1Magister Program, Gajayana University, Malang, Indonesia 2,3Lecturer at Gajayana University, Malang, Indonesia Corresponding author: Ratna Sri Hartatik E-mail: nana_psik12@ub.ac.id Article History: Received: 2022-01-18 Revised: 2023-02-24 Accepted: 2023-03-17 Abstract: Implementing higher education for tertiary institutions involves all existing elements to achieve the quality of higher education internal services. One of the essential components in the management of the tertiary education system is the academic community, which has a scientific tradition and academic culture. The academic community in a tertiary institution consists of teaching staff (lecturers), students, education staff and stakeholders, where each has different main tasks and functions to improve the quality of tertiary services. This study aims to examine and analyze the influence of work behavior, competence and job satisfaction on the performance of non-civil servant (non-PNS) permanent education staff at the Faculty of Engineering, University of Brawijaya Malang. This study uses qualitative data types from primary data, namely, 70 employees as respondents. The data analysis method used in this study is Partial Least Squares (PLS) with a new Structural Equation Modeling (SEM) approach, which can predict the relationship between the dependent and independent variables. The results of this study provide information that there is a positive and significant influence between behavior and competence on performance. This study also provides information that there is a positive and significant influence between behavior and competence on job satisfaction. In addition, the job satisfaction of non-PNS permanent staff of FT UB Malang has a positive and significant effect on performance as well as a positive and significant effect on work behavior and competence of non-PNS permanent education staff on performance through job satisfaction. Keywords: Work Behavior, Competence, Job Satisfaction and Performance. Cite this as: HARTATIK, R., HADIYATI, E., GUNADI. (2023). “The Influence of Work Behavior and Competence on Performance Through Job Satisfaction in Permanent Education Workers Non PNS Faculty Of Engineering Brawijaya University Malang” International Journal of Environmental, Sustainability, and Social Science, 4 (2), 570 - 584. INTRODUCTION Academic staff is an element of planners, thinkers and movers in a tertiary institution who work together to achieve tertiary goals. According to Law no. 20 of 2003, in implementing the Tri Dharma of Higher Education, educational staff have an important role, namely the management of learning administration for students and teaching staff in a tertiary institution. The role of academic staff in administrative management at tertiary institutions must be distinct from the current development of information technology. These two things go hand in hand with educational administration, research and community service. The performance of educational staff at a tertiary institution is a real behavior displayed by each academic staff as work performance in accordance with their role. Performance is a quality component of human resources that must be continually improved since good knowledge, skills, and attitudes will foster a positive educational environment and high service standards. mailto:nana_psik12@ub.ac.id 571 Therefore, it is necessary to strengthen the quality and quantity of existing education staff to create a good performance that can compete in the world of work. Based on the Regulation of the Chancellor of Brawijaya University Number 25 of 2020 concerning Organizational Structure and Work Procedures, the Faculty of Engineering, Universitas Brawijaya Malang, hereinafter abbreviated as FT UB Malang, is a part of the higher education institution under Universitas Brawijaya which has an organizational structure with support from human resources. Academic staff at the Faculty of Engineering Universitas Brawijaya are divided into 9 sections, namely the Dean, Department of Civil Engineering, Mechanical Engineering, Water Engineering, Electrical Engineering, Architecture, Regional & City Planning, Industrial Engineering, and Chemical Engineering, which play an essential role in each section. Especially in each department, where each department consists of many students and lecturers, it is necessary to strengthen the quality and quantity of existing education staff to create a good performance. PNS, non-PNS academic staff and employees appointed by the chancellor who has administrative functions are in charge of managing work related to administrative and operational services in education and offices. In contrast, those who have functioned as functional staff, namely educational laboratory institutions, are tasked with managing and providing educational services in educational laboratories. Based on Rector Regulation No. 31 of 2021, Article 63, performance appraisal for UB's permanent staff, both civil servants and non-PNS, is regulated in accordance with statutory regulations. In this case, performance evaluation refers to Government Regulation Number 30 of 2019. The existence of non-PNS academic staff at the Faculty of Engineering, Universitas Brawijaya, greatly influences the smooth running of the existing information system because the tasks carried out are more than just supporting. However, the main tasks and functions have a value of responsibility. Therefore, competence and expertise in order to improve services in tertiary institutions are very important. In improving services to lecturers, academic staff, students and stakeholders, Universitas Brawijaya has many information systems collected in one gate. In this case, the competence of education staff must master information technology supported by skills and knowledge of technology. Judging from the many existing information systems in UB, it is expected that educational staff will be able to carry out their duties as academic staff based on existing competencies. Therefore, the competence of each educational staff is essential in carrying out