Abstract

This research has a purpose to examine the factors that influence the success of the accounting information system implementation in private universities (PU) in Bali. This is to measure the extent of success of the accounting software or application in terms of their usage in 8 PU in Bali. The data based on the source is primary data, collected using questionnaires which were directly distributed to the respondents. The sample were selected using the purposive sampling technique. The total research respondents are 55 people. The data analysis technique used is the Partial Least Square (PLS) with the help of the SmarPLS 3.0. program at a 5 percent level of significance. The research results state that system quality, information quality and the importance of the system have an influence on the usage and satisfaction of AIS users. In this research, the empirical evidence attained showed that system usage and system satisfaction have an influence on the net benefit gained by the private universities in Bali.

Introduction

An improvement in this research is the modification of ( Seddon & Kiew, 1996) and ( Delone & McLean, 2003) information system success model, bringing up the internal conflict between the foundation and the management of the PU which arise due to the financial transaction management being not transparent and a poorly executed administration. This occurs because the PU do not use an accounting information system. ( Seddon & Kiew, 1996) stated that if the system user assumes that an information system is important, it shows that there are benefits that can be attained from the system. The benefit attained from the usage of the system will influence the satisfaction level of system users. This is in line with the characteristics of a PU’s accounting information system. The characteristics are accounting information system is only used for the interest of one party. Thus, the researcher attempts to evaluate this by replacing the service quality variable with system importance ( Latifa, 2011). Afterwards, the researcher also used the usage variable because the researcher wants to assess how are the behavior of users towards the accounting information system (AIS) ( Purwanto & Suharno, 2017).

Information system-related studies have been previously conducted by several researchers, such as ( Rai, Lang, & Welker, 2002), ( Livari, 2005), ( Radityo & Zulaikha, 2007), ( Istianingsih & Wijanto, 2008), ( Wang & Liao, 2008), (Al-adaileh, 2009), ( Budiyanto, 2009), ( Darmawan, 2010), ( Sudarmadi, 2010), ( Gowinda, 2010), ( Wahyuni, 2011), (Urbach & Müller, 2012), (Al-Hiyari, Al-Mashregy, & Mat, 2013), ( Groho, Winarno, & Permanasari, 2014), ( Tan, Suyatno, & Aliyah, 2015), and ( Noviyanti, 2017).

The research model utilized in this research is as follows:

Figure 1. Research Model

An improvement in the system quality will result in the increase in system usage. The research results by ( Tan et al., 2015) became an empirical evidence which showed that system quality significantly influence the usage of a system. This is consistent with the research by ( Wahyuni, 2011), ( Budiyanto, 2009), and ( Darmawan, 2010). Based on the research results of previous studies, the first hypothesis is as follows:

H1: System quality has an influence on AIS use in PU.

The better the system and system output quality provided, for example by improving the speed of access and system output usage, the less reluctant the users are to reuse it. Thus, the system usage intensity will increase, vice versa. DeLone and McLean stated that system quality is influenced by the satisfaction of users. If the system has characteristics that are in line with their expectations, the usage will increase the satisfaction of users. This research result is consistent with the research conducted by ( DeLone & Mclean, 1992; 2003), ( Budiyanto, 2009), ( Bal, et al., 2012), (Al-Khower, et al., 2016), (Al Athmay., et al. (2016) and ( Azwar, Amriani, & Subekan, 2016). Consistent with this explanation, the second hypothesis developed is as follows:

H2: System quality has an influence on the users satisfaction of AIS in PU.

In their study, DeLone and McLean revealed that information quality focuses on the alignment of the product or result of the information system (output) with the expected characteristics. Mason ( 1978) stated that system quality and information quality determines the attitude of system users as the information receiver. Several studies, which were conducted by ( DeLone and McLean, 1992; 2003), ( Wahyuni, 2011), ( Halawi et al., 2007), and ( Rai et al., 2002), found that a high information quality will have an influence on the increase in system usage.

H3: Information quality has an influence on the AIS use in PU.

