The Impact of Big Data Analytics on Audit Procedures: Evidence from the Middle East

Abstract

The goal of this study was to see how big data analytics (BDA) affected external audit procedures in the Middle East. The measurement model and structural model of this investigation were evaluated using PLS-SEM (3.3.3). The study sample members were (361) auditors who work in auditing companies in Kuwait, Saudi Arabia, the United Arab Emirates, Jordan, Bahrain, Egypt, Lebanon, and Iraq. A questionnaire was chosen to the study sample members electronically, and the study sample members were (5093) auditors who work in auditing companies in Kuwait, Saudi Arabia, the United Arab Emirates, Jordan, Bahrain, Egypt, Lebanon, and Iraq. To choose the sample, the researchers used a stratified random sampling procedure. The findings show that BDA has an impact on audit procedures at all phases of the auditing process, where it contributes to information delivery that helps auditors understand the client's internal and external environments, which in turn influences the choice to accept the audit assignment. Furthermore, by providing essential information, BDA enables auditors to simply run analytical procedures, estimate client risks, and understand and evaluate the internal control system. As a result, auditors must develop their abilities in the BDA field, as it adds to the creation of additional value for both auditors and their clients.

1. Introduction

The external audit has received a lot of attention in recent decades, notably after a series of scandals involving a group of international businesses like Enron and WorldCom, as well as other audit scandals (Aledwan et al., 2017; Al-Beshtawi et al., 2014). Several parts of the audit process include planning the audit, completing a client risk assessment, executing internal control tests, gathering evidence, and sharing information with appropriate parties (Appelbaum et al., 2017; Zraqat et al., 2021). Auditor jobs have become increasingly sophisticated in recent years, with auditors now providing business analysis and external reports in addition to audit tasks (Eulerich et al., 2019). Due to the challenges of the information age and the need for audit clients, audit companies have been pushed to transition from traditional audits to data-driven audits in the future (Kend & Nguyen, 2020; Lohapan et al., 2021), which allows for improved audit quality (Zureigat, 2015; Chae et al., 2020).

Existing technology firms, such as start-ups at Google or FinTech, may seize the opportunity to enter the audit services market if accounting and auditing firms do not take advantage of opportunities or neutralize threats made possible by emerging technologies such as artificial intelligence and big data analytics (BDA), according to the audit community. This will lead to increased competition among accounting firms to provide audit services (Richins et al., 2017). This necessitates audit firms continuing to develop and incorporate technology into audit procedures for auditors to gain a better understanding of how to use modern information technology at various stages of the audit process (Salijeni et al., 2019).

To complete audit engagements, auditors rely on data analytics (Appelbaum et al., 2017). Auditors employ current technology to better understand audit customers and conduct risk assessments (Bauer & Estep, 2019). They must gain an overview of the client’s risks as well as objective methods to gather adequate information to offer a professional opinion on management’s financial statement declarations (Zureigat, 2014; Ji et al., 2020; Salijeni et al., 2021), Data analytics simply refers to the processing of data provided to the auditor to generate useful information that aids the auditor in decision-making, resulting in an improvement in the audit’s quality and efficiency (Salijeni et al., 2019).

Over the past few decades, the world has witnessed an increasing impact of big data on business (Alotaibi et al., 2021). Many corporations have spent significant resources on big data to generate value. Because big data is projected to provide valuable economic benefits to audit firms and audit clients, audit firms must use BDA proactively in audit procedures to reap the benefits. It is vital for the audit profession to keep pace with these changes and to be proactive in understanding how new technological trends can impact audit procedures, ” says the Institute of Chartered Accountants in England and Wales.

Big data comprises traditional financial and non financial data sources, including email communication, phone conversation logs, and text messages from private and business social media platforms, among others (Appelbaum et al., 2017). With the advancement of technology, all forms of data may now be recorded, saved, and measured (Zraqat, 2020). The difficulty that auditors confront in filtering huge amounts of data to find useful information for audit procedures is as follows (Brown-Liburd et al., 2015; Hussien et al., 2021). As a result, audit companies have used BDA to help auditors uncover abnormalities and extract meaningful information from data essential or related to the audit subject through analysis, modeling, and visualizations for planning and performing audit engagement (Salijeni et al., 2019).

