ResearchHub · Scientist Ranking Academy
Researcher Data Confidence: High, Medium and Low Evidence
Why source quality and identity certainty should be reviewed separately from impact scores.
Confidence concerns evidence
A data-confidence assessment asks how strongly a source supports an attribution or metric. It is not a value judgment about the scientist. A verified identifier match generally provides stronger identity evidence than a loosely similar author name.
Multiple uncertainties coexist
A publication DOI may be clear while its author identity is disputed. An identity may be verified while a citation feed is incomplete. Responsible systems track these uncertainties separately and avoid hiding them inside a single impressive-looking ranking score.
What improves confidence
Stable source identifiers, consistent bibliographic metadata, documented retrieval dates and reviewable affiliation evidence make records more trustworthy. Manual corrections should record the decision and supporting material so future imports do not restore old mistakes.
Use it before recognition
A certificate or other public claim should be based on sufficiently documented evidence. Where identity, source coverage or integrity flags need review, a provisional status is more informative than an unjustified definitive badge.
