Description
TitleThree essays on the impacts of emerging technologies on accounting and auditing
Date Created2022
Other Date2022-05 (degree)
Extent227 pages : illustrations
DescriptionThis dissertation examines the adoption and impacts of emerging technologies on accounting and auditing. The first essay (Chapter 2) responds to an increasing need for research on Robotic Process Automation (RPA) in external auditing, especially research that concerns auditors’ roles in an RPA-enabled audit workflow. Since more than half of the audit tasks require certain levels of auditors’ judgment and cannot be fully automated, audit automation should include attended automation, in which auditors work alongside and interact with automation routines. This chapter adopts the Design Science Research (DSR) approach and proposes an Attended Process Automation (APA) framework that guides the implementation of attended automation in audits. This paper also demonstrates the APA framework by applying it to the planning process for Single Audits, a government-required external audit for beneficiaries of funding. The APA framework emphasizes auditors' vital role in an automated audit workflow in providing professional judgments that are currently irreplaceable by automation.The second essay (Chapter 3) adopts Machine Learning (ML) to examine predictive factors of audit failure. Researchers have identified a broad set of explanatory variables for audit quality. However, little is known about the effectiveness of these audit-related variables (ARV) in predicting audit failure (i.e., low-quality audits) and which are the most predictive. Understanding the predictive power of ARV can help researchers, regulators, and practitioners evaluate whether these provide practical value in red-flagging audit failure. Using machine learning techniques, we find that ARV have acceptable predictive power and that they outperform benchmark financial variables in predicting audit failure. The most predictive ARV reflect auditor competence, independence, effort, incentive, and the quality of the audited financial reports. We synthesize predictive ARV into a score that can significantly outperform existing academic measures in incrementally associating with audit failures. Our study informs researchers, regulators, practitioners, and investors about the usefulness of ARV in predicting audit failure and provides them with a list of predictive audit features and a score that can be used to predict the likelihood of an audit failure.
The third essay (Chapter 4) explores whether the implementation of artificial intelligence (AI) in firms’ operations is associated with improved accuracy of management earnings forecasts. We identify non-technology firms that have implemented AI in their operations from 2014 to 2018. We find that AI is associated with more accurate management earnings forecasts after its implementation and that it improves management forecast accuracy indirectly through improving the performance of firms’ operations. However, this indirect effect is small compared to AI’s direct effect. We also find evidence that AI more profoundly improves management forecast accuracy when the forecast horizon is longer, and that ML is the primary AI technology that contributes to the improvements in management forecast accuracy. In contrast, we find no evidence that AI is associated with changes to the precision, frequency, or bias of management forecasts. We contribute to the literature by providing initial archival evidence about the association between AI implementation and improvements to the accuracy of management earnings forecasts.
Overall, this dissertation adopts a diverse set of research mythologies to examine the prevailing issues regarding the usage of RPA, ML, and AI in accounting practice and research.
NotePh.D.
NoteIncludes bibliographical references
Genretheses
LanguageEnglish
CollectionGraduate School - Newark Electronic Theses and Dissertations
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.
Statistical Profile