Analysis of User Satisfaction Factors of E-Kinerja Application as Utilization of the Paperless Office System: A Case Study in Regional Civil Service Agency, North Sulawesi Province

User Satisfaction is a crucial factor to the success of the Information System which will have extensive impact on the benefits for individuals and organizations as well as users’ continuance intention. Paperless office system is the use of e-government which aims not only to reduce paper usage but to increase effectiveness and efficiency in public services. E-Kinerja is a method provided and supported by the government with a purpose of improving employee productivity and performance. Through the integration of Extended Expectation Confirmation Model-IT (EECM-IT) and updated Delone and McLean IS success model, this research examines the relationship between each variables adopted and then identify factors intended to influence user satisfaction as well as building significant impact towards implementation of E-Kinerja application. Using data collected from a research of 553 respondents users’ E-Kinerja applications through the Partial Lest Square Structural Equation Model (PLS-SEM) method, interestingly results have shown that User Confirmation, Perceived Ease of Use, and Perceived Usefulness are strongly positive and create significant effect towards User Satisfaction. Whereas User Satisfaction had impacted positively to the Use and Net Benefit. However data also indicate that User Satisfaction is critical and bring no effect towards Continuance Intention to Use. In addition, Use and Net Benefit convey important impact positively to the users’ continuance intention.


Introduction
The use of e-government by citizen becomes a tendency in the long term because a growing number of people realize the benefits of e-government which have an impact on the efficiency, effectiveness and convenience of its users [1]. Paperless office appears to save the environment by reducing usage paper as a work tool. Paperless office system is the use of e-government which aims not only to reduce paper usage but to increase effectiveness and efficiency in public services. The e-government project in North Sulawesi Provincial Government implemented to realize good governance through improving the quality of performance and public services by the regional government agency in a professional, integrated, effective, efficient, transparent, accountable and realizing the performance optimization in achieving the vision and mission of the provincial government [2]. However, citizens' satisfaction in implementing e-government was not always up to their expectations [3]. Satisfaction analysis has been 2. Literature Review

E-Kinerja
North Sulawesi Provincial Regional Civil Service Agency is an agency that has the task of assisting the Governor in implementing the formulation and regional policies in the area of regional staffing [22], has formulated an information technology-based regional performance measurement in an effort to support e-government with a purpose of improving professionalism, productivity and discipline of Civil Servants. Performance is the entire result of the implementation of tasks in the form of work standards, goal or targets which are carried out during a certain period and previously agreed upon [23]. E-Kinerja at the North Sulawesi Provincial Government is an IS developed as a measurement tool for evaluating the performance of candidate of Civil Servants and Civil Servants in the context of payment of additional income by utilizing electronic media networks [24]. Additional income is given to civil servants who hold senior leadership positions, pratama high leadership positions, administrator positions, supervisory positions, certain functional positions, executive positions and candidate of Civil Servants based on an assessment consisting of discipline and performance elements [24]. The E-Kinerja application which is the focus of this study was used since April 2019 by civil servants in the  [25]. Referring to the number of users, it is considered necessary to do research to analyze user satisfaction which is also the first study in this application.

Updated DeLone and McLean
The updated DeLone and McLean IS success model [11] has six constructs: three dimensions of quality ("system quality", "information quality", and "service quality") that affect subsequent "use" and "user satisfaction", as a results "net benefit" will occur. The updated DeLone and McLean model is one of the most widely used models in researching the success factors of an information system [26].

Extended Expectation Confirmation Model -IT (EECM-IT)
According to Hong et al. [12] conceptually TAM [13] and ECM [7] have similarities. For example, both models have a perceived usefulness variable, so they combine the two models into a hybrid model by adding perceived ease of use of TAM to the ECM-IT. As theorized in TAM, perceived ease of use is expected to have a significant impact on perceived usefulness and continued IT usage intention [12]. Furthermore, the perception of ease of use is applied to the relationship between confirmation and perceived usefulness in ECM-IT that have an impact on continued IT usage intention. So that when the user gets a confirmation experience, the user's perceived ease of use will become more real and updated. This concept provides a more complete explanation of the use of continued IT usage intention [12].

