The students’ perception towards the audit using Audit Command Language (ACL) software

In line with the upsurge of computer application in companies in processing their information, auditor’s skill of computerized audit becomes very important. This study aimed to examine the effect of Perceived Usefulness and Perceived Ease of Use on the Computer Self- Efficacy. Samples were taken from a population of the ACL software users in Perbanas Surabaya. The data in this study was obtained through questionnaires. Of the 130 questionnaires distributed, 123 respondents were obtained. The data analysis that used in this study is the classical assumption test and multiple regressions. The result shows that perceived usefulness and perceived ease of use have no significant effect on computer self-efficacy, because there are other factors that affect the computer self-efficacy.


INTRODUCTION
Audit is a process which is used to evaluate the performance of an organization and to determine whether it has been in line with the plan and the existing productivity performance standards. The person in charge of the audit process is referred to as an auditor. In addition to audit the financial statement and the performance of an organization, there is also a special auditor who handles special object anyway. In this case, the auditor concerned should also have cetain qualifications of expertise, one of which is the ability to conduct the audit related to information technology.
Recently Indonesia still requires a lot of reliable auditors who are able to uphold the principles of a good auditor. This is related to the educational system and curriculum in Indonesia. The world of education today still rarely provides insight into how the profession of an auditor and even the need for auditors to have certain expertise, for example, the audit in these three areas.
In the on going education today, there are still many universities that have not applied or provided their teaching and training on audit in the specialized field such as information technology of audit. (Nieschwietz, Panyb, & Zhangb 2002) argued that it is important to incorporate existing information technology into the classroom of audit. It can reinforce students' understanding on the basic concepts of audit and prepare them to use the technology in the workplace.
The study that examines the Computer-Assisted Auditing Techniques (CAAT), especially in the application of audit using Audit Command Language (ACL) software, is rarely conducted. Many Public Accounting Firms have applied ACL in performing their audit, but there are also many junior auditors who do not understand or do not even know what ACL is. Therefore, this study is intended to find out what students' perceptions of Audit Command Language (ACL) software and whether the audit using computer software, especially ACL, is helpful for the students in understanding Computer-Assisted Auditing Techniques (CAAT). The aim of this study is to determine what the students' perceptions of the computer-assisted auditing, especially the audit using ACL software, and to develop the quality of education. At present there are still very few universities and entities that provide knowledge about the technology of audit using common audit software such as ACL. The audit using computers has been widely used in the world of work now, so it allows the researcher to determine what the students' perceptions of the audit using computer software such as ACL.

THEORETICAL FRAMEWORK AND HYPO-THESIS Technology Acceptance Model (TAM)
The model that describes the level of technology acceptance is called Technology Acceptance Model (TAM). The main purpose of TAM is to provide an explanation of the determination of computer acceptance in general and to provide an explanation of the behavior and attitude of the users in a population about the computer acceptance (Davis 1989). TAM has been modified using two main constructs, namely: 1. Perceived Usefulness According to (Hartono 2007), perceived usefulness is the extent to which a person believes that using a technology will improve his job performance. Perceived usefulness has an effect on the behavior of interest and perceived ease of use.

Perceived Ease of Use
According to (Hartono 2007), perceived ease of use is the extent to which a person believes that using a technology will be free from effort. Perceived ease of use has an effect on the behavior of interest. If a person feels confident that the information system is easy to use, then he will use it.

Perception
The perception of an individual is influenced by some factors that lead the individual to be able to give different responses when viewing something. According to (Jalaludin 2004), the factors that influence the perception are: 1. Functional factors, derived from the needs, past experiences, motivations, expectations and desires, attention, emotion and mood, and other things that are included in the personal factors. 2. Structural factors, derived from the nature of the physical stimuli and neural effects thereof on the nervous system of an individual. 3. Cultural factors, derived from the culture of a place where the individual grows and develops. Perception can also be influenced by factors of experience, learning process, horizon and knowledge of an individual. Factors of experience and learning process or socialization give shape and structure to what is seen. While factors of knowledge and horizon give meaning to the object which is observed.

