Thursday, 10 November 2016

University Library : Bibliometrics: Journal, article and author metrics - Durham University


Bibliometrics: Journal, article and author metrics

Bibliometrics is the process of extracting measurable data through
the statistical analysis of texts, and information about how the texts
are being used. For researchers this might be:

  • to determine how many times a researcher's work has been cited in key literaturE;
  • (SCHOLARLY IMPACT) to explore the impact of their field, the impact of a set of researchers, or the impact of a particular paper
  • (EVALUATION) a means of measuring patterns of authorship, publication, and the use of literature to aid in evaluation and review
    to help identify if an author is attracting citations from outside
    their main field of study, and thus highlight potential opportunities
    for cross-disciplinary collaboration.
  • (DISSEMINATION) to help guide or influence publication strategies (eg deciding where to publish research in to obtain maximum visibility)
tools, such as citation analysis, remain controversial as proxy
indicators of the quality of published research. For example, citations
do not measure the extent or particiaption in multi-authored
publications, and negative citations will contribute as positive
indicators of impact.

Traditional bibliometric tools can also
potentially underestimate or completely fail to highlight the true
impact of a given piece of research as they tend to rely on data sets
restricted to academic publications and citation indexes and lack a
measure of the 'immediacy' of any initial impact which may not
necessarily translate into published citing articles.

Metrics (or Altmetrics) are developing branch of metrics which look at
non-scholarly as well as scholarly sources such as social media and
non-academic online sources.

This page looks at some of the
tools, resources and training available for you as a researcher in
exploring the various journal, article and author metrics you come
across or wish to use.

Citations, citation counts and citation indexes


Citations are a reference to a source or underpinning set of data for
the purpose of acknowledging their relevance to the topic of
discussion. Different publication cultures in different disciplines mean
that in some subject areas you are more likely to be expected to cite
extensively the work that your scholarship and research is building
upon, than in others.

Citation counting

number of citations an article receives is one indicator of the
"academic impact" of the article, based upon its popularity in terms of
how many people have read and then refered to that research. A high citation count
is not an indicator of high quality. Indeed, an article may be widely
cited as an example of bad practice. It should also be noted that there
are wildly varying traditions of citiation in different disciplines and
therefore what is highly cited in one field may only represent an
average citation count in another. There are various tools available
which enable you to track academic debate based on who has cited a given
work. This might be useful for:

  • Tracking the impact of
    your own research (for means of engaging with other researchers,
    identifying opportunities for collaboration or simply identifying more
    fruitful publictaion within which to publish;
  • Keeping up to date with how a particular piece of research you are interested in is received and treated by others.

Citation indexes

order to monitor citations, you first of all of course need a dataset
recording citations across a broad enough range of publications to make
any collection, counting and analysis in any way meaningful.

JSTOR provides an indication of the number of times an article has been
cited, but only counts those citations in journals which also appear on
the JSTOR platform. For many subject areas, this might only be a very
small proportion of the full range of academic titles where citations
might have appeared.
There are three main datasets
available which each have multi-disciplinary coverage of available
research and the citations associated with that research.

  • Web of Science and related products (provided by Thomson Reuters)
  • Scopus and related producst (provided by Elsevier) [Not available at Durham University]
  • Google Scholar
These services allow you to:
  • count citations - find out how many others have cited a given work
  • cited reference search - find out which articles have cited a previously published work
  • citation alerts - find out when a newly published article cites a previously published work
  • citation reports - for a variety of statistics relating to an author
  • citation maps -find related material and shared citations [Web of Science]
For further information, see training available.

Journal Citation Reports® and Journal Impact Factors

Journal Citation Reports®

Every year, Thomson Reuters publishes the Journal Citation Reports® which are available via Web of Science
for Science and Social Science journals indexed within the service.
They give an indication of the "scholarly impact" of a particular
journal using quantitive analysis of citations for articles published in
that journal. They should always be considered in conjunction with
advice from peers, and it is not possible to do a comparison of journals
across discipline boundaries where citation cultures might vary

Journal Impact Factors (JIF)

The Journal Impact Factor (JIF)
are probably the best known aspect of these reports. An impact factor
represents the average number of citations in a year given to those
papers in a journal that were published during the two preceding years.
It is therefore a reflection of the current impact, or popularity, of a

Alternative ways of measuring journal impact include the Eigenfactor and the data produced by SCImago (which utilises data from the Scopus database, rather than Web of Science).

The presentation below provides a brief overview, or for further information, see training available.

Author metrics

In addition to counting the number of citations a particular research
paper has attracted or looking at the 'impact' of a particular journal
in your field of research, bibliometrics can also be used to assess the
output and scholarly impact of an individual author's research. These
might be taken into account for:

  • applications for tenure
  • assessments for promotion or probation
  • applications for research funding awards
is useful therefore to be aware of the various metrics that can be
used, how they are generated and some of the issues around them.

  • Total number of papers / total number of citations
  • The (Hirsch) h-index of an author is the number of their publications (h) which have been cited at least h times.
  • The (Egghe) g-index of an author is a adaptation of the h-index which gives more weight to an authors more highly cited papers.
The presentation below provides a brief overview, or for further information, see training available.


Traditional bibliometrics have many limitations, not the least of
which is that they focus solely on citations in scholarly publications

  • May not be available til some time after the original article is published;
  • Only
    provide information on who has cited a work in a peer-reviewed article,
    not who else is reading, sharing, discussing or commenting outside of
    traditaional publications;
  • Are limited to the coverage of the particular citation data set used.
researchers share ideas and research at all stages of the research
process online, and other researchers as well as a wider, non-academic
audience, can engage with research outputs via a variety of different
media such as twitter, blogs, youtube, news coverage etc.

means that the 'impact' of you research can be seen much more quickly,
and across a much broader scope, than than might be reflected using
traditional bibliometric data.

Altmetrics seek to measure the
impact of articles by counting mentions by social media sites and other
digital communications. There are various tools available which you
might be interested in exploring, such as:

with traditional bibliometrics, there are still reasons to treat
altmetrics with a level of caution (for example, it is easier to "game"
altmetric data) but they can be useful to inform decisions on impact,
publication strategy and potential collaboration opportunities.
are planning to provide guidance as part of the Researcher Development
Programme in 2014, alongside sessions on traditional bibliometric

Bibliometrics Training

Training on bibliometrics, Journal Impact Factors, author level metrics and open access are currenlty provided as part fo the Researcher Development Programme, and scheduled to run during the Michaelmas and Epiphany terms.

session is also delivered as part of the Leading Research Programme,
aimed at early career researchers who have recently submitted, or are
about to submit, their first research award application.

For any additional help or guidance, or a short one-to-one sessionyou can also contact James Bisset.

University Library : Bibliometrics: Journal, article and author metrics - Durham University

1 comment:

  1. It is nice blog Thank you provide important information and i am searching for same information to save my time Data Science online Training