their duties. In addition to work behavior and competence of academic staff, job satisfaction is essential to study to improve performance. In FT UB, the average education staff is young. There is a tendency that young people, namely the age range of 31 to 40 years, to be more satisfied at work due to a high level of productivity so that they can complete work targets quickly. Widakdo (2021) in his research stated that age and education level had a positive and significant effect on performance. Apart from age, in FT UB, there are differences in compensation policies between PNS employees and non-PNS permanent employees. In addition to competence, work behavior and job satisfaction, performance is also to achieve organizational goals. The importance of Human Resources in companies or institutions is a fact that cannot be denied. Any greatness owned by the agency will only be meaningful with human resources who can manage it properly. Professional and qualified human resources are the primary key to achieving the progress and prosperity of a nation. HR is one of the elements responsible for government organizations, be it planners, implementers, movers, or supervisors. Therefore, in carrying out their main tasks and functions, employees are required to improve their services in quality and quantity. Professional and quality. A government organization must always try to improve employee performance to produce professional and quality human resources. It is known that there are still many e-complaints that are entered on the website of the Faculty of Engineering, Universitas Brawijaya. It indicates the dissatisfaction of external parties with the 572 performance of the services provided. The many e-complaints that exist illustrate the decline in academic staff performance. This decline in performance is possible due to a mismatch between the level of ability possessed by educational staff, and the development of needs and the dynamics of the problems faced are increasingly complex and competitive. All complaints in every scope of existing complaints are a form of dissatisfaction with the services provided. Seeing the percentage of complaints on e-complaints shows that the service is still not going well, so it needs to be repaired, one of which is related to human resource management, namely non-PNS education staff as one of the drivers of service administration in the education sector. From the phenomena described above, the researchers aim to investigate further the variables of work behavior, competency, job satisfaction and performance. METHODS This research was carried out from June to July 2022 at FT UB Malang. The number of samples used was 70 non-PNS academic staff. This type of research is causality research, meaning research that aims to know the effect on variables. This research method uses a survey, in which questionnaires are distributed to obtain answers to research questions online. The data obtained were processed with tools using Microsoft Excel computer software and Structured Equation Model (SEM) analysis using the Partial Least Square (PLS) approach. SEM-PLS is an analytical method that is not based on many assumptions. The data used in SEM-PLS does not have to be normally distributed. Multivariate/indicators (scale, ordinal, interval, ratio) can be used for samples using the SEM-PLS approach. Besides that, SEM-PLS does not require large samples (Ghazali, 2015). The components in SEM-PLS consist of a Structural Model (inner model) and a measurement model (outer model). Research Conceptual Framework. The performance of educational staff at a tertiary institution is a real behavior displayed by each academic staff as work performance in accordance with their role. Performance is a quality attribute of human resources that must be consistently improved since having good knowledge, skills, and attitudes will foster a positive and healthy learning environment and have high service standards. Mafra's research, Nisa' Ulul (2016), proves that work behavior affects performance. Higher education currently relies heavily on Information Technology to support the educational process. This information system can support human resource management and work activities to be integrated to achieve organizational goals. Information technology functions to support the performance of education staff, which includes: information function, communication function, data storage function, and learning function. His research by Amar, H. & Fikri, Z. (2020) proves that competency affects performance. From the formation of work behavior and competence in each employee, a feeling or sense of work will be formed, resulting in a feeling of satisfaction or dissatisfaction with employees at work. Employee satisfaction at work is essential in carrying out work because job satisfaction is a feeling of pleasure in carrying out work. After all, what is expected is in accordance with reality. Job satisfaction is a positive attitude shown by employees towards their work so that employees can work happily without feeling burdened with the work and provide optimal results for the company. Design. Based on the explanation above, the conceptual framework of the research can be described as follows. 