A STUDY ON INFORMATION SYSTEMS SUCCESS: EXAMINING USER SATISFACTION OF ACCOUNTING INFORMATION SYSTEM

The purpose of this study is to empirically examine the quality of information systems, information, and service influence on user satisfaction on accounting information systems. The study model used to examine the accounting information systems succession was the model modified from DeLone & McLean (2003) and the research developed by Gable, Sedera, & T (2008) The data was collected through questionnaires given to 84 respondents who were users of the accounting information system software (SIMDA/SIPKD) that worked as an analyst in city/regency governments all over West Java Province. The data was analyzed using a multiple regression analysis through the SPSS program. The results show that the variables' quality has a partially positive and significant influence on accounting information systems' user satisfaction. It implies that an increase in the quality of information systems, information, and services implemented will increase user satisfaction. The test results also reveal that the variables' quality positively and significantly affected the users' satisfaction with the accounting information system in city/regency governments throughout West Java Province.


INTRODUCTION
Regional financial management should be done in a professional, open, and responsible manner to realize good governance in implementing regional autonomy. Law No. 17 of 2003 on State Finances and Law No. 1 of 2004 on State Treasury require regional governments and Regional Work Units (SKPD) as budget users to prepare financial reports as a form of accountability for their financial management. The Government has also issued regulations regarding the mandatory use of Information Technology (IT) in Government Regulation (PP) No. 65/2010, an amendment from PP No. 56/2005 about Regional Financial Information Systems.
Information technology combines computers with high-speed communication lines that carry data, voice, and video. A system supported by Information Technology will give the organization value-added if they design it effectively and efficiently. The Accounting Information System (AIS) implementation in integrated computer-based regional financial management can help the regional government prepare and present high-quality financial reports by increasing accountability and control (Fatmawati & Setiawan, 2018). Using an accounting information system will help users achieve a competitive advantage (Wuriasih & Mokodompit, 2017).
Several previous studies stated that the succession of implementing a system depends on how the system is operated, the ease of use, the utilization, and whether the costs incurred are proportional to the final results that users will receive. The failure of implementing an information system in the public sector can be caused by project failure, system failure, or user failure (M & Howcroft, 2002).
Based on the opinions obtained from the examination of Regional Government Financial Reports (LKPD) in 2012-2016, city/regency governments' performance in West Java Province has increased. However, the LKPD also indicated that there were still some significant findings in several regions. One of them was about the implementation of the accounting information system with 2,156 total cases. The Summary of Semester Examination Results (IHPS) in the first and the second semesters of 2016 also revealed that one of the problems with the Internal Control System in local governments is accounting reporting controls' weakness. This is indicated by the entity's delay in submitting its financial reports and implementing an accounting information system that is not yet supported by sufficient human resources.
Furthermore, the examination results by the Supreme Audit Agency of the Republic of Indonesia (BPK RI) in the first semester of 2017 throughout Java Island showed that West Java Province was in first place for the inadequate report and accounting information systems with 31 cases. Thus, it can be concluded that the reporting of local government financial information in Indonesia is still far from the quality characteristics required in Government Regulation No. 71/2010. Various IT applications have been designed to support both individual performance enhancement and public service improvement. For example, the government created SIMDA and SIPKD, a regional financial management information system, to improve the government sector's financial reporting quality. Jaafreh (2017) said that user satisfaction could be used to assess the success of system implementation. If the system user is satisfied, it means that the system implementation has successfully supported the user's performance.
This study used the modification model of DeLone & McLean (2003) and Gable et al. (2008). Delone & McLean argued that the success of information systems is multidimensional. They also stated that researchers do not need to use all information system success factors but combine Besides, Saputro, Budiyanto, & Santoso (2015) stated that the Delone & McLean model could measure the success of e-government in Pekalongan City. This study contributes to the literature by examining how the three success factors of information systems apply to the local government sector. This study provides a new perspective in seeing system user satisfaction from an empirical model's explanatory perspective. After considered various aspects, it can be concluded that the measure of user satisfaction in information systems is reflected by the quality of the information system, information, and service. This study aimed to determine how much those variable's quality affects the user satisfaction of the SIMDA/SIPKD. This article introduces system quality, accounting information quality, service quality, and its relationship to application user satisfaction. The second part is an explanation of the theoretical framework and hypothesis development. Next is the methodology, results, and research discussion. In the last section, conclusions, limitations, and suggestions will be discussed.

