Potential Adverse Effect of Exposure to Females: An Analysis of a Microscopic Peer Relationship in Teacher Evaluation
How to cite
Keller, Tamas. 2025. "Potential Adverse Effect of Exposure to Females: An Analysis of a Microscopic Peer Relationship in Teacher Evaluation." AEA RCT Registry. AEARCTR-0016301. https://doi.org/10.1257/rct.16301-1.0
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
Girls are widely believed to deliver beneficial peer effects due to their higher school performance, greater effort, and better school behavior. This project explores a potential negative peer effect of being exposed to girls in a microscopic peer relationship that is present in nearly every educational setting: being evaluated immediately after a girl.
The study asks how teacher evaluation of a focal student differs when this student is assessed after a female student. This is a relevant research question, as teachers often evaluate students’ written assignments sequentially, one after the other.
It is hypothesized that being evaluated after a female student raises teachers’ evaluation standards, potentially lowering the grade received by the focal student.
The hypothesis is tested in a vignette study in which teachers evaluated six students’ essays, randomized as good, average, or poor quality. The gender of each student was visible via first names, randomly assigned to be either male or female with equal probability. The gender of each vignette student was randomized with replacement, meaning there was no predetermined sequence or pattern between the focal student and the one evaluated before or after.
By regressing the teacher-assigned grade on the focal student’s gender and the gender of the student evaluated immediately before (the treatment), we can identify whether the previously evaluated student’s gender influences the evaluation of the focal student.
Intervention
The treatment is the gender of the student previously evaluated by the same teacher, coded as 1 if the student was a girl and 0 if the student was a boy.
Experimental design
In the vignette study, teachers evaluate six students’ essays, randomized as good, average, or poor quality. The gender of each student was visible via first names, randomly assigned to be either male or female with equal probability. The gender of each vignette student was randomized with replacement, meaning there was no predetermined sequence or pattern between the focal student and the one evaluated before or after.
Design details
Cited by
No citation data available for this registration yet.