Written and maintained by CASRAI Editorial Board
Last updated
The Eigenfactor Score and the Article Influence Score are Journal Citation Reports (JCR) metrics that rank a journal by the influence of the journals citing it, not simply by how many citations it receives. Both are calculated using an algorithm modeled on the same eigenvector-centrality logic behind Google’s PageRank: a citation from a heavily-cited, highly-influential journal counts for more than a citation from a marginal one. That makes them structurally different from the Journal Impact Factor (JIF), which simply divides a raw citation count by a raw article count. This guide covers how each score is actually calculated, what counts as a high or low value, and how they compare to the JIF, SJR, and CiteScore.
What Eigenfactor Score and Article Influence Score Measure
Both metrics are published annually in Clarivate’s Journal Citation Reports (JCR), alongside the JIF, JIF 5-Year, and the field-normalized Journal Citation Indicator (JCI). They measure two related but distinct things:
- Eigenfactor Score estimates a journal’s total influence on the scholarly literature — how much of the citation “traffic” in the network flows to that journal, weighted by the prestige of the journals sending it. Because it is a measure of total influence, a journal that publishes more citable content each year tends to accumulate a larger Eigenfactor Score, all else equal.
- Article Influence Score (AIS) takes the Eigenfactor Score and divides it by the journal’s share of articles published, producing a per-article figure that is directly comparable across journals of very different sizes — closer in spirit to what the JIF is trying to measure, but calculated from the same network-weighted citation data as the Eigenfactor Score rather than from a raw citation ratio.
Who Created Them, and How They Ended Up in JCR
Both metrics come from the Eigenfactor Project, co-founded in January 2007 by Carl Bergstrom and Jevin West at the University of Washington (the Bergstrom Lab in the Department of Biology and the West Lab in the Information School). The project has been incorporated into what was then Thomson Reuters’ JCR since 2007, and has remained part of JCR through its acquisition and rebrand as Clarivate. The relationship is a data-sharing arrangement rather than an in-house Clarivate metric: Clarivate supplies the underlying citation data from its indexed journals, and the Eigenfactor Project calculates and supplies the Eigenfactor Score and Article Influence Score back for publication in JCR. Both scores remain part of the current JCR metric roster as of the 2026 release.
How the Eigenfactor Score Is Calculated
The Eigenfactor Score is computed from a citation network covering the five years immediately preceding the JCR year, with two features that distinguish it clearly from the two-year JIF window:
- Five-year citation window. Every citation counted is drawn from the preceding five years of the network, not two — closer to the JIF 5-Year variant in timeframe, though the resemblance stops there since the JIF 5-Year is still a raw citation ratio, not a network-weighted score.
- Journal self-citations are excluded. Citations from a journal to its own previously published articles are stripped out of the network before the algorithm runs, which the standard two-year JIF calculation does not do at the same level (Clarivate applies separate self-citation review to flagged JIF outliers instead — see how the Journal Impact Factor is calculated for how that works).
- Citations are weighted by the influence of the citing journal. This is the PageRank-like part: a citation from a journal that is itself frequently cited by influential journals contributes more to the cited journal’s Eigenfactor Score than a citation from an obscure, rarely-cited journal. The algorithm runs iteratively across the whole citation network to converge on each journal’s influence score simultaneously, rather than crediting citations one at a time.
Because Eigenfactor Scores are a measure of a journal’s total share of citation influence across the entire indexed network, individual scores are rescaled each year so that the Eigenfactor Scores of every journal in the network sum to 100. In practice this means even the most prolific, highest-influence journals typically post Eigenfactor Scores in the low single digits, while the great majority of indexed journals post scores well under 1 — the number is a share of total network influence, not a citations-per-paper figure, and reading it as one is the most common misinterpretation.
How the Article Influence Score Is Derived
The Article Influence Score corrects for the size effect built into the Eigenfactor Score. It is calculated by dividing a journal’s Eigenfactor Score by its fractional share of the total articles published across the JCR-indexed network in that period, then rescaling the result so that the mean article across the entire database has an Article Influence Score of 1.000. That normalization makes the number directly interpretable without needing to know the journal’s publishing volume:
- An AIS above 1.00 means the journal’s articles are, on average, more influential than the average article across the whole JCR network.
- An AIS below 1.00 means below-average per-article influence.
- An AIS of exactly 1.00 is, by construction, the network average — the same anchoring logic the JCI uses for its own category average, though JCI’s 1.0 baseline is computed separately within each subject category while AIS is not restricted to a single category baseline.
That last distinction matters in practice: because Eigenfactor Score and Article Influence Score are computed once across the full network rather than re-based within each JCR subject category, a large, heavily-cited multidisciplinary journal will tend to post a higher AIS than a narrow specialist journal of comparable rigor, purely because the specialist journal’s citing community is smaller. The network weighting corrects for some cross-field differences in citation density — a citation from a small but influential field counts more than a citation from a large but low-influence one — but it does not guarantee the kind of clean field-to-field comparability that JCI’s explicit category normalization is built to provide.