To enable appropriate data extraction from clients and other third parties, BDA demands a large investment in technology, software, and skills development (Dowling & Leech, 2014; Lee, 2021). The Big Four audit firms have recently invested large sums in purchasing or developing new technologies for evaluating corporate performance, such as Deloitte’s text mining technology for extracting key information from unstructured data. To provide additional value to its clients, KPMG has tended to develop solutions for evaluating big data. Ready-made applications have also been used by medium and small auditing firms to improve their data analysis capabilities.

Many studies have also suggested that auditors’ use of BDA may increase the effectiveness and credibility of audit reports (Gepp et al., 2018; Austin et al., 2020; Salijeni et al., 2019; Serag & Al-Aqiliy, 2020; Dagilien & Klovien, 2019; Salijeni et al., 2018), where it was suggested that auditors’ use of BDA may increase the effectiveness and credibility of (Eilifsen et al., 2020). Auditors can perform analysis on an entire community, and assessing business performance through the use of BDA enables auditors to improve risk assessments, objective procedures, and the internal control test, in addition to the potential effects in reducing the cost of auditing and enhancing the bottom line of audit firms because unlike traditional audit methodology, auditors can perform analysis on an entire community, and assessing business performance through the use of BDA enables auditors to improve risk assessments, objective procedures, and the internal control test (Austin et al., 2021).

Professional accounting organizations have taken notice of corporations’ growing usage of BDA and have issued recommendations on how to use it in audits (Eilifsen et al., 2020). “In a complex and high-volume environment, the use of technology and data analytics offers opportunities for the auditor to obtain a more effective and robust understanding of the entity and its environment, enhancing the quality of the auditor’s risk assessment and response, ” according to the International Audit and Assurance Standards Board (IAASB, 2013). Certified professional accountants in Canada think that the use of information technology by audit clients necessitates the use of BDA by auditors.

Hence, this study came to identify the impact of the use of BDA on external audit procedures in the Middle East. There are many motives for doing this study. First, exploring the impact of BDA on audit procedures is important for all audit stakeholders. Second, the impact of BDA on the audit profession has been explored in developed country contexts, and there are a limited number of studies conducted in developing country contexts. Third, this study is important because it is considered one of the first studies that dealt with the impact of BDA on all audit procedures. This study can be important for regulatory and legislative bodies when developing standards and setting training programs for auditors, as it provides field evidence from several countries.

2. Literature Review and Hypothesis Development

Several studies have found that little is known about how BDA is used in the audit profession and how it influences auditing (Gepp et al., 2018; Austin et al., 2020; Salijeni et al., 2019; Serag & Al-Aqiliy, 2020; Dagilien & Klovien, 2019; Salijeni et al., 2018). Auditors employ BDA to identify risks related to the audit client’s business, according to Salijeni et al. (2019). Austin et al. (2020) discovered that while auditors employ BDA, they face considerable challenges when doing so. According to Brown-Liburd (2019), using BDA enhances audit quality. Zraqat (2020) Essen claims that using BDA enhances financial reporting quality and that using business intelligence tools improves auditors’ capacity to offer a quality judgment.

Serag and Al-Aqiliy (2020) found that using BDA by audit companies at various phases of the audit process improves audit quality, as evaluated by the audit quality index (input, operation, and output). The application of BDA in audit methodology has an impact on the execution of audit engagement activities and also contributes to the acquisition of more skills and knowledge in the field of auditing, particularly in relation to information technology. Data analytics will save time in the advanced stages of the audit process, as auditors use analytics to gain more ideas about client activities and communicate these ideas to clients, and data analytics generate fact-based audit evidence, allowing auditors to visualize and analyze audit evidence to guide their professional judgment and decision-making.