Extended Expectation Confirmation Model -IT (EECM-IT) dan DeLone dan McLean
A hybrid model integrating EECM-IT and DeLone and McLean was also considered in this study. With the similarity of user satisfaction component in these two models, integration was carried out. In EECM-IT [12] user satisfaction variable is the impact of confirmation, perceived ease of use and perceived usefulness. Whereas in updated DeLone and McLean [11] user satisfaction variable is the impact of all existing variables. But both of these models explain user satisfaction is one of the critical success variables in information system. Given the evidence of the significant impact of user satisfaction on continued IS usage intention [7] from previous studies, continuance intention to use is added to the updated DeLone and McLean. Similar to user satisfaction, net benefit is expected to have a significant impact on users' continuance intention. Through user satisfaction, individual and organizational impacts will occur that it is expected to have an impact on users' continuance intention. Likewise, the use variables that are defined for attitudes, links with behaviour, and nature of use [11] are expected to have a positive influence on continued IS usage intention.

Theoretical Framework
As mentioned previously, this research integrates EECM-IT [12] and the updated DeLone and McLean models [11]. EECM-IT has five variables: Confirmation, Perceived Usefulness, Perceived Ease of Use, User Satisfaction and Continuance Intention to Use, while DeLone and McLean have six variables: Information Quality, System Quality, Service Quality, User Satisfaction, Use, and Net Benefit. However, in this study we only adopted three DeLone and McLean variables: User Satisfaction, Use, and Net Benefit where these three variables are integrated with the EECM-IT model and this hybrid model is rarely explored by other researchers. The three quality variables are not adopted anymore because the indicators for these three variables have been represented by the Confirmation, Perceived Usefulness and Perceived Ease of Use variables. Figure 1 represents research model and hypotheses examined in this study. In the sub-section below, the relationships between variables and the hypothesis formulation process in the proposed model will be explained further.

Relation between Confirmation to User satisfaction, Perceived Usefulness, and Perceived Ease of Use
According to Bhattacherjee [4], confirmation has a significant impact on user satisfaction and perceived usefulness by users of information systems. In some previous studies [27,28,29,30] also showed a positive influence on confirmation of user satisfaction and perceived usefulness. Furthermore, Hong et. Al [12] added perceived ease of use which became one of the salient beliefs in determining IT acceptance. The user's perceived ease of use will become more concrete after the confirmation experience which has a positive impact on user satisfaction. E-Kinerja is a government program specifically for civil servants to increase work productivity [24], so it is expected that this application can be used in the long term where users can adjust their perceived ease of use, perceived usefulness and satisfaction based on confirmation experience. H1. Confirmation (CON) has a positive and significant impact on Perceived Ease of Use (PEU) H2. Confirmation (CON) has a positive and significant impact on Perceived Usefulness (PUS) H3. Confirmation (CON) has a positive and significant impact on User Satisfaction (USN)

Relation between Perceived Usefulness to User Satisfaction and Continuance Intention to Use
Bhattacherjee [4] adopted the perceived usefulness of Davis et al. [31] which mentions perceived usefulness as a major part of determining users' intention to use a computer. Perceived usefulness is also an important expectation of user satisfaction after using information systems [4]. From several previous studies [15,28,29,30] the perceived usefulness have a positive impact on user satisfaction and continuing intention to use. In implementing E-Kinerja applications, perceived usefulness is a determinant of user satisfaction because it is related to employee productivity which impacts on the continued IS usage intention.

