Presentation
Calibrating Trust and Reliance in Variable-Reliability Automation
DescriptionTrust plays a crucial role in human-automation collaboration, influencing team performance through over-trust and under-trust. Factors like self-confidence, automation capabilities, and workload affect both user behavior and trust in automation. Prior research shows that trust in automation is dynamic and influenced by its initial performance. For example, automation that shows moderate reliability may gradually build trust, while initially unreliable systems struggle to gain trust, even if their reliability improves. Users might still follow recommendations from automation they distrust, highlighting a complex relationship between trust and reliance. This study investigates these dynamics by examining users' reliance behavior and self-confidence under varying automation reliability. Additionally, it explores how feedback on performance and task order affects trust calibration, aiming to enhance our understanding of human-automation interaction to inform future system design.
Event Type
Lecture
TimeThursday, September 12th3:40pm - 4pm MST
LocationFLW Salon B
Human Performance Modeling