This study examines how audit theory and cognitive psychology can be combined to gain a greater understanding of the regulatory, competency and cognitive challenges auditors face in the transition to more data-driven audits by using modern data analysis. The regulatory challenge is analyzed based on the international ISA-standards. According to the analysis performed, auditors disagree how the standards should be interpreted when applying data analysis and the automated data analytics tools of whole populations. Some auditors interpret data analysis as a sophisticated substantive analysis procedure, while others argue it should only be used as a risk assessment procedure, which might be strengthened by the upcoming revision of ISA 315, making it the only standard to mention data analysis directly. However, the analysis showed that the standards do not prohibit the use of data analysis in either way, nor do they provide guidance on how to apply it. Therefore, it is currently the responsibility of the auditor to interpret how the standards apply to data analysis and make professional judgments around its application, which sets requirements for competences within both data analysis and audit. For the auditor to be able to evaluate the probative value of audit evidence (Gronewold, 2006), the auditor must be able to reconstruct a non-observable reality. Data analysis can be a tool to make these inferences about this relevant reality, if the right competences are in place. Here the analysis showed that auditors did not have confidence in the current level of competence within the industry to fulfill client expectations, and that a significant portion of auditors find it more difficult to judge when appropriate and sufficient audit evidence was obtained with data analysis and evaluate when to consult an expert for assistance. The implementation of more standardized data analytics tools and stronger inclusion of data analysis experts could pose potential solutions to some of these challenges. In particular data analytics can automate parts of the analysis and provide guidance for the user, while data analysis experts can advise and develop customized analyses either for or together with the auditor. By combining the two solutions, auditors can create more valuable analyses with a lower competence level and learn through practical application and guidance from experts. To support the auditors in utilizing data analytics and experts, I introduced a decision model that could guide auditors in their decision process through an evaluation of the complexity of the analysis and the auditors’ own competence. In addition to challenges based on level of competence, the auditor also faces human cognitive limitations in when performing professional judgements and decision-making related to the data analysis. The analysis showed a significant portion of auditors faced information overload or overwhelming demand on their skills and adaptability, which according to cognitive psychology can impair the ability to make deliberate and rational decisions (Kahnemann & Tversky, 1974). The analysis exemplified the risks of making wrong inferences from data visualizations, with more than 90% of respondents indicating signs of bias and introduced potential risks of other types of bias with a possible impact on data analysis in auditing. In the transition to data analysis, auditors must therefore be aware of the potential risks of bias and automated decisionmaking. The risks of bias can also to some extent be mitigated through the application of more data analytics tools, which however also introduce new risks of bias, as the auditor’s deliberate decisionmaking might decrease due to the automated aspect of the tools. Based on the results of the analysis it has become evident that further exploration of cognitive psychology in relation to data analysis in auditing can contribute to developing the right competences needed to improve professional judgments desired to lift the quality of audits which obtain audit evidence through data analysis. I therefore suggest further analysis and tests of the potential impact of bias on data analysis in auditing, and developing guided tools in cooperation with experts on cognitive psychology and data analysis, as well as a continuous focus on developing auditor's level of competence, as the auditor is required to make correct evaluations of the presented data and have a basic understanding of the pitfalls that the application represents.
|Uddannelser||Cand.merc.aud Regnskab og Revision, (Kandidatuddannelse) Afsluttende afhandling|