It's not Paul, it's a robot: The impact of linguistic framing and the evolution of trust and distrust in a collaborative robot during a human-robot interaction

An appropriate level of human-robot trust is widely acknowledged as a crucial factor for successful human-robot interaction (HRI). Previous research has addressed anthropomorphic design but mostly neglected linguistic framing as a means to influence trust by altering the perceived human-likeness of a robot. Similarly, time -dependant patterns of trust and distrust during an HRI have rarely been investigated, hindering the develop-ment of a coherent theoretical framework on framing effects, their formation conditions, and their relation to trust and distrust evolution in the HRI context. A previous online study suggested that linguistic framing mod-ulates inexperienced factory workers' trust in a collaborative robot at the workplace. This article presents a follow-up study with a sample of students and an realistic HRI setting. Despite using similar framing stimuli, the study failed to replicate significant framing effects on trust, suggesting their context-sensitivity and dependence on individual characteristics of the recipients of the framing. Additionally, a significant increase in trust and decrease in distrust during the HRI highlights the need for time-dependant considerations. The results call for a more fine-grained investigation of linguistic framing effects as well as of trust and distrust evolution in order to develop a comprehensive theoretical framework applicable to the HRI context.

Automated summary

An appropriate level of human-robot trust is crucial for successful human-robot interaction. Previous research has focused on anthropomorphic design but overlooked the role of linguistic framing in influencing trust. The time-dependent patterns of trust and distrust during human-robot interaction have also been underexplored. This study failed to replicate significant framing effects on trust, indicating the influence of context and individual characteristics. The findings highlight the need for further investigation into linguistic framing effects and the evolution of trust and distrust in order to develop a comprehensive theoretical framework applicable to the HRI context.