Innovative Accounting Interviewing: A Comparison of Real and Virtual Accounting Interviewers

Abstract

Recent technological advances make it possible to create automated, virtual interviewers, called embodied conversational agents (ECAs). We study how an ECA compares to a human interviewer in three experiments. In experiment 1, we show that two theoretically motivated factors-making the ECA facially and vocally similar to the interviewee-result in the ECA performing similarly to or better than human interviewers for six antecedents of disclosure quality. In two additional experiments, we show that employees are on average 21 to 32 percent more likely to disclose violating internal controls to an ECA than to a human, even if the human interviewer has significant interviewing experience. These findings contribute to the ECA design literature by showing that similarity-enhancing features of ECAs increase the antecedents of disclosure. The findings also contribute to the accounting literature by demonstrating that ECA technology can increase the scope of interviewing in accounting without reducing interview quality.

New Frontiers for Internal Audit Research*

“…Thus, it is unclear if auditees' disclosure behavior, auditors' skepticism, and the likelihood of follow-up questions in these situations would be more similar to face-to-face interactions or email communications. There is some prior research that shows auditees are more likely to disclose information to a virtual interviewer (i.e., an avatar) than to a human internal auditor (Pickard et al 2020); however, in the case of remote auditing, the auditee is still interacting with and disclosing to another live human. Thus, it is unclear that the results from this prior research would generalize more broadly to remote auditing as it is conducted today.…”

A Framework for Using Robotic Process Automation for Audit Tasks*

“…Our study makes several important contributions to prior research and to practice. From a broad perspective, we contribute to the expanding line of research that is testing the effectiveness of emerging technologies within accounting settings (e.g., Dowling and Leech 2014;Pickard et al 2020;Hodge et al 2021;Austin et al 2021;Christ, Emett, et al 2021). This line of research has found mixed results regarding technology effectiveness, leaving many to question whether these emerging technologies have failed to achieve their expected benefits within accounting settings (Chen et al 2014;Han et al 2016;Fersht 2017;Reed 2018;Information Age 2019;Eulerich, Masli, et al 2021).…”

The Attenuating Effect of Intelligent Agents and Agent Autonomy on Managers' Ability to Diffuse Responsibility for and Engage in Earnings Management

“…Further, since this is among the first study to explore IAs in accounting, our operationalization of the IAs in our setting was kept very simple, with a short description of an IA and a picture of a desktop computer. A large literature exists on several other attributes of intelligent agents, including physical presence, anthropomorphism, and dynamic interaction, among others (Pickard, Schuetzler, Valacich, and Wood 2019). Future research could investigate how one or more of these attributes impact manager decision-making and could decrease managers' "explainability" for the results produced by an IA, which may affect aggressive financial reporting.…”