Audit and Computer Audit
Audit is a critical and sytematic inspection or examination conducted by an independent party on the financial statements that have been prepared by management, as well as accounting records and supporting evidence, with the aim to be able to give an opinion about the fairness of the financial statements.
The audit procedure is related to information technology and is not much different from the general audit procedures. An audit using information technology, or commonly called as the audit using the Electronic Data Processing (EDP) requires the auditor to understand the concepts that exist in the EDP. Therefore, the auditors are not only required to master the science of the audit, but also demanded to master computer or computer programs that become the basis of EDP science.

Audit Command Language (ACL)
Audit Command Language (ACL) is one of the common audit software used. ACL greatly assists auditors in performing the audit. But today there are still very few auditors who have special insight into the ACL. This special insight is very important for special auditors who examine the object specifically anyway.

Definition of Computer Self-Efficacy
Computer self-efficacy can be interpreted as a person's ability to operate a computer which is sup-ported by an adequate intellectual ability, either obtained through innate talent or a way of learning. The computer self-efficacy is the user's abilities in terms of computer applications, computer operating systems, files and hardware handling, data storage and the use of keyboard keys (Indriantoro 2000).
Based on the social cognitive theory developed by (Bandura 1986), self-efficacy can be defined as the belief that one has the ability to perform a particular behavior. It shows that the key characteristics of self-efficacy are components of skill and ability in terms of organizing and carrying out an action.

Aspect of Behaviour in the Development of Information Technology
The user involvement is a very important factor for the development of IT. Organizational performance can be improved through the use of technology when it is supported by the skills in operating computer. Therefore, to achieve the improved performance requires the support of a variety of management instruments and the individual of computer users.
The important factor that contributes to IT acceptance is the aspect of user's attitudes (Igbaria 1989). Each individual will think positively for the presence of computer technology, if he perceives the usefulness of computer technology to improve performance and productivity. The perceived usefulness of the computer users apperars because of the ability of each individual in operating a computer, or skills of operation.

The Relationship between Perception and Audit Command Language (ACL)
The individual's perception of the use of information technology, especially the Audit Command Language (ACL), has an influence on the behavior of interest. If the technology user perceives that the technology system he uses is useful and easy, he will have a high interest to use this technology. Similarly, the perceived ease of use also has an influence on the behavior of interest. Despite the effect on behavior of interest, perception can also affect perceived ease of use. The user will use the system if it is useful, even though it is difficult to use. The difficult system will still be used if the user perceives that the system is useful. H1 : The effect of perceived usefulness of the students in performing audit using Audit Command Language (ACL) software on the computer selfefficacy.
H2 : The effect of perceived ease of use of the students in performing audit using Audit Command Language (ACL) software on the computer selfefficacy.

RESEARCH METHOD
This research applies quantitative methods using primary data obtained directly from the object of the research through questionnaires.
The variables used in this research are the computer self-efficacy in using Audit Command Language (ACL) and the students' perceptions. The dependent variable in this research is the computer self-efficacy in using Audit Command Language (ACL) software. Audit Command Language (ACL) is software commonly used in audit. While, selfefficacy is an induvidual's ability to use computer softwares, one of them is ACL.
The independent variable used in this research is the perception of students. Perception is one's response to the use of computers and computer programs reception.
The measurement of the data used is a fivepoint Likert scale. The Likert scale which is used in this research is ranging from 1 meaning 'strongly disagree', 2 meaning 'disagree', 3 meaning 'neutral', 4 meaning 'agree' and 5 meaning 'strongly agree'. The measurement of variables in this research is different from that of the previous research. The previous research uses a 11-point Likert scale, while this research uses a 5-point Likert scale.
Hypothesis testing is performed by using regression analysis. Regression analysis is used to examine the relationship between the two variables. This research uses regression analysis to examine the effect between the dependent variable and the independent variable. The regression model in this research is formulated as follows: Y =ß 0 + ß 1 X 1 + ß 2 X 2 + e. (1)

DATA ANALYSIS AND DISCUSSION
This study involves undergraduate students of STIE Perbanas Surabaya who administer the subject of Electronic Data Processing -Accounting Information Systems (PDE-SIA). The data obtained through the questionnaires and the results of this data collection will be analyzed in accordance with the needs of the research. The questionnaires distributed are as many as 130 questionnaires. From the results of the survey, the total number of questionnaires distributed is 130 questionnaires, but only 128 questionnaires are returned. After the initial screening, it is known that 5 respondents are categorized not feasible, because the answers in the questionnaire are found incomplete.

