Abstract

Could a simple pair of glasses really fool us into thinking Superman and Clark Kent are two different people? Here, we
investigated the perception of identity from face images with a task that relies on visual comparison rather than memory. Participants
were presented with two images simultaneously and were asked whether the images depicted the same person or two different
people. The image pairs showed neither image with glasses, both images with glasses, and ‘mixed’ pairs of one image
with and one without glasses. Participants’ accuracies, measured by both percentage correct and d′ sensitivity, were significantly
lower for ‘mixed’ trials. Analysis of response bias showed that when only one face wore glasses, people tended to respond ‘different’.
We demonstrate that glasses affect face matching ability using unconstrained images, and this has implications for both disguise
research and authenticating identity in the real world.