|
|
Charlotte CORCHETE , Sciences Po Paris
This study examines racial bias in French middle-school teachers’ decisions and explores whether structured grading rubrics and teacher experience can mitigate such bias. While ethnic inequalities in education are well documented in the French context, few experimental studies have investigated whether teachers' judgments contribute to these disparities. Drawing on stereotype activation/attenuation theory, we conducted a nationwide between-subjects survey experiment with French literature teachers, using realistic student vignettes and an essay from national exams. Participants (N = 631 so far, target N = 2,024) were randomly assigned to evaluate one student profile varying on three dimensions: student origin (French vs. North African name), behavior (positive vs. disruptive), and grading condition (with or without rubric). Teachers graded a real essay and rated the student’s academic potential and recommended track. Preliminary results show no significant difference in assigned grades by student origin. However, students with North African–sounding names were 72% less likely to be perceived as capable of passing the national exam and 41% less likely to be recommended for the general track, compared to otherwise identical students with French-sounding names. The gap in perceived exam success was significantly larger among novice teachers and tended to narrow with experience. While rubrics slightly lowered grades overall, they may help reduce disparities, though interaction effects were not statistically significant. These findings suggest that teacher bias affects critical outcomes beyond grading. They underscore the importance of structured assessment tools, equity-focused teacher training, and more strategic teacher assignment policies to avoid overexposing minority students to inexperienced educators.
Presented in Session 88. Migrant Populations and Ethnoracial Discrimination