Contents
Services · Journal selection · High-impact journals
High-Impact Academic Journals: Citation Metrics and Their Limitations
The phrase has no definition of its own. It means scoring well on one of several metrics with different owners, different databases and different time windows.
What High-Impact Means for Academic Journals
The phrase has no definition of its own. In practice it means a journal that scores well on one of a handful of citation metrics — and since those metrics have different owners, different source databases and different time windows, the phrase means something slightly different depending on which one the speaker has in mind.
That is not a pedantic point. A journal can rank highly on one metric and not appear on another at all, because the metric is calculated from a database that does not cover it. So the first question about any impact claim is: which metric, from whose database.
Journal Citation Metrics and Who Calculates Them
Table 1. The five metrics you will meet
| Metric | Calculated by | Citation window | How it works |
|---|---|---|---|
| Journal Impact Factor (JIF) | Clarivate, in Journal Citation Reports | Two years | Citations in the JCR year to items published in the previous two years, over the citable items in those two years. |
| CiteScore | Elsevier, calculated by Scopus | Four years | A four-year window for both the citations and the documents, recalculated annually. |
| SJR (SCImago Journal Rank) | SCImago, on Scopus data | Three years | Weights each citation by the standing of the citing journal, using an algorithm SCImago likens to PageRank. |
| Eigenfactor | The Eigenfactor Project, University of Washington; distributed in JCR | Five years | Weights citations by the influence of the citing journal and excludes journal self-citations. |
| h-index | Not owned by one body; computed from any citation database | All years | The number of items with at least that many citations each. Originally proposed for individual researchers. |
Two of those windows deserve attention. The impact factor looks back two years; CiteScore looks back four. A journal in a field where citations accumulate slowly will look materially better under CiteScore, and the difference is an artefact of the window rather than of the journal.
A sourcing note. The owners' own pages state ownership clearly but do not all serve their methodology as readable text: Elsevier's CiteScore pages render their methodology headings with no body, so the four-year window and Eigenfactor's five-year window are cited here to a university library guide6 rather than to the metric owners. Treat those two windows as secondary.
The Journal Impact Factor
Clarivate's metric, published in Journal Citation Reports. Its own reference guide defines it: “The JIF is defined as citations to the journal in the JCR year to items published in the previous two years”1, divided by “the total number of citable items, (i.e., Articles and Reviews)” published in those two years1. On the numerator specifically: “The numerator of the JIF consists of any citation from material published in the JCR year to material the journal published in the prior two years.”1
So, written out: citations received in the current year, to items the journal published in the two preceding years, over the count of citable items — articles and reviews — published in those two years.
Where it came from. Garfield devised it as a tool for selecting journals for citation indexing, and his own 2006 account in JAMA sets out its origin, calculation and weaknesses7. Archambault and Larivière trace how its components were settled and argue that the two-year window, the treatment of self-citations and the asymmetry between citable and non-citable items were largely arbitrary, concluding that “the indicators presented in the JCR cannot be considered robust objective measures”8
Those two references are the right ones for a critique of the measure and the wrong ones for how it is calculated — for the calculation, Clarivate's own guide is the source, and it is cited above.
CiteScore, SJR, Eigenfactor and the h-index
CiteScore is Elsevier's, and its support pages state that it is “Calculated by Scopus”2 and “calculated once a year”2. Its description of itself: “CiteScore is a simple way of measuring the citation impact of sources, such as journals”2 The four-year window is what distinguishes it from the impact factor6.
SJR is produced by SCImago using Scopus data, and it differs from the two above in kind rather than in window: it weights each citation by the standing of the citing journal. SCImago's own description is that it “measures journal visibility using an algorithm similar to Google's PageRank”3 and “Represents the average number of weighted citations received in the selected year”3, averaged over the three preceding years.
Eigenfactor works on the same weighted principle and over a five-year window, excluding journal self-citations. It comes from the Eigenfactor Project, co-founded in 2007 by Carl Bergstrom and Jevin West at the University of Washington, and its scores are also distributed inside Clarivate's Journal Citation Reports4.
The h-index is the odd one out, because it was proposed for people rather than journals. Hirsch defined h as “the number of papers with citation number higher or equal to h”5 — so an h of 20 means twenty items with at least twenty citations each. Citation databases also compute it for journals. Because any database can calculate one, the same journal or author has different h-indices in Scopus, Web of Science and Google Scholar, and a figure quoted without its source is uninterpretable.
We found no authoritative source stating how the meaning of the h-index differs when it is applied to a journal rather than an author, so this guide makes no claim on that point.
