A data availability statement tells a reader where the data behind your results lives and under what conditions it can be accessed. Journals ask for it because a result that cannot be traced to its underlying material is hard to verify, reuse, or build on. This article gives you a submission-ready structure, a worked example, a checklist, and the completion criteria a reviewer or supervisor will implicitly apply.
The core statement has three parts: what exists, where it lives, and under what conditions someone can get it. Most rejections or revision requests come from one of these parts being vague. “Data available on request” can work, but only when you say who to contact, what will be shared, and how long it will take. A statement that names a repository, a persistent identifier, and a clear access route is stronger and needs no back-and-forth.
Where Do I Start If My Data Is Scattered?
Start by listing the materials your paper actually depends on. For a course paper or thesis chapter, this is usually a folder of cleaned spreadsheets, interview transcripts, survey exports, or analysis scripts. Write down the smallest set a reader would need to reproduce each figure and table.
Then sort the list into three groups: data you can share, data you can share only in a modified form, and data you cannot share under your ethics approval or institutional rules. UK Data Service guidance points to organized files, meaningful names, documented versions, and open or suitable formats as the preparation step that makes data understandable and reusable. Do that sorting before you write, because the statement must reflect what is actually shareable, not what you hope to share later.
If your material is still spread across folders and drafts, gather it in one place first. Save the relevant papers and supporting files in My Library so the papers you cite and the notes you took about your own data sit next to each other while you write the statement.
Which Access Model Fits My Project?
Pick one of four access models and state it plainly.
- Open repository: data is public, with a persistent identifier and a license. This is the strongest option and is the norm for quantitative work with no privacy risk.
- Repository with controlled access: data is deposited but requires an application. This fits de-identified clinical or sensitive social data.
- Available on request: data is not deposited but can be shared if a specific person is contacted. Include the contact, the request process, and the expected response time.
- Not available: the reader is told why, such as participant consent restrictions or commercial ownership. This must be a reason, not a preference.
A common revision request is “please specify the repository and accession number” after a statement only said “available on request.” Avoid that loop by naming the actual system and identifier when one exists. The FAIR principles treat persistent identifiers, rich metadata, provenance, and clear usage conditions as the features that make research objects findable, accessible, interoperable, and reusable. A named repository with an identifier satisfies the reviewer’s real question without extra explanation.
What Does a Strong Statement Look Like in Practice?
Worked example for a thesis chapter that used cleaned survey data and analysis scripts:
The de-identified survey dataset supporting the findings in this chapter, together with the R scripts used for cleaning and analysis, is deposited in the Open Science Framework under DOI 10.17605/OSF.IO/EXAMPLE. Files are organized by analysis step, with a README describing variable definitions, version history, and the original questionnaire. Data are available under a CC-BY 4.0 license.
This works because a reader can act on it: they know the repository, they know the identifier, and they know that documentation exists. If the data could not be opened, the same structure still works:
The interview transcripts analyzed in this study contain information that could identify participants and cannot be shared under the approved ethics protocol. De-identified summaries and the coding framework are available from the corresponding author on reasonable request; requests are typically answered within two weeks.
Here the restriction is attached to a named reason, and the request route is concrete.
How Do I Match the Statement to My Funder or Journal?
Check the target journal’s instruction to authors first, then any funder policy that applies to the work. The NIH data management and sharing policy expects planning around data management and sharing, with attention to related tools, software, standards, preservation, access, oversight, and responsible reuse. A journal statement is shorter than a data management plan, but it should be consistent with the plan you made.
When the journal template has fixed wording, use it exactly and add your specific repository, identifier, and restriction detail where the template allows. When there is no template, the three-part structure above is acceptable in most fields.
How Do I Turn My Scattered Files Into Shareable Assets?
A statement is only trustworthy if the named materials exist. Walk through each item in your statement and confirm the reader can reach it. Rename files so the content is clear without opening them, record version changes in a README, and convert proprietary formats where an open alternative preserves the information. If you used a proprietary qualitative coding file, export the codebook to a portable format and state which parts are shared.
For a winter-break cleanup or January project start, this is the highest-value hour: make the folder match the statement. If you need to ask questions about papers you collected while preparing your own dataset, Library QA lets you query materials saved in My Library or added from search results, which helps when you are reconciling your methods text with the papers you cited.
Which Tools Help Me Reconstruct What I Actually Did?
The statement should match the methods section and the analysis you ran. If you cleaned and analyzed data several months ago, reconstruct the exact steps before you write. Re-run the scripts if they exist, and note where a figure came from. When a script or dataset came from a published workflow, check the original paper’s own data availability statement to see how the authors described it.
If you are starting a new project in January and want the statement to be easy to write later, create a dedicated space now. Use Research Projects to set up the project where you will keep the papers, notes, and dataset descriptions together from the first analysis rather than reconstructing them at submission.
What Do Reviewers and Supervisors Actually Check?
Supervisors and reviewers check for internal consistency, not elegance. They compare your statement against three places in the manuscript: the methods section, the figure captions, and the supplementary materials. A statement that mentions a repository but a figure caption that says “data not shown” creates doubt. The same applies to a statement that promises scripts but the appendix contains none.
They also check that a restriction is attached to a reason. “Data available on request from the authors” is weak when the paper involves no human participants and no third-party data, because there is no explanation for why the data is not simply deposited.
To find that mismatch yourself, search your paper for every place a result appears and trace it back to the statement. When the trace is unclear, fix either the statement or the methods text.
How Do I Know the Statement Is Done?
Use these completion criteria. The statement is ready when every one of these is true.
- Every dataset, script, and instrument that a figure or table depends on appears somewhere in the statement or in a clearly linked repository record.
- Each shared item has a persistent identifier or a repository name plus a file path a reader can follow.
- Any restriction names the reason, the contact route, and the expected response time.
- The statement matches the methods section, figure captions, and supplementary materials.
- The wording follows the journal template or funder policy when one exists.
- A stranger reading only the statement could find or request the data without reading the rest of the paper.
Run a final consistency pass with your paper open next to the statement. When a cited reference is involved, verify the context before you rely on it in the methods text. If you are checking BibTeX entries for a submitted manuscript, TrueCite screens BibTeX against real academic databases to flag hallucinated references, so a statement describing a paper’s data is not built on a citation that does not exist.
Reusable Readiness Checklist
| Area | Check | Ready when |
|---|---|---|
| Inventory | List every dataset, script, and instrument behind each figure and table | Nothing in the results lacks a trace |
| Access model | Choose open, controlled, on-request, or not available | The choice fits the data’s sensitivity level |
| Repository | Name the repository, persistent identifier, and license | A stranger can reach the files |
| Documentation | Provide a README with variables, versions, and provenance | Files make sense without verbal explanation |
| Restrictions | Name the reason, contact, process, and response time | The restriction is traceable to ethics or ownership |
| Consistency | Compare statement with methods, captions, and supplementary materials | No contradictions remain |
| Policy fit | Match journal or funder wording and requirements | The statement survives a compliance check |




