Differences in social expectations about robot signals and human signals

Abstract

Decision-making processes are involved in a large portion of humans’ everyday life and often occur in social situations. Although social decision-making has received increasing interest in literature, the influence of communicative signals, like social cues and feedback, on social decision-making is still poorly understood. In particular, the issue of whether social signals exhibited by non-human agents influence decision making has not yet been addressed. This question is of great importance today, in the new era in which artificial agents, such as robots or avatars start interacting with humans on a daily basis. This study aimed to examine whether robot non-verbal communicative behavior has an effect on human decision-making and can be perceived as a social signal. To this end, we implemented a 2-alternative-choice task with a between-subjects design where a robot acted as a game partner. We manipulated robot’s cues before participant’s decision and robot’s feedback after the decision. We found that manipulating robot signals affected participants’ performance. In particular, participants were slower in the condition where cues were mostly invalid and the robot reacted positively to wins. We show that this effect could not be attributed to attentional mechanisms, feedback expectation, or cognitive control. Instead, our results suggest that the effect is rather due to the violation of expectations about the incongruence between social signals. Our findings also indicate differences in these social expectations when they are about robot signals compared to human signals.