Limitations of Journal Citation Metrics
All five are journal-level or author-level averages built from citation counts. Three limits follow, and the first is stated by the metric's own owner.
They say nothing about an individual article. Clarivate's reference guide puts it directly: “Article citation performance as a part of scholarly assessment is not reflected in the JIF.”1 Citation distributions within a journal are heavily skewed, so most papers in a high-scoring journal receive fewer citations than its average implies.
They should not be used to judge people. Garfield wrote that “The use of journal impacts in evaluating individuals has its inherent dangers.”7 The San Francisco Declaration on Research Assessment makes it a formal recommendation: “Do not use journal-based metrics, such as Journal Impact Factors (JIFs),” “as surrogate measures of the quality of individual research articles,” “to assess an individual scientist's contributions, or in hiring, promotion, or funding decisions.”9
They are not comparable across fields or across metrics. Citation practices differ enough between disciplines that an impact factor of 3 may be outstanding in one field and unremarkable in another, and the different windows mean a journal's impact factor and its CiteScore cannot be set side by side as though they measured the same thing.
Choosing a Journal
Metrics belong late in this sequence, not first.
Table 2. What to settle, in order
| Question | Why it comes where it does |
|---|---|
| Does the journal publish work like mine? | Scope mismatch is the commonest cause of rejection without review, and no metric compensates for it. Read the last six months of the journal, not its aims page. |
| Will the readers I want see it? | The audience that matters for a clinical paper is not the same as the one that maximises citations. |
| Is it indexed where my readers and my institution look? | This is a separate question from its metrics, and often the more consequential one. |
| What is the time to first decision? | Often stated by the journal, and often the deciding factor for a thesis or a grant deadline. |
| What will it cost? | Article processing charges, and whether your institution has an agreement. |
| How does it score, against the shortlist? | Useful for ranking candidates that already pass the questions above. Not useful for building the shortlist. |
Journal Indexing Versus Journal Impact
Indexing and impact are often confused, including in guidance that sets out to explain one of them. Indexing means a named curated database selected the journal against its published criteria. A metric is a number calculated from citation data by a particular organisation.
The practical consequence: a journal indexed in Scopus but not in the Web of Science Core Collection will have a CiteScore and no impact factor, and that tells you which databases cover it rather than anything about its quality. Our page on indexed and non-indexed journals sets out how to verify an indexing claim against each database's own list.
If your question is instead about where to search the literature rather than where to publish, the databases themselves have been compared on coverage and search capability10 — which is a different question again, and the reason that comparison is cited here and not in the metrics section.
How to apply this to your own work
- Name the metric and its source whenever you quote one. “An impact factor of 4.2” is incomplete; “a 2024 Journal Impact Factor of 4.2, from Journal Citation Reports” is checkable.
- Do not compare an impact factor with a CiteScore. Different owners, different databases, two-year against four-year windows.
- Build your journal shortlist on scope fit, then rank it on metrics — not the other way round.
- Check indexing separately from metrics, against each database’s own list, using the journal’s ISSN.
- Never use a journal metric to argue for the quality of your own paper, in a thesis, a grant application or a promotion case. DORA recommends against it and reviewers know it.
- If you quote an h-index, say which database it came from, because the same journal has different values in different ones.
- Expect a journal to have some metrics and not others, and treat that as information about database coverage rather than about quality.
Take-home message
“High-impact” has no definition of its own; it means scoring well on one of several metrics with different owners, databases and windows. The Journal Impact Factor is Clarivate’s, published in Journal Citation Reports, over a two-year window. CiteScore is Elsevier’s, calculated by Scopus, over four years. SJR is SCImago’s and weights citations by the citing journal; Eigenfactor does likewise over five years; the h-index was proposed for individuals and varies by database. None of them says anything about an individual article — Clarivate says so itself — and DORA recommends against using them to judge people at all.
Clarivate, and it is published in Journal Citation Reports. That is worth stating because a great deal of guidance names the impact factor without ever saying whose metric it is. The calculation is citations in the JCR year to items the journal published in the previous two years, divided by the number of citable items — articles and reviews — in those two years.
Different owners and different windows. The impact factor is Clarivate’s, calculated over a two-year window from Web of Science data. CiteScore is Elsevier’s, calculated by Scopus over a four-year window. So the two are not comparable numbers, and a journal can have one and not the other depending on which database covers it.
For choosing where to submit, not necessarily, and for judging a paper, no. Clarivate’s own reference guide states that article citation performance is not reflected in the JIF. Garfield, who devised it, wrote that using journal impacts to evaluate individuals has inherent dangers. And DORA recommends explicitly against using journal metrics as a surrogate for the quality of individual articles or to assess a scientist’s contribution.
