Data Governance
How we source, normalize, score, and maintain the UQLI™ dataset.
Data Sources
Methodology Summary
Normalization: Winsorized (p2/p98) min-max normalization; floor score 2.0 (no city scores 0)
Dimension aggregation: Arithmetic mean of normalized indicator scores within each dimension
Composite score: Weighted geometric mean across 10 dimensions (0–100 scale)
Confidence rating: Based on indicator coverage — High (≥80%), Medium (50–79%), Low (<50%)
Full methodology documentation is available at /methodology.
Data Cycle
UQLI scores are refreshed annually. The current active cycle is 2024. Historical versions are retained and accessible via their methodology version DOI. Each cycle includes a full re-normalization against the current year's data.
Quality Controls
- Automated drift detection: score changes >15% trigger manual review
- Peer review required for any methodology version change
- All source data timestamped and archived at ingestion
- Confidence ratings computed per-city based on indicator completeness
Reproducibility & Citability
All methodology versions are archived with a DOI (format: 10.UQLI/ISO3-YEAR-vVERSION). Scores are DOI-citable and can be verified at universalqli.com/verify.
Data Corrections
To report a potential data error, email data@universalqli.com with the city, indicator, and supporting evidence. Corrections are reviewed and processed within 30 days.
Open Data Commitment
All composite scores, global ranks, and confidence ratings are publicly accessible. Dimension-level scores are available to Research+ subscribers. Raw indicator data is subject to source licensing agreements and is not redistributed.