573 Source: Author Figure 1. Conceptual Framework RESULT AND DISCUSSION Descriptive Data. Research conducted on non-PNS permanent education staff at FT UB Malang involved 70 respondents. The following table 2 relates to the demographics of the respondents dominated by: Table 1. The Highest Domination of Respondent Demographics No Category Criteria Percentage Total Respondent 1 Working Period 11 to 20 years 67% 47 2 Class III/b 56% 39 3 Age 31 - 40 years old 74% 52 4 Last Education level Bachelor’s Degree 54% 38 5 Sex Women 64% 45 Source: Data Processed 2022 Validity test. The validity test is a test to determine whether each statement item is valid or not used in the questionnaire. The item is declared valid if the r count item value exceeds the r table value. The test was carried out using the Pearson Correlation test with the following results after distributing a questionnaire to 30 respondents: Table 2. Validity Test Results Variable r count r table desc. Work Behavior (X1) Service Orientation X1.1 0.847 0.361 Valid X1.2 0.858 0.361 Valid Commitment X1.3 0.827 0.361 Valid X1.4 0.814 0.361 Valid X1.5 0.845 0.361 Valid X1.6 0.900 0.361 Valid Work Behavior (X1) Competence (X2) Job Satisfaction (Y1) Performance (Y2) H3 H 4 H1 H2 H5 H6 H7 574 X1.7 0.863 0.361 Valid Work Initiative X1.8 0.834 0.361 Valid X1.9 0.885 0.361 Valid X1.10 0.780 0.361 Valid Team Work X1.11 0.871 0.361 Valid X1.12 0.848 0.361 Valid X1.13 0.825 0.361 Valid X1.14 0.839 0.361 Valid X1.15 0.834 0.361 Valid Leadership X1.16 0.538 0.361 Valid X1.17 0.656 0.361 Valid X1.18 0.667 0.361 Valid X1.19 0.546 0.361 Valid Competency (X2) Knowledge X2.1 0.863 0.361 Valid X2.2 0.857 0.361 Valid Skill X2.3 0.889 0.361 Valid X2.4 0.839 0.361 Valid Attitude X2.5 0.889 0.361 Valid X2.6 0.824 0.361 Valid X2.7 0.870 0.361 Valid X2.8 0.783 0.361 Valid Work Satisfaction (Y1) Productivity Y1.1 0.845 0.361 Valid Y1.2 0.875 0.361 Valid Work Level Y1.3 0.851 0.361 Valid Y1.4 0.758 0.361 Valid Y1.5 0.853 0.361 Valid Corporation Organization Size Y1.6 0.882 0.361 Valid Y1.7 0.869 0.361 Valid Y1.8 0.889 0.361 Valid Performance (Y2) Quality Y2.1 0.912 0.361 Valid Y2.2 0.890 0.361 Valid Y2.3 0.891 0.361 Valid Quantity Y2.4 0.789 0.361 Valid Y2.5 0.722 0.361 Valid Y2.6 0.829 0.361 Valid On-Time Y2.7 0.863 0.361 Valid Y2.8 0.900 0.361 Valid Y2.9 0.878 0.361 Valid Y2.10 0.877 0.361 Valid Cost Y2.11 0.862 0.361 Valid Y2.12 0.890 0.361 Valid 575 Source: Data Processed 2022 The results of the validity test with the Pearson Correlation to the work behavior questionnaire (X1), competency (X2), job satisfaction (Y1), and performance (Y2) show that all of these items are valid, provided that the value of r count is more than r table (r hit > r table) so that all of these items can be used in research. Reliability Test. The reliability test is a test to determine whether all statement items used in the questionnaire are reliable. The item is declared valid if the r count item value exceeds the r table value. Tests were carried out using the Cronbach Alpha test with the following results. Table 3. Reliability Test Results No Variable Total Item Cronbach Alpha Value Limit desc. 1 Work Behavior (X1) 19 0.979 0.600 Reliable 2 Competency (X2) 8 0.946 0.600 Reliable 3 Job Satisfaction (Y1) 8 0.945 0.600 Reliable 4 Performance (Y2) 12 0.967 0.600 Reliable Source: Data Processed 2022 The results of the reliability test with Cronbach Alpha on the work behavior questionnaire (X1), competency (X2), job satisfaction (Y1), and performance (Y2) showed that all of these items were reliable, provided that the Cronbach Alpha value was more than 0.600 so that all of these items can be used in research. Description of Respondents' Answers to Research Variables. This research consists of four variables, namely work behavior (X1), competence (X2), job satisfaction (Y1), and performance (Y2). The answer choices consist of 5 choices with a value between 1 to 5, or the answer choices Strongly Disagree to the answer choices Strongly Agree. The description of each variable is explained with a frequency distribution table obtained from the score tabulation of the respondents' answers. The results of the descriptive analysis are the frequency and percentage of answers, as well as the average (mean) value of each item in the measurement statement for each variable. Table 4. Variable Description No Variable Mean SD Min Max Description 1 Work Behavior 3.354 0.962 1 5 Fair 2 Competency 3.327 0.905 1 5 Fair 3 Job Satisfaction 3.332 0.910 2 5 Fair 4 Performance 3.296 0.933 1 5 Fair Source: Data Processed 2022 The description of work behavior consisting of 5 indicators with a total of 19 statements obtained an average variable of 3.354 with a standard deviation of 0.962. The lowest answer score given by the respondent is a score of 1, and the highest answer score is a score of 5. These results indicate that the respondent's perception of work behavior is in the sufficient category. The average answer is from 5 indicators. The lowest is on the service orientation indicator with the spirit of helping others, both internal and external. The competency description consisting of 3 indicators with a total of 8 statements obtained an average variable of 3.327 with a standard deviation of 0.905. The lowest answer score given by the respondent is a score of 1, and the highest answer score is a score of 5. These results indicate that the respondent's perception of competence is in the sufficient category. The average answer is 576 from 3 indicators. The lowest is on the knowledge indicator with work items according to field competence. The description of job satisfaction consisting of 3 indicators with a total of 8 statements obtained an average variable of 3.332 with a standard deviation of 0.910. The lowest answer score given by the respondent is a score of 2, and the highest answer score is a score of 5. These results indicate that the respondent's perception of job satisfaction is sufficient. The average answer is from 3 indicators. The lowest is on the age indicator, with items giving good ideas at work. The performance description consisting of 4 indicators with a total of 12 statements obtained an average variable of 3.296 with a standard deviation of 0.933. The lowest answer score given by the respondent is a score of 1, and the highest answer score is a score of 5. These results indicate that the respondent's perception of performance is in the sufficient category. The average answer is from 4 indicators. The lowest is on the on-time indicator, with items carrying out work without delaying time. Analysis Results. Structural Equation Modeling (SEM) is a method used to cover the weaknesses in the regression method. Data management in this study will use SmartPLS 3.2.9 Software. Analysis of the results of the study using the Partial Least Square (PLS) test was carried out to examine the influence of work behavior (X1), competence (X2), job