Relation between Perceived Ease of Use to User Satisfaction and Continuance Intention to Use
Perceived ease of use is added in this model given its significant impact on perceived usefulness and continuance intention to use that has been proven in several previous TAM-based studies [13], with the expectation this variable can also have a positive influence on user satisfaction [12]. According to Davis et al. [31] perceived ease of use is a secondary determinant after perceived usefulness towards

Relation between User Satisfaction to Continuance Intention to Use, Use and Net Benefit
Expectation-Confirmation Theory identifies IS continuation intentions determined by their satisfaction with IS use [4]. According to Delone and McLean [11] an increase in user satisfaction will lead to the intention to use and thus used. As a result of user satisfaction and use a net benefit will be achieved.
Previous studies have shown that user satisfaction is highly correlated with usage [32,33], which is also related to net benefits [32,34], and IS continuance intention [4,16,36]. Likewise, expectations for the users' satisfaction of E-Kinerja applications will have a positive impact on their use and can provide benefits to individuals and organizations so that it positively influences the intention to use E-Kinerja applications continuously.

Relation between Use to Net Benefit and Continuance Intention to Use
As in previous studies [32,34] use has a positive impact on net benefits. Then we explore the relationship between use and continuity of intention to use, with reference to the definition of use not only the attitudes, links with behaviour but also nature of use [10] so that is expected to have an impact on continuance intention to use.

Relation between Net Benefit to Continuance Intention to Use
Considering the effect of net benefits has a significant impact on organizational success [35] so we explore the relationship between benefits and intentions for continuing SI and it is expected that if users perceive these benefits, the intention to continue using E-Kinerja application will be achieved.
H13. Net Benefit (NET) has a positive and significant impact on Continuance Intention to Use (CIU)

Research Methodology
Stages of research to obtain the results and conclusions of this study are shown in Figure 2. While for the stages of problem determination, literature review and theoretical framework have been explained in previous sections.

Design and Procedure
This study uses a quantitative approach through a questionnaire, using a Likert scale [30] starting from level 1 (strongly disagree) to 5 (strongly agree). This questionnaire contains two parts: respondent demographics and research questions. Before distributing the questionnaire, a readability test was conducted on five E-Kinerja application users. The questionnaire was distributed online using https://survey.ui.ac.id/web/ which was then distributed to civil servants within the North Sulawesi Provincial Government as the user of the E-Kinerja application. A total of 553 people participated in filling out the questionnaire. Then the data is processed using PLS-SEM together with SmartPLS 3.0 as its software to analyze the results of the data that has been received.

Respondent Demographics
The number of respondents in this study were 553 out of 5,601 users of the E-Kinerja application. As shown in Table 1, 51% of respondents are women, 97% of respondents use the E-Kinerja application every workday, and 76% of respondents have an executive position.

Result
There are two stages of model measurement in this study, first, evaluation of the measurement model (outer model) and second, evaluation of the structural model (inner model). In the outer model, Cross Loading Factor (see Table 2) and AVE Reliability Test, Composite Reliability, Cronbach's Alpha (see Table 3) are carried out.
After building the initial path diagram it is necessary to test the validity of the indicators seen in the loading factor (LF). According to Yamin [37] LF values between 0.5 -0.6 can still be accepted so that it is declared valid. In this study the LF value on all indicators was declared valid because it was above 0.5. Furthermore, in discriminant validity testing using cross loading factors, Table 2 shows that each indicator on the research variable has the largest cross loading value on the variable it forms compared to the cross loading value on other variables. That means the indicators in this study have good discriminant validity in arranging their respective variables [21]. Then evaluate the value of Construct Reliability by observing the value of Cronbach Alpha (CA) and Composite Reliability (CM) as shown in Table 3. If all CA values are greater than 0.7 it can be concluded that the indicators are consistent in measuring the construct, while the AVE value is required greater than 0.5 for a good model and the CM value must be greater than 0.6.  The structural model evaluation is also called the inner model, determining the relationship between independent and dependent latent variables to find out whether a hypothesis is accepted or rejected. Structural model with R 2 to determine the effect of the independent (exogenous) latent variable on the dependent (endogenous) latent variable as shown in Table 4. It can be seen that CIU is influenced by 62% of PUS, PEU and USN, while USN is influenced by 65% of CON, PEU and EFA. The structural model testing is done by bootstrapping then Path Coefficients. By observing the tstatistic value where the t-statistic value greater than equal to 1.96 is declared valid [12] as shown in Table 5.  Table 5, of all the proposed hypotheses there is only one hypothesis that is rejected, that is H8, it means that user satisfaction does not significantly influence the intention to use IS continuously. This differs from the ECM theory which argues that there is a significant influence of satisfaction on IS continuance intention.

