Wednesday, 19 April 2017

Talking to your researchers about the h-index


Talking to your researchers about the h-index

By Jenny Delasalle, Freelance Librarian | May 24, 2016

h-graph for David Neal

The h-index is an indicator of research impact based on
citation measurement. It is an attempt to measure a researcher’s
productivity and the impact of their published documents. It was
introduced by Jorge E. Hirsch in 2005 and has become one of the most
common research impact metrics. It can be used for an individual author,
or any collection of documents, e.g., for a journal or a research
group’s outputs.

How is h-index calculated?

If a researcher has an h-index of 4, it means that 4 of
his/her documents have been cited at least 4 times as shown below. If
document 5 in the list accrues one more citation then the author’s h-index would increase to 5.

Professor X has a total of 10 documents:

Document 1: 50 cites

Document 2: 18 cites

Document 3: 11 cites

Document 4: 7 cites

----------------------------------------------------------- h-index: 4

Document 5: 4 cites

Document 6: 3 cites

Documents 7, 8, 9, 10: 0 cites

What we can also see is that though the author has received an
impressive 50 citations on Document 1 that is not reflected in the h-index.

h-index will vary based on data source

The calculation for h-index is the same regardless of where it’s found; however, the h-index
could vary depending on the data source. For example, if one database
indexes and draws data from a larger pool of journals then it may have
more citations to include in the calculation. Therefore, when comparing h-indexes, it is important to compare from the same data source. Three of the most common are:

  • Scopus

  • Web of Science

  • Google Scholar

Benchmarking an author’s h-index

It’s important not only to use the same data source when benchmarking
authors, but also to account for discipline and career stage. This is
also a good time to have a discussion with your researcher on whom they
want to compare themselves to, which will likely change over time. Your
researcher’s co-authors or colleagues might be a place to start.

You can find sources that attempt to quantify what a good h-index
might be in a certain discipline. For example, on the London School of
Economics and Political Science Impact Blog they discuss average
h-scores for 120 social science academics, showing average h-scores by
discipline and position.

LSE Impact Blog table - average h-scores for 120 social science academics


In comparison, the following table of highly cited Spanish scientists
demonstrates the wide range from clinical medicine to mathematics.

Scaling the h-index for different scientific ISI fields - Juan E. Iglesias and Carlos Pecharromán


Criticisms of the h-index and suggested alternatives

Criticism of h-index


h-index can only be as high as the number of documents published.

g-index (number of citations factors into calculation)

An emeritus professor may have a higher h-index than an
early career researcher, even if the emeritus professor has not
published in years and the ECR has one of the most popular research
papers of the year.

- contemporary h-index (weights recent citations more heavily)

- h5-index (outputs from last five years; Google Scholar)

There are many other h-type indexes that can be used based on your researchers’ specific situation.

Other criticisms of the h-index include that it does not
take into account the authors’ roles or the context of the citation. For
example, what if a co-author has only played a minor role in authoring
the paper, yet they get the same value out of the citations within their
h-index calculation as a lead author. Or perhaps the paper was
cited in a negative way, yet the citation still counts toward a
positive impact metric.

Improving a researcher’s h-index

As a librarian, you can discuss with your authors ways in which they might improve their h-index and other measures based upon citations. These include: 

  • Publish a review article. These articles receive more citations than
    other papers (in general). Caveat: Some data sources may not include
    review articles as a document type in the
    h-index calculation.

  • Don’t be too modest: sometimes you can self-cite. Even if the
    researcher’s self-citation has been stripped out of a calculation, it
    might lead to other citations that are counted.

  • List documents out in order of the number of citations they have
    received. Is there one document that lacks only a few citations that
    could boost the score? Has the citation-tracking source missed any
    citations for that paper?

  • Ensure author profiles are consolidated and accurate in databases such
    as Scopus, Web of Science and Google Scholar. Are all of your
    researcher’s papers attributed to him/her? Having multiple profiles in
    one data source will likely reduce your

  • Register for an ORCID iD and link it to the other profiles to maintain a global view in one place.

For more information, see “Effective Strategies for Increasing Citation Frequency.”


Remind your researchers that the h-index is only one measure
of productivity and impact, and that they should be looking at a more
comprehensive picture that includes other scholarly activity and
commentary, and social and mass media activity.

Additional resources

Watch the Library Connect webinar “Research impact metrics for librarians: calculation & context," where Jenny Delasalle discuss the h-index among other topics, along with Andrew Plume.

Explaining the g-index: trying to keep it simple, May 25, 2016 post on Jenny Delasalle's blog

Read about Scopus and the h-index on the Scopus blog, "5 facts about Scopus and the h-index."

For more information, see “Effective Strategies for Increasing Citation Frequency.”

Talking to your researchers about the h-index

No comments:

Post a Comment