Dominant or Submissive? Exploring Social Perceptions Across the Human-Robot Spectrum

Abstract

The study investigates the perception of humanlike robots, exploring the influence of behavior, particularly dominance or submissiveness, on social perceptions. The focus is on the Uncanny Valley Hypothesis (UVH), which suggests that as robots become more humanlike, they initially elicit positive responses but eventually lead to discomfort and negative reactions. The study analyzed data from 330 participants and utilized two sets of 11 stimuli representing humanlike-robot 3D morphs with either dominant or submissive behavior to examine emotions. The method involved an online setup where participants viewed movie clips featuring the morphs and completed questionnaires assessing dominance, categorization (human or robot), eeriness, likability, and perceived threat. The results indicated that humanlike robots, aligning with the UVH, evoked more negative emotions, although likability differences were not significant. Contrary to expectations based on previous studies, the results reveal that the negative perception of dominant robots is not solely attributable to their categorization as robots. Furthermore, the study explored the influence of the Uncanny Valley on social perception. While the results lean towards considering the Uncanny Valley as a potential side effect of perceptual processing, the evidence is inconclusive, pointing to the need for further exploration. The outcomes also underscore the challenges in categorizing humanlike morphs, emphasizing the ambiguity in the concept of a robot.