Social robots are becoming increasingly relevant in education, for example, by using them as tutors. To create a more empathetic and engaging learning environment, it is important to consider the anthropomorphism of these social robots. However, an ethnic perspective on the use of anthropomorphization is still lacking when it comes to improving learning gains. Therefore, this research focuses on whether personalized, ethnicity-based anthropomorphization of a robot can enhance learning gains. To this end, history lessons were tutored with a Furhat robot, with groups of participants interacting with a Furhat whose face matched the ethnicity of the participants, in an experimental setting. Our results showed that participants who interacted with the robot displaying the personalized, ethnicity-based anthropomorphization learned more than participants interacting with a robot displaying a robotic appearance. These findings highlight the importance of incorporating cultural diversity into educational technologies to foster more effective and inclusive learning environments.