What impact factors and quartiles tell you about a journal

A citation metric becomes meaningful only when its label is complete. “Impact factor 4” leaves out the provider and year. “Q1 journal” leaves out the ranking system, subject category and year. Those omissions can turn a correct number into a misleading statement.

For an author choosing a journal, the practical task is modest: understand what the indicator describes, record it accurately if it is relevant, and keep it separate from the assessment of the paper itself. You do not need to reproduce a commercial database to do this well. You do need to resist treating different metrics as interchangeable measures of quality.

Read the name before the number

Start by identifying the exact indicator. The Journal Impact Factor, CiteScore, Journal Citation Indicator and other measures have different definitions. A number described vaguely as an “impact score” should not be assumed to be Clarivate’s Journal Impact Factor.

Locate the provider’s own journal record. Match the journal title and ISSN rather than relying on a screenshot in a publisher brochure. The ISSN helps identify the publication, but it does not certify the quality or validity of its contents. 5

Write down the displayed metric name before copying the value. Keep abbreviations attached to their definitions in your notes. If the record shows a percentile and a quartile, do not enter one into a form asking for the other. If a department asks for a specific indicator, clarify whether alternatives are accepted rather than substituting a more favourable figure.

Treat an unfamiliar metric as something requiring explanation. Ask who calculates it, what material it uses and where its method is documented. The appropriate response to an unverified number is “not verified”, not a guessed conversion to a familiar scale.

What the Journal Impact Factor measures

Clarivate’s current glossary defines the JIF using citations in a given JCR year to items from the preceding two years, divided by the journal’s scholarly items from those two years. The numerator and denominator are not simply identical sets of document types, so it should not be read as the citation count an average future paper is guaranteed to receive. 1

For intuition, consider a deliberately simplified arithmetic illustration: dividing 600 qualifying citations by 200 denominator items gives 3. This is not data from a real journal and is not a reproduction of all the database’s processing rules. It demonstrates a ratio. It does not mean each of the 200 papers received three citations, or that a paper submitted today will receive three.

The distinction matters because a journal-level aggregate can conceal very different distributions. A few heavily cited papers and many lightly cited papers can produce the same ratio as a more even distribution. When available, inspect the citation distribution rather than assuming the headline value describes every article. Clarivate’s own introductory guidance encourages this contextual reading. 4

A citation is also not a direct observation of methodological quality. Before making a claim about the strength of a specific study, read its question, methods, evidence and limitations. The journal’s metric cannot perform that assessment for you.

Keep the metric year separate from the release year

The year printed beside an indicator is part of its identity. A report released in one calendar year may describe citation activity in an earlier reporting year. Copy the provider’s displayed year exactly; do not replace it with the year you happened to download the page.

Keep three fields if you need an auditable record: metric year, date accessed and, where relevant, release or edition. This helps explain why two collaborators obtained different figures. One may be looking at a different year rather than making a calculation error.

Do not describe a historical maximum as the current indicator. If a trend is relevant, show the sequence with its years and explain what changed in the comparison. A lone favourable value selected from an earlier edition gives an incomplete account.

For institutional reporting, ask which year’s indicator the rule uses. It might be tied to submission, acceptance, publication or an assessment exercise. This guide does not supply that rule. Obtain the applicable written policy and preserve the interpretation given by the office responsible for applying it.

A quartile belongs to a category

A quartile places a journal within a particular ranked category. Clarivate’s glossary describes separate category ranks and quartiles for JIF and for JCI. A journal assigned to more than one category can therefore have more than one relevant category result. 1

This makes “Q1” an incomplete description. Record the indicator used for ranking, the category and the year. Do not silently choose whichever category gives the most attractive result when the question concerns a specific discipline.

The category is not a perfect description of your manuscript. A cross-disciplinary journal may sit in several categories, while your paper addresses a particular research community within it. Use the category to interpret the metric, then return to the journal’s actual scope and recent articles to judge fit.

Avoid calculating a quartile yourself from a casual search result. Filters, ties, omitted records and the provider’s ranking method can affect interpretation. Use the displayed official result and its explanatory notes. If no quartile is shown for the combination you need, record that absence rather than inferring one from another metric or year.

A move from one quartile to another also needs context. Ask whether the rank, category composition, value or classification changed. The label alone does not identify the cause, and it does not establish that the quality of every published paper suddenly changed at the category boundary.

CiteScore is a different measure

Scopus provides CiteScore and related source metrics on its source pages. Its documentation describes a four-year CiteScore window and specifies the document types used. That differs from the JIF’s two-year construction, so the numerical values should not be compared as if they shared a scale. 2

Do not write “impact factor” when the verified value is CiteScore. Keep the provider and indicator explicit in a CV, grant document or shortlist. Similarly, a quartile reported by another ranking service should retain that service’s name and method rather than being relabelled as a JCR quartile.