The influence of information usage (output) will influence the satisfaction of users. ( DeLone & McLean, 1992), ( Budiyanto, 2009), ( Radityo & Zulaikha, 2007), Darmawan ( 2010), (Urbach & Müller, 2012), (Al-Khower, et al., 2016) and ( Al Athmay, et al., 2016) stated that an improvement in information quality will increase the satisfaction of users. In his study, ( Sudarmadi, 2010) stated the opposite, in which system quality does not have a significant influence on the satisfaction of information system users.

H4: Information quality has an influence on the users satisfaction of AIS in PU.

If the system user perceives that the information system quality is good, the perceived benefit will be high ( Seddon & Kiew, 1996). This is also consistent with the research by ( Darmawan, 2010). The system importance is an important element of the success measure of an information system. The system importance shows the level of dependency of the user towards the system. An opinion regarding the importance of the system is the result of various positive factors felt which can assist the users in their job. If the system is not important, it will not have an impact on the work of its users.

H5: System importance has an influence on the use of AIS in PU.

( Li, 1997) stated that if the system does not disturb the work of the system users, the system is important ( Sudarmadi, 2010). The perception regarding the importance of the system as a predictor of usefulness and user satisfaction is based on the empowerment and user involvement aspect in a system. The importance of a system is related to the information quality attained as a part of the relevant system to make decisions. This is consistent with the research by ( Seddon & Kiew, 1996), ( Darmawan, 2010), ( Triono, et al., 2013), and ( Laksana, Subroto, & Baridwan, 2017).

H6: System importance has an influence on the users satisfaction of AIS in PU.

Repeated usage would imply that the usage has a benefit for the user. The high degree of benefit attained would result in users becoming more satisfied. ( Delone & McLean, 2003) stated that the relationship between usage and user satisfaction is a reciprocal relationship. This research result is consistent with the study by ( Istianingsih & Wijanto, 2008), ( Floropoulos, Spathis, Halvatzis, & Tsipouridou, 2010), ( Jiang and Ji, 2015), ( Muharor, Busaini, & Fitriah, 2015) , and ( Laksana et al., 2017). Based on the results of previous studies and available theories, the hypothesis is formulated as follows:

H7: The use has an influence on the users satisfaction of AIS in PU.

Shannon and Weaver ( 1949) stated that in the theory of communication, effectivity level focuses on the influence of information on the information receiver. Effectivity level was further divided by Mason ( 1978) into several parts. This result is in line with the statement by ( Livari, 2005), ( Hussein, Karim, & Selamat, 2007), ( Muliono, 2006), and ( Latifa, 2011) the increase in user satisfaction will cause the intensity of usage to be higher. Based on the results of previous studies and the available theories, the hypothesis is formulated as follows:

H8: The users satisfaction has an influence on the use of AIS in PU.

( Noviyanti, 2017)stated that as users use the information system more frequently, there will be a higher level of learning the information system. This is consistent with the research by ( Mastan and Winarno, 2013), ( Salim, 2014), ( Tan et al., 2015), and ( Wang & Liao, 2008). Based on the results of several studies and the available theories, the hypothesis is formulated as follows:

H9: The system use has an influence on the net benefit of the AIS in PU.

( Delone & McLean, 2003) stated that if the system is more frequently used, it will cause the net benefit of a system to be more easily achieved. If the system implemented is able to satisfy the users, it will also cause the net benefit of a system to be more easily achieved. Their research is consistent with ( Purwanto & Suharno, 2017), ( Mulyono, 2008), ( Groho et al., 2014), and ( Tan et al., 2015). Based on the results of several previous studies and theories, the hypothesis is formulated as follows:

H10 : User satisfaction has an influence on the net benefit of AIS in PU.

