Ascribing emotions to robots: Explicit and implicit attribution of emotions and perceived robot anthropomorphism

Abstract

Despite growing literature on various aspects of robot perception by humans, little is still known about how robots' anthropomorphism is perceived at various degrees of conscious perception. In the current study, in two experiments, we assessed whether explicit and implicit attributions of primary and secondary emotions towards robots predict participants' perception of robots' anthropomorphism. In Experiment 1, participants explicitly evaluated whether robots were likely to experience different primary and secondary emotions, followed by the assessment of the level of attributed anthropomorphism towards robots. In Experiment 2, participants completed an adapted version of the Implicit Association Test (IAT), which aimed at implicit evaluation of participants' attribution of different primary and secondary emotions to robots, followed by the assessment of the level of attributed robot anthropomorphism. At the explicit level, attribution of secondary emotions to robots was linked to higher perception of robots' warmth and competence. At the implicit level, participants’ perception of conceptual similarity between robots and humans (i.e. lower difference between attribution of primary and secondary emotions to humans and robots), predicted more anthropomorphic warmth perception and less discomfort towards robots. The current results reveal that implicit and explicit attribution of primary/secondary emotions towards robots predict specific aspects of robot anthropomorphism, with the warmth dimension consistently being predicted across both levels of evaluation. These results shed light on a novel approach to predict various aspects of robot anthropomorphism, which in the future might be useful for evaluations of social robots for their suitability in human-robot interactions.