satisfaction (Y1), and performance (Y2) among the variables. Outer Model Test Results. The outer model analysis is carried out to ensure that the measurement used is feasible to be used as a measurement (valid and reliable). This model analysis specifies the relationship between latent variables and their indicators. The results of testing the outer model include convergent validity, discriminant validity, and construct reliability. Convergent Validity Test. Convergent Validity is an indicator assessed based on the correlation between the item score/component score and the construct score, which can be seen from the standardized loading factor, which describes the magnitude of the correlation between each measurement item (indicator) and the construct. An individual reflexive measure is considered high if it correlates > 0.7 with the construct you want to measure. According to Chin, quoted by Imam Ghozali, an outer loading value between 0.5 - 0.6 is considered sufficient. Table 5. Convergent Validity Test Results No Indicator Loading Factor T count P Desc. Work Behavior (X1) 1 Service Orientation 0.880 29.957 0.000 Valid 2 Commitment 0.905 44.923 0.000 Valid 3 Work Initiative 0.898 38.005 0.000 Valid 4 Team Work 0.915 44.816 0.000 Valid 5 Leadership 0.886 35.814 0.000 Valid Competency (X2) 1 Knowledge 0.929 51.299 0.000 Valid 2 Skill 0.902 42.030 0.000 Valid 3 Attitude 0.896 29.591 0.000 Valid Job Satisfaction (Y1) 1 Productivity 0.913 48.162 0.000 Valid 2 Working Level 0.888 31.502 0.000 Valid 3 Corporation Organization Size 0.937 72.377 0.000 Valid Performance (Y2) 1 Quality 0.905 40.895 0.000 Valid 2 Quantity 0.889 30.276 0.000 Valid 3 On-Time 0.911 43.731 0.000 Valid 4 Cost 0.879 28.958 0.000 Valid Source: Processed Research Data (2022) 577 The results of the convergent validity test on work behavior (X1), competence (X2), job satisfaction (Y1), and performance (Y2) with the reflective measurement model obtained a factor loading value of an indicator of more than 0.700 so that the indicator of the variable has met convergent validity. Table 6. AVE Test Results No Variable AVE Value Limit Desc. 1 Work Behavior (X1) 0.804 0.500 Fulfilled 2 Competency (X2) 0.827 0.500 Fulfilled 3 Job Satisfaction (Y1) 0.833 0.500 Fulfilled 4 Performance (Y2) 0.803 0.500 Fulfilled Source: Processed Research Data (2022) The results of the construct validity test using the Average Variance Extracted value obtained that the test results met the test criteria of more than 0.50 so that each variable of work behavior (X1), competency (X2), job satisfaction (Y1), and performance (Y2) has fulfilled the validity construct. Discriminant Validity Test. Discriminant Validity is a measurement model with reflexive indicators assessed based on cross-loading measurements with constructs. If the construct's correlation with the measurement items is more significant than the size of the other constructs, then it shows that their block size is better than the other blocks. Meanwhile, another method to assess discriminant validity is by comparing the square root value of the average variance extracted (AVE). Table 7. Discriminant Validity Test Results Using Cross Loading No Indicator X1 X2 Y1 Y2 Desc. Work Behavior (X1) 1 Service Orientation 0.880 0.750 0.734 0.737 Fulfilled 2 Commitment 0.905 0.801 0.759 0.835 Fulfilled 3 Work Initiative 0.898 0.810 0.792 0.764 Fulfilled 4 Teamwork 0.915 0.836 0.775 0.808 Fulfilled 5 Leadership 0.886 0.833 0.810 0.814 Fulfilled Competency (X2) 1 Knowledge 0.815 0.929 0.759 0.794 Fulfilled 2 Skill 0.798 0.902 0.799 0.782 Fulfilled 3 Attitude 0.841 0.896 0.764 0.814 Fulfilled Job Satisfaction (Y1) 1 Productivity 0.798 0.806 0.913 0.822 Fulfilled 2 Work Level 0.750 0.713 0.888 0.741 Fulfilled 3 Corporation Organization Size 0.815 0.808 0.937 0.834 Fulfilled Performance (Y2) 1 Quality 0.752 0.787 0.805 0.905 Fulfilled 2 Quantity 0.807 0.825 0.789 0.889 Fulfilled 3 On-Time 0.819 0.812 0.808 0.911 Fulfilled 4 Cost 0.787 0.713 0.739 0.879 Fulfilled Source: Processed Research Data (2022) The results of the discriminant validity test using the cross-loading value obtained the loading factor in the column of each variable (bold) is the highest value compared to the cross-loading value of other variables so that the indicators of each variable work behavior (X1), competence (X2), job satisfaction (Y1), and performance (Y2) have met discriminant validity. Table 8. Discriminant Validity Test Results Using the Fornell-Larcker Criterion No Variable X1 X2 Y1 Y2 Desc. 578 1 Work Behavior (X1) 0.897 0.800 0.864 0.884 Fulfilled 2 Competency (X2) 0.800 0.909 0.852 0.877 Fulfilled 3 Job Satisfaction (Y1) 0.864 0.852 0.913 0.877 Fulfilled 4 Performance (Y2) 0.884 0.877 0.877 0.896 Fulfilled Source: Processed Research Data (2022) The results of the discriminant validity test using the AVE root value obtained the AVE root value (bold) more significant than the correlation value between latent variables so that work behavior (X1), competency (X2), job satisfaction (Y1), and performance (Y2) has met discriminant validity. Construct Reliability Test. Composite reliability is an indicator to measure a construct that can be seen given the latent variable coefficients. Internal consistency and Cronbach's alpha are measurement techniques used to assess composite dependability. In these measurements, if the value achieved is > 0.70, the construct has high reliability. Cronbach's Alpha is a reliability test that is carried out to strengthen the results of composite reliability. A variable can be declared reliable if it has a Cronbach's Alpha value > 0.70. Table 9. Construct Reliability Test Results No Variable Cronbach's Alpha Composite Reliability Value Limit Desc. 1 Work Behavior (X1) 0.939 0.954 0.700 Fulfilled 2 Competency (X2) 0.895 0.935 0.700 Fulfilled 3 Job Satisfaction (Y1) 0.899 0.937 0.700 Fulfilled 4 Performance (Y2) 0.918 0.942 0.700 Fulfilled Source: Processed Research Data (2022) The results of the reliability test using the Composite Reliability value and the Cronbach Alpha value obtained the value of the test results meeting the