Automation Biases in Human-Robot Trust Interactions
Abstract
This final report summarizes the examination of possible human biases in trust between human-human and human-robot teaming. The studies found humans do hold biases against robot partners. Specifically, robot partners were suffered greater trustworthiness declinations after a trust violation than did human partners but only when lower trustworthiness could be ascribed to specific aspects. That is, there were differences between the types of manipulation, such that performance/ability violations were much stronger than benevolence or integrity violations. Additionally, the perfect automation schema (PAS) construct significantly predicted slope variance in benevolence perceptions following a trust violation.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 2021
- Accession Number
- AD1159286
Entities
People
- Gene M. Alarcon
Organizations
- 711th Human Performance Wing