Trust Calibration in Voice Assistants
Voice assistants speak with the same steady, confident tone whether they're right or wrong, and people tend to believe them anyway. This project looks at how tone, pacing, and phrasing shape how much trust a listener extends to a machine's advice, and where that trust outruns the system's actual accuracy. The goal is to understand when confident-sounding AI earns more credibility than it deserves, and what design changes could help people calibrate their trust more accurately instead of defaulting to blind confidence in a friendly voice.