Research assessment / Practical reference

Journal metrics: choose the measure that fits the question

JIF, CiteScore, SJR and h5 describe different citation patterns. Altmetric describes attention to a research output. None is a shortcut for judging the quality of your paper or your likelihood of acceptance.

Sources and scope · Guide reviewed September 11, 2026

Coverage: Six metric families; definitions and interpretation, not a ranking of journals
Source basis: Official provider documentation and DORA. Sources linked beside each definition; the Eigenfactor reference is explicitly historical.

Prepared by the Manusights editorial team. Publisher and institutional instructions take precedence.

Start with the decision

Choosing where to submit: Check scope, recent relevant articles, article type, fees, access requirements and editorial process first. Add a dated citation metric only if it answers a remaining question.
Comparing journal citation patterns: Choose one provider, metric year and comparable subject context. Keep the document set and category visible; do not create a blended prestige score.
Assessing a researcher or paper: Read the work and document its contributions and use. A journal-level metric is not evidence that an individual output is strong.
Understanding public attention: Inspect the mentions behind an attention indicator. Separate discussion, criticism, uptake and endorsement rather than calling all of them impact.

The decision guidance is Manusights' synthesis. It follows DORA's recommendation to assess research on its own merits rather than using the journal as a proxy.

Six definitions, with their limits

These are different measures, not six competing answers to one question. Follow the linked provider definition before reproducing a calculation.

Journal

Journal Impact Factor (JIF)

Citations in one year to the preceding two publication years

A ratio of citations in the reporting year to journal items from the previous two years, divided by the citable items from those publication years. The numerator and denominator are not simply an identical set of document types.

Use it to: Describe citation patterns for a journal in a specified year and subject context. Retain the JCR year and category when reporting a rank or quartile.
Do not infer: It does not measure an individual article's quality or predict its citations. Do not substitute the journal average for an assessment of the paper.

Clarivate: the JIF numerator

Serial title

CiteScore

The same four-year span for publications and citations

Citations received during four years by eligible documents published during those four years, divided by those documents. The five included types are articles, reviews, conference papers, book chapters and data papers, not every document type.

Use it to: Compare titles using the same Scopus metric edition and subject category. Check the underlying publication set when explaining a difference from JIF.
Do not infer: CiteScore is not a converted JIF. Database coverage, eligible documents and the time window differ; a larger score is not evidence that one method is more accurate.

Elsevier: measuring a journal's impact

Journal or other serial

SCImago Journal Rank (SJR)

Citation-network measure; retain the metric year

A citation-network indicator that weights citations according to characteristics of the source, including its subject field and influence. It is not a count of citations or a percentage of accepted papers.

Use it to: Use it as an additional view of citation-network position. Record the year, subject category and provider alongside any quartile.
Do not infer: Do not treat weighting as permission to rank unrelated disciplines without context. A Q1 label without its category, year and provider is incomplete.

Elsevier: Scopus metric definitions

Journal

Eigenfactor and Article Influence

Current-year citations to the previous five publication years

Eigenfactor weights citations through the journal citation network and excludes journal self-citations. Article Influence provides a per-article normalization, scaled to a mean of 1. It is not simply the raw Eigenfactor score divided by an unnormalized article count.

Use it to: Distinguish total network influence from the normalized per-article measure. Record which score and database edition you used.
Do not infer: A size-sensitive score is not a guarantee that any named large journal outranks a specialist title. Neither score establishes the merit of an individual article.

JCR reference card: Eigenfactor metrics, page 4 (2015 definition)

A defined publication set

Journal h-index and h5-index

State the period; Google Scholar h5 uses five complete calendar years

The h-index is the largest number h for which at least h items each have at least h citations. This can describe a journal as well as an author. Google Scholar's h5-index restricts the publication set to five complete calendar years.

Use it to: Inspect the articles contributing to the index, not just the headline number. Record the database, publication window and citation snapshot.
Do not infer: Different coverage and time windows produce different indices. Do not compare an all-career author h-index with a journal h5-index, or assume a missing title has no citations.

Google Scholar Metrics help

Research output, not a journal quality score

Altmetric Attention Score

Accumulated tracked attention; record the observation date

A weighted indicator of attention from tracked online sources. The source breakdown and the mentions themselves are necessary context for understanding the number.

Use it to: Investigate who is discussing an output and what they are saying. A policy reference or a substantive discussion is more informative than an unexplained total.
Do not infer: Attention is not evidence of scientific quality, beneficial impact or endorsement. This score is not interchangeable with a citation count or another provider's metric.

Altmetric: attention score and source breakdown

Original teaching examples / Fictional data

A correct calculation can still answer the wrong question

Example 1: one paper changes an average

Five fictional papers have 60, 4, 3, 2 and 1 citations. Their mean is 14, but their median is 3. Saying that a new paper should receive 14 citations would mistake a group summary for a prediction. This is an illustration of skew, not a reproduction of the JIF formula.

Example 2: compute the h-index

The same ordered set has h = 3: three papers have at least three citations each, but the fourth has only two. If these were the papers in a defined five-year publication window, the calculation would illustrate an h5-index. It would not tell you whether the uncited parts of the work are useful or rigorous.

Comparison decision: do not pick a journal because its CiteScore is numerically larger than another journal's JIF. First align the metric, edition and subject context. Then return to the audience and publication requirements that matter to your manuscript.

Make your comparison reproducible

Keep one record for each observation. If any field is missing, say so in the conclusion instead of filling it from memory.

Identity: journal title and ISSN, or the exact research output.
Measure: provider, metric name, value and metric year.
Comparison set: subject category, document types and publication window.
Evidence: source URL, date checked and any definition or coverage limitations.
Decision: the specific question this observation helps answer, plus what remains unresolved.

Reviewed September 11, 2026. This revision corrects the earlier CiteScore document-type description, journal h-index distinction and Article Influence simplification, and removes unsupported career and cross-field claims. Definitions are not a new measurement of current journal scores. Report a correction.

Cite this resource

Citation formats

Free to use and cite, with attribution. See the reuse policy for what you may republish.

APA: Manusights. (2026). Journal metrics: choose the measure that fits the question. Updated September 11, 2026. https://manusights.com/resources/journal-metrics-guide
MLA: Manusights. "Journal Metrics: Choose the Measure That Fits the Question." Manusights, 11 Sept. 2026. https://manusights.com/resources/journal-metrics-guide
Vancouver: Manusights. Journal metrics: choose the measure that fits the question [Internet]. 2026 Sep 11. Available from: https://manusights.com/resources/journal-metrics-guide