CiteScore vs. Impact Factor — Why Everyone Still Uses the IF

CiteScore and the Impact Factor (IF) are both metrics for how many citations the average paper in a journal receives. The numbers look alike, but when colleagues discuss research assessment or where to submit a paper, the IF is almost always the only one mentioned. The differences between the two come down to the formulas themselves and to how long each has been around.

Start with the formulas

The IF's mismatched numerator and denominator

The most debated feature of the IF calculation is the asymmetry between its numerator and denominator. Citations received by editorials, letters and news items count toward the numerator, yet those items are excluded from the denominator. Structurally, then, the more non-original content a journal publishes, the more it can inflate the numerator without adding to the denominator. The same logic explains why publishing more review articles tends to raise a journal's IF.

CiteScore removed that mismatch. It uses the same document types in both the numerator and the denominator, and editorials and letters are counted in neither. It is designed so that the calculation can be traced: on Scopus you can drill down to the individual papers and check the arithmetic yourself. The JIF, by contrast, is calculated in a database specific to the JCR, and even with a Web of Science subscription you cannot reproduce the same figure — a transparency problem that has drawn criticism for years.

2 years vs. 4 years

Clinical research papers often have not finished accruing citations two years after publication. The IF's 2-year window cuts that tail off, while CiteScore's 4-year window captures more of it.

The important point is that comparing an IF with a CiteScore is meaningless. Compare IFs with IFs and CiteScores with CiteScores. When a journal advertises "CiteScore 12.3" in large type, that does not amount to an IF of 12.

Why the IF is the one everyone uses

CiteScore has the more straightforward formula, yet in practice the IF is almost all anyone uses. There are several reasons.

A 40-year head start

The IF has been published annually as the JCR since 1975, and Eugene Garfield first set out the idea in 1955. CiteScore, by contrast, appeared only in December 2016, and in June 2020 its methodology changed substantially (a 3-year window became a 4-year window), with all past values recalculated. CiteScores from the old and new methods are not comparable. The metric has only been settled for a few years.

It is built into assessment systems

Thesis defenses, departmental publication tallies, grant applications, job-application dossiers. Criteria such as "total IF" or "N first-author papers in journals with an IF above X" have a long history in Japanese academic medicine, and once a metric is written into institutional rules, a better metric does not displace it easily. There is also the fact that reviewers have an intuitive feel for what IF values mean, and for nothing else.

Journals advertise their IF

What appears in large type on a journal's home page is almost always the IF. As long as the IF is the first number an author sees when considering where to submit, authors will go on framing things in terms of the IF.

The provider is a publisher

Elsevier, which produces CiteScore, is also one of the world's largest academic publishers. From the metric's launch, critics have pointed to the conflict of interest in a publisher ranking every journal, its own included, with its own metric. Clarivate is not a publisher and is therefore less exposed to that criticism. In terms of perceived neutrality, CiteScore started at a disadvantage.

Being free works against it

The JCR cannot be viewed without a subscription. It is hard to deny that being "a number only certain people can see" has itself lent the IF an air of authority. CiteScore can be checked by anyone for free — but that also makes it a number with no institutional badge behind it.

How to use it in practice

CiteScore earns its keep in situations like these.

1. When you want to place a journal that has no IF. Scopus indexes more titles than the JCR, so a journal without an IF may still have a CiteScore you can look up. Newer journals, and journals that publish mainly case reports, fall into this group.
2. When you do not have a JCR subscription. Search for a journal by name at <a href="https://www.scopus.com/sources" target="_blank" rel="noopener">scopus.com/sources</a> and you get the CiteScore and the journal's percentile within its field, free of charge. No subscription to Scopus itself is needed.
3. When you are assessing a field where citations accumulate slowly. Thanks to the 4-year window, the value can be closer to reality than the IF.

Conversely, when you are filling in a grant application or a publication list, check what the recipient is asking for. Entering a CiteScore in a field that specifies "IF" can make the magnitude look wrong. It is always safer to state which metric it is and which year's edition.

Quartiles and percentiles

The IF comes with Q1–Q4 quartiles. CiteScore comes with a percentile, presented as "top X% within the field." You can read the top 25% as the equivalent of Q1 in IF terms. The field classifications themselves differ, however, between the JCR (254 categories) and Scopus (334 categories), so the same journal can rank differently depending on how the field is drawn.

Summary

• As a piece of design, CiteScore is the more coherent calculation (matched numerator and denominator, a longer window, and free to verify)
• The IF nonetheless remains in use for reasons unrelated to the quality of the metric: 40 years of history, its embedding in assessment systems, and the neutrality of its provider
• The two are different rulers, so do not compare their values directly
• CiteScore is worth knowing about as a way to look up journals for which no IF is available

The listings on this site carry the IF and quartile from the JCR. When you are considering a journal that is not listed, try checking its CiteScore on Scopus.