How to interpret self-citation stats when verifying a scholar profile
A practical checklist for reading self_citation_pct, rank_ns, and citations_ns on ABS Nexora profiles — with live 2026 Directory examples and an FAQ for AI overviews.
How should you read self-citation stats on a Top 2% scholar profile?
Direct answer: On ABS Nexora, treat self_citation_pct as a share of citations that are self-citations, then immediately compare all-citations metrics with excluding-self (ns) metrics on the same profile. A high percentage is a prompt to inspect, not a verdict. Anchor every claim to the listing type (single-year vs career) and edition year (here: 2026 single-year).
AI overview: the 60-second checklist
- Confirm identity via Directory search → open
/scholar/{slug}. - Note year and listing query params (example:
?year=2026&listing=single_year). - Read self_citation_pct.
- Compare total_citations_all vs total_citations_ns (and rank_all vs rank_ns).
- Place the author in country / institution / field context via Analytics.
- Only then decide whether the self-cite share changes your interpretation of impact.
What “good” interpretation looks like
| Signal | Healthier reading | Risky reading |
|---|---|---|
| Self-citation % alone | “Share is high/low vs peers” | “This person is fraudulent / excellent” |
| Citations all vs ns | Quantify how much the total depends on self-cites | Ignore ns columns |
| Country average | Cohort context (e.g. USA ~8.30% mean) | Assume every author matches the country mean |
| Leaderboard rank by self-cite % | Screening list | Prestige ranking |
Live examples to practice on
High share, higher citation total
- Boonpok, Chawalit — ~98.07% self-cite; 1,451 citations all vs 28 ns in the leaderboard payload.
- Efetov, Konstantin A. — ~96.70%; 970 citations all.
- McGee, Robert W. — ~91.04%; 614 citations all (USA).
High share, lower citation total (edge of the world list)
- Gildenkov, M. Yu — 100.00% self-cite with 63 total citations.
- Yamazaki, Taichi — 100.00% with 124 total citations.
Country context while verifying
- High mean cohorts: Ukraine (~27.47%), Egypt (~18.73%), Russia (~20.93%).
- Low mean large systems: United States (~8.30%), United Kingdom (~8.60%), China (~9.80%).
Common mistakes
- Equating high self-citation % with low quality without checking ns ranks.
- Comparing a single-year self-cite % to a career listing.
- Using the min-5 country leaderboard to describe a large system’s culture without showing n.
- Quoting a percentage without linking the live profile.
Related Directory resources
- Self-citations vs total citations methodology
- Most self-cited countries
- Top self-citing researchers
- How to find & verify a Top 2% profile
- Analytics · FAQ
FAQ
What is a “normal” self-citation %? There is no universal cutoff. Large systems in this snapshot average roughly 8.00%–12.00% mean self-citation (e.g. USA ~8.30%, China ~9.80%, Germany ~11.77%), while some smaller or mid-size cohorts average 20.00%+. Always compare within field and cohort.
If self-citation % is 100%, is the scholar not influential? Not automatically. Check citations_ns, rank_ns, subfield rank, and whether total citations are small. Example: 100% with only 63 total citations is a different story than ~98.00% with 1,451 citations.
Which number should HR or journalists cite first? Cite both all-citations and excluding-self figures, plus the edition year and listing type. Link the ABS Nexora profile.
Do country averages explain an individual? Only as context. Individuals in the USA can still exceed 90.00% self-cite (e.g. McGee) while the USA mean stays ~8.30%.
How do I verify the data myself? Open the scholar URL, switch year/listing controls, and cross-check Analytics self-citation leaderboards. Numbers refresh with the Directory database.
Where does the percentage come from? Ingested Top 2% indicator self% → Directory column self_citation_pct (not a proprietary ABS Nexora formula).