In Hirsch’s own words, the number of papers with a citation count higher than or equal to h. It was proposed in 2005 as a way to quantify an individual researcher’s output, and it is also applied to journals by citation databases. Unlike the other metrics here it is not owned by one body: any citation database can compute one, and the value depends on which database you use.
Metrics are one input and a weak one on their own. Scope fit is the first question, because a mismatched manuscript is rejected without review regardless of anyone’s impact factor. Then audience, then whether the journal is indexed where your readers and your institution look, then time to decision, then cost. A metric is useful for ranking a shortlist you have already built on fit.
No, and the two get conflated constantly. Indexing means a named database selected the journal against its published criteria; a metric is a number calculated from citation data. A journal can be indexed in Scopus with a CiteScore and have no impact factor at all, because the impact factor comes from a different database. Our page on indexed and non-indexed journals covers the indexing side.
References and Sources
- Clarivate. Journal Citation Reports Reference Guide 2024. read 15 October 2026. View on ClarivateClarivate's own reference guide, and the source for the Journal Impact Factor's definition and calculation window. Also the source for Clarivate's own statement that the JIF does not reflect article-level citation performance.
- Elsevier. What is CiteScore? Scopus support. read 15 October 2026. View on ElsevierElsevier's own page confirming CiteScore is calculated by Scopus. Its methodology sections render as headings with no body text, so the four-year window is cited to a library guide rather than to Elsevier.
- SCImago. About us. SCImago Journal & Country Rank. read 15 October 2026. View on SCImagoSCImago's own description of the SJR indicator, which is built on Scopus data and weights citations by the citing journal's own standing.
- Fred Hutchinson Cancer Center Library. Eigenfactor.org journal ranking. read 15 October 2026. View on Fred Hutchinson Cancer Center LibraryA library guide describing the Eigenfactor Project, co-founded in January 2007 by Carl Bergstrom and Jevin West at the University of Washington. Eigenfactor scores are also distributed inside Clarivate's Journal Citation Reports.
- Hirsch JE. An index to quantify an individual's scientific research output. Proceedings of the National Academy of Sciences. 2005. doi:10.1073/pnas.0507655102 read 15 October 2026. View on arXivThe paper that introduced the h-index. The arXiv record carries the DOI but no journal reference field, so the volume and pages are not printed here.
- Queen's University Library. Research Impact: Journal Metrics. read 15 October 2026. View on Queen's University LibraryA library guide, cited for the calculation windows that the metric owners' own pages do not serve as readable text: CiteScore's four years and Eigenfactor's five. Labelled as secondary wherever it is used.
- Garfield E. The history and meaning of the journal impact factor. JAMA. 2006;295(1):90–93. read 15 October 2026. View on University of PennsylvaniaGarfield's own account of the impact factor's origin in his citation-indexing work, its calculation, and its weaknesses. The DOI is not printed on the copy we opened and is not given here.
- Archambault É, Larivière V. History of the journal impact factor: contingencies and consequences. Scientometrics. 2009;79(3):635–649. doi:10.1007/s11192-007-2036-x read 15 October 2026. View on the authors' post-printTraces how the impact factor's components emerged, arguing the two-year window, the handling of self-citations and the citable-item asymmetry were largely arbitrary. The right source for a critique of the measure, and the wrong source for how it is calculated. One repository record prints the pages as 639-653; the post-print carrying the DOI prints 635-649, which is what is used here.
- San Francisco Declaration on Research Assessment (DORA), General Recommendation 1. read 15 October 2026. View on Stellenbosch University LibraryThe declaration's first general recommendation, which is the authoritative statement against using journal metrics to judge individual papers or people. DORA's own site was not reachable, so no release date is given.
- Falagas ME, Pitsouni EI, Malietzis GA, Pappas G. Comparison of PubMed, Scopus, Web of Science, and Google Scholar: strengths and weaknesses. The FASEB Journal. 2008;22:338–342. read 15 October 2026. View on WorldCatA comparison of four literature databases' coverage and search capability. It is about where to search, not about journal metrics, so it is cited here only against a claim about database coverage. The issue number and DOI are not displayed on the record we opened and are not given.
Pubrica AcademyWritten and maintained by the subject matter experts of Pubrica who do this work. Articles are revised when the standard underneath them moves, and each one carries the date it was last checked.
How Pubrica helps
Pubrica’s journal selection service builds a shortlist on scope fit, audience and indexing, with each metric attached and attributed to its source, and checks each candidate against the database’s own records rather than a third-party list. If you have a manuscript but no target journal, a reasoned shortlist helps you avoid submitting to a journal outside the paper’s scope.