ResearchHub · Scientist Ranking Academy
Ten Common Data Errors Behind Unreliable Scientist Rankings
An introduction to the metadata defects that can distort researcher scores and certificate eligibility.
Identity and duplication errors
Homonymous authors may be merged, one scholar may be split across profiles, and preprint or publisher versions may be counted twice. All three problems can significantly shift apparent productivity and impact.
Bibliographic and temporal errors
Wrong publication years, missing DOIs, inconsistent document types and old citation snapshots can invalidate comparisons. A paper with incomplete metadata should not be silently assigned a convenient date or source.
Affiliation and integrity errors
Historic organisations may be shown as current employers, and corrections or retraction statuses can be missing or misapplied. Incorrect integrity labels require especially prompt human review because of reputational consequences.
What prevents recurrence
Use persistent identifiers, source-linked evidence, deduplication rules, versioned snapshots and documented correction workflows. Ranking equations cannot rescue an untrustworthy scholarly graph; metadata quality must come first.
