Challenges in experimentally demonstrating racial bias in school discipline: Insights from a multi-study replication

In an effort to better understand the ways in which racial biases might play out within classrooms among teachers and students, we attempted to replicate an experimental paradigm by Okonofua and Eberhardt (2015). The original study provided persuasive evidence of the ways in which teachers' racial biases might accumulate over the course of multiple interactions, which can serve to damage student-teacher relationships and give rise to racial discipline disproportionality. In replicating their study procedures, participants provided their perceptions and recommendations after reviewing a disciplinary file in which a student misbehaved twice. The race of the student was manipulated as a between-subjects factor. Across four studies with diverse samples of both teachers and undergraduates, we found limited evidence of replication for the core findings of Okonofua and Eberhardt (2015). Although research suggests that racial disparities are pervasive in teachers' disciplinary actions, we did not find evidence of racial biases using this experimental paradigm. Namely, participants did not feel more troubled by the behavior, recommend more severe punishment, more readily label the student as a troublemaker, or suggest that they would be more likely to suspend the Black student, relative to the White student. We discuss the practical and theoretical implications of these non-replication findings and emphasize the importance of replication attempts to understand the boundary conditions of an effect. By sharing and making sense of non-replication findings within the broader literature, researchers have an opportunity to refine theoretical mechanisms of change, formulate the conditions under which an effect occurs, and, ultimately, enrich the science of bias and bias reduction within school settings.