According to Kend and Nguyen (2020), the usage of BDA has a positive impact on auditing because it diverts auditors’ attention away from manual activities and allows them to focus on more important tasks such as assessment and judgment. According to Dagilien and Klovien (2019), there are two types of motivators for the use of BDA in the audit profession: client-related motivators and audit firm related institutional motivators, and the use of BDA plays an important role in changing audit procedures at all stages of the audit and providing greater value to audit clients.

According to Salijeni et al. (2018), data analytics has an impact on audit procedures, particularly in the implementation of routine audits and analytical procedures, and like audit evidence by integrating data that was previously outside audit considerations but can be used in a diagnostic way to alert auditors to potential problem areas in financial statements, and the use of data analytics and big data guides auditors in making prudent decisions. According to Yoon et al. (2015), the evidence obtained by the auditor through BDA is more dependable than other types of evidence because it cannot be tampered with due to its large size, and the auditor acquires it from outside sources.

According to Richins et al. (2017), using BDA to collect evidence will minimize the auditor’s responsibility. As a result, before a BDA can be accepted in the audit profession, audit standards must be amended. Gepp et al. (2018) investigated the extent to which BDA is employed in the audit profession. They discover that BDA is not as common in the audit profession as it is in other professions. This is due to auditors’ aversion to using techniques that are far ahead of those used by their clients. According to Tiberius and Hirth (2019), the audit profession has both difficulties and opportunities as a result of digitization in the age of big data. Technological advancements will drive the audit profession toward continuous auditing, but new technologies will aid auditors.

According to Manita et al. (2020), digital technology will improve auditors’ ability to provide new services, and BDA will allow auditors to examine all consumer data. According to Austin et al. (2021), BDA provides auditors with a strategic advantage in giving business-related insights to their clients. De Santis and D’Onza (2021) feel that the BDA still requires time to obtain legitimacy in the field of auditing, arguing that one of the most significant barriers to the adoption of BDA in the audit profession is audit customers’ failure to keep up with technological changes. Traditional audit evidence, according to Appelbaum (2016), is no longer sufficient due to the changing nature and efficiency of audit evidence, and that using BDA allows auditors to access evidence that aids them in all future audit stages.

Based on the previous literature review, we find that there is an increasing trend towards the application of BDA in the audit profession, and accordingly, the following hypotheses were developed:

H0: There is no effect of using BDA in the external audit procedures (accepting the audit task, planning the audit process, evaluating the internal control system, performing the preliminary analytical review procedures, determining the initial levels of materiality and audit risk) in the Middle East.

The following sub-hypotheses are derived from this hypothesis:

H01: There is no effect of using BDA in accepting the audit task.

H02: There is no effect of using BDA in planning the audit process.

H03: There is no effect of using BDA in evaluating the internal control system.

H04: There is no effect of using BDA in performing the preliminary analytical review procedures.

H05: There is no effect of using BDA determining the initial levels of materiality and audit risk.

3. Research Methodology

As part of the data collection process, the field survey was used to fill out the questionnaire for this study. As a result, this study is considered an outstanding field study in terms of its purpose, as it will look into the impact of employing BDA in external audit procedures in the Middle East. The current study employed a descriptive-analytical strategy because it is appropriate for the phenomenon under investigation, and it also refers to an attempt to obtain accurate knowledge of the phenomenon’s elements by gathering data from a group of respondents who are familiar with the phenomenon. Primary data was collected using the questionnaire as the primary tool for the study to address the analytical components of the research issue. The questionnaire comprised several terms that reflected the study’s aims and queries, as well as the variables and dimensions of the investigation, which covered both BDA and BDA.

The study sample members were (361) auditors who work in auditing companies in Kuwait, Saudi Arabia, the United Arab Emirates, Jordan, Bahrain, Egypt, Lebanon, and Iraq. The questionnaire was chosen to the study sample members electronically, and the study sample members were (5093) auditors who work in auditing companies in Kuwait, Saudi Arabia, the United Arab Emirates, Jordan, Bahrain, Egypt, Lebanon, and Iraq. The sample for this study was chosen using a stratified random sampling procedure, because stratified sampling ensures that the sample would have particular qualities that the researcher desires (Creswell & Poth, 2016). The study population was separated into strata, and then a simple random sample was taken from each stratum. As shown in Table 1, the questionnaires were issued to 361 auditors, and 342 replies were received, accounting for 94.7 percent of the total number of questionnaires distributed.

