Newest Issue of JSDSE Focuses on Sharing Deidentified Data, Code

The July 2024 issue of the Journal of Statistics and Data Science Education starts with an editorial co-authored by Nicholas J. Horton and Sara Stoudt that discusses the journal’s requirements (instituted in 2022) for authors to share deidentified data and code that underlies their papers. These changes—prompted by an increased focus on reproducibility and open science—are intended to facilitate validation, replication, and reproducibility and are being instituted by many other journals, including several published by the American Statistical Association.

Horton and Stoudt also discuss why the requirements for code and data sharing were instituted, summarize ongoing trends and developments in open science, describe options for data and code sharing, and share advice for authors.

Key take-aways for authors include the following:

• Authors are asked to share a minimal deidentified data set that can replicate all study findings reported in the article.
• In addition to the deidentified data, all code used to analyze the data and generate results from the manuscript should be shared.
• Deidentified data and code should be deposited in a repository such as the Open Science Framework that fully implements FAIR (Findable, Accessible, Interoperable, and Reusable) data principles. (GitHub does not fully implement the FAIR data principles.)
• Other meta-data and supplementary materials (e.g., survey instruments, codebooks) can be included in the repository.
• The Open Science Framework allows anonymization of author names to allow the link to the data and code for a reproducibility check during the review process.
• A data availability statement should be included at the end of the manuscript before the references and in the metadata for the paper in the submission system (the editorial includes sample examples of wording).
• “Share upon request” is not an acceptable option for data sharing.
• For qualitative data projects, it may be appropriate to share the deidentified coded data rather than video or transcripts.
• The journal’s policies acknowledge there is a need to balance both transparency and reproducibility with data privacy. The editorial discusses ways to address sensitive data and describes the journal’s policy to request a waiver of data sharing.
• Researchers are encouraged to carefully review their institutional review board protocols to ensure data sharing plans are clearly described.

The editors close by noting best practices to foster reproducibility and replicability are fast changing and require additional efforts by authors, reviewers, and editors. The requirements to share deidentified data and code are a necessary but insufficient requirement to foster improved computational reproducibility. However, the editors think these changes can and should be undertaken and will help balance privacy and sharing in a way that fosters better science.

Other papers published in the issue include the following:

• “The Teaching of Introductory Statistics: Results of a National Survey” (Chelsey Legacy, Laura Le, Andrew Zieffler, Elizabeth Fry, and Pablo Vivas Corrales)
• “In Pursuit of Campus-Wide Data Literacy: A Guide to Developing a Statistics Course for Students in Nonquantitative Fields” (Alexis Lerner and Andrew Gelman)
• “Can You Trust Your Memory?” (Jeff Witmer)
• “Culturally Relevant Data in Teaching Statistics and Data Science Courses” (Travis Weiland and Immanuel Williams)
• “Implementation of Alternative Grading Methods in a Mathematical Statistics Course” (Brenna Curley and Jillian Downey)
• “Questions (and Answers) for Incorporating Nontraditional Grading in Your Statistics Courses” (Brenna Curley, Jillian Downey, Katherine M. Kinnaird, Adam Loy, and Eric Reyes)
• “Teaching Reproducible Methods in Economics at Liberal Arts Colleges: A Survey” (Anthony Underwood, Aidan Sichel, and Emily C. Marshall)
• “Data Analytics and Programming for Linguistics Students: A SWOT and Survey Study” (Dennis Tay)
• “Building Capacity for COVID-19 Surveillance: A Statistics Course for Health Officials in Seven Low- and Middle-Income Countries” (Isabel R. Fulcher, Donald Fejfar, Nichole Kulikowski, Jean-Claude Mugunga, Michael Law, and Bethany Hedt-Gauthier)
• “Interview with Hollylynne Stohl Lee” (Allan Rossman)