The Effect of Social Robotic Embodiment on Persuasion in Collective Moral Decision-Making

Abstract

From day-to-day impressions such as marketing and social media, to political debates, people try to influence and persuade each other. Social cues such as speech, gazing, and listening are important to effectively convince people. But can robots apply the same persuasion techniques, and what is the effect of their appearance on it? In order to study the effect of appearance on persuasion, this study compares two types of robot appearances: a humanoid and a non-humanoid. Eighteen groups (N=54) engaged in moral collective decision-making scenarios. During the trials, it was measured how persuasive the robot was by comparing individual choices, robot arguments and collective decisions. There was no effect of appearance on persuasion across both conditions. However, the non-humanoid robot was considered to have a higher level of autonomy, contradicting most studies on robot appearance and autonomy. Regardless of appearance, participants conformed to the robot and the most dominant human equally. Although the robot was objectively persuasive, the robot was not perceived to be by the participants. This discrepancy might have been caused by an effect seen in other HRI studies. The perceived different levels of autonomy might have been caused by the expectations set by the humanoid and non-humanoid robot. Future research should focus on validating objective persuasion measures, and discovering other important types of social cues for persuasive technology.