What is it about?
This article challenges the default use of U.S. Census race categories in research and data collection, arguing that these categories are not scientific but rooted in political and colonial histories. The authors offer concrete guidance for data librarians, educators, and researchers on how to recognize and disrupt the harms caused by reductive or habitual use of racial data. Using a critical framework known as QuantCrit (Quantitative Critical Race Theory), the paper outlines how racial categories can obscure disparities rather than clarify them, and how data consultations and education can help shift those patterns.
Featured Image
Photo by Pietro Jeng on Unsplash
Why is it important?
Researchers often include race as a variable without reflecting on what it means or where the categories come from. This can lead to misleading findings, reinforce stereotypes, and overlook real sources of disparity. This paper urges a shift away from "default" demographic datasheets and calls on data professionals to advocate for intentional, reflective research design. It also introduces a practical model, based on the Model of Domain Learning (MDL), to help researchers move from rote data habits to informed, ethical practices. The article doesn’t just critique, it provides a roadmap for constructive change.
Perspectives
This work is both personal and professional for me. As a data consultant and someone who has seen firsthand how unchecked data practices perpetuate harm, I believe we have a responsibility to do better. We can’t build equitable research on colonial foundations. This paper is a call to action, especially for data librarians and research consultants, to ask better questions, educate gently but firmly, and help research teams rethink what race data is really for.
Dr Sam Leif
Read the Original
This page is a summary of: Data Consultations, Racism, and Critiquing Colonialism in Demographic Datasheets, Journal of eScience Librarianship, November 2021, University of Massachusetts Medical School,
DOI: 10.7191/jeslib.2021.1213.
You can read the full text:
Contributors
The following have contributed to this page
