Trusting the Algorithm: The Impact of Evidence Framing on Algorithm Aversion in Auditing
Keywords
No Thumbnail Available
Authors
Issue Date
2025-06-16
Language
en
Document type
Journal Title
Journal ISSN
Volume Title
Publisher
Title
ISSN
Volume
Issue
Startpage
Endpage
DOI
Abstract
The use of digital technologies to transform audit processes has gained significant interest in the audit profession, particularly with the recent rise of more advanced tools such as Artificial Intelligence (AI). In this research, I examine whether auditors are prone to algorithm aversion and whether the framing of audit evidence influences the extent of this aversion during the decision-making process. I use a 2 × 2 between-subjects experimental design, manipulating the source of audit evidence and the framing of the audit evidence.
Overall, my results suggest auditors will propose smaller adjustments to the clients’ management estimates when evidence is received from an AI system instead of a human specialist. I also find that receiving evidence from an AI system, compared to a human specialist, reduces the level of proposed adjustments to a greater extent when the evidence is negatively framed compared to positively framed evidence. This finding highlights the disordinal interaction effect between the framing of audit evidence and the source of the evidence on proposed adjustments. In the AI system condition, negatively framed evidence leads to smaller proposed adjustments compared to positively framed evidence. In contrast, in the human specialist condition, the opposite pattern occurs since negatively framed evidence increases proposed adjustments compared to positively framed evidence.
Description
Citation
Supervisor
Faculty
Faculteit der Managementwetenschappen