Table 1: The Distribution and Response Rate of Respondents by Each Country

The measurement model and structural model of this investigation were evaluated using PLS-SEM (3.3.3). PLS-SEM can evaluate both the theoretical structural model and the measurement model at the same time. Furthermore, PLS-SEM is an ideal tool for complex models with hierarchical constructs (full disaggregation approach), mediating and moderating effects (full disaggregation approach) (Chin et al., 2003). Figure 1 depicts the evaluation of a measurement model.

Figure 1: Measurement Model Assessment

The measuring model was evaluated using convergent and discriminant validity tests. As shown in Table 2, item loadings range from 0.741 to 0.931, with Cronbach’s alpha and composite reliability values greater than 0.7. In terms of average variance, all variables had values greater than 0.5. As a result, the model used in this investigation demonstrates convergent validity, according to Hair et al. (2016).

Table 2: Convergent Validity

Regarding discriminant validity, Table 3 shows the HTMT test, where all variables achieved values between 0.393 to 0.748, which falls within the recommended range by Hair et al. (2016).

Table 3: HTMT Test

Using bootstrapping statistics in SmartPLS (3.3.3), the structural model assessment has been tested, as shown in Figure 2.

Figure 2: Structural Model Assessment

Then, P-values and T-values were created to conclude whether the hypotheses are statistically significant or insignificant. Table 4 shows the hypotheses test.

Table 4: Hypotheses Testing

Note: *p < 0.001.

H0: Using BDA in external audit procedures (accepting the audit task, planning the audit process, evaluating the internal control system, performing preliminary analytical review procedures, determining the initial levels of materiality and audit risk) has no effect in the Middle East, as shown in Table 4. As a result, H0 was not supported (Path Coefficient = 0.635; T-value = 16.991; P-value = 0.000; 95 percent LL = 0.336; 95 percent UL = 0.524).

This result is the consequence of Middle Eastern audit firms responding to the challenges of the information age by shifting from traditional auditing to data-based auditing, as the use of BDA in audits allows the auditor to collect information that helps in decision-making at all stages of the audit. This finding suggests that audit firms in the Middle East keep up with the changes in the workplace and invest appropriately in BDA methodologies. This outcome also reflects audit companies’ perceptions of the role of BDA procedures in improving the effectiveness and credibility of audit reports in the Middle East. This is in line with the literature (Kend & Nguyen, 2020; Salijeni et al., 2019; Dagilien & Klovien, 2019.

The following sub-hypotheses are derived from this hypothesis:

H01: Using BDA has no influence on accepting the audit task. H01 was not supported (Path Coefficient = 0.435; T-value = 8.908; P-value = 0.000; 95 percent LL = 0.270; 95 percent UL = 0.469). Before accepting the audit assignment, the auditor assesses the feasibility of auditing clients, and at this point, the auditor must follow auditing standards in accepting or rejecting possible audit clients, as well as retaining existing clients. The auditor must also make decisions about the audit scope, time, and fees, all of which necessitate collecting information about the audit client. At this point, BDA assists the auditor in obtaining the relevant information. As a result, BDA has an impact on the auditor’s acceptance of the audit assignment. This finding is consistent with the literature (Bauer & Estep, 2019; Salijeni et al., 2019, 2021; Manita et al. 2020).

H02: The use of BDA to plan the audit process has no effect. H02 was not supported (Path Coefficient = 0.374; T-value = 7.431; P-value = 0.000; 95 percent LL = 0.341; 95 percent UL = 0.528). The auditor is required by auditing standards to design a plan for carrying out audit activities. For the auditor to build an acceptable plan, the data accessed by BDA is required. BDA assists the auditor in identifying all of the resources needed to complete the audit. Auditors need the information to identify audit risks and understand the client’s industry when building an audit plan, and BDA plays a critical part in the auditor’s capacity to acquire information that aids in the development of an effective plan. This finding is in line with previous research. (Serag & Al-Aqiliy, 2020; Manita et al., 2020; Austin et al., 2021).

