Self and Other Accuracy

Participants' faces were covertly recorded while they rated the attractiveness of people, the decorative appeal of paintings, and the cuteness of animals. Ratings employed a continuous scale. The same participants then returned and tried to guess ratings from 3second videotapes of themselves and other targets. Performance was above chance in all three stimulus categories, thereby replicating the results of an earlier study (North, Todorov, & Osherson, 2010) but this time using a more sensitive rating procedure. Across conditions, accuracy in reading one's own face was not reliably better than otheraccuracy. We discuss our findings in the context of "simulation" theories of face-based emotion recognition (Goldman, 2006) and the larger body of accuracy research.