Unveiling bias: the impact of blinded marking on university examen evaluation
How to cite
Pacheco, Tomas; Rossi, Martín; Llamas, Paola. 2024. "Unveiling bias: the impact of blinded marking on university examen evaluation." AEA RCT Registry. AEARCTR-0013553. https://doi.org/10.1257/rct.13553-2.0
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
This experiment at the University of San Andrés investigates the effects of blind versus non-blind grading methods on the persistence of grades across undergraduate courses, which are randomly distributed into control and treatment groups. In the control group, instructors can easily identify their students during grading, as students are required to write their full names on their exams. Conversely, in the treatment group, students use anonymized ID numbers, which obscures their identities from the graders. The study evaluates various outcomes, including the overall level of grades, fluctuations between midterm grades, alignment with historical GPA, and the incidence of repeating grades. Additionally, the research includes an analysis of how different grading methods influence these outcomes in relation to student characteristics such as gender, age, financial aid status, and special educational needs (SEN).
Intervention
This experiment involves all undergraduate courses at the University of San Andrés. Courses are randomly assigned to two groups, treatment and control, using a computer. This experiment aims to analyze the persistence of grades within courses. That is, we want to analyze whether students' grades on the second midterm are positively related to the grades they received on the first midterm.
Control Group: Teachers will have the following information about their students: full name, email, major, and student photo. In these courses, when taking the first and second midterms, students will write their full names on their papers. Thus, teachers can fully identify their students while grading the exams. When grading the second midterm, they can also fully identify the students. That is, both assessments in the course are traditional (or non-blind).
Treatment Group: In courses assigned to the treatment group, teachers will have access to a complete list of their students from the beginning of the semester with the following information: name, email, major, and photo. In these courses, when taking the first midterm, students will write their student ID number (identifier) on their papers. Thus, teachers cannot fully identify their students while grading the exams. When grading the second midterm, students will write a different ID number on their papers. Thus, teachers cannot identify their students during both assessment instances.
This analysis will include an examination of differences in grades based on observable student characteristics such as gender, age, financial aid status (full scholarship), and special educational needs (SEN). We will show econometric results for all courses and without math courses.
Experimental design
We will take all undergraduate courses at Universidad de San Andrés and conduct a random assignment without stratification so that half of the classes must grade the blind exam and the other half do not. For this intervention, we will have N=x classes to draw, so x/2 will belong to the treated group and x/2 to the control group.
Design details
Cited by
No citation data available for this registration yet.