The research was described in Evolutionary Psychology
and it examines how beauty biases are applied based on ethnic background compatibility.
Maria Agthe conducted two studies with her colleagues. The first experiment tested the hypothesis that beauty biases will be generalizable across ethnicities when people are evaluating members of their own ethnic group.
A total of 583 international students in Germany participated in study 1 and were randomly assigned to one of four experimental conditions (based on own sex, target sex, and attractiveness levels). In all conditions, subjects examined materials describing the experience and skills of a target person along with a photo of the target.
Photographs were previously rated and varied by attractiveness. Following the viewing, they were asked to rate the social competency of and desire for interaction with the target individual. Results showed that beauty biases were generalizable across ethnic groups, though less so among Asians evaluating same-sex targets, perhaps indicating a cultural difference that deserves further exploration.
Experiment 2 was designed to evaluate the relationship of beauty biases to judgements of targets from other ethnic groups. Participants included 2557 Caucasian Germans that were recruited from universities across Germany. The study design was identical to that of the first investigation, with the lone change being that evaluations and ratings were made on targets from multiple ethnic groups. It was discovered that the participants did not apply the typical beauty biases when judging people of other ethnicities. However, the results cannot be generalized across cultures since the entire subject pool shared ethnicity and country of origin.
Gender biases among Caucasians are known to give attractive people benefits when being evaluated by the opposite sex, while they are withheld from attractive members of the same sex (who may instead face increased oppression). According to the research described above, this effect is only present when the target being judged is also Caucasian, although people belonging to other ethnicities did tend to evaluate targets from their own groups in a comparable manner to Caucasians. More investigation will be required to see if this cross-ethnic effect extends to non-Caucasian groups, as they were not tested in this manner.
source - http://www.psypost.org