Whitepaper
Research Funding Impact System
This system evaluates research funders through an evidence-first workflow. It organizes funder identity, source evidence, structured findings, scoring assessment, and reviewed profile cards into one transparent process.
Step 1
Entity Registry
Database
The system starts with one canonical row per funder-like entity: name, aliases, entity type, country or region, website, domain, durable identifiers, evidence path, status, and provenance.
Step 2
Evidence Collection
Evidence Room
Each entity gets an evidence package with source register, search plan, website pages, combined evidence markdown, and manifest metadata. Missing sources and failed searches are recorded as findings.
Step 3
Feature Blocks
Feature Gallery
Evidence is converted into structured feature blocks with summaries, evidence level, claim confidence, source URLs, positive findings, missing findings, and reviewer notes.
Step 4
Scoring And Assessment
Impact / $ Leaderboard
The scoring layer selects or flags Funding Capacity, computes approved impact-dollar metrics, calculates completeness, and emits evidence notes.
Step 5
Review And Entity Card
Entity Cards
The reviewer pass checks source traceability, Funding Capacity, entity-boundary risks, and missing evidence before an entity card is used.
Approved Scoring Outputs
Rather than asking readers to trust a single opaque number, the system separates the score into inspectable parts: Funding Capacity, points per dollar, impact-per-dollar score, completeness, evidence notes, score notes, and Funding Capacity.
Funding Capacity
The dollar capacity value used for efficiency analysis. It can come from grants paid, cumulative funding, assets, endowment, revenue, donations received, expenses, public appropriations, awards, or another sourced money amount.
Points Per $1M
Impact numerator points normalized by Funding Capacity in millions.
Impact-Per-Dollar Score
A discounted efficiency score derived from points per $1M, evidence confidence, and funding comparability.
Completeness Index
IPD Completeness: Funding Capacity quality, impact proxy evidence quality, and identity/context/source traceability on a 0-100 scale.
Evidence Notes
Explicit notes for missing Funding Capacity, source coverage, entity-boundary ambiguity, untraceable claims, or scoring review needs.
Feature Blocks
Feature blocks are reusable evidence summaries for display, filtering, completeness assessment, and scoring interpretation. Each block should carry source URLs, confidence, evidence level, findings, missing findings, and reviewer notes.
Review Rules
Every display claim must trace back to a source URL or an explicit missing-evidence note.
Entity identity and boundary uncertainty must remain visible when it affects funding or output interpretation.
Feature blocks are structured evidence summaries, not final public scores.
The scoring layer should compute only the approved current score outputs.
Impact Per Dollar quality should be communicated through Confidence.
Product Surfaces
The interface separates source review, feature interpretation, and score inspection so users can see where a conclusion came from before trusting a ranking.
Entity Cards
Funder profiles that summarize identity, evidence quality, source highlights, missing evidence, and scoring outputs.
Impact / $ Leaderboard
A ranking and review surface for Funding Capacity, points per dollar, completeness, and confidence.
Feature Gallery
A structured view of the evidence blocks behind each card, including confidence, evidence level, findings, and missing evidence.
Evidence Room
A source-review area for evidence packages, source registers, excerpts, search plans, and provenance checks.
Database
The entity registry and workflow coverage view for searching funders, identifiers, regions, entity types, and review status.
What The User Should Trust
A useful entity card says what the funder is, what evidence was found, what evidence is missing, what Funding Capacity was selected, what outputs or outcomes were observed, how complete the evidence package is, and what evidence supports the score.
What The System Should Not Hide
Funding evidence gaps, failed source search, self-reported-only source, untraceable claim, or ambiguous entity boundary is part of the result. The interface should surface those limits instead of smoothing them away.