David King Hall, Arch Lab
October 15, 2021, 10:00 AM to 12:00 PM
Robots that engage face-processing are better able to transmit social signals that improve human-robot interactions. However, although robots are often crafted with face-like displays, research has yet to examine whether they actually engage face-like processing. To this end, we examined across seven studies how robot’s machine-like appearance and machine-like nature modulate causes and consequences of face-processing with robots. In a first set of studies, we demonstrate that robots generally engage configural processing (study 1; causes of face perception) less strongly than human faces but do elicit configural processing when robots contain more physical facial features (studies 2) that are distinctly human (study 3). In a second set of studies, we demonstrate that, even when controlling for physical differences between humans and robots, dehumanizing stereotypes associated with robots’ machine-like nature, alone, modulates configural processing with robots (studies 4 & 5). In a final set of studies, we find that perceptual discrimination of robots (consequences of face perception) is also modulated by a robots’ machine-like physical features (study 6), but not their machine-like nature (study 7). Results are discussed in terms of theoretical and practical application for human-robot teaming.