Research data and reproducibility
Most people meet these topics as a form to fill in. The form is easier, and the project is better, when you know what the requirement is actually protecting against.
Updated
A data management plan is usually written the week a grant is due, by someone who has never lost data and cannot yet imagine how. The result is a document that satisfies the funder and helps nobody.
These pages try to make the requirements legible. What a reviewer is checking in a data management plan and how long it needs to be. What FAIR asks for beyond the acronym, and why FAIR does not mean open. Which file naming habits still work three years in, when the project has four collaborators and two rewrites behind it.
Two pages go further than compliance. Preregistration is treated as a research design question rather than a form, including what it does and does not prevent you from doing afterwards. Anonymizing qualitative data is treated as the genuinely hard problem it is, because removing names is the easy half and rarely the part that identifies someone.
How do I write a data management plan?
What a data management plan has to cover, what funders actually check when they read one, and how long the document needs to be to satisfy a review panel.
Read the answer →What are the FAIR data principles?
What FAIR data actually requires beyond the acronym, what making data FAIR involves in day-to-day practice, and why FAIR does not mean open.
Read the answer →How do I organize research data files?
How to organize research data files so the structure still works in year three, what makes a good file naming convention, and how to version files without chaos.
Read the answer →What is preregistration in research?
What preregistration involves, whether it stops you exploring your data, and how a registered report differs from an ordinary preregistered study.
Read the answer →How do I anonymize qualitative interview data?
How to anonymize interview transcripts properly, the difference between anonymization and pseudonymization, and whether anonymized transcripts can be shared.
Read the answer →