Attitudinal predictors of relative reliance on human vs. automated advisors

Stephanie Merritt, University of Missouri–St. Louis
Ruchi Sinha, Indian School of Business
Paul Curran, Michigan State University
Daniel Ilgen, Department of Psychology, Psychology Building, 316 Physics-Room 340A, East Lansing, MI 48824, USA


Trust and liking are attitudes with important implications for automation reliance in single-advisor settings; however, the extent to which their relationships with reliance generalise to settings in which the user receives conflicting advice from a human and automation is unknown. Participants completed an X-ray screening task and received simultaneous advice from what they believed was another human and an automated aid. High disuse was found for both advisors. Among participants who relied on advice, those with greater relative liking for the automation than for the human significantly increased their reliance on the automation relative to the human during the first half of the task. No significant relationships were found between relative trust or relative liking with reliance in the later part of the task, suggesting that reliance processes in dual-advisor settings may differ from those in single-advisor settings.


Repository URL