testing criteria of more than 0.70 so that each variable of work behavior (X1), competence (X2), job satisfaction (Y1), and performance (Y2) has meet construct reliability. Inner Model Test Results. Inner model analysis is usually called (inner relations, structural model and substantive theory) which describes the relationship between latent variables based on substantive theory. The inner model analysis can be evaluated using R-square for the dependent construct. Evaluating the inner model with PLS (Partial Least Square) begins by looking at the R- square for each latent dependent variable. Then the interpretation is the same as the interpretation of the regression. Changes in the value of the R-square can be used to assess the effect of certain independent latent variables on the latent dependent variable and whether it has a substantive effect. The results of the inner model test include the value of the coefficient of determination (R-square). Table 10. Determination Coefficient Test Results No Independent Variable Dependent Variable R Square R Square Adjusted 1 Work Behavior (X1) Job Satisfaction (Y1) 0.775 0.769 Competency (X2) 2 Work Behavior (X1) Performance (Y2) 0.845 0.838 Competency (X2) Job Satisfaction (Y1) Source: Processed Research Data (2022) The results of the coefficient of determination of the influence between work behavior (X1) and competence (X2) on job satisfaction (Y1) obtained an R square value of 0.775, which means that 579 the effect on job satisfaction (Y1) can be explained by 77.5 percent by work behavior (X1) and competence (X2). In contrast, the rest is explained by other variables. The results of the coefficient of determination of the influence between work behavior (X1), competence (X2), and job satisfaction (Y1) on performance (Y2) obtained an R square value of 0.845, which means that the effect on performance (Y2) can be explained by 84.5 percent by behavior work (X1), competency (X2), and job satisfaction (Y1). In contrast, the rest is explained by other variables. Research Model Development. Based on testing the latent variable model in this study, it is grouped into two groups: exogenous and endogenous variables. Exogenous variables are work behavior and competence, while endogenous variables are job satisfaction and performance. The model is said to be good if empirical data theoretically support the development of the hypothetical model. Work behavior (X1), competence (X2), job satisfaction (Y1), and performance (Y2) have a positive relationship between variables. The work behavior variable (X1), as measured by the 5 indicators, shows that the cooperation indicator (X1.4) is the dominant indicator in forming work behavior variables with a loading factor value of 0.915. The competency variable (X2), as measured from the 3 indicators, shows that the knowledge indicator (X2.1) is the dominant indicator in forming competency variables with a loading factor value of 0.929. The job satisfaction variable (Y1), as measured from 3 indicators, shows that the company organizational size indicator (Y1.3) is the dominant indicator in forming the job satisfaction variable with a loading factor value of 0.937. The performance variable (Y2), as measured from the 4 indicators, shows that the timely indicator (Y2.3) is the dominant indicator in forming performance variables with a loading factor value of 0.911. Hypothesis test. In testing, the hypothesis can be seen from the value of the t-statistics and the probability value. For hypothesis testing, namely by statistical values, for alpha 5%, the t-statistic value used is 1.960. So that the criteria for accepting or rejecting the hypothesis are that Ha is accepted and H0 is rejected when the t-statistic is > 1.96. To reject/accept the hypothesis using probability, Ha is accepted if the p-value <0.05. Based on the empirical data used in this study, it is possible to test the proposed hypothesis. The following presents the results of hypothesis testing based on the path coefficient values and T-Statistics / P-value. Table 11. Hypothesis test Results No Influence Path Coeff Stdev T Count P Desc. 1 X1 -> Y2 0.311 0.148 2.103 0.036 Significant 2 X2 -> Y2 0.287 0.141 2.042 0.042 Significant 3 X1 -> Y1 0.513 0.128 4.004 0.000 Significant 4 X2 -> Y1 0.390 0.135 2.883 0.004 Significant 5 Y1 -> Y2 0.363 0.105 3.456 0.001 Significant 6 X1 -> Y1 -> Y2 0.187 0.064 2.905 0.004 Significant 7 X2 -> Y1 -> Y2 0.142 0.068 2.076 0.038 Significant Source: Processed Research Data (2022) H1: work behavior (X1) affects performance (Y2). The hypothesis of the influence of work behavior (X1) on performance (Y2) obtained a path coefficient of 0.311 with a t-value of 2.103 and a significance value of 0.036. These results show that the t value is more than the t table (t count> 1.960). The significance value is less than 0.05 (sig < 0.05), so that work behavior (X1) has a positive and significant influence on performance (Y2), meaning the higher the work behavior (X1) will effect, the higher the performance (Y2). These results indicate that the hypothesis is proven correct. It shows that the work behavior of non-PNS permanent education staff at FT UB Malang significantly influences performance. Work behavior is influenced by having a stake in supporting performance. For example, in work life, work behavior is an essential part. Factors Influencing Work 580 Behavior, according to Notoatmodjo (2003:25), is the factor of people (individuals), namely factors that come from the person/individual himself. For example, Education, age and years of service. From table 2, the age of non-PNS permanent education staff is mostly young, namely 31 to 40 years. At this age, employee performance will increase. It is in line with the opinion of Aprilyanti (2017), namely that age can be in a productive period having a higher productivity level when compared to workers who are old. Hence, their physical condition becomes weak and limited in their activities. According to Robbins (2008: 45), translated by Molan (2010: 46), the relationship between performance and age is very