Data Rot is a Central Issue of Publicly Accessible Open State School Education Indicator Datasets

• In Event: Data, AI, and Power: Critical Perspectives on Equity and Influence in Education

Abstract

Public education agencies publish datasets that measure education indicators associated with student opportunities and outcomes. Datasets published by federal and state sources ignore Findable, Accessible, Interoperable, and Reusable (FAIR) data principles, making them susceptible to “link rot,” which threatens data accessibility and reuse. Dataset URLs from one federal and eight state education data repositories were collected and manually checked after several months to observe which dataset URLs remained active. This study finds that of the 3,724 dataset URLs collected, 2,228 URLs rotted, for a total of 61.7% link rot.

Authors

• Alex J. Bowers, Teachers College, Columbia University

• Ghipsel Cibrian, Teachers College, Columbia University

• Maeghan Sill, Teachers College, Columbia University

• Yeonsoo Choi, Teachers College, Columbia University