This paper shows that the same Asian-European multiracial faces are rated more "European" if they are associated with European names and more "Asian" if they are associated with Asian names.
This should come as no surprise to participants in various online anthropology boards, where the "Guess his origin/classify" kind of topic receives widely different responses depending on how much information is released about a given subject.
This result may seem "irrational" at first: why should the exact same stimulus be perceived differently depending on an associated name?
However, I would argue that people are acting like good Bayesians in this case:
Everyone has a mental model of P(appearance | European) and P(appearance | Asian), i.e., the distribution of phenotypes expected from persons of European or Asian ancestry. A person's P(European | appearance) opinion will not only depend on the above, but also on the P(European) and P(Asian) prior, and this will, of course, be influenced by knowledge that, e.g., a person is named Ng Yat Ho or David Smith.
Perception doi: 10.1068/p6255
Barack Obama or Barry Dunham? The appearance of multiracial faces is affected by the names assigned to them
Kirin F Hilliar, Richard I Kemp
Does semantic information in the form of stereotypical names influence participants’ perceptions of the appearance of multiracial faces? Asian-Australian and European-Australian participants were asked to rate the appearance of Asian-Australian faces given typically Asian names, European-Australian faces given typically European names, multiracial faces given Asian names, and multiracial faces given European names. Participants rated the multiracial faces given European names as looking significantly ‘more European’ than the same multiracial faces given Asian names. This study demonstrates how socially derived expectations and stereotypes can influence face perception.