Funding dimension: absent metadata ≠ absent funding

[aadivar]

Problem

The funding dimension (20/100 points) assumes all published research has external funding. This systematically disadvantages humanities, arts, and social science publishers whose content is legitimately unfunded.

Crossref's coverage API reports "X% of DOIs have funder metadata" but makes no distinction between:

• Missing metadata — funded research where the publisher didn't deposit acknowledgements (a real gap)
• Unfunded research — work that genuinely had no external funding (not a gap)

A humanities publisher with perfect metadata everywhere else caps at 80/100.

Why this is hard to fix with current data

• No discipline signal at the member level. The Crossref Member API doesn't include subject classifications. Journals have optional ASJC codes, but publishers often span multiple disciplines.
• No "unfunded" assertion in the Crossref schema. Publishers can deposit funder info when it exists, but can't declare "no funding received."
• Per-content-type breakdowns help but don't solve it. We can see that book chapters score differently from journal articles, but a philosophy monograph and a biomedical monograph get the same treatment.

This is structurally similar to #4 (ORCID penalizing institutional authors) — the scoring assumes a model of individual, funded academic authorship that doesn't hold across all scholarly publishing.

Possible approaches

1. Content-type-aware weight adjustment — Reduce funding weight for content types structurally less likely to be funded (monographs, book chapters, edited books), redistribute points to other dimensions.
2. Threshold-based normalization — If funding coverage is near-zero (<5%) across all content types, treat the dimension as N/A and normalize the score out of 80.
3. External enrichment — Use OpenAlex or ASJC subject classifications at the journal level to infer discipline mix and adjust weights accordingly.
4. Methodology note — At minimum, document this limitation on the site.

Related

• Funding: enable both FundRef and ROR as indicators #2 (FundRef/ROR overlap in funding indicators)
• Data lacks context #3 (broader data context concerns, including funding)
• People dimension: account for institutional authorship #4 (ORCID penalizing institutional authors — same structural issue)

References

• Barcelona Declaration on Open Research Information — notes funding metadata as a unique challenge because funding agencies don't provide standardized metadata
• Crossref funding schema: https://www.crossref.org/documentation/schema-library/markup-guide-metadata-segments/funding-information/

Source

Raised by Andy Byers on LinkedIn

[AhemNason]

Again, lots of humanities research is funded.

[AhemNason]

I think the bias here might be in your perception of humanities research and publishing.

[aadivar]

You're right, and I framed this wrong. The issue title itself encoded the bias you're pointing out — "humanities bias" implies funding is a humanities problem, when it's not. Plenty of humanities research is funded, and plenty of science isn't.

The real issue is simpler: absent funding metadata doesn't mean absent funding. That's true across every discipline. Renamed this issue to reflect that.

Where I'm leaning right now:

• Funding weight at 20 feels too high — something closer to 10 might be more honest, with those points going toward provenance and access
• Publisher pages should surface context: a low funding score may reflect unfunded research or undeposited metadata, not a quality gap

This connects to the broader point you raised in #3 about weights and transparency — responding there too.