"Are You Really Sure?" Understanding the Effects of Human Self-Confidence Calibration in AI-Assisted Decision Making

• DOI: 10.48550/arxiv.2403.09552
• Publication Date: 2024-03-14
• Authors: Xiaojuan Ma; Xinru Wang; Lei Ying; Chuhan Shi; Ming Yin; Xiaojuan Ma
• Keywords: Self-confidence

In AI-assisted decision-making, it is crucial but challenging for humans to achieve appropriate reliance on AI. This paper approaches this problem from a human-centered perspective, "human self-confidence calibration". We begin by proposing an analytical framework highlight the importance of calibrated human self-confidence. our first study, we explore relationship between appropriateness and appropriateness. Then in second propose three calibration mechanisms compare their effects humans' user experience. Subsequently, third study investigates decision-making. Results show that calibrating enhances human-AI team performance encourages more rational AI (in some aspects) compared uncalibrated baselines. Finally, discuss main findings provide implications designing future decision-making interfaces.