closely related. The reason is the widespread belief that performance declines with increasing age. Older employees are also considered less flexible and reject new technology. From the description above, work behavior influences the performance of non-PNS permanent education staff at FT UB. This supports Mafra's previous research, Nisa' Ulul (2016), which states that behavior significantly affects employee performance. H2: competence (X2) affects performance (Y2). The hypothesis of the influence of competence (X2) on performance (Y2) obtained a path coefficient of 0.287 with a t-value of 2.042 and a significance value of 0.042. These results show that the t value is more than t table (t count> 1.960). The significance value is less than 0.05 (sig < 0.05), so competency (X2) has a positive and significant effect on performance (Y2), meaning that the higher competence (X2) will affect the higher performance (Y2). These results indicate that the hypothesis is proven correct. It shows that the competence of non-PNS permanent education staff at FT UB Malang significantly influences performance. Ahsan in Mulyasa (2003: 38) argues that competence is defined as knowledge, skills and abilities that are controlled by someone who has become part of himself so that he can perform behaviors on him in terms of knowledge, attitude and treatment (cognitive, affective and psychomotor) properly, meaning that by having a good attitude that is owned by non-PNS FT UB permanent education staff, it will support their performance in completing a job. From table 2, the non-PNS permanent education staff, on average, have received education up to university (bachelor's degree), reaching 54%. It implies that someone with extensive knowledge and insight and highly educated reflects a high professional value at work. Being someone who has knowledge is closely related to the spirit of seeking knowledge. Therefore, with broad insight and experience will increase, the non-PNS permanent staff of FT UB Malang will be more skilled, and the ability to work will be better. From the description above, competence influences the performance of non-PNS permanent education staff at FT UB. It supports previous research from Putri, E. N., & Ardiana, I. D. K. R. (2019), which states that competency affects employee performance. H3: work behavior (X1) influences job satisfaction (Y1). The hypothesis of the influence of work behavior (X1) on job satisfaction (Y1) obtained a path coefficient of 0.513 with a t-value of 4.004 and a significance value of 0.000. These results show that the t value is more than t table (t count> 1.960) and the significance value is less than 0.05 (sig < 0.05), so that work behavior (X1) has a positive and significant influence on job satisfaction (Y1). Higher work behavior (X1) will affect job satisfaction (Y1). These results indicate that the hypothesis is proven correct. It shows that the work behavior of non-PNS permanent education staff at FT UB Malang significantly influences job satisfaction. Gray (2002, pp. 401) in Sumartik (2019, pp. 54) argued that men and women must know that gender differences can affect their work behavior. Table 2 shows more female non-PNS permanent education staff than men, reaching 64%, which is 45 employees. Unnoticed by men and women, many words or behaviors considered normal by each gender can offend the feelings and self-esteem of the opposite sex. It, of course, can lead to conflict, affecting work behavior and disrupting a comfortable working atmosphere. Gray (2002: 403) in Sumartik (2019: 54), to create good work behavior, one must pay attention to male and female communication and feelings at work, set boundaries in each work behavior, and remind the differences that exist. In this case, the 581 work behavior between men and women is different. Understanding work behavior according to gender requires full communication and understanding so that it does not result in conflict at work. From the description above, work behavior influences the job satisfaction of non-PNS permanent education staff at FT UB. It supports previous research from Bakri, M. and Yadi, R. (2020), which states that work behavior has a significant effect on job satisfaction. H4: competency (X2) influences job satisfaction (Y1). The hypothesis of the influence of competence (X2) on job satisfaction (Y1) obtained a path coefficient of 0.390 with a t-value of 2.883 and a significance value of 0.004. These results show that the t value is more than t table (t count> 1.960). The significance value is less than 0.05 (sig < 0.05), so competency (X2) has a positive and significant effect on job satisfaction (Y1), meaning the higher the competency (X2) will effect, the higher the job satisfaction (Y1). These results indicate that the hypothesis is proven correct. It shows that the competence of non-PNS permanent education staff at FT UB Malang significantly influences job satisfaction. One of the factors influencing work behavior, according to Notoadmodjo (2003:25), is the person factor (individual), namely the period of service. Non-PNS FT UB permanent education staff have an average working period of 11 to 20 years, which is 67%. The longer employees work in one agency, the more work behavior increases. In addition, the ability to work will be more skilled, so working more skillfully will result in better performance, and employees will feel satisfied at work. From the description above, competence influences the job satisfaction of non-PNS permanent education staff at FT UB. It supports previous research from Nugraha et al. (2022), which stated that competency influences job satisfaction. H5: job satisfaction (Y1) affects performance (Y2). The hypothesis of the effect of job satisfaction (Y1) on performance (Y2) obtained a path coefficient of 0.363 with a t-value of 3.456 and a significance value of 0.001. These results show that the t value is more than t table (t count> 1.960). The significance value is less than 0.05 (sig < 0.05), so job satisfaction (Y1) has a positive and significant effect on performance (Y2), meaning