If a journal displays several indicators, resist choosing the largest number simply because it looks strongest. Ask what each measure contributes to the decision. A metric that is not relevant to the audience, institutional requirement or analytical question can be left out.

For a comparison across candidate journals, use the same indicator and year where possible. If equivalent records are unavailable, show that limitation. A table with mixed years and mixed measures can look precise while making a comparison that the underlying data do not support.

What metrics cannot decide for your paper

DORA recommends against using journal-based metrics as substitutes for assessing individual research contributions or researchers. Its concerns include variation across fields and the limitations of journal averages. 3

For submission planning, apply that principle concretely. A citation metric cannot tell you whether the journal accepts your article type, whether its readers need your finding, whether the editors can evaluate the method, or whether the publishing route meets your obligations. Those questions require different evidence.

Nor can the indicator predict acceptance. Do not turn a lower value into an assumption that review will be easier, or a higher value into proof that a rejection establishes the paper is weak. Editorial criteria, manuscript quality and journal priorities are not reducible to a single citation statistic.

A metric may still be relevant to a documented institutional requirement. Acknowledging that practical constraint does not require endorsing the metric as a measure of individual merit. Keep the two statements separate: “This meets the office’s stated criterion” and “This is an appropriate venue for this paper.” Each needs its own reasoning.

A worksheet for recording metrics accurately

Use one row for each indicator–year–category combination. Do not combine several categories into a single “best quartile” field unless that is explicitly what the reporting requirement asks for.

Field	Entry to complete
Journal title	___
ISSN or electronic ISSN	___
Provider and database	___
Exact indicator name	___
Metric year	___
Displayed value	___
Category name	___
Category rank and denominator, if shown	___
Quartile or percentile, with label	___
Direct source record	___
Date accessed	___
Requirement this evidence addresses	___
Qualifications or missing information	___

Before sharing the table, ask a colleague to reconstruct one row using only the recorded information. If they cannot reach the same record or tell which year you used, add what is missing. The test is reproducibility, not whether the table looks impressive.

For a manuscript shortlist, add a separate field for recent relevant articles. This prevents the metric sheet from becoming the whole journal evaluation. Use the journal-selection worksheet to compare audience, methods, costs and requirements alongside the metric evidence.

Make comparisons with the decision in view

Before sending a metrics table to collaborators, add a sentence describing what comparison is being made. Are the journals in the same category? Are all values from the same year? Is the table intended to document an institutional condition or to explore citation patterns? The reader should not have to infer the purpose from the largest number.

Then list what the table does not contain. It may have no evidence about readership, review quality, publication speed or the suitability of a particular method. Leaving those gaps visible prevents the numerical comparison from acquiring authority over questions it never examined.

If two journals have very similar displayed values, do not invent a meaningful precision beyond the source. Return to the factors that directly affect the submission: audience, article type, reporting expectations and a feasible publishing route. You can choose between reasonable candidates without claiming that a small numerical difference proves one is scientifically superior.

For a formal report, preserve the underlying entries rather than only the final chart or ranking. A later reader should be able to see the identifiers, years and categories used, and determine whether the comparison still answers the original question.

Resolve apparently conflicting figures

When the journal website and the provider disagree, first compare the indicator name and year. Then check journal identity, category and whether the page is presenting a historical figure. Many apparent conflicts can be narrowed by identifying what each number actually describes.

If the same indicator and year still disagree, retain both links and ask the provider or library for clarification. Do not average the values. Do not select the larger one because it benefits an application. Mark the row unresolved until you have an explanation suitable for the purpose of the record.

Be particularly careful with screenshots circulated in messaging groups. They can omit the year selector, category name or surrounding qualifications. Ask for the underlying source link. If access is restricted, request help from the library rather than relying on a cropped image whose context cannot be recovered.

A missing value should remain missing. It is not necessarily zero, and it is not evidence of misconduct. State exactly what was available and what was not. If a formal application demands an entry, ask the responsible office how to represent an unavailable indicator.

How to write the result in ordinary language

A useful template is: “In [provider’s report], [journal title, ISSN] had a [year] [indicator] of [value] and was in [quartile] for [category], checked on [date].” Fill it only with verified information. If reporting several categories, list them rather than implying there was one universal rank.

When explaining a choice to co-authors, lead with the reason the journal fits the work. The metric can follow as context or as evidence of an external requirement. For example, the blank structure might read: “We prioritised this journal because [reader and scope evidence]. We also checked [specified institutional criterion] using [dated source].”

Do not use the metric to describe a paper you have not assessed. “Published in a journal with…” is a statement about a venue. “High-quality evidence that…” is a judgement about a study and needs study-level support. Preserving that distinction makes your reporting both more accurate and more useful to readers.