The Quality of Information Systems
The quality of the system significantly influences the decision made by top management. In this research, system quality refers to the quality of an accounting system. The focus is on the system's performance, which refers to how well the hardware, software, policies, and procedures of an information system can provide the information needed (DeLone & McLean, 2003). There is a positive relationship between system quality and decision-making. The success of information systems is expected to overcome the failures in an organization because the existing systems will influence the decisions made by top management, so the better the quality of a system, the better the decision-making will be. That is why the success of a system is a critical thing for an organization.
System quality is a variety of constructs designed to determine how a system performs when viewed from an engineering and design perspective (Gable et al., 2008). The quality of information systems is a combination of hardware and software used in a system. The better the quality of a system and the output is given, the better the system performance. In measuring the quality of the system, indicators are needed to be achieved. Assessing system quality can be done by measuring several indicators such as ease of use, reliability, response time, flexibility, and security. This indicator has a crucial role because its quality is a latent variable that cannot be measured directly. Based on the description above, we propose the following hypothesis: H1: System quality has a positive effect on accounting information system user satisfaction.

The Quality of Accounting Information
DeLone & McLean (2003) stated that information quality reflects the quality of the information system output. Especially the output in the form of reports. Meanwhile, according to Gable et al. (2008), information quality is related to the quality of information generated by the system in the form of reports and on-screen displays. Information quality influences the success of a system. It has a significant positive impact on increasing decision-making capability, work effectivity, and work quality (DeLone & McLean, 2003). The better the quality of information results in more appropriate decisions. However, if the information produced is not of high quality, it will have a negative effect on user satisfaction (Istianingsih & Wijayanto, 2008). Jaafreh (2017) has tested the impact of information quality on information system user satisfaction, and the results show that information quality is positively related to information system end-user satisfaction. Five indicators can be used to measure the quality of information: completeness, information presentation, relevance, accuracy, and timeliness. Based on the description above, the researchers propose the following hypothesis: H2: Information quality has a positive effect on accounting information system user satisfaction.

The Quality of Service
Service quality is a quality standard that should be understood in providing a service. The quality of service is mostly determined by how big the gap is in comparing customer expectations and the perceived service they will receive in real terms. Service quality is seen from the consumers' perspective regarding its advantages and disadvantages and its services. Service quality would meet the customers' expectations if the service obtained is as satisfying as expected and vice versa. (2003) argued that a system service's quality directly affects the intended use and user satisfaction. They also mentioned that three components affect service quality; assurance quality provided by the system, system empathy, and system responsiveness to users' actions. Meanwhile, to measure service quality, five indicators can be used, including service reliability, responsiveness, assurance, empathy, and evidence of tangibles. Based on the description above, we propose the following hypothesis:

DeLone & McLean
H3: Service quality has a positive effect on user satisfaction accounting information systems.

User Satisfaction
Based on previous research (W. H. DeLone & McLean, 1992;Doll & Torkzadeh, 1998;Jawad, 1997), information system user satisfaction has a significant impact on the application systems' continuity of use. User satisfaction is a measure of the information systems' success because when the user is satisfied and feels the benefits of the application, the user will use it repeatedly.
According to DeLone & McLean (2003), user satisfaction is often used as a substitute measure for information systems' effectiveness. An effective system is defined as a system that provides value-added to the company. Information systems are expected to provide satisfaction then positively impact their users, such as increasing productivity, decision making, etc.
Information system user satisfaction can be assessed from the technological aspect by assessing the content, accuracy, format, time, and ease of use. Jaafreh (2017) states that user satisfaction is a feeling of pleasure or displeasure based on an information system's benefits. User satisfaction is the extent to which a system succeeds or fails to meet each user's aspirations. In measuring information system user satisfaction, four indicators can be used: adequacy, effectiveness, efficiency, and overall satisfaction (Jaafreh, 2017). These indicators can measure user perceptions about their satisfaction in interacting with information systems, such as meeting information needs, fulfilling efficiency, fulfilling effectiveness, and overall satisfaction. Based on the description above, we propose the following hypothesis: H4: System quality, accounting information quality, and service quality have a positive effect on user satisfaction in accounting information systems.

Success Model Information Systems
Every implementation of an information system in the entity is expected to be successfully implemented. However, the question arises "What are the parameters of an information system's success?" and "What's needed to be done to achieve that success?". Many previous studies were conducted to identify the success factors of information systems. Some of the well-known studies are the research conducted by DeLone & McLean (1992) and Gable et al. (2008).
DeLone and McLean developed a parsimony model called the DeLone and McLean IS Success Model, while Gable et al., Developed a new model called the "A Priori Model." "A Priori Model" is a measurement model to assess information systems' success using five dimensions: system quality, information quality, satisfaction, individual impact, and organizational impact. The study of the information system success model results from the DeLone and McLean model development, which focuses on the impact of information system implementation (Seddon & Kiew, 1996). Until now, this model's application is still often used as a tool to measure the success of information systems.

