Description
TitleContinuous audit data analytics and internal control intelligence
Date Created2020
Other Date2020-10 (degree)
Extent1 online resource (ix, 190 pages)
DescriptionEmerging data and technologies have fundamentally changed how humans and technologies work together and created new benefits and conflicts. The PCAOB calls for studies to obtain insights into data analytics and related emerging technologies in auditing (PCAOB, 2019). In response to the PCAOB, this dissertation explores how to develop internal control intelligence with Continuous Audit Data Analytics (CADA).
The first essay develops a framework, Social Construction of Technology (SCOT), useful for adopting emerging data and technologies in developing internal controls. This paradigm can explain and also guide how to adopt emerging technologies more coherently. The SCOT describes the institutionalization in four phases: problem definition, interpretative flexibility management, stabilization, and social construction. We derived six propositions from a participatory case study that we developed a rule-based Continuous Monitoring System (CMS) for a Procure-to-Payment process at a state university.
The second essay tests a prototype of the proposed analytics. The case study validates that CADA can enhance the rule compliance investigation and discover potential control risks that do not appear in the current internal control system. A shareable data platform, which digitizes the entire internal control system, acts as the function to fit CADA to the Committee of Sponsoring Organization (COSO) framework. The proposed analytics holds its advantage after embedding Audit Data Analytics in the CMS. Theoretically, this arrangement provides flexible analytical techniques to handle domain constraints to extract relevant intelligence from other sources around this domain. Specifically, the study demonstrated two combining analytics. First, if management already has clear business rules to manage risks, the combination of prescriptive analytics and diagnostic analytics can ensure exhaustive monitoring in these areas. Second, suppose management has uncertainties about control risks. In that case, the analytics discovers potential risky controls by combining a risk test with the result from the diagnostic analytics.
The third essay intends to manage ethical issues for algorithmic audit analytics. The interaction between the emerging technology and audit context intensifies the uncertainty of the consequences of CADA. The study explores the root cause of potential ethical issues based on the unique feature of algorithmic analytics components. We theorize eleven ethical dimensions from the technology level, artifact level, and application level. Then we construct a four-phase ethics assessment to evaluate and deal with these moral problems. The four stages include the series assessment about the data source, the training data selection, the algorithm traceability, and the output interpretation.
The dissertation contributes to the literature from the following four points. First, we develop a theoretical framework to avoid the technology usage trap and promote CADA in auditing. Second, this study guides the development of flexible combining analytical schema. It exemplifies how to embed Audit Data Analytics schema in the CMS to provide audit evidence for assessing internal controls. Third, the proposed CADA schema provides enhanced auditing performance. Auditors can use the enhanced analytic schema to implement SOX404. Last but not least, the dissertation offers a practical roadmap to solve ethical issues in data-driven CADA.
NotePh.D.
NoteIncludes bibliographical references
Genretheses, ETD doctoral
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