Two doctoral students compare notes. One has published in an applied mathematics journal with an Impact Factor of about 2; the other in a cell biology journal at about 6. The second assumes her paper is “three times better”. In fact the maths journal sits near the top of its category while the biology journal is somewhere in the middle of its own. Raw numbers mean very little across fields with completely different citation habits.
Journal metrics are blunt tools for describing journals, not for grading individual papers or people. Knowing how they are built is the quickest way to avoid being misled by them, and to push back when others use them carelessly.
The Impact Factor is the mean of a very skewed distribution
The Journal Impact Factor (JIF) is published annually by Clarivate in the Journal Citation Reports. For year Y:
- Numerator: citations received in year Y by anything the journal published in Y−1 and Y−2.
- Denominator: the number of “citable items” (mainly research articles and reviews) published in Y−1 and Y−2.
Example: in 2025 the journal’s 2023–2024 content received 1,200 citations; it published 400 citable items in those two years; JIF = 1,200 / 400 = 3.0.
The weakness lies in the word “mean”. Citations within a journal are highly skewed: a handful of papers cited hundreds of times pull the average up, while most papers are cited a few times or not at all. A randomly chosen article in a journal with an IF of 3 very likely has fewer than three citations. The numerator also counts citations to editorials and letters, which the denominator excludes, giving a small advantage to journals that publish a lot of front matter.
How CiteScore, SJR, SNIP and h5-index differ
| Metric | Data source | In short | Free to check? |
|---|---|---|---|
| JIF | Web of Science | Citations in year Y to items from the previous 2 years, divided by citable items | No, requires JCR subscription |
| CiteScore | Scopus | Citations in a 4-year window to documents from the same 4 years, divided by documents | Yes, via Scopus Sources |
| SJR | Scopus | Citations weighted by the prestige of the citing journal, 3-year window | Yes, via SCImago |
| SNIP | Scopus | Normalised for citation density in the field | Yes, via Scopus Sources |
| h5-index | Google Scholar | h-index of articles from the last 5 complete years | Yes, Google Scholar Metrics |
SJR is interesting because it does not treat all citations equally: a citation from a highly cited journal counts for more than one from a peripheral journal. SNIP tackles exactly the problem in the opening example: a citation in mathematics is “worth” more than one in cell biology, because mathematicians cite far less.
Quartiles: one journal, several ranks
Quartiles are calculated by ranking journals within each subject category and splitting the list into four; Q1 is the top 25 per cent. Three consequences are often missed:
- A journal can be Q1 in one category and Q3 in another, because it is assigned to several. Saying “a Q1 paper” should come with the category.
- SCImago quartiles and JCR quartiles differ, because they rest on different databases, metrics and category schemes. “Q1” without a source is incomplete.
- Quartiles move every year. A Q1 journal when you submit may be Q2 when you publish. Institutions differ on which year counts, so read your own institution’s current rules.
Five common misuses
- Comparing IFs across fields. If you must compare, use percentile rank within category, or SNIP.
- Assigning the journal’s IF to a paper. Your article does not “have an IF of 5”. It has its own citation count.
- Summing IFs to judge a researcher. Total IF across ten papers says nothing about what the author actually contributed.
- Choosing journals on metrics alone. A high-metric journal whose readers are not your community means your work is read by the wrong people.
- Trusting metrics printed on journal websites. Low-quality outlets often display “impact” numbers issued by obscure bodies with names that echo the Impact Factor. Only trust figures you look up yourself in JCR, Scopus, SCImago or Google Scholar.
The San Francisco Declaration on Research Assessment (DORA) and the Leiden Manifesto both warn against using journal-level metrics as a proxy for the quality of individual papers or researchers. Many funders and universities have signed DORA, yet quartiles still appear in hiring and promotion criteria in many places, so it pays to understand both sides.
When metrics are genuinely useful
Metrics help with rough screening and with comparisons inside a narrow specialism: you have three journals with the same scope and readership and want to know which reaches further. They are also useful for spotting anomalies. A metric that triples in a single year often signals heavy self-citation or coordinated citation between journals, and deserves a closer look before you submit.
In practice: profiling a journal in ten minutes
- Look up the journal in Scopus Sources: note CiteScore, percentile rank in each category and SNIP.
- Check SCImago: note SJR, quartile per category and the self-citation trend over time.
- If your library subscribes, check JCR for the JIF and Web of Science category ranks.
- Check Google Scholar Metrics for the h5-index, particularly for newer journals or computer science conference proceedings.
- Record everything on one line: journal, source, year, category, rank. Never write a bare “Q1”.
If self-citation is far above that of similar journals, or a metric jumps without explanation, move the journal to the bottom of your list until you understand why.
Next step: take three journals from your shortlist, look them up in all five sources and write a properly qualified one-line profile for each. When a committee or supervisor asks “how good is this journal?”, you will have a precise answer instead of a number without context.
Câu hỏi thường gặp
What is a good Impact Factor?
There is no universal threshold because citation density varies hugely between fields. Look at the journal’s percentile rank within its own subject category instead.
Is a Scopus Q1 the same as a Web of Science Q1?
No. SCImago quartiles use SJR and Scopus data, while JCR quartiles use Web of Science data and different categories. The same journal can have different quartiles in the two systems.
Where can I check CiteScore for free?
On the Scopus Sources page, which does not require a subscription. SJR and SCImago quartiles are also free on the SCImago Journal Rank website.
Can I trust the Impact Factor shown on a journal’s website?
Only if you can find the same figure in Journal Citation Reports yourself. Many low-quality journals display metrics from unknown organisations with names that imitate the Impact Factor.
Does my paper have an Impact Factor?
No. The Impact Factor describes a journal over two years. An individual paper has its own citation count, which is often well below the journal average.