RESEARCH METHOD
There are two types of data used in this study, primary data, and secondary data. Primary data was obtained through direct observation, interviews, and distributing questionnaires to treasurers and SIMDA/SIPKD operators in city/regency governments in West Java Province. Meanwhile, secondary data were obtained through a literature study. After the data is obtained, the results will be described descriptively and analyzed to test the hypothesis proposed at the beginning of the study. This study systematically describes, analyses, and concludes the relationship between the collected data and existing theories.
The method used in this research is a descriptive-analytical method with a survey approach. The research model used is a modified multidimensional measurement model with interdependencies between success factors. In this study, the analysis unit was the city/regency governments in West Java Province. Thus, this study's population were users of the Accounting Information System software (SIMDA/SIPKD) in the city/regency governments throughout West Java Province. This study's sampling technique is saturated (census) because the sample used is the entire population. In each analysis unit, the observation/respondent unit studied were treasurers and SIMDA/SIPKD Operators at the Financial and Asset Management Agency (BPKA), which generated 28 analysis units or 84 respondents.
In understanding the characteristics of each research variable, certain techniques are needed to produce empirical indicators. This step begins by breaking down each variable into an operational definition by observing and measuring the variable under study. In this study, variables are classified into independent variables and dependent variables. The variables used are accounting information system user satisfaction (Y) as the dependent variable, accounting information system quality (X1) as the independent variable, accounting information quality (X2) as the independent variable, and service quality (X3) as the independent variable. In clarifying each variable used, all of these variables are translated into several questions using an ordinal measurement scale.
Furthermore, data analysis was carried out using multiple regression analysis through the SPSS program to obtain a comprehensive picture of the relationship between the independent and dependent variables. In ensuring that the model meets the assumptions of data normality and is free from classical statistical assumptions, data testing was conducted consisting of validity tests, reliability tests, normality tests, and classical assumptions tests. Subsequently, then we analyzed the data using descriptive analysis and hypothesis testing. The descriptive analysis method was employed to systematically obtain descriptions and data relating to the quality of accounting information systems, the quality of accounting information, service quality, and user satisfaction. So that the data processed and presented is accurate and can be justified for its truth.
In contrast to qualitative data, all statistical calculations and analyses are carried out using the SPSS program. The analysis results from data processing were then used to test the hypothesis. In the final stage, conclusions based on the hypothesis testing were laid down.

Respondent Characteristics
In this study, 84 questionnaires were distributed, but only 80 questionnaires were returned (95%), indicating a good response rate. The remaining four unanswered questions could be due to unavailable relevant data. Based on the results of data collection, it is known that 48 people use SIMDA (60%) and 32 people use SIKPD (40%). Based on gender, most respondents were male, as many as 41 people (51.28%), and the rest were women as many as 39 people (48.75%). Most of the respondents' ages are between 36 to 40 years, as many as 36 people (40%), and the least are respondents aged between 41 to 50 years as many as 21 people (26.25%). Based on the latest education data, most respondents with the latest education of D4/S1 were 50 people (62.5%). The least were the respondents whose last education was high school, as many as seven people (8.75%). Meanwhile, based on the working period, it is known that the majority have worked for 6 to 10 years as many as 33 people (41.25%) and the least is the respondent whose work period is one year to 5 years as many as 17 people (21.25%).

