Research on ethnic segregation in schools regularly assumed that ethnic homophily—the tendency to befriend same-ethnic peers, above and beyond other mechanisms of tie formation—is associated with salient ethnic boundaries. We devise a more direct test of this assumption based on a novel measure of ethno-racial group perceptions. In a network study of more than 3000 students in 39 schools of a metropolitan region in Germany, we asked students to indicate which cliques they perceived in their school grade and to describe these groups in their own words. We find that ethno-racial labels are more likely directed at larger cliques that include a higher share of Muslim students or more students with stronger ethnic identification. Still, ethno-racial labels are rarely employed, both absolutely and relative to other modes of classification. Moreover, net ethnic segregation in friendships (“ethnic homophily”) and the reverse pattern in dislike relations (“ethnic heterophobia”) are not associated with a more frequent use of ethno-racial labels. Our results have substantive and methodological implications for the study of social networks and diversity in educational settings.

Close

关于学校种族隔离的研究通常认为，种族同质性——超越其他关系形成机制，与同族同龄人交朋友的倾向——与显着的种族界限相关。我们基于一种新的民族群体观念衡量标准，设计了对这一假设的更直接的测试。在一项针对德国大都市区 39 所学校的 3000 多名学生的网络研究中，我们要求学生指出他们在学校年级中感知到的派系，并用自己的语言描述这些群体。我们发现，民族种族标签更有可能针对更大的派系，其中包括更高比例的穆斯林学生或更多具有更强种族认同的学生。尽管如此，无论是绝对的还是相对于其他分类模式，民族-种族标签都很少被使用。而且，友谊中的净种族隔离（“种族同质性”）和不喜欢关系中的相反模式（“种族异性恐惧症”）与更频繁地使用种族标签无关。我们的结果对于社交网络和教育环境多样性的研究具有实质性和方法论意义。

关闭