Research data
A practical route to responsible research-data sharing
Useful data availability begins with planning: what can be shared, under which safeguards, and how the published claim can be verified.
Classify before sharing
Identify whether files contain personal, clinical, confidential, proprietary, culturally sensitive or security-relevant information. Remove direct identifiers, assess re-identification risk and respect consent, law, contracts and institutional obligations.
Open publication is not a reason to expose participants. When data cannot be public, describe the legitimate restriction and any controlled route through which qualified access may be considered.
Make reusable material intelligible
A table without a data dictionary may be technically available but scientifically unusable. Define variables, units, codes, missing values, transformations, software versions and the relationship between raw, processed and analytic files.
Keep analysis code beside environment information and instructions when sharing is appropriate. Avoid placing passwords, access tokens, hidden identifiers or licensed material in a public package.
Write an accurate availability statement
State what is available, where it can be found, which version supports the article and under what conditions it may be used. If nothing can be shared, explain the reason without promising access that cannot be delivered.
Before publication, test the link and compare the deposited files with the final methods, results, figures and tables. Update the record when a corrected dataset or code version is released.
Practical checklist
Practical checklist
- Sensitivity and consent reviewed
- Identifiers and re-identification risk assessed
- Variables, units and missing values documented
- Code and software environment described
- Availability statement matches the actual route
- Final links and versions tested
Official readings
Official readings
Selected sources for further reading.