A statement is only the beginning
A data availability statement (DAS) is often treated as a small section near the end of a manuscript. It may say that data are available in a repository, included in supplementary files, accessible on reasonable request or restricted for ethical or legal reasons. Yet the statement can determine whether readers and reviewers can inspect the evidence behind a paper.
Table of Contents
That is why journal platforms should treat data availability as a workflow, not a checkbox. A platform can help editors and authors create structured declarations, validate identifiers, confirm that a repository record can be found, capture reviewer access, and route exceptions to the right person. These checks can improve transparency and reduce avoidable follow-up work.
Major publishers already set expectations that go beyond a generic sentence. Nature Portfolio requires data availability statements for original research and asks authors to explain access to the minimum dataset needed to interpret, verify and extend the work [1].
Why structured data availability declarations matter
A free-text field is flexible, but it makes consistent checks difficult. One author may write “data available at DOI,” another may paste a long URL, and another may describe a controlled-access process in a paragraph. Editorial staff must interpret every statement manually, often under deadline pressure.
A structured form can capture the information behind the statement: where the data are held, the persistent identifier, whether access is public or controlled, the release date, any embargo, the responsible access committee, the reason for restrictions, and whether code or materials are also available. The platform can then generate a readable statement while retaining the underlying fields for checks and reporting.
The form should not force every paper into a public-repository model. Some research data cannot be openly released because of participant privacy, consent, security, contractual restrictions or third-party terms. PLOS explicitly notes that human participant data must not be shared in ways that compromise privacy and that authors should consult ethics bodies before sharing [4].
Five layers of data availability verification
The first layer is statement completeness. The system checks that the required declaration exists and that the selected route includes the expected information. A public-data route should include a repository name and persistent identifier or stable URL. A controlled-access route should explain the restriction and provide a process for requesting access. “Available from the corresponding author” may need additional details under the journal’s policy.
The second layer is identifier and link validation. The platform can resolve a DOI, check whether the landing page responds, and confirm that the identifier is not obviously malformed. Where DataCite metadata is used, its public REST API allows users to retrieve and search DOI metadata records [5].
The third layer is metadata consistency. Compare the title, creator, repository, publication year and resource type in the dataset record with the manuscript’s declaration. Flag possible mismatches for a human to review. A paper may legitimately cite a dataset created by a collaborator or use a versioned dataset, so automated checks should produce warnings and evidence rather than make an unsupported rejection decision.
The fourth layer is access verification. Can an editor or reviewer use the supplied private link? Is the record under embargo? Does the page explain controlled access? Does the repository require a login or approval?
The fifth layer is policy and privacy review. The platform should identify statements that mention personal, confidential, proprietary or third-party data and route them for appropriate assessment. A machine can flag terms and missing fields, but it cannot determine whether consent permits sharing or whether de-identification is adequate. Editorial policy, research ethics and, where necessary, specialist review remain essential.
Design the workflow around decision points
At submission, ask authors to select a data-sharing route: public repository, supplementary files, controlled access, third-party data with restrictions, data included in the article, or no new data generated. Each route should have a clear explanation and tailored fields. For public repositories, capture the DOI or accession number and allow a private reviewer URL where supported.
At editorial screening, run automated checks and show results in a concise dashboard. Use statuses such as “identifier resolved,” “metadata needs review,” “reviewer link inaccessible,” or “restricted-access explanation incomplete.” Avoid a single opaque score. Editors need to know what the system checked, when it checked it and what evidence produced the result. Every override should have a reason and a time-stamped audit record.
During peer review, provide reviewers with the data-access route appropriate to the manuscript and the journal’s confidentiality rules. Private links should have defined permissions and expiry periods. The platform should avoid exposing sensitive datasets to reviewers who are not authorized to see them.
At acceptance and publication, recheck data availability at the public destination and ensure that the final statement matches the released dataset. Where a repository supports embargoes, coordinate article publication with data release. If a DOI is still provisional, make the publication dependency visible so the production team knows what must be resolved before the article goes live.
A practical scenario: a restricted health dataset
Imagine a journal receives a clinical study based on patient records from a hospital network. The authors cannot make the raw records public because the consent terms and local ethics approval restrict sharing. They submit a DAS stating that de-identified data may be available through the hospital’s data-access committee, subject to an approved proposal and a data-use agreement.
A more capable platform offers a controlled-access route. It asks for the reason for restriction, the committee contact, the materials required for a request, the criteria for approval and any expected response period. It flags missing information, checks that the contact route is present, and sends the statement to an editor for review.
The platform records the reviewer’s access status and the final published statement. If the access committee later changes its process, the journal can update the guidance and identify papers that use an outdated request route. This illustrative scenario shows why verification must recognize legitimate restrictions rather than treat open release as the only acceptable outcome.
What platforms should measure
Useful measures of data availability include the percentage of submissions with complete statements, DOI resolution rates, metadata mismatch rates, reviewer-link failures, average time to resolve data queries, the share of statements requiring manual escalation, and the number of post-publication corrections to access information. Segment the results by journal, discipline and sharing route.
Start implementation with a small policy-aligned pilot. Work with editors, research-integrity staff, repository partners, production teams and privacy specialists to define routes and decision rules. Test common edge cases: provisional DOIs, embargoes, multiple repositories, third-party datasets, code repositories, supplementary files, controlled access and no-data studies. Make the system explain each flag and allow authorized human review.
The platform should also be transparent about the limits of its data availability checks. It can confirm that a DOI resolves and that a dataset record is discoverable. It may check selected metadata and whether a reviewer link opens. It cannot infer from a link alone that the files are scientifically adequate, unaltered in every respect or capable of reproducing the reported results.
The strategic takeaway
A data availability declaration is valuable only when readers can understand and use it. Journal platforms can turn a static statement into a traceable process by collecting structured information, checking identifiers and metadata, managing reviewer access, and routing restrictions for human review. The result is not a machine that certifies a paper’s truth.
Frequently Asked Questions
What is a data availability statement?
It tells readers whether and how the data supporting a study can be accessed, including repository details or reasons for restrictions.
Can a journal platform verify that a dataset is scientifically valid?
Not fully. A platform can check links, identifiers and selected metadata, but assessing whether data support the claims requires human scientific and editorial judgment.
Should every dataset be publicly available?
No. Privacy, consent, security, ethics, legal obligations and third-party agreements may require controlled access or other restrictions. The statement should explain the route clearly.
What can a platform check automatically?
It can check required fields, DOI or URL resolution, repository metadata, selected consistency rules, embargo information and whether a supplied reviewer link is reachable.
Why are private reviewer links useful?
They can give editors and peer reviewers access to a dataset before public release, including when a repository uses a provisional DOI or embargo workflow.
What should happen when an automated check fails?
The platform should show the evidence, create a tracked task and route it to the right editor or specialist. A failure should not automatically imply misconduct.
References and Further Reading
1. Nature Portfolio, “Reporting standards and availability of data, materials, code and protocols”
2. PLOS, “Data Availability” policy
3. PLOS, guidance on repository integration and private reviewer URLs
4. PLOS, “Data Availability” and privacy guidance for human research data
5. DataCite, REST API for DOI metadata retrieval
6. IOP Publishing, Research Data Availability Policy
Related reading: learn about preserving editorial history during peer-review platform migration.
