The primary paper usually gets authorship right, because everyone is present and paying attention.
The problems arrive eighteen months later, when funding has stopped, people have dispersed, and someone writes a secondary analysis using the shared dataset.
Agree the after-rules in advance
- Who may use the dataset after the project ends, and for what.
- Whether notification is required before a new analysis starts.
- Default authorship expectations for later papers.
- How long these arrangements remain in force.
- Write it into the original agreement, not into an email later.
Why credit erodes over time
- People who moved on are harder to contact.
- Whoever still has funded time does the writing and becomes lead author.
- Original contributors are remembered less precisely as time passes.
- Nobody intends the erosion; it is a product of drift.
Practical protections
- Maintain a contribution record while the project runs.
- Keep an up-to-date contact list including personal addresses.
- Agree that data generators are offered co-authorship on analyses using their data.
- Set a notification requirement before new work starts.
If you are the one writing later
- Contact everyone who contributed to the data, even if it is inconvenient.
- Offer authorship where their contribution warrants it.
- Accept a decline gracefully, and acknowledge them.
- Send the draft before submission, not after.
If your contribution is being overlooked
- Raise it directly and early, in writing.
- Refer to the original agreement.
- Most cases are oversight and resolve once raised.
- If it does not resolve, institutional processes exist on both sides.
One thing worth remembering
Keep personal contact addresses for everyone on the project, not only institutional ones.
Institutional email dies when people move, and "we could not reach them" is the most common and most convenient explanation for why someone was left off a paper written three years after the data were collected.
Câu hỏi thường gặp
When do authorship problems typically arise?
Eighteen months after the project ends, when funding has stopped, people have dispersed, and someone writes a secondary analysis using the shared dataset.
What after-rules should be agreed in advance?
Who may use the dataset afterwards and for what, whether notification is required before a new analysis, default authorship expectations for later papers, and how long the arrangements last.
Why does credit erode over time?
People who moved on are harder to contact, whoever still has funded time does the writing and becomes lead author, and original contributions are remembered less precisely.
What should you do if writing a later paper?
Contact everyone who contributed to the data even if inconvenient, offer authorship where warranted, accept a decline gracefully with acknowledgement, and send the draft before submission.
Why keep personal contact addresses?
Because institutional email dies when people move, and we could not reach them is the most convenient explanation for leaving someone off a paper written years later.