Results
To find out the relationship between system quality, information quality, and service quality on accounting information system user satisfaction at the City/Regency Government in West Java, the researchers conducted multiple linear regression analysis consisting of classical assumption tests, multiple linear regression equations, analysis of the coefficient of determination, and hypothesis testing by first testing the research instrument. Based on Figure 2, the above test shows the result related to user satisfaction (i.e., "Kepuasan Pengguna". It is known that the residual values (dots) obtained still follow the diagonal line, which indicates that the residual variable in the data is normally distributed. In other words, it can be said that the assumption of data normality is met. The next test is the multicollinearity test, which identifies the correlation between the independent variables. Based on the results in table x, it is known that the three variables have a tolerance value greater than 0.10 and a VIF value less than 10. These results indicate no multicollinearity problem between the independent variables in the model so that the regression model that will be formed fulfills one of the assumptions for regression testing.  Figure 3 shows the results of heteroscedasticity testing using the scatter plot graph method. From this figure, it can be seen that the points obtained formed an irregular random pattern spread above and below the zero (0) on the Y-axis so that in the regression model, there are no heteroscedasticity violations. In other words, the residual variance is homoscedasticity. a. The constant of -4.617 means that when the three dependent variables are zero (0), and there are no changes, user satisfaction is predicted to be worth -4.617.
b. X1 variable is the quality of information systems that has a regression coefficient value of 0.311. It indicates that if the quality of information systems is improved, it is predicted that user satisfaction will increase by 0.311 times. c. X2 variable is the quality of information that has a regression coefficient value of 0.209. It indicates that if the quality of information improves, it will increase user satisfaction by 0.209 times. d. X3 variable is service quality that has a regression coefficient value of 0,244. It shows that if the service quality is improved, it will increase user satisfaction as many 0,244 times as predicted. e. The coefficient of determination simultaneously using R square shows 76.6%. It means that the three independent variables tested contribute 76.6% to user satisfaction, while the remaining 23.4% describe the contribution of other variables that are not examined. f. The partial determination coefficient value shows that the quality of the information system (X1) gives the most dominant contribution to user satisfaction (Y) with a contribution of 39.1%, followed respectively by the quality of information (X2) of 16.4% and service quality (X3) of 21.1%.

Hypothesis Testing and Discussion
In determining the significant effect of information system quality, information quality, and service quality simultaneously on user satisfaction, the researchers conducted hypothesis testing. From the values in table 1.1, it can be seen that the F-value is 82.908, which is greater than the F-table value of 2.725. Following the hypothesis testing criteria, H0 is rejected, and Ha is accepted. This means that simultaneously the quality of information systems, information, and service has a significant positive effect on user satisfaction.
The positive influence shows that higher quality leads to higher user satisfaction of accounting information systems in Cities/Regencies throughout West Java Province. The positive influence of information system quality, information quality, and service quality on user satisfaction follows the descriptive analysis results, which shows that the average percentage value obtained from the quality of accounting information systems and service quality is 82.67% and 80.97%, respectively. This indicates that the quality of SIA and service quality at the City/Regency Government of West Java Province is considered good. Meanwhile, the average percentage value obtained from the quality of accounting information and user satisfaction is 85% and 86.53%, respectively, which indicates that the quality of accounting information and user satisfaction at city/regency governments in West Java Province is considered very good.
From table x, it can be seen that the value of T-value obtained is 6.020 > T-table value is 1.665. Following the hypothesis testing criteria, H0 is rejected, and Ha is accepted. It means the quality of information systems has a positive effect on user satisfaction partially.
Besides descriptive analysis based on the respondents' responses, there are findings in several cities/regencies that the main problem in using applications at city/regency governments in West Java Province is SIMDA application that is still computer-based and not connected to the web. Also, SIMDA is not made uniformly and should be adjusted to the needs of each local government. Unlike SIMDA, the SIPKD application has been designed as a web-based one to be accessed and used by other entities.
Based on the survey results, several aspects of the accounting information system implementation in West Java Province still score low, including integration. The survey results prove that several cities/regencies have not integrated the Regional Work Unit (SKPD) server with the Regional Financial and Development Supervisory Agency (BPKAD). They are West Bandung Regency, Indramayu Regency, Banjar Regency, Depok City, Bandung City, Subang City, and Purwakarta City. This shows that cities/regencies in West Java do not yet have fully integrated and consistent data/information in their financial information systems.
Although some aspects still score low, overall, the accounting information system's implementation by the city/regency government in West Java Province gets 82.67%. This shows that the quality of accounting information systems in city/regency governments in West Java Province is considered high. Most SIMDA/SIPKD application users are proficient in using the city/regency government system in West Java Province. This also shows that the city/regency government in West Java Province has paid attention to components in developing system quality based on the Gable et al. (2008) model, including data accessibility, data currency, database contents, ease of use, ease of use. Learning, access, user requirements, system features, system accuracy, flexibility, reliability, efficiency, sophistication, integration, customization. It can be concluded that the model development focuses on the impact of the implementation of information systems. The Gable model's implementation can be used to measure the impact of the accounting information system implementation.
The results of this study are in line with previous research conducted by DeLone & McLean (1992); Sirsat & Sirsat (2016); DeLone & McLean (2016); Jaafreh (2017) and Istianingsih & Wijayanto (2008), who concluded that there is a positive relationship between system quality and user satisfaction. Thus, it can be concluded that based on AIS users, the higher the quality of accounting software will increase user satisfaction with the software. However, this study's results are not in line with the research results by Zai & Dewi (2014), which concluded that system transformation, job transitions, and user interests could hinder the adaptation process of users to a new system. Likewise, the results of research by Tan, Suyatno, & Aliyah (2015) state that system quality does not affect user satisfaction.
For the quality of information, the T-value is 3.076, higher than the T-table value of 1.665. Following the hypothesis testing criteria, H0 is rejected, and Ha is accepted. This means that partially the quality of information has a positive effect on user satisfaction. Overall, the average percentage value obtained by the quality of accounting information is 85%. These results indicate that accounting information quality is considered very high, and the information generated helps users complete their work. City/regency governments already paid attention to develop accounting information quality, including importance, availability, usability, understandability, relevance, format, content accuracy, conciseness, timeliness, and uniqueness. Thus, system users can provide complete, accurate, and credible financial reports.
Information quality is the quality of output in the form of information generated by the information system used. The better the quality of information, the more precise the decisions will be made. If the information produced is not of high quality, it will have a negative effect on user satisfaction. Jaafreh (2017) has tested the impact of information quality on information system user satisfaction. Their test results indicate that information quality is positively related to information system user satisfaction.
This study's results are in line with previous research conducted by (Istianingsih & Wijayanto (2008), which provides empirical evidence that the quality of information significantly affects user satisfaction. Likewise, with Zai & Dewi's (2014) research, which states that this positive influence is caused by most system users being able to use the system's information, the system's information is beneficial for users to work. On the other hand, the research results of Tan, Suyatno, & Aliyah (2015) and Arifin & Pratolo (2012) state that the quality of information generated from SIKD does not have a significant positive effect on user satisfaction.
For service quality, the T-count value is 2.907, higher than the T-table value of 1.665. Following the hypothesis testing criteria, H0 is rejected, and Ha is accepted. This means that partially service quality has a positive effect on user satisfaction. This study's results were supported by 80 respondents' responses regarding the information quality variable statements given. The smallest percentage value obtained is 69% concerning the statement "application providers take the time to respond to user requests." Besides the descriptive analysis, the interviews' findings indicate that application providers at each Financial and Asset Management Agency (BPKA) in cities/regencies in West Java Province have not provided helpdesk, so there are difficulties for treasurers and application operators in delivering existing problems. However, as a whole, the average percentage value obtained for service quality is 80.97%. This shows that the quality of service to city/regency governments throughout West Java Province is considered good. The SIMDA and SIPKD application developers/providers have paid attention to developing the accounting information system services quality.
In their study, Prybutok, Zhang, & Ryan (2008) stated that the quality of service, system, and information influence user satisfaction. If the user of the information system feels that the quality of service provided by the accounting information system application provider is good, then he will tend to feel satisfied in using the system. This study's results are in line with the research of Rifanti et al. (2014), which affirms that service quality has a positive effect on user satisfaction. However, the results of this study are not in line with the results of Tan, Suyatno, & Aliyah's (2015) research because they said SIKD developers pay less attention to components in the system development such as flexibility of the system, integration of the system, response/turnaround time, convenience of access, recovery, and user friendly.