Research Method

There are 60 private universities with an active status in Bali. The population in this research are eight private universities that participated in the AIS training which was held by KOPERTIS VIII of the private universities on the 21st of March 2017. They have attended and stated that they have an accounting information system which is integrated with the private universities’ academic system. Sampling in this research was conducted using the Purposive sampling technique. The sample selected are 10 employees as the respondents from each private university. The sampled employees comprise the financial bureau staffs, financial system operator, and the financial statement users from the eight private universities, with a total of 80 people. In distributing the questionnaires, there were 20 employees who had double positions, thus, the total number of respondents is 60 people. After the questionnaires were returned, there was 55 processable questionnaires, while the remaining 5 questionnaires which were not processable. The model is analyzed utilizing the SmartPLS 3.0 program at a 5 percent significant level. In the test using the Partial Least Square (PLS) variance, the inner and outer model test is conducted. The inner model is the relationship between the variables in the research model. The model developed identifies the research variables. The indicators of each variable can be seen in Table 1.

Source:Computed Data, 2018

Discussion of the Research Results

There were 60 questionnaires distributed to the sampled private universities. 5 questionnaires were not returned, thus, there are 55

questionnaires collected and computed. The descriptive statistics results in this research can be seen in Table 2. as follows:

Source: computed data ( 2018)

Based on Table 2, the information attained regarding the description of each research variable consists of the minimum value, maximum value, average value, standard deviation and the three boxes average method, with a total sample of 55. For the system quality (KS) variable, information quality variable (KI), the importance of the system (PS), Usage (P) variable, User Satisfaction variable (KP), and the Net Benefit (NB) variable, the respondents assessed that each AIS utilization

variable, using the three box method, is categorized as high.

The inner model or the structural model evaluation is conducted by assessing the relationship between variables, significance value, and R-square of the research model. The inner model test results is able to assess the relationship between constructs by comparing the significance value with the R-Square of the research model (Ghozali, 2015:42). The structural model is shown in Table 4.

Source: Computed Data, 2018

The higher the R-Square value is, the greater the ability of the exogenous variable in explaining the endogenous variable, and the better the structural equation. Evaluation is subsequently conducted to calculate the Q-square predictive relevance, as follows:

Q model 1 = 1- (1- Ry1 )(1- Ry2 ) (1- Ry3 ) = 1-(1- 0.723 ) (1- 0.762 )(1- 0.641 ) = 1- (0.124) = 0.875

The Q2 value is above zero which means that the model designed has predictive relevance ( Ghozali, 2015).

Source: computed data, 2018

It can be seen from Table 5. that the hypothesis test is conducted by using the resampling bootstrap method which was developed by Geiser dan Stone. The t-statistics value of the inner weight output result is compared with the P-Value with a significance level of 5 percent. System quality has an influence on the use and user satisfaction

The test results provide empirical evidence that H1 is supported with a positive coefficient. This means that with a higher system quality, there will be an increase in the usage of the system. The implementation of AIS can increase the intention of users to use and recommend the AIS for the accounting process of PU in Bali. This research result is consistent with previous studies conducted by ( Budiyanto, 2009), ( Darmawan, 2010), ( Tan et al., 2015), and ( Wahyuni, 2011).

The test results provide empirical evidence which showed that H2 is supported with a positive coefficient. This means that system quality has an influence on the satisfaction of users, because the system has characteristics which meets user expectations and the usage increases the satisfaction of users. This research result is consistent with the research conducted by (DeLone & McLean, 1992:2003), ( Budiyanto, 2009), ( Bal, et al., 2012), (Al-Khower, et al., 2016), ( Al Athmay, et al., 2016) and ( Azwar et al., 2016).

Empirically, the system has given satisfaction to the AIS users and operators in performing their tasks. However, the features in AIS can be developed to minimize repeated posting process.

