The Effect of Advice Valence on the Perceived Credibility of Data Analytics

Authors

Clara Xiaoling Chen, University of Illinois at Urbana-ChampaignFollow
Ryan M. Hudgins, Western UniversityFollow
William F. Wright, University of Illinois at Urbana-ChampaignFollow

Document Type

Article

Publication Date

2021

Journal

Journal of Management Accounting Research

URL with Digital Object Identifier

https://doi.org/10.2308/JMAR-2020-015

Abstract

We use an experiment to examine how advice valence (i.e. whether the advice suggests good news or bad news) affects the perceived source credibility of data analytics compared to human experts as a result of motivated reasoning. We predict that individuals will perceive data analytics as less credible than human experts, but only when the advice suggests bad news. Using a forecasting task in which individuals are seeking advice from either a human expert or data analytics, we find evidence consistent with our prediction. Furthermore, we find that this effect is mediated by the perceived competence of the advice source. We contribute to the nascent accounting literature on data analytics by providing evidence on a potential impediment to successfully transitioning to the use of analytics for decision-making in organizations.