The Impact of a Third-Party on Trust of a Human-Robot Dyad

Abstract

Research has mainly focused on dyadic trust relationships between individual humans and robots. However, in real-world applications, multiple humans interact with the same robot over time, creating opportunities for third-party influences on trust. This study investigates whether a human third party can influence a new human trustor’s perceived trust in a robot trustee. We conducted a between-subjects experiment with 24 participants, using a humanoid robot. Results show that the experimental group exhibited significantly higher trust levels after the robot successfully completed the task again. Interestingly, we did not find a significant difference directly after the intervention, which we used after a failure to repair trust. These findings suggest that trust in human-robot dyads is not isolated but can be influenced by social dynamics. This has important implications for both research, highlighting the need for standardized researcher behavior, and practice, indicating that trust in robots may be interconnected across workforces.