Maintenance
Part of Science and technology announcements: a reporting guide
Science announcement reporting checklist
Science announcement checklist for source status, methods, sample, comparison, uncertainty, funding, conflicts, data, practical readiness, headlines, and updates.
What to take away
- Confirm the underlying record and version.
- Translate the main result with its denominator and uncertainty.
- Check funding, affiliations, patents, and data access.
- State the tested setting and readiness stage.
- Schedule checks for corrections and later versions.
Use this list before publishing a research, engineering, standards, or mission announcement. A missing field can be labeled unavailable, but it should not be replaced by a guess. The fields distill the science announcement reporting guide.
Source status
- Release issuer and publication date are recorded.
- Paper, preprint, dataset, report, patent, standard, or award is open.
- Title, authors, identifier, and version match the release.
- Peer-review status is stated.
- Corrections, withdrawals, or retractions were checked.
Methods and result
- Research question or engineering objective is clear.
- Sample, material, dataset, or test unit is defined.
- Comparator or benchmark is described.
- Primary outcome is separated from secondary analysis.
- Effect size has a unit, denominator, and uncertainty.
- Failed tests and exclusions are not hidden.
Numeric fields here mirror data-release work; the economic release reporting checklist is the parallel discipline.
Methods and result checks
- Research question or objective clear
- Sample, material, dataset defined
- Comparator or benchmark described
- Primary outcome separated
- Effect size has unit and uncertainty
- Failed tests and exclusions not hidden
Interests and provenance
- Funding sources are named.
- Author employment and relevant patents are checked.
- Declared conflicts are summarized.
- Data ownership and collection source are clear.
- Institutional support is not described as endorsement of every claim.
NSF's award search explains that award abstracts are public records while proposals remain confidential intellectual property, and that listed publications can have separate copyright. The NSF Award Search overview supports distinguishing an award abstract from the underlying proposal, research outputs, and publisher record.
Record the award number and reporting period when funding is material. Then check whether the announced result was produced under that award or merely concerns a related topic. Multiple funders can support different parts of one project. A public abstract may describe planned methods that later changed, so it should not replace the final paper's method section.
For a company collaboration, identify who supplied equipment, data, personnel, or analysis. Ask whether the institution controlled publication and whether agreements limited data access. These facts help readers assess provenance without assuming that financial support invalidates or validates the result. Provenance questions start with the release itself; how to read a science press release orders them.
Readiness and use
- Simulation, laboratory, field, pilot, or deployed setting is named.
- Scale and test duration are included.
- Reliability, safety, cost, power, maintenance, and training limits are checked.
- Required approval or standardization is identified.
- A proposed application is not written as an available product.
When readiness is far off, the story selection checklist may point to a narrower piece or none at all.
Data and reproducibility
- Data and code availability statements were read.
- Repository, identifier, license, and version are recorded.
- Independent replication is distinguished from repeated tests by the same team.
- Documentation permits the claimed analysis.
NASA's Planetary Data System guidance for data proposers explains that archive content and documentation undergo scientific peer review for validity and usability and must meet defined archive standards. This supports checking documentation and archive status rather than treating a mission team's chart as the full reusable dataset.
Headline and update
- The headline names the actual evidence stage.
- First, proves, solves, and breakthrough are supported or removed.
- Limitations appear before distant applications.
- A correction and version-monitoring date is assigned.
- Media captions identify simulations, illustrations, and file images.
Common questions
Does unavailable data invalidate a story?
Not automatically. State the access limit and narrow claims that cannot be independently checked.
Is an award abstract the same as a proposal?
No. It is a public summary; proposals can contain confidential intellectual property.
What is the most important engineering field?
The tested conditions and comparison. Performance without them is hard to interpret.
When should the article be revisited?
At journal publication, correction, data release, replication, standards approval, major test, or deployment decision.