Descriptive Analysis
Based on the data obtained and processed, it can be seen that no respondent who filled in a questionnaire with the same NIM, each questionnaire is filled by different respondents for a total of 130 respondents. There are two questionnaires which are not returned, and 5 respondents who do not complete the questionnaires. So there are 123 questionnaires eligible to be processed. Based on the results of respondents' responses, the identity of the respondents can be explained below. The characteristics of the respondents are identified by gender. Based on the responses collected, the results of the identification of the characteristics of respondents are presented in Table 1.
Based on Table 1, it is known that the majority of the respondent gender is females, as many as 105 people or 85.4%, while male is as many as 18 people or 14.6% of the total respondents of 123 people.

The Analysis of the Respondents' Responses to the Perceived Usefulness Variable
The measurement of perceived usefulness variable consists of nineteen items of question with the measurement scale from 1 to 5. In this variable there is no unfavorable valuation so that there are no reversed scores of each respondent's answer. Table 2 is the result of the respondents' responses to each question item of perceived usefulness variable.
Based on the results of the questionnaire responses in Table 2, it can be seen that the respondents' answers to the perceived usefulness are above the theoretical average with a mean value of 3.9662 and a standard deviation of 0.40618 with the tendency of respondents' responses spread into five categories, where the high level of response to the perceived usefulness is included in the category of 'agree'. It can be seen through the mean value of the whole question of 3.9662. Thus the respondents perceive that in general they state "agree" that the Audit Command Language (ACL) software is useful for the audit capabilities possessed, making it  easier to do the audit work.

The Analysis of the Respondents' Responses to Perceived Ease of Use Variable
The measuremant of perceived ease of use variable consists of seventeen items of question with the measurement scale ranging from 1 to 5. At this variable there is unfavorable valuation so that it enables this variable to have two valuations; the favorable valuation and unfavorable valuable. As for the unfavorable question, the data is entered by way of reversing the score of each respondent's answer. Table 3 shows the results of respondents' responses to each question item of Perceived Ease of Use variable: Based on the results of the respondents' responses in Table 3, it can be seen that the total of average response is 3.2678 and a standard deviation of 0.50485 with the tendency of respondents' responses spread into five categories. Where, the level of response to the perceived ease of use is included in the category of "doubtful", viewed from the indication of the mean value of all questions 3.2678. Thus the respondents perceive that in general they state "doubtful" that the Audit Command Language (ACL) software is easy to use according to the expertise possessed.

The analysis of the Respondents' Responses to Computer Self-Efficacy Variable
The measurement of computer self-efficacy variable consists of nineteen items of question with measurement scale ranging from 1 to 5. At this variable there is no unfavorable valuation, so no reversed scores of each respondent's answer. Table 4 shows the results of the respondents' responses to each question item on computer self-efficacy variable.
Based on the results of the respondents' responses in Table 4, it is noted that the total of average response is 3.1065 and a standard deviation of 0.70044 with the tendency of respondents' responses spread into five categories. Where, the response rate to the computer self-efficacy is included in the catagory of "doubtful", viewed from the indication of mean value of overal questions of 3.1065. Thus the respondents perceive that in general they state "doubtful" on their expertise on Audit Command Language (ACL) software.

Validity Test
In this study, the validity test is cunducted by correlating scores on the questions whose validity is measured by the total score of all questions in one variable. If the correlation result obtained shows significant results (sig <0.05), the point of the question is valid.
The results of the study sample validity test on the perceived usefulnes variable show that of all question items on this variable have significant value of Pearson Correlation (r) 2 tailed below 0.05 (5%). Based on the indication, the question items are valid and all items will be retained for further testing processes.
The results of the study sample validity test on the perceived ease of use variable show that the entire question items on this variable have a signifi- cant value of Pearson Correlation (r) 2 tailed below 0.05 (5%). Based on the indication, the question items are valid and all items will be retained for further testing processes. The results of the study sample validity test on the computer self-efficacy variable show that of all question items on this variable have a significant value of Pearson Correlation (r) 2 tailed below 0.05 (5%). Based on the indication, the question items are valid and all items will be retained for further testing process.

