Data availability statements: a template for every situation.

A data availability statement names the data behind your results and tells readers how to reach it. The formula is one sentence per dataset: what the data is, which repository holds it, the accession number or DOI, and any condition on access. When the data cannot be shared, the statement explains why and names the institutional body that handles requests. Most authors write this sentence inside the submission form, five minutes before the deadline, and it shows.

What a data availability statement must cover

Four things: what data exists, where it lives, how a reader gets it, and under what conditions. A statement that answers all four in two or three sentences is finished.

The statement is one of the standard declarations every submission system asks for, alongside funding, ethics, and author contributions; the first-submission checklist walks through the full set. This one differs from its siblings in a single way: editors can verify it. A funding declaration is taken on trust. A data statement contains an identifier anyone can paste into a browser, and increasingly someone does. Springer Nature now requires the statement in original research articles across its journals, and other large publishers ask for it on the submission form whether or not the manuscript template mentions it.

Cover every dataset the results depend on, not only the one you generated. If you analysed a public dataset, cite its accession number too. If your figures come from simulation output, the code that produced them counts as data for this purpose and needs its own sentence. And when part of the data is shareable while part is not, split the statement: one sentence for the deposited portion, one for the restricted portion. Journals accept mixed statements without complaint. What they push back on is vagueness.

Template: data in a public repository

Name the repository, give the persistent identifier, and state the licence or access level if the repository offers a choice. One sentence usually carries it.

The datasets generated and analysed during this study are available
in the [repository name] repository at [DOI or accession number].

If everything a reader needs sits inside the paper itself, say so, but only when it is literally true:

All data supporting the findings of this study are available within
the article and its supplementary information files.

Code gets the same treatment, with one extra precaution. Point the statement at an archived snapshot with its own DOI, through Zenodo or a similar archive, rather than at a live repository where the default branch can move after acceptance:

The analysis code is available at [archive DOI], corresponding to
version [tag] used to produce the reported results.

Specificity is the standard. Springer Nature's own example bank shows the level it expects: "PRO-Seq data were deposited into the Gene Expression Omnibus database under accession number GSE85337." A named database and a resolvable number. No interpretation required.

Template: restricted or sensitive data

State the restriction, then route requests to an institutional body rather than to yourself. Those are the two parts editors look for.

The data contain potentially identifying information about study
participants and cannot be shared publicly under the terms of our
ethics approval. De-identified data are available to qualified
researchers from the [institution] Data Access Committee at
[committee contact address].

The routing detail matters more than authors expect. PLOS puts it plainly in its data availability policy: "it is not acceptable for an author to be the sole named individual responsible for ensuring data access." The reasoning is practical. Your email address goes stale when you change institutions; a committee persists. In PLOS's words, a non-author institutional contact "helps guarantee long term stability and availability."

Notice that the template names the reason. A bare "cannot be shared for ethical reasons" invites a query asking which reasons. Identifying participant information under the terms of a named ethics approval closes the question before it is asked. Where a data protection law applies, cite the law itself.

Why is available upon request losing acceptance?

Because journals learned that the requests go unanswered. The bare sentence "data are available from the corresponding author upon reasonable request," offered with no reason and no institutional route, is now the weakest statement you can submit, and at some journals it is not accepted at all.

PLOS moved first. Since its data policy took effect in 2014, its journals have not accepted a plain on-request statement; restricted data must instead go through a named data access or ethics committee, with the restriction explained in the statement. Springer Nature still lists an on-request model sentence, but even that example names its justification: data "not openly available due to reasons of sensitivity." On-request wording now survives only where a restriction justifies it, and it holds up best when a controlled-access repository or a committee stands behind the promise instead of one person's inbox.

If nothing restricts your data and you simply have not deposited it, the fix is to deposit. A generalist repository such as Zenodo, Figshare, or Dryad takes an afternoon and returns a DOI you can paste straight into the first template above. A statement that promises nothing checkable belongs on the list of problems an editor can spot in the first five minutes.

Match the statement to what you actually deposited

Before you submit, open every identifier in the statement and confirm it resolves to the version of the data your results use. Five minutes, and it can save a full revision round.

The common mismatches repeat across fields. An accession number for a deposit still under embargo, with no reviewer access link supplied. A statement promising all underlying data while the repository holds only the processed tables. Code archived at a commit that predates the final analysis. A supplementary file renamed after the statement was written. None of these is misconduct. Each one briefly reads like it to a reviewer who clicks the link.

The statement also has to agree with the rest of the manuscript. If the methods section describes a survey instrument the statement never accounts for, or the statement cites a repository that appears nowhere in the references, someone will notice, whether editorial staff at intake or a reviewer three months later. Cross-document consistency of exactly this kind is what a pre-submission review against the target journal's own rules checks mechanically, before an editor gets the chance to.

The deeper fix is scheduling. Write the statement when you write the methods section, not on submission day. By then the deposits exist, the identifiers resolve, and the sentence describes something real rather than something planned. The statement written last is the one written worst.