Source: Computed data 2018

The quality construct is measured using 4 indicators, namely system Flexibility X1.1, system Integration X1. 2, comfortability in access X1. 3, and Language X1. 4 with the three box scale method. Table 6. shows that a number of respondents gave good (4) and very good ( 5) for their response. However, more than 20% of the respondents gave neutral ( 3) and bad ( 2) for item 1, namely the AIS application you use is flexible and is able to facilitate the changes needed to be related with the institution’s accounting (such as updating database, username and password). Furthermore, less than one percent of the respondents gave very bad for their response, specifically for item 2. Comparing the score value of each indicator in Table 6, indicator X1.4 is perceived as the best indicator with a score of 222. On the other hand, two indicators X1.1, X1.2 are perceived by respondents as the worst with a score of 215. This shows that if the respondents are able to easily understand the accounting language used in the AIS, along with the increase in AIS quality or the AIS is developed in line with the needs of the PU, the users will have a greater intention to use the AIS. An increasing AIS usage will have an influence on the satisfaction of users that feel AIS provides economic benefits.

Information quality has an influence on the use and user

Satisfaction.

The test results provide empirical evidence which shows that H3 is supported. This means that information quality has a significant influence on AIS users. This empirical evidence has an implication in which with the high information quality, system users feel that the output or accounting report of the institution still must be developed, so that it can be easily understood and is in line with the institution’s needs. This empirical evidence is consistent with the research by (DeLone & McLean, 1992: 2003), ( Wahyuni, 2011), ( Halawi et al., 2007), and ( Rai et al., 2002).

The test results provide empirical evidence that H4 is supported by a positive coefficient. This means that the increase in information quality will influence the satisfaction of users. In other words, information quality has a positive relationship with the satisfaction of its users. The high quality of information produced by system users that are satisfied with the AIS application is able to provide accurate information for the institution’s accounting process. This research result is consistent with the research conducted by ( DeLone & McLean, 1992), ( Budiyanto, 2009), ( Radityo & Zulaikha, 2007), ( Darmawan, 2010), (Urbach & Müller, 2012), (Al-Khower, et al., 2016) and (Al Athmay, et al., 2016.

Source: Computed data 2018

The quality construct is measured using 4 indicators, namely completeness X 2.1, Accuracy X2.2, Reliability X2. 3, and Reliability X 2. 4, with the three box method. Table 7. shows that a portion of the respondents gave good (4) and very good ( 5) for their response, while around 16-19 percent of the respondents gave neutral ( 3) or bad ( 2) for item 1, which states that AIS provide complete information in the institution’s accounting process. From the comparison of the score value for each indicator in Table 7, indicator X2.2 is perceived by respondents to be the best indicator with a score of 219, while two other indicators X2.1, X2.3 is perceived by respondents as the worse with a score of 217. This shows that respondents highly prefer information quality that is more accurate, in line with the financial statement characteristics, namely easily understandable, relevant, reliable, and comparable. Thus, information quality also influences the satisfaction of users, especially for those that feel that the AIS in the PU produces accurate information.

3. System importance has an influence on use and user satisfaction

The test results provide empirical evidence that H5 is supported with a positive coefficient. This means that an increase in system importance influences the usage. The implication of the empirical evidence is that if the system users perceive that the information system quality is good, the usefulness of the system will be high. The system usage has an influence on the dependency level of the system user. This research result is consistent with the research conducted by ( Darmawan, 2010), and ( Seddon & Kiew, 1996).

The test results provide empirical evidence which shows that H6 is supported with a positive coefficient. This means that with a higher system importance, there will be an influence on the user satisfaction. The empirical evidence has implications in which an increase in the system importance will have an influence on the level of user satisfaction. The benefit attained by operators and the users of the AIS will increase their satisfaction and the dependency of the user towards the system. This research result is consistent with the studies by ( Seddon & Kiew, 1996), ( Armstrong, et al., 2005), ( Darmawan, 2010), ( Triono, et al., 2013), and ( Laksana et al., 2017) which stated that system importance has a positive effect on the satisfaction of the users.