Reliability Test
Based on Table 5, it can be seen that the magnitude of Cronbach's alpha on all variables is greater than 0.07 (> 0.07), this means that all of the study variables are considered reliable so that they can be used in research for further analysis.

Normality Test
In testing the normality, the researcher uses nonparametric statistical test Kolmogorov-Smirnov (KS). Table 6 shows the results of non-parametric statistical test with the Kolmogorov-Smirnov (KS).
Based on the result of normality test, the Kol-mogorov-Smirnov value is 1.049 and significant value of 0.221. This means that the residual data is normally distributed. The data is said to be normally distributed when the result of Asymp. Sig.
(2-tailed) is above the confidence level of 0.05. This means that the regression model of this study indicates that the confounding variable is normally distributed.

Regression Analysis
This analysis is used to determine the effect of perceived usefulness and perceived ease of use on the computer self-efficacy. The formulation used in the linear regression is: (2) The initial step in proving the hypothesis that has been made is to do a regression on the model. Table 7 shows the test result of regression model that has been developed.

Coefficient Determination
Based on Table 7, it is known that the correlation of the dependent variable is 22.1% (R = 0.221 a ). It indicates that the correlation is at the weak stage. Then the value of R 2 in the research model is 0.049 or

F Test
Based on the result of multiple linear regression analysis listed in Table 7, to test the Goodness of Fit of a model, it can be concluded that the significant value of F is 0.05 (Sig F = 0.05 ≤ 0.05). It can be concluded that the regression model in this study is 'Fit'.

T Test
Based on Table 7, it can be seen the relationship of each variable. So based on the hypothesis that has been made, it can be described as follows: 1. Perceived usefulness variable. Based on the calculation, it is obtained β = 0.202 with a significance level (sig t = 0.253> 0.05), so H0 is accepted and H1 is rejected at the significance level of 5%. It can be concluded that perceived usefulness does not affect the computere self efficacy. 2. Perceived ease of use variable. Based on the calculation, it is obtained β = 0.192 with a significance level (sig t = 0.177> 0.05), so H0 is accepted and H1 is rejected at the significance level of 5%. It can be concluded that perceived ease of use does not affect the computer self efficacy.

CONCLUSION, IMPLICATION, SUGGES-TION, AND LIMITATIONS
The perceived usefulness and perceived ease of use have no effect on computer self efficacy. This result  might be caused by students' perceptions of the usefulness of Audit Command Language (ACL) software which is influenced by other factors that are more important. It can be seen from the perception based on social factors, such as attitudes, interests and behavior. The perceived ease of use has no effect because it might be caused by the respondents who are not confident with their own expertise, so they think that ACL software is difficult and confusing. The recent finding in the research on ACL shows that there is no relationship between perceived usefulness and perceived ease of use. It might appear because most respondents are still doubtful with their own expertise. So the result indicates that perceived usefulness and perceived ease of use have no relationship with the computer self efficacy. There are many other factors that influence the expertise, in addition to be viewed from the perceived usefulness and perceived ease of use, such as interests, experiences, needs, memory, mood, and attention. Another thing that affects the outcome of this study is the sample used. Because the TAM research with the main constructs of perceived usefulness and perceived ease of use cannot be used on students. As described in the study of Darsono (2005) in (Hartono 2007), that the TAM research requires professional samples in the sense that they already understand the implications of the shoftware they use. Those who are not professional still need to learn and do not fully understand the usefulness and importance of the software. It will surely affect the results of the research conducted.
The limitation of this study can be seen from the results of Goodness of Fit performed which indicates weak result. Since the significance value of F test is exactly at 0.05, although it is still acceptable but considering its condition, it can be concluded that the Goodness of Fit is in a weak condition. Another weakness of this study is the frame-works that only test the effect of perceived usefulness and perceived ease of use on the computer self-efficacy, while there are still other external factors that influence it, not just the perceived usefulness and perceived ease of use.
Suggestion for future research is to add other variables in perception, such as the perception from the viewpoint of internal and external factors or even other factors underlying TAM. In addition, future studies are suggested to test the respondents who have been using the software for a long time, so it can determine whether the software is useful or not.