H-Index

Sort your publications by citation count (highest first). H-index is the largest number h where you have at least h papers each cited ≥h times. Example: if your papers are cited [50, 40, 30, 25, 20, 10, 5], then h=5 (you have 5 papers with ≥5 citations; the 6th paper has only 10 citations, but you need 6 papers with ≥6 citations). If papers are [100, 50, 30, 20, 10, 5, 2], then h=5 (5 papers with ≥5 citations). Databases automate this calculation; if calculating manually, use a spreadsheet or online h-index calculator.

Different databases index different publication sets and publication types. Web of Science and Scopus limit to indexed journals/articles; Google Scholar includes preprints, books, theses, and non-indexed sources. Coverage dates differ (Scopus starts ~1996; WoS earlier). Citation counts differ due to indexing scope. This causes h-index variations typically 1–5 points between databases. Report h-index with the database: 'h=18 [Scopus, June 2024]' is more informative than just 'h=18'. For hiring/evaluation, accept a range (e.g., 'h approximately 18–20 across databases') rather than a single number.

Typical active researchers gain 0.5–1.5 h-index points per year, depending on field, publication rate, and citation patterns. Faster growth (2+ points/year) indicates accelerating impact (highly-cited papers published). Slower growth (<0.5/year) indicates stable but not expanding impact. Early-career researchers (first 5 years) may have irregular growth if citations accumulate in clusters. Compare your h-index growth to peers at similar career stage. Sudden plateaus may indicate declining citation impact or publication output; surges indicate breakthrough research.

Yes, h-index reflects field-specific citation cultures. Physics and biomedical research cite frequently, enabling h-indices of 30+ for leading researchers. Mathematics, philosophy, and engineering cite less frequently; h-index of 10–15 is exceptional. Researchers in low-citation fields are disadvantaged by h-index-based evaluation. Use field-normalized h-indices or discipline-specific comparisons. Compare your h-index to peers in your field, not globally. When evaluating researchers across disciplines, use other metrics (impact in discipline, peer assessment, contribution significance).

Yes, h-index typically includes self-citations (when you cite your own work). This is both a strength and a weakness. Self-citation is legitimate (acknowledging prior own work builds on it) but can be inflated (citing own papers to boost h-index artificially). Some databases offer 'h-index excluding self-citations,' which may be more conservative. Be aware if comparing h-indices with/without self-citations. For evaluation, prefer recent citations from other researchers (more indicative of independent impact) over self-citations.

H-index: largest h where researcher has h papers with ≥h citations. Balanced, robust metric. G-index: addresses h-index weakness by emphasizing highly-cited papers. G-index is always ≥h-index and grows faster. I10-index (Google Scholar): count of papers cited ≥10 times. Simple but ignores citation distribution (a paper with 11 citations and 100 citations count equally). Use h-index for general career assessment, g-index if you have highly-cited papers, i10-index for simplicity. H5-index (last 5 years) shows recent impact independent of older work.

H-index improves by publishing papers that accumulate citations. Legitimate approaches: (1) Publish more papers (increases pool of citable work); (2) Publish in visible venues with wide readership; (3) Publish on timely topics with high citation demand; (4) Promote your work through presentations, collaborations, and citing it appropriately in new papers. Avoid illegitimate approaches: mandating journal self-citations, citation cartels (coordinated mutual citation), or fabricating citations. Legitimate self-promotion (networking, disseminating findings) is appropriate; manipulation is unethical.