Source: Computed data, 2018

The system importance construct is measured with 5 indicators, namely, sense of ownership X 3.1, interested to work on it X3. 2, develop abilities X3.3, enhances confidence X3. 4, and important to be used X 3. 5, using the three box method. Table 8. shows that some of the respondents gave good ( 4) and very good ( 5) as their response. However, around 18- 2 3 percent of the respondents gave a neutral (3) and less than 1 percent gave bad ( 2) as their response for item 1 and item 3 which states that “I have a sense of ownership towards the institution’s AIS” and “by using the AIS application, I can develop my abilities in completing my report”. Assessed from the comparison of the score value of each indicator in Table 8., the indicator X3.5 is the indicator perceived by respondents as the best with a score of 231, while two other indicators X3.1, X3.3 are perceived by respondents as the worst with a score of 216. This shows that respondents perceive that system importance has an influence on the usage of AIS, which also influences the satisfaction of the users.

4. Use has an influence on the users satisfaction

The test results provide empirical evidence that H7 is supported with a positive coefficient. This shows that usage has an influence on the satisfaction of users. The empirical evidence has implications in which an increase in usage entails an increase in the satisfaction of users. AIS can be useful for its users, and users will feel satisfied, as it is expected to be. This research result is in line with Delone and McLean’s theory of information system success which states that an increase in information system usage will increase the satisfaction of users. This research result is consistent with the study conducted by ( Istianingsih & Wijanto, 2008), ( Floropoulos et al., 2010), ( Jiang and Ji, 2015), ( Muharor et al., 2015), and Laksana, et al. ( ( 2017).

5. User satisfaction has an influence on use

The research results provide empirical evidence which shows that H8 is supported with a positive coefficient. This means that with a higher level of user satisfaction, there will be a higher level of usage. This empirical evidence has an implication in which an increase in user satisfaction will influence the usage, the more the AIS user feels that they gain decent benefit from the usage of the system in finishing their work, the less reluctant the user will be in reusing. Thus, the usage intensity of the system will increase. This result is in line with what was stated by ( Livari, 2005), ( Hussein et al., 2007), ( Muliono, 2006), and ( Latifa, 2011), who also attained results which showed that an increase in user satisfaction is followed by the increase in usage intensity

Source: Computed data, 2018

The construct is measured using three indicators, namely, usage frequency Y1.1, system selection Y1. 2, and time duration of usage Y1. 3, with the three box method. Table 9 shows that a portion of the respondents gave good ( 4) and very good ( 5) as their response. However, around 17- 28 percent of the respondents gave neutral ( 3) as their response and more than 3 percent gave bad ( 2) for

their response to item 2, system usage, and item

3, time duration of usage.

Viewed from the score value comparison of each indicator in Table 9, indicator Y1.2 is perceived by respondents as the best indicator with a score of 217, while indicator Y1.1 is perceived by respondents as the worst with a score of 213. This also shows that respondents perceive that usage is highly dependant on how much the users want to recommend AIS which has a positive influence on the satisfaction of users.

Source: Computed data 2018

The construct is measured using 2 indicators, namely satisfaction towards the system and information quality Y 2.1, satisfaction towards the system facilities and features Y2.2 with the three box scale method. It is shown in Table 10 that a number of respondents gave good ( 4) and very good ( 5) as their response. However, there were around 21-22 percent of the respondents that gave a neutral ( 3) assessment and more than 4 percent gave a bad ( 2) grade for response on item 2, the satisfaction towards the facilities and features of the system. Based on the score value comparison of the indicators in Table 10., indicator Y2.1 is perceived by respondents as the best indicator with a score of 211, while indicator Y2.2 is perceived by the respondents as the worst variable with a score of 210. In assessing user satisfaction, respondents highly focus on their satisfaction towards the system and information quality which produces relevant accounting reports. AIS user satisfaction also influences the intention of users to reuse the AIS.

6. Use has an influence on net benefit (SI)

The test results provide empirical evidence which shows that H9¬ is supported with a positive coefficient. This means that system usage has an influence on the net benefit. The existence of AIS becomes a positive stimulus and challenge for individuals in the organization to work better, which in turn will affect the organizational performance. This reaction can be in the form of new motivation to compete and increase performance. ( Mulyono, 2008) proved that usage has an influence on individuals. A high frequency of information system usage by users is usually followed by the increase in degree of learning the information syst