Robots Asking for Favors: The Effects of Directness and Familiarity on Persuasive HRI

Abstract

As robots work in increasingly collaborative settings with humans, scenarios will arise where robots need to successfully request favors to be effective. Communication directness and familiarity have been shown to be important factors to persuasion. However, these two critical factors have not yet been jointly investigated in human-robot interaction (HRI). This letter explores how they can be used by social robots to request favors from people. We present a social HRI study that uniquely investigates the effects of a robot's communication directness and its familiarity on persuasiveness, trustworthiness, and a person's willingness to help the robot. In the study, we present participants with scenarios where two different robots (one familiar, one unfamiliar) ask participants for favors using either direct or indirect requests. Our results show that a familiar robot is more persuasive, trustworthy, and people are more willing to help it. Furthermore, indirect requests are perceived to be more trustworthy and encourage greater willingness to help than direct requests, regardless of the robot's familiarity. Further discussions of these results highlight key considerations for collaborative social robots, particularly when robots request assistance or favors from people.