CONCLUSIONS
Based on the results of the research and hypothesis testing that has been done, we can conclude that partially the quality of the accounting information system, accounting information, and services has a significant positive effect on user satisfaction at the city/regency governments in West Java with each contribution of 39.1%, 16.4%, and 21.1%. After that, the three variables' quality simultaneously has a significant positive effect on user satisfaction with a contribution of 76.6%, while the remaining 23.4% is the contribution of other variables not examined.
Several limitations on this study were encountered. First, the data collection method is obtained using a questionnaire, so it is likely to be subjective. Second, regional coverage only focuses on government work units in West Java Province. Third, the only variables tested are system quality, information quality, and service quality, while many other factors can affect user satisfaction.
Several suggestions can be considered for city/regency government agencies, in particular West Java for improving the quality of information systems, accounting information, and services to increase user satisfaction. First, the government must pay more attention to the quality of their system and service to support SIMDA/SIPKD user satisfaction in using the application. It is because the contribution of user satisfaction is substantial in generating relevant and reliable financial reports. Second, improve the quality of data/information integration and consistency that will be used. The government should understand the importance of integration between applications because it will make BPKA easier to consolidate and reconcile the data/information. So, it is better to choose applications that are already integrated. Third, the application provider should pay more attention and respond to application users' interests to improve service quality. It would be better if the application provider and local government stated in the contract that it is necessary to provide a help desk when the application is put into use in each BPKA in the City/Regency government.
We also propose several suggestions. It is advisable for future research to conduct interviews to complement the information obtained through the questionnaire, expanding the coverage of respondents both from the area coverage and the type of work unit, and conducting research with different objects and indicators.