Journal Impact Factor: What It Measures and Why It's Hated
A journal's impact factor is a simple average: how many times papers it published in the last two years were cited, divided by how many papers it published. It tells you nothing about whether a particular paper is good, true, or important. Thousands of researchers have signed declarations asking the scientific community to stop using it to judge individual scientists and their work.
The actual calculation
Take a journal. Look at every paper it published in the previous two years. Count how many times those papers were cited by other papers in the current year. Divide that total by the number of papers the journal published in those two years. The result is the impact factor.
So if a journal published 500 papers over two years and those papers were cited a combined 2,500 times this year, the impact factor is 5. It is an average citation count per paper, nothing more exotic than that.
The figure is produced annually by a company called Clarivate, which owns the database used to track citations. It appears in a report that researchers, universities, and funding bodies have treated as a kind of league table of journals for decades.
What it was built for
The metric was invented in the 1960s by Eugene Garfield, a scientist who wanted a tool for librarians. The problem he was solving was practical: libraries have limited shelf space and budgets, and someone has to decide which journals are worth subscribing to.
For that narrow purpose, average citation counts are a reasonable proxy. A journal whose papers get cited a lot is probably being read and used by other scientists, which is useful information when you are choosing between subscriptions.
Garfield never intended the number to say anything about individual papers or individual researchers. He said so explicitly, more than once, as the metric's use expanded far beyond library shelving decisions.
Why researchers rely on it anyway
Despite its narrow original purpose, the impact factor became a shorthand for quality almost everywhere in science. Hiring committees, grant panels, and tenure boards often use it as a quick filter: did this person publish in high impact factor journals, yes or no.
The appeal is obvious. Reading and judging someone's actual papers takes time and expertise. A single number attached to the journal name lets a committee skip that work. It is a shortcut, and shortcuts spread fast in systems with too many decisions and not enough hours.
Once the number mattered for careers, it started shaping behaviour. Scientists began choosing where to submit based on a journal's score rather than its fit for the work, and some editors began making decisions designed to push the number up rather than to serve readers.
Why the average misleads
An average hides the shape of the data behind it. Citations to papers in any journal are heavily skewed. A small number of papers get cited enormously, and most papers get cited rarely or not at all. The average can be pulled high by a handful of blockbuster papers while the typical paper in that same journal attracts barely any attention.
This means a high impact factor tells you almost nothing about the paper you are actually holding. A paper in a journal with an impact factor of 30 is not thirty times more likely to be right, or useful, or well conducted, than a paper in a journal with an impact factor of 3. It might be one of the rarely cited ones dragging the average down from the other side.
The two year citation window also creates odd incentives. Fields that cite slowly, such as mathematics, look artificially low scoring compared with fast moving fields like molecular biology, even when the underlying work is equally solid. The number compares journals across fields as if citation speed were a universal currency, and it is not.
The campaign against it
The campaign against it
In 2012 a group of journal editors and publishers drafted a document called the San Francisco Declaration on Research Assessment, usually shortened to DORA. It calls on universities, funders, and publishers to stop using journal impact factor as a stand in for the quality of individual articles or individual scientists.
DORA has been signed by a large number of institutions and tens of thousands of individual researchers worldwide, including many prominent scientific societies and funding agencies. Its core recommendation is specific: judge research on its own merits, read the actual paper, and if you need numbers, use ones that describe the paper itself rather than the journal it happened to land in.
Some funders and universities have changed their formal policies as a result, stating that they will not consider journal impact factor when assessing grant applications or promotion cases. Other institutions have been slower to change, partly because impact factor remains an easy number to put in a spreadsheet, and alternatives require more judgement and more time.
What to use instead
What to use instead
If you want to judge a specific paper, the most direct approach is to read it, or at least read how other scientists have responded to it in print. Citation counts for that individual paper tell you more than the journal's average, since they measure attention to the actual work rather than to everything published nearby.
Some researchers look at measures like the h-index, which tries to capture a scientist's overall output and influence across their career rather than any single journal's score. These measures have their own flaws, but they at least attach to the person or the paper rather than to the publication venue.
The simplest practical move, for a reader rather than a hiring committee, is to treat the journal name as a weak signal at best. A paper's methods, its sample size, whether it has been replicated, and what independent experts say about it carry far more information than which journal's masthead sits at the top of the page.
Common questions
- Does a high impact factor mean a journal has stricter peer review?
Not necessarily. Peer review standards vary by journal and field, and a high impact factor reflects citation volume, not the rigour of the review process. Some well reviewed journals in smaller fields have modest impact factors simply because fewer people work in that area and cite each other.
- Why do universities still use it if scientists dislike it?
Institutional habits change slowly, and impact factor is easy to look up and compare across candidates with minimal effort. Replacing it means committees have to actually read and judge research quality directly, which takes more time and expertise than checking a number.
- Can a journal manipulate its impact factor?
Yes, to some extent. Editors can encourage authors to cite other papers from the same journal, publish more review articles which tend to attract more citations, or adjust what counts as a citable item in ways that shift the ratio favourably.
- Is there a better single number that replaces it?
No single number has replaced it cleanly, which is part of the point researchers behind DORA make. They argue that no single metric can substitute for actually reading the work, though article level citation counts and expert assessment come closer than a journal average.
- Does impact factor differ a lot between scientific fields?
Yes, substantially. Fields with fast publishing cycles and large numbers of active researchers, such as parts of biomedicine, tend to produce higher impact factors than slower moving fields like pure mathematics or some areas of the humanities adjacent sciences, even when research quality is comparable.
Lucent, once a week
We read the new research and write up what actually holds, in the same plain language as this page. No jargon, no hype, three papers a week.
Read the latest issueRead next
- What a Retraction Means and How to Check a Paper
A retraction is a journal's formal withdrawal of a published paper because its findings can no longer be trusted, and checking its status takes under a minute.
- Open Access Publishing: The Models and Who Pays
Open access publishing makes research free to read, but someone still pays, usually the author's grant or institution, through a fee called an article processing charge.
- Why Scientific Papers Are Paywalled, and How to Read Them Free
Researchers write and review papers for free, but publishers charge for access, though preprints and repositories let you read most papers legally at no cost.