--insinuated through the use of typical first names: Max and Murat.
203 pre-service teachers (69% female, all with classroom experience) were given either a poor or an average performance third grade paper, and asked to assess both the number of errors committed by the student and the overall quality of their work.
The hypothesis was that the first task, a rule-based judgment, would leave little room for personal interpretation; while the second would be open to subconscious biases because of the necessity to integrate different components, and thus to resort to intuition.
As expected, the dictation was graded less favorably (about 12%) when the student appeared to have a migrant background.
Interestingly, the effect was stronger for the poor performance paper, and did not affect the error count. Participants thus gave different grades to papers they agreed were identical, purely based on the name of the student.
What is more, the effect was also stronger for preservice teachers with more positive attitudes towards Turkish Germans--as measured by a specially created Implicit Attitudes Test.
The reason, the authors hypothesized, coudl be a feeling of disappointment compared to their expectations.
Source: Bonefeld and Dickhauser (2018)