Discussion
Information system user satisfaction is the impact of ease of use even after experiencing confirmation (the reality of the expected benefits), and the perceived usefulness. This is in accordance with [12] and in this study each variable that affects user satisfaction has a positive and significant influence. So it was found that user satisfaction with the implementation of the E-Kinerja application has been achieved where the user feels this application is easy to use and operate even after going through this application experience is better than expected by the user, so users can feel the benefits of using this application. But the success of information systems is not only measured by user satisfaction but also from the use of information systems in the long term [7]. From the results of the analysis of this study prove the application of E-Kinerja provides benefits for individuals and organizations that have an impact on the use of applications in the long term. Likewise, when viewed from the usage factor, the results of the analysis prove the overall use of this application has a positive and significant impact on the user's intention to continue using the E-Kinerja application. But it is different from user satisfaction. The results of the analysis prove the users satisfaction of E-Kinerja applications does not significantly influence the user's intention to use the application so this is contrary to ECM theory [7].

Implications
Paperless office systems can have implications to support the realization of good governance by exploiting information technology, and can increase the effectiveness and efficiency. The theoretical implications of the results of this study indicate that user satisfaction does not directly impact the users' continuance intention of IS but through net benefits. This contradicts ECM theory which is generally accepted in the context of the continued IS usage intention [7]. Therefore, by conducting a hybrid model between ECM and DeLone and McLean, we found other factors that influence the continued IS usage intention. The integration of these two models opens up another point of view for researchers that the factor of use and net benefits can significantly influence the users' continuance intention. As a practical implication, for researchers who want to expand the IS success model can be done by integrating variables that have similarities both conceptually and the resulting impact. Furthermore, this model can be used for future research, especially in other e-government implementations. This research can also be explored again by adding several new variables to make the context of this study seen from various aspects.
While the government is advised to maintain the continuity of the information system, several factors must be considered: first, perceived ease of use of the system which is more emphasized on how users operate the system easily, the use of the system indicated by the frequency of users accessing and using the system, secondly, the perceived usefulness where users can feel the results of the use of information systems provide benefits for performance and productivity, and thirdly, net benefits that have an impact on individuals, industries and organizations. By maintaining the continued IS usage intention, e-government utilization is successful.
For Regional Civil Service Agency, North Sulawesi Province this study can provide recommendations that user satisfaction with the implementation of E-Kinerja applications in the North Sulawesi Provincial government has proven successful as shown by the results of the analysis conducted. However, in order to maintain user satisfaction, it is necessary to periodically evaluate other supporting factors that influence user satisfaction.

Conclusion
Reducing the use of paper in the office is one concrete action to create a green office. The E-Kinerja application as an embodiment of a paperless office system is a way to support e-government. This 10 study examines the factors that influence user satisfaction that affect the benefits for individuals and organizations that have an impact on the continued intention to use the E-Kinerja application. From the results of this study it can be concluded the factors that influence user satisfaction E-Kinerja application are perceived ease of use, confirmation that the realization of the results of using the E-Kinerja application is better than user expectations, and perceived usefulness of E-Kinerja application. However, user satisfaction is not enough to determine the success of the information system, so we need other factors that support so that this application can continue to be used in the long term. The results of this study prove the benefits of E-Kinerja applications to individuals and organizations, and the frequency of use of E-Kinerja applications has a positive and significant impact on the continued intention of using E-Kinerja applications. While user satisfaction with this application has not been proven to affect the user intention to use E-Kinerja application continuously.