that higher job satisfaction (Y1) will affect the higher performance (Y2). These results indicate that the hypothesis is proven correct. It shows that the job satisfaction of non-PNS permanent education staff at FT UB Malang significantly affects performance. Job satisfaction also affects a person's psychological condition. Employees who do not get job satisfaction never reach psychological maturity and, in turn, will become frustrated. Meanwhile, employees who get job satisfaction usually have better attendance and turnover records, are less active in union activities, and (sometimes) perform better at work than employees who do not get job satisfaction. Therefore, job satisfaction has important meaning for both employees and companies or organizations, especially because it creates positive conditions in the work environment of an agency, according to Robbins (2002: 181), one that is related to performance that determines or encourages job satisfaction, namely wages/salaries. Salary is the main factor influencing employee satisfaction, so when agencies or organizations feel they have provided sufficient salaries, they feel that their employees are satisfied. Table 2 shows that non-PNS UB Malang permanent education staff are already in class III/b. The salary for class III/b is already above the Working Minimum Wage (UMK) in Malang City according to the Decree of the Governor of East Java Number 188/803/KPTS/013/2021 concerning Regency/City Minimum Wage in East Java in 2022. The UMK of Malang city is Rp. 2,994,143. Apart from getting a salary, non-PNS FT UB permanent education staff also receive incentives, meal allowance and other income regulated by the decision of the UB Chancellor. From the description above, job satisfaction affects the performance of non-PNS permanent education staff at FT UB. It is supported by previous research from Sudiyanto, E. (2020), which states that job satisfaction significantly affects performance. H6: work behavior (X1) influences performance (Y2) by mediating job satisfaction (Y1). The hypothesis of the influence of work behavior (X1) on performance (Y2) by mediating job satisfaction 582 (Y1) obtained a path coefficient of 0.187 with a t-value of 2.905 and a significance value of 0.004. These results show that the t value is more than t table (t count> 1.960). The significance value is less than 0.05 (sig < 0.05), so that work behavior (X1) has a positive and significant effect on performance (Y2) by mediation job satisfaction (Y1), meaning that the higher work behavior (X1) will affect the higher job satisfaction (Y1), which in turn will have an indirect effect on higher performance (Y2). These results indicate that the hypothesis is proven correct. It shows that work behavior through job satisfaction of non-PNS permanent education staff at FT UB Malang significantly influences performance. Work behavior has contributed to supporting performance. For example, in work life, work behavior is a significant part. In accordance with the explanation in hypothesis 1, that work behavior is influenced by the age factor Notoadmodjo (2003: 25). When compared to the workforce who are older, the non-PNS FT UB education employees who are young may be considered to be in the productive period and have a higher degree of productivity, causing the physical possession to weaken and become limited to activities. According to Robbins (2008: 45), translated by Molan (2010: 46), the relationship between performance and age is very closely related. The reason is the widespread belief that performance declines with increasing age. Older employees are also considered less flexible and reject new technology. From the 5th hypothesis that one of the factors that can affect performance satisfaction is wages or salary (Robbins (2002: 181), salary is the main factor affecting employee satisfaction, so when agencies or organizations feel they have provided sufficient salary, they feel that their employees are satisfied From this explanation, non-PNS permanent education staff at FT UB at a young age will improve their performance supported by sufficient salaries which have exceeded the UMK in 2022 Malang City will feel satisfied, thereby increasing the enthusiasm of non-PNS permanent education staff at FT UB The results of this study support previous research from Sujarwo, A and Wahjono (2017), which stated that there is a mediating effect of job satisfaction from the influence of innovative behavior on employee performance. H7: competency (X2) influences performance (Y2) by mediating job satisfaction (Y1). The hypothesis of the influence of competence (X2) on performance (Y2) by mediating job satisfaction (Y1) obtained a path coefficient of 0.142 with a t-value of 2.076 and a significance value of 0.038. These results show that the t value is more than t table (t count> 1.960). The significance value is less than 0.05 (sig < 0.05), so competency (X2) has a positive and significant effect on performance (Y2) mediated by satisfaction work (Y1), meaning that the higher competency (X2) will affect the higher job satisfaction (Y1), which in turn will have an indirect effect on the higher the performance (Y2). These results indicate that the hypothesis is proven correct. It shows that the competence of non- PNS permanent education staff at FT UB Malang significantly influences performance. Ahsan in Mulyasa (2003: 38) competence is defined as knowledge, skills and abilities that are mastered by someone who has become part of himself so that he can carry out behaviors on him in terms of knowledge, attitude and treatment well, meaning by having an attitude The good education staff of Non-PNS FT UB will support their performance in completing a job. According to the explanation in hypothesis 2, that competence is influenced by educational factors. Education is an indicator that reflects a person's ability to complete a job (Hasibuan, 2011: 45). In this study, the Non-PNS FT UB academic staff have received education up to university (bachelor's degree). It implies that someone who has broad knowledge and insight