Eigenfactor and Article Influence Score vs. the Journal Impact Factor
| Dimension | Eigenfactor Score / AIS | Journal Impact Factor |
|---|---|---|
| Citation window | 5 years | 2 years (JIF 5-Year variant also exists) |
| Self-citations | Excluded from the network | Included, with separate outlier review |
| Weighting | By influence of the citing journal (PageRank-like) | None — every citation counts equally |
| What it captures | Total network influence (Eigenfactor) or per-article influence (AIS) | Raw citations per citable item |
| Normalization | Rescaled to a fixed total (100) or network mean (1.000) | Not normalized; raw ratio |
A journal can rank very differently on the two scales. A journal with a modest JIF but a strong network position — heavily cited by other influential, well-cited journals — can post a comparatively strong AIS, and vice versa for a journal with an inflated JIF driven by a small number of highly-cited outlier papers or heavy self-citation, since Eigenfactor strips the latter out entirely.
Eigenfactor and Article Influence Score vs. SJR and CiteScore
Eigenfactor Score and AIS are not the only network-weighted alternative to the JIF. SJR (SCImago Journal Rank) uses the same basic PageRank-like principle — citations weighted by the prestige of the citing source — but is computed from Scopus data by the independent SCImago Research Group rather than from Web of Science data by the Eigenfactor Project, and uses its own citation window and self-citation cap rather than Eigenfactor’s five-year, fully-excluded approach. CiteScore, by contrast, is closer in spirit to the JIF than to Eigenfactor or SJR: it is Elsevier’s Scopus-based metric, and like the JIF it is an unweighted count of citations over a window (four years) divided by a document count, with no network weighting at all. See Impact Factor vs. SJR, CiteScore vs. SJR, and CiteScore vs. Journal Impact Factor for the full side-by-side breakdowns.
Where to Find Eigenfactor and Article Influence Scores
Both metrics are published as part of the annual JCR release for every journal that receives a JIF-equivalent evaluation, accessible to institutions with a JCR subscription through Clarivate’s Web of Science platform. Unlike CiteScore (free at scopus.com/sources) or SJR (free at scimagojr.com), Eigenfactor Score and Article Influence Score are not published as a free public lookup outside of a JCR subscription — a researcher without institutional JCR access typically has to go through a library that subscribes to JCR, or ask a librarian to run the lookup, rather than checking the number directly online. See how to look up a journal’s Impact Factor for the general JCR-access workflow, which applies the same way to these two metrics.
A Caution That Applies Here Too
Eigenfactor Score and Article Influence Score are journal-level metrics, exactly like the JIF, JCI, SJR, and CiteScore — they describe a journal’s aggregate citation position, not the quality of any specific article or the researcher who wrote it. The San Francisco Declaration on Research Assessment (DORA) and the Coalition for Advancing Research Assessment (CoARA) both name journal-level metrics broadly — not just the JIF specifically — as inappropriate proxies for evaluating individual researchers in hiring, tenure, and funding decisions. That caution applies to Eigenfactor Score and AIS exactly as it applies to the JIF, even though their network-weighted design corrects some of the JIF’s known distortions. Author-level alternatives such as the h-index carry their own, different limitations — see Impact Factor vs. h-index for how a journal-level and an author-level metric differ in what they can and can’t tell you.
Frequently Asked Questions
Is a higher Eigenfactor Score always better?
Not on its own. Because Eigenfactor Score reflects total network influence, it rewards publishing volume as well as prestige — a journal that publishes far more articles per year will tend to post a higher Eigenfactor Score than a smaller, equally-influential journal. Use the Article Influence Score, not the Eigenfactor Score, when the comparison needs to control for journal size.
Can Eigenfactor Score or Article Influence Score be negative or zero?
No. Both are non-negative by construction — a journal that receives no incoming citations from the network in the relevant window will post a value at or near zero, but the algorithm does not produce negative scores.
Why doesn’t my journal have an Eigenfactor Score even though it has a JIF?
Coverage for Eigenfactor Score and Article Influence Score follows the same JCR-indexing rules as the JIF itself — a title newly added to Web of Science Core Collection, or one that has not yet accumulated a full five-year citation history in the network, may not yet have a stable, reportable Eigenfactor Score even if it already qualifies for a JIF under a shorter window.
Do Eigenfactor Score and Article Influence Score replace the Journal Impact Factor in JCR?
No. JCR reports the JIF, JIF 5-Year, JCI, Eigenfactor Score, and Article Influence Score side by side for each qualifying journal — none of them has been retired in favor of another, and Clarivate presents them as complementary views of a journal’s citation performance rather than a single ranking.
