AUTHOR=Noguchi Yohei , Kamide Hiroko , Tanaka Fumihide TITLE=Weight Shift Movements of a Social Mediator Robot Make It Being Recognized as Serious and Suppress Anger, Revenge and Avoidance Motivation of the User JOURNAL=Frontiers in Robotics and AI VOLUME=9 YEAR=2022 URL=https://www.frontiersin.org/journals/robotics-and-ai/articles/10.3389/frobt.2022.790209 DOI=10.3389/frobt.2022.790209 ISSN=2296-9144 ABSTRACT=

Humans can become aggressive during text messaging. To maintain a healthy interpersonal relationship through text messaging, our negative mental states, such as anger, have to be well-controlled. This paper discusses the use of a handheld social robot deployed as a mediator in text messaging between humans. The robot is equipped with a movable weight inside its body. By controlling the movement of the internal weight during the time when the robot speaks out messages received from a human sender, we hypothesize that the psychological state of a receiver who holds the robot can be affected (for example, he/she will listen to the messages more seriously). In a controlled study (n = 94), in which participants were manipulated to be frustrated by using a context scenario, we studied the effect of three dialogue scripts with/without weight shifts. Results showed that introducing weight shifts together with the robot speech suppressed on average 23% of the user’s anger. However, only 3.5% of the anger was suppressed when the weight shifts were not applied. Additionally, in cases where the robot showed empathy to the user in words with weight shifts, the user’s revenge urge was successfully reduced by 22%. There was almost no effect confirmed when the weight shifts were not applied. A similar effect was also found in avoidance motivation: 15% of the avoidance motivation was reduced if weight shifts were applied. The reductions in revenge and avoidance motivation are considered important factors for human forgiveness. Therefore, our findings provide experimental evidence that weight shifts can be an effective expression modality for mediator robots, from the perspective of not only suppressing the user’s anger but also by inducing forgiveness during messaging.