Face Identification in the Laboratory and in Virtual Worlds

Investigations into human cognition typically control variables tightly in the laboratory or relinquish systematic control in field studies. Virtual Reality (VR) can provide an intermediate approach by facilitating research with complex but controlled environments. However, understanding of the correspondence between VR and laboratory paradigms is still limited. This study addresses this issue by comparing established laboratory tests of face identification with passport control at a VR airport. We show that test characteristics transcend comparison of the laboratory tests and VR and demonstrate consistent correlations between these tasks. However, person identification in VR was also marked by bias to accept mismatching identities. These findings support correspondence between laboratory tests of face perception and VR but also highlight the importance of understanding human behaviour under more complex conditions. This problem arises in many areas of psychology, and our study shows that VR offers a solution by providing complex but controlled environments. General Audience Summary Psychological experiments into human cognition either tend to study behaviour in the laboratory, where the conditions under which research is conducted are simplistic but tightly controlled or relinquish such control in field studies, where behaviour is examined in natural environments in which additional factors can be at play. Both approaches have some disadvantages that the development of Virtual Reality (VR) can bridge by facilitating behavioural research in environments that are both complex and controlled. However, how research in VR corresponds to traditional experimental approaches is still unknown. This study investigates this issue by comparing established laboratory tests of face identification with person identification at a VR airport, in which participants take on the role of passport control officers. We demonstrate that person identification in laboratory tests is linked to the same behaviour in VR. However, we also find that person identification in VR is marked by a tendency to incorrectly accept travellers who bear the identity documents of another person. These findings demonstrate the importance of understanding human behaviour under conditions that more closely mimic real life and show that VR can facilitate such research.