Presentation
Trust in Uncertain Automation: Risk and Framing Effects
DescriptionThe development and adoption of automation have drastically increased in recent years. From automated decision support systems (DSS) to machine learning algorithms and AI, automation is entrenched in many aspects of life. With its adoption into many industries, it is essential that automation is used appropriately. Thus, automation may be designed to communicate information to the operator via automation transparency. Automation transparency may occur in the form of communicating its uncertainty or certainty in its recommendation. Its effects on operator behaviors such as trust and accuracy are varied and inconsistent (Bhaskara et al. 2020), making the extent to which uncertainty communication impacts operators’ perceptions and behaviors elusive. In risky contexts, communicating the systems’ certainty or uncertainty in its suggestions may affect an operator’s decision to continue using the automated system. Consequently, this study aimed to understand how framing automation transparency as certain or uncertain impacts operator behaviors and perceptions of automation under varying levels of risk.
Event Type
Lecture
TimeThursday, October 22nd3pm - 3:20pm PDT
Location
Similar Presentations