Research metrics are quantitative indicators used to explore the influence and reach of scholarly publications.
In broader terms, bibliometrics measure usage of research publications based on citation counts and patterns in academic literature, while alternative metrics, track attention beyond scholarly publications, including references in news media, policy documents, patents, and social media platforms. Both attempt to quantify how research outputs are engaged with, but they capture different types of activity.
When used thoughtfully, metrics can offer useful insights into research activity in different disciplines. However, their use also has limitations. Applied without appropriate context, metrics can distort the evaluation of research and researchers. Over-reliance on quantitative indicators, particularly when they replace expert judgement, risks overlooking the complexity, diversity, and qualitative contributions that cannot be captured through numerical measures alone.
Responsible use of metrics
Research metrics must be used responsibly and in line with internationally recognised principles, including those set out in the Leiden Manifesto and The Metric Tide. Key principles include:
- Robustness – using reliable, accurate, and appropriate data sources.
- Humility – recognising that metrics should support, not replace, expert judgement.
- Transparency – ensuring that methods and data used in evaluation are open and verifiable.
- Diversity – recognising differences across disciplines and using a range of indicators that reflect varied research outputs and career paths.
- Reflexivity – understanding the potential effects of metrics and reviewing their use as research practices evolve.
Most Commonly Used Metrics
- Article-level Metrics
-
Article-level metrics provide insight into how individual research outputs are being accessed, shared, and cited. Unlike journal-level indicators, these metrics focus on the reach and engagement of a specific publication.
Common article-level indicators include:
Citation Count
The total number of times a publication has been cited by other scholarly works.
Pros
- Provides a direct indication of how often a publication is referenced in academic literature.
- Easy to interpret when comparing outputs of the same type and age within the same field.
Cons
- Citations take time to accumulate (often several years), meaning newer publications may appear to have limited impact.
- Citation practices vary significantly between disciplines.
- Citation counts alone do not indicate the quality or context of the citation (not all citations are positive!).
Field-Weighted Citation Impact (FWCI - Scopus) or Category Normalized Citation Impact (CNCI – Web of Science)
FWCI and CNCI measure the number of citations received by a publication compared with the expected number of citations for similar publications. A value of 1.0 indicates performance at the global average; 1.50 indicates that the publication has been cited 50% more than expected.
This metric compares outputs with similar publications of the same type, age, and discipline, based on the Scopus (FWCI) or Web Of Science (CNCI) dataset.
Pros
- Compares publications on a like-for-like basis by accounting for discipline, publication type, and publication year.
- As this is a normalised measures it allows for comparisons across different research fields.
Cons
- May disadvantage intentionally interdisciplinary research that spans multiple subject areas.
- In disciplines with lower average citation rates, a small number of highly cited papers can skew results.
Highly Cited Publications (Top Percentile Metrics)
This indicator measures the proportion of publications that fall within the top 1%, 5%, 10%, or 25% most cited publications for their field and year of publication, based on data of a specific dataset (e.g. Scopus, Web of Science, OpenAlex or Dimensions).
A publication’s percentile is calculated by ranking all similar publications, those from the same year, subject area, and document type, based on how many citations they have received. The papers are ordered from the lowest to the highest number of citations, and the position of each paper is used to determine its percentile.
For example, if a paper has a percentile of 99, it means that 99% of comparable papers have received fewer citations, placing it among the top 1% most cited publications in its field for that year and document type.
Percentile thresholds are calculated annually, allowing outputs to be compared with other publications from the same year.
Pros
- Highlights research outputs that are performing exceptionally well compared with global benchmarks.
- Field-normalised, enabling comparisons across disciplines.
- Useful for distinguishing performance between research groups or institutions with similar output or citation averages.
- Percentage-based indicators allow fairer comparison between entities of different sizes.
Cons
- Metrics are more robust with larger sample sizes; they are less meaningful when applied to individual researchers or very small groups.
- Comparisons should be made between entities of similar scale to avoid misleading conclusions.
- Journal-level Metrics
-
Journal metrics provide an overview of a journal’s visibility, influence, and editorial standards, including peer review, transparency, and ethical practices. They should not be used to evaluate individual articles or authors, as they reflect the journal as a whole rather than specific research outputs.
Journal Impact Factor (JIF - Clarivate)
Measures how often the average article in a journal is cited over a two-year period. Calculated by dividing the total citations in a year to articles published in the previous two years by the total number of scholarly items (articles, reviews, proceedings) published in that period.
