CoolFace
Datasetpublic

mheilimo/grant-application-evidence-review-register

Grant Application Evidence Review Register Grant applications are easier to compare when every review item stays attached to the same criterion definition, version, evidence, date, state, reviewer and human decision. This open CSV resource gives public funding and grant evaluators a 20-item register for a consistent, documented first pass. The first twelve rows organise commercial evidence. The final eight keep programme-specific policy, eligibility, public-value and… See the full description on the dataset page: https://huggingface.co/datasets/mheilimo/grant-application-evidence-review-register.

sourceHugging Facecc-by-4.0updated 2mo agoView on Hugging Face
0likes26downloads
Dataset Card

Grant Application Evidence Review Register

Grant applications are easier to compare when every review item stays attached to the same criterion definition, version, evidence, date, state, reviewer and human decision. This open CSV resource gives public funding and grant evaluators a 20-item register for a consistent, documented first pass.

The first twelve rows organise commercial evidence. The final eight keep programme-specific policy, eligibility, public-value and accountability review outside that commercial layer. Programme rules differ. Replace the starter wording with the call's published criteria, rubric and current version before use.

Files

  • grant-application-evidence-review-register.csv contains twenty blank review items in the missing and unreviewed states.
  • fictional-example.csv demonstrates all six evidence states using synthetic applicant and programme data.
  • data-dictionary.csv defines the eighteen fields and their allowed values.
  • LICENSE.md contains the CC BY 4.0 licence notice.

Evidence States

  • supported: the cited evidence supports the criterion under the recorded definition and is current for this review.
  • partial: some required support exists, but a named part is absent.
  • missing: no reviewable evidence reference is recorded.
  • conflicting: two or more reviewed sources disagree.
  • stale: the evidence falls outside the selected review window.
  • not_applicable: a named reviewer has recorded why the item does not apply.

An evidence state is not an award decision. human_decision and decision_rationale keep the accountable evaluator or panel decision separate.

Suggested Workflow

  1. 1.Copy the blank register into the organisation's protected case system.
  2. 2.Replace every starter criterion with the call's exact published wording, rubric reference and current version.
  3. 3.Keep confidential applicant material out of a public copy. Use internal references or properly redacted artefacts.
  4. 4.Use one shared definition and version across reviewers. Duplicate the rows per reviewer when inter-reviewer comparison is required.
  5. 5.Record a specific gap or clarification request for every state that is not supported.
  6. 6.Keep commercial evidence, programme policy fit and the final human decision separate through the whole process.

Limits

This register does not verify evidence, decide eligibility, rank applicants, award funding, calculate a DDScore or replace policy assessment, procurement or legal review, an accountable evaluator, a panel or full due diligence. The fictional example is synthetic and contains no real applicant, company, reviewer or personal data.

For a separate commercial-viability first pass, DDScore turns submitted private-company materials into a structured 0–100 Due Diligence Score and full written report across 12 dimensions. It checks relevant claims against current public sources and shows confidence and gaps. The evaluator remains responsible for policy fit and the final decision.

Disclosure: I work on DDScore at Playful Pixels Oy.

See DDScore for evaluators