In the article, “Navigating a social world with robot partners: A quantitative cartography of the Uncanny Valley,” Maya B. Mathur and David B. Reichling argues that human-robot interactions are more complicated than they outwardly appear.
In the article, the authors used several experiments to assess the validity of the presence of an uncanny valley. The idea of the uncanny valley is rooted in the fact that machines that are created in the likeness of humans, but have small imperfections, can be very unsettling. This provides a problem in a society where we are becoming more and more reliant on machines. In these experiments, they gathered a body of subjects and had them rate different photographs of androids that varied in their realism and facial expressions and features. These tests accessed the human-resemblance, the likability, and the trustworthiness of different robot faces. They authors graphed and assessed their findings from these tests in order to see if human reactions to android robots actually create an uncanny valley effect that would influence the degree of trust that humans put in robot social partners. They found that, “humans often have unexpectedly uncomfortable reactions to android robots that were designed to have pleasant social interactions with humans.”
Overall, this study suggested that the uncanny valley has a real influence on the way that humans perceive robots as social partners, and it impacts conscious assessments of trust.