SHARE 2.0 · METHODS & RESULTS

Better-described data does more science.

Start with the evidence a dataset makes visible at deposit time. Then inspect the results and compare the submitted instrument directly with SHARE 2.0.

See how it works Compare 1.0 vs 2.0 Open live metrics

SHARE 1.0 vs 2.0

Same records. Same scientific question. Cleaner score.

The reconciled comparison puts 154,569 records in both columns, traces the SourceOf event overlap, and labels every result by how direct the comparison really is.

154,569

records in both

362

shared events

9 → 11

repositories

See the full 1.0 vs 2.0 evidence

Methods

Start with the method.

SHARE scores the evidence a dataset makes visible at deposit time, then tests whether that evidence predicts what happens later. The outcome never earns a point.

candidate concepts

Start from established metadata practice.

The deck selection process reviewed 64 candidate concepts drawn from four established metadata standards. These are candidate concepts, not scored records.

Every scored signal must be

Depositor-controlled

Known at deposit time

Publicly machine-checkable

Meaningful for FAIR or reuse

Freeze the evidence at deposit time

The frozen SHARE 2.0 adapter was applied to 1,328,100 Zenodo records across Stewardship, Harmonization, Access, Reusability, and Engagement.

1,024,298 records carried observed file-evidence inputs.

Measure observed downstream reuse

Separate reuse by new teams

Follow researcher trajectories forward

Three metrics, one foundation

Dataset evidence becomes a durable research record.

S-Index

S-Impact

25 × 4

Score the evidence in one dataset.

SHARE is the equal-weight sum of 25 deposit-time evidence checks. It operationalizes evidence aligned to FAIR; it is not a complete FAIR maturity assessment.

Boundary

Later downloads, citation luck, field size, and reuse outcomes do not earn SHARE points.

Results

What changed when the method was tested?

Results release 0.1.0 · SHARE 2.0

Observed downstream reuse

Top-quartile SHARE datasets were 4.86× more likely to generate direct downstream reuse.

Direct reuse

3-year derivative

Direct downstream reuse

4.86×

1× no difference

95% CI 3.62–6.52

Comparison	Estimate	Observed events	Cohort
Top versus bottom SHARE quartile	4.86× (per +10: 3.29×)	471	154,570 deposits

The neutral line is 1×. These are relative associations, not an absolute reuse forecast.

How was this result measured?

More results from the v0.51 validation portfolio

Higher SHARE predicted a strictly verified later derivative.

Odds ratio per +10 SHARE points · 95% CI 1.42–3.92 · 96 strict events.

2.36×

The association persisted across one-, three-, and five-year windows.

Odds ratios per +10 SHARE points, with each clock starting at the dataset deposit date.

2.37× · 2.35× · 2.43×

The derivative association remained after field adjustment.

Pooled odds ratio per +10 SHARE points; strongest in health and biomedical at 2.59×.

1.81×

Higher S-Index identified later high-standard output.

Incidence-rate ratio per doubling · 95% CI 1.72–2.80 · 1,196 returning researchers.

2.19×

S-Impact was associated with future scientific contribution.

Future datasets shared; the deck also reports +8% senior-author works, +4.7% publications, and +4.1% top-10% cited works per doubling.

+17%

The pattern appears in NIH’s ClinicalTrials.gov registry.

Per +10 SHARE points: odds of declaring IPD sharing 1.71× (95% CI 1.55–1.89; 98,306 trials) and of recorded IPD availability 3.55× (95% CI 2.54–4.97; 151,597 trials).

1.71× · 3.55×

Equity that directs improvement

Make the path to high-quality sharing visible.

SHARE does not use funding status, country, or language as score inputs. This view identifies the repository conditions that can improve.

3.08%

high SHARE in selected group

Access is a practical improvement target.

3.08% compared with 0.01% among restricted records.

Operational next move

Make the access path clear, supported, and measurable.

What’s next

Score. Explain. Improve. Re-score.

The standard gets stronger when its evidence, changes, and future tests remain visible to the people who rely on it.

Score

A public deposit is assessed from the evidence it exposes at deposit time.

Explain

The record receives a receipt that shows the evidence behind the score and practical deposit-time fixes.

Improve

Creators and repositories can strengthen the metadata, access path, provenance, and reusable files they control.

Re-score

After evidence changes, the record can be re-scored and the improvement becomes inspectable.

Validate across the data ecosystem

Extend the frozen, evidence-first study pattern across more repositories, research fields, and reuse pathways.

Next evidence artifact

A repository-by-repository validation extension using the same frozen protocol.

Who can act now

Repositories can provide public metadata and relation pathways; collaborators can propose additional reuse signals.

Track quality sharing over time

Turn equity measures into repository action

Keep every point inspectable