H03: There is no effect of using BDA in evaluating the internal control system. (Path Coefficient = 0.439; T-value = 9.160; P-value = 0.000; 95% LL = 0.454; 95% UL = 0.626), therefore H03 was not supported. The auditor needs the information to be able to assess the client’s internal control environment. BDA contributes to enhancing the auditor’s ability to determine the level of reliability of the internal control system in the client company, and thus determine whether the procedures they follow generate sufficient evidence.

H04: When performing preliminary analytical review procedures, using BDA has little effect. H04 was not supported (Path Coefficient = 0.544; T-value = 12.446; P-value = 0.000; 95 percent LL = 0.633; 95 percent UL = 0.773). Initial analytical procedures require the auditor to collect data from within the client organization and compare it to data collected from outside sources (Dagilien & Klovien, 2019). BDA enables the auditor to filter and process massive amounts of data collected from outside sources, allowing them to better comprehend the client’s activities and assess risks. This finding is consistent with the literature (Salijeni et al., 2019; Austin et al., 2021).

H05: Using BDA to determine the initial levels of materiality and audit risk has no effect. H05 was not supported (Path Coefficient = 0.706; T-value = 19.835; P-value = 0.000; 95 percent LL = 0.336; 95 percent UL = 0.524). To aid him in planning the collection of evidence, the auditor decides the level of materiality. The BDA assists auditors in acquiring additional ideas and producing fact based audit evidence, as well as allowing them to visualize and analyze audit evidence to guide their professional judgment and decision-making. BDA also allows the auditor to concentrate on the most important evaluations rather than wasting time sifting through a vast amount of data.

4. Conclusion

The goal of this study was to determine the influence of BDA on external audit procedures in the Middle East. There are numerous reasons for conducting this research. The impact of the BDA on the audit profession has primarily been studied in developed countries, with only a few studies undertaken in underdeveloped countries. This research is significant since it is one of the first to look into the influence of BDA on all audit procedures. Because it includes field evidence from multiple nations, this study can be useful to regulatory and legislative organizations when developing standards and training programs for auditors.

The measurement model and structural model of this investigation were evaluated using PLS-SEM (3.3.3). PLS-SEM can evaluate both the theoretical structural model and the measurement model at the same time. The study sample members were (361) auditors who work in auditing companies in Kuwait, Saudi Arabia, the United Arab Emirates, Jordan, Bahrain, Egypt, Lebanon, and Iraq. The questionnaire was chosen to the study sample members electronically, and the study sample members were (5093) auditors who work in auditing companies in Kuwait, Saudi Arabia, the United Arab Emirates, Jordan, Bahrain, Egypt, Lebanon, and Iraq. The sample for this study was chosen using a stratified random sampling procedure because stratified sampling ensures that the sample would have particular qualities that the researcher desires.

Auditors in the Middle East appear to rely on big data technology to a moderate extent, according to evidence. Data science has become a major focus in the audit industry, and the study’s findings show that BDA has an impact on the audit process at every stage. Where BDA contributes to auditors’ understanding of the audit client’s internal and external environment, which helps them decide whether or not to accept or continue the audit assignment. By providing pertinent information, BDA also makes it simple for auditors to undertake analytical procedures, analyze client risks, and understand and evaluate the internal control system. As a result, auditors must improve their abilities in the use of BDA, as it adds to the creation of additional value for both auditors and audit clients.

This research was conducted from the perspective of auditors. Clients and regulators may hold similar or opposing viewpoints. Hence, conducting a similar study from the perspective of audit clients and regulators may be beneficial. A study could be undertaken to look into the impact of auditor skill level on the perceived advantage of employing a BDA in the audit profession.