and is highly educated reflects a high value of professionalism at work. Being someone who has knowledge is closely related to the spirit of seeking knowledge. Therefore, with broad insight and experience, it will increase, permanent non-PNS FT UB Malang educational staff will be more skilled, and the ability to work will be better so that working more skillfully will result in better performance and employees will feel satisfied in their 583 work. Work. The results of this study support previous research from Umamy (2021), which stated that there is a mediating effect of job satisfaction from competence to employee performance. CONCLUSION The work behavior of non-PNS permanent staff of FT UB Malang has a positive and significant effect on performance. The competence of non-PNS permanent education staff of FT UB Malang has a positive and significant effect on performance. The work behavior of non-PNS permanent staff of FT UB Malang has a positive and significant effect on job satisfaction. The competence of non-PNS permanent education staff of FT UB Malang has a positive and significant effect on job satisfaction. Job satisfaction of non-PNS permanent education staff of FT UB Malang has a positive and significant effect on performance. The work behavior of non-PNS permanent staff of FT UB Malang has a positive and significant effect on performance through job satisfaction. The competence of non- PNS permanent education staff of FT UB Malang has a positive and significant effect on performance through job satisfaction. REFERENCES Amar, H. (2020). Pengaruh Kompetensi Dan Perilaku Kerja Terhadap Kinerja Pemeriksa Di Inspektorat Kabupaten Bangka. Kemudi: Jurnal Ilmu Pemerintahan, 4 (2), 275 – 291. https://doi.org/10.31629/kemudi.v4i2.1932 Bakri, M., & Rahmad, Y. (2020). Pengaruh Perilaku Kerja, Lingkungan Kerja Dan Interaksi Sosial Terhadap Kepuasan Kerja Pegawai Pada UPT Kementerian Kehutanan Provinsi Aceh. Serambi Konstruktivis, 2 (4), 151 – 165. https://doi.org/10.36665/jusie.v4i01.183 Edison, E., Anwar, Y., & Komariyah, I. (2016). Manajemen Sumber Daya Manusia: Strategi dan Perubahan Dalam Rangka Meningkatkan Kinerja Pegawai dan Organisasi. Cetakan Kesatu. Bandung: CV. Alfabeta. Ghozali & Latan. (2015). Partial Least Squares: Konsep, Teknik dan Aplikasi Menggunakan Program SmartPLS 3.0. Semarang: Badan Penerbit UNDIP. Hasibuan, M. S. P. (2011). Manajemen Sumber Daya Manusia (Edisi Revisi). Jakarta: Bumi Aksara. Mafra, N. U. (2016). Pengaruh Perilaku Dan Disiplin Kerja Terhadap Kinerja Pegawai Pada Unit Pelaksana Teknik (UPT) Universitas Pgri Palembang. Jurnal Media Wahana Ekonomika, 12 (4), 104 – 120. Mangkunegara, P. A. (2013). Manajemen Sumber Daya Manusia Perusahaan. Bandung: PT Remaja Rosdakarya. Mulyasa. (2003). Kurikulum Berbasis Kompetensi. Bandung: Remaja Rosda. Notoatmodjo, S. (2003). Pengembangan Sumber Daya Manusia. Jakarta: PT. Rineka Cipta. Nugraha, D. A., Nadeak, B., Martini, N., & Suyaman, D. J. (2022). Pengaruh Pengembangan Karir, Kompetensi, Dan Kepuasan Kerja Terhadap Kinerja Pegawai Negeri Sipil Dinas Kelautan Dan Perikanan Provinsi Jawa Barat. At-Tadbir: Jurnal Ilmiah Manajemen, 6 (1), 81 – 93. https://doi.org/10.31602/atd.v6i1.5886 Peraturan Pemerintah Nomor 30 Tahun 2019 tentang Penilaian Kinerja Pegawai Negeri Sipil. Peraturan Rektor Universitas Brawiijaya Nomor 31 Tahun 2021 Tentang Kepegawaian. Peraturan Rektor Universitas Brawijaya Nomor 25 Tahun 2020 Tentang Susunan Organisasi dan Tata Kerja. Putri, E. N. (2019). Pengaruh Motivasi Kerja, Kompetensi, Dan Kompensasi Terhadap Kinerja Tenaga Kependidikan Melalui Kepuasan Kerja Di Lingkungan ITS Surabaya. Global, 3 (2), 28– 36. https://doi.org/10.31629/kemudi.v4i2.1932 https://doi.org/10.36665/jusie.v4i01.183 https://doi.org/10.31602/atd.v6i1.5886 584 Robbins, P. S. (2008). Organizational Behaviour, Tenth Edition (Perilaku Organisasi Ke Sepuluh), alih bahasa Drs. Benyamin Molan. Jakarta: Salemba Empat. Robbins, S. (2003). Perilaku Organisasi (terjemahan: Dr.Hadayana Pujaatmaka). Jakarta: Prenhalindo. Robbins, S. P. (2002). Prinsip-prinsip Perilaku Organisasi (5th ed). Jakarta: Airlangga. Sidik, W. W., Samdin, & Syaifuddin, D. T. (2018). Pengaruh Lingkungan Kerja Dan Perilaku Kerja Terhadap Kinerja Pegawai (Effect of Employment Ability, Employment Engagement and Work Motivation on Employee Performance). JUMBO (Jurnal Manajemen, Bisnis, dan Organisasi), 2 (1), 32 – 45. Sudiyanto, T. (2020). Pengaruh Kepuasan Kerja dan Beban Kerja Terhadap Kinerja Pegawai Pada Dinas Tenaga Kerja dan Transmigrasi Banyuasin. Jurnal Media Wahana Ekonomika, 17(1), 93 – 115. https://doi.org/10.31851/jmwe.v17i1.4338 Suharno, P., & Dara, D. (2017). The Impact of Work Motivation and Competence on Employee Performance Through Service Quality in Administrative Staff Of universitas Negeri Jakarta, Indonesia. RJOAS, 1(61), 160 – 171. https://doi.org/10.18551/rjoas.2017-01.16 Sujarwo, A., & Wahjono. (2017). Pengaruh Motivasi Kerja Dan Perilaku Inovatif Terhadap Kinerja Karyawan Dengan Kepuasan Kerja Sebagai Variabel Mediasi (Studi Kasus pada LKP ALFABANK Semarang). INFOKAM, 13 (3), 63 – 71. Sumartik. (2019). Perilaku Organisasi. Sidoarjo: Universitas Muhammadiyah Sidoarjo Press. https://doi.org/10.21070/2019/978-602-5914-62-1 Umamy, S. H. (2021). The Role of Job Satisfaction in Mediating the Effect of Competency and Perceived Organizational Support on Education Performance: An Evidence from the Muhammadiyah University of Jember, Indonesia. International Journal of Management, Entrepreneurship, Social Science and Humanities (IJMESH), 4 (2), 11 – 27. https://doi.org/10.31098/ijmesh.v4i2.513 Undang-Undang No. 20 Tahun 2003 tentang Sistim Pendidikan Nasional. Widakdo, D. S. W. P., Holik, A., & Iska, L. N. (2021). Efek Usia dan Tingkat Pendidikan terhadap Kinerja Tenaga Bantu Penyuluh Pertanian. Jurnal Penyuluhan, 17 (01), 52-59. https://doi.org/10.25015/17202131614 https://doi.org/10.31851/jmwe.v17i1.4338 https://doi.org/10.18551/rjoas.2017-01.16 https://doi.org/10.21070/2019/978-602-5914-62-1 https://doi.org/10.31098/ijmesh.v4i2.513 https://doi.org/10.25015/17202131614