Pros
- Widely recognized and easy to understand.
- Gives a quick indication of a journal’s visibility in its field.
Cons
- Biased towards fields with higher citation rates.
- Can be skewed by a small number of highly cited articles.
- It can be influenced by editorial practices, self-citations, or citation “cartels.” For example, citations may be artificially increased if editors:
- Encourage authors to cite other articles from the same journal
- Publish more highly citable content, such as review articles
- Increasing the number of “non-citable items (e.g. editorials, letter)” in the JIF calculation
Five-Year Journal Impact Factor
Similar to JIF, but based on citations over a five-year period, providing a broader perspective on a journal’s influence.
Pros
- Smooths out short-term fluctuations in citations.
- Useful for fields where citations accumulate more slowly.
Cons
- Still does not account for differences between disciplines.
- It can be manipulated.
CiteScore (Scopus)
CiteScore is based on all citations received in a given year to documents published in the previous four years, divided by the total number of documents published in those four years. It provides a broader indication of a journal’s citation impact.
Pros
- Covers a wide range of document types and disciplines.
- Transparent calculation and methodology.
- Provides an alternative to JIF.
Cons
- Can be influenced by highly cited review articles.
- Citation practices vary between disciplines, which may affect comparability.
Scimago Journal Rank (SJR)
Measures the average number of weighted citations received by a journal’s publications over the past three years. Citations from prestigious journals carry more weight.
Pros
- Considers both the quantity and type of citations.
- Highlights journals with strong influence within citation networks.
Cons
- More complex to calculate and interpret.
Journal Acceptance Rate
The percentage of submitted manuscripts accepted for publication, reflecting selectivity and editorial standards.
Pros
- Provides insight into a journal’s competitiveness and selectivity.
- Can indicate editorial rigor and quality control.
Cons
- Varies widely by discipline and journal size.
- Low acceptance rate does not necessarily mean higher quality content.
- Author-level Metrics
-
Author metrics are quantitative indicators used to assess the citation impact of an individual researcher’s publications. They should never be used in isolation or to directly compare researchers, due to differences across disciplines, career stages, and publication practices.
H-index: Measures both productivity and citation impact. An h-index of h means the author has h publications that have each been cited at least h times.
i10-index: Created by Google Scholar, it counts the number of publications with at least 10 citations.
G-index: An alternative to the h-index that gives more weight to highly cited articles, highlighting major contributions.
Important Considerations
- Citation metrics are field- and career-dependent; outputs and citation patterns differ across disciplines and over time.
- Metrics fluctuate annually and may change depending on the source database (e.g., Web of Science, Scopus, Google Scholar).
- Recent papers, which may have high impact or relevance, often contribute little to cumulative indices.
- These metrics should always be used alongside qualitative assessment, such as peer review, research contributions to policy, teaching, or societal impact.
- Alternative Metrics
-
Alternative metrics provide a complementary view of research impact, working alongside traditional citation metrics to give a broader picture of how research is being noticed and discussed worldwide. They track mentions on social media, blogs, news outlets, policy documents, and other online platforms to measure attention around a publication, researcher, journal, or institution.
Media Mentions
The number of times a publication is referenced or discussed in sources such as social media, news outlets, blogs, policy documents, or online platforms.
Pros
- Captures early attention and engagement shortly after publication.
- Reflects interest beyond academic audiences, including policy and public discourse.
- Can highlight societal relevance and broader research engagement.
Cons
- Attention does not necessarily reflect research quality.
- Susceptible to short-term trends or media visibility.
- Coverage varies widely across disciplines and topics.
Views and Downloads
The number of times an article has been viewed or downloaded from publisher platforms, repositories, or other online sources.
Pros
- Provides an early indicator of readership and interest.
- Available soon after publication.
- Useful for understanding engagement with open access research.
Cons
- Views or downloads do not necessarily indicate that the work has been read or used.
- Metrics may vary across platforms depending on how they are recorded.
- Not directly comparable across publishers or repositories.
Saves (e.g., Mendeley Readers)
The number of times a publication has been saved to personal libraries on reference management platforms such as Mendeley.
Pros
- Indicates interest among researchers who may intend to read or cite the work later.
- Often appears earlier than citation counts.
- Can provide insight into engagement within research communities.
Cons
- Represents intention to read rather than confirmed use.
- Dependent on the popularity of specific reference management tools.
- Coverage varies across disciplines.
Metrics Toolkit : Learn more about specific metrics, and what they do and do not measure