A citation-based metric complementing the S-Index, measuring how much research datasets are actually reused by the community.
A researcher has an S-Impact of k if k of their datasets have at least k citations each.
Like the h-index for publications, the S-Impact rewards researchers whose shared datasets are consistently cited and reused by the community.
Example Calculation
Researcher has 6 datasets with citation counts: Sorted descending: [12, 8, 5, 3, 1, 0] Position: 1 2 3 4 5 6 Check each position: Position 1: 12 >= 1? Yes ✓ Position 2: 8 >= 2? Yes ✓ Position 3: 5 >= 3? Yes ✓ Position 4: 3 >= 4? No ✗ S-Impact = 3 (3 datasets with 3+ citations each) Citation: S-Impact(v1.0) = 3 (as of 2026-02-06, n=6 datasets)
Complements S-Index
S-Index measures input quality (how well datasets are shared). S-Impact measures output quality (how much datasets are reused). Together they give a complete picture.
Citation-Based
Counts formal citations from publications that reference datasets. Sourced from DataCite, Crossref, and repository-native citation tracking.
Gaming Resistant
Like the h-index, you need k datasets each with k+ citations. Self-citation alone cannot significantly inflate S-Impact.
Repository-Agnostic
Citations are tracked across all sources regardless of which repository hosts the dataset. Cross-repository impact is fully captured.
| S-Impact | Badge | Description |
|---|---|---|
| 5+ | Elite | Exceptional citation impact across datasets |
| 3-4 | High Impact | Strong evidence of dataset reuse |
| 2 | Emerging | Growing citation footprint |
| 1 | Initial | First cited dataset |
| 0 | Uncited | No tracked citations yet |
When combined with the S-Index, S-Impact creates four quadrants that characterize a researcher's data sharing profile:
High S-Index + High S-Impact
10
researchers
High S-Index + Low S-Impact
1,693
researchers
Low S-Index + High S-Impact
992
researchers
Low S-Index + Low S-Impact
30,829
researchers
| Property | S-Index (Sharing Quality) | S-Impact (Citation Impact) |
|---|---|---|
| Definition | n datasets with SHARE score >= n each | k datasets with citations >= k each |
| What it measures | Data sharing quality and consistency | Dataset citation impact and reuse |
| Focus | Metadata completeness (depositor effort) | Community uptake (outcome metric) |
| Scale | 0-100 (bounded by SHARE score max) | Unbounded (driven by citation counts) |
Key Insight: The S-Index measures input quality (how well data is shared), while S-Impact measures output quality (how much data is reused). A researcher in the “Excellence” quadrant excels at both.
401,281
Researchers Analyzed
5
Highest S-Impact
33,524
With 3+ Datasets
10
Excellence Quadrant