Bibliometric review

NOTE: this post is very much under development, and is largely just notes for myself. 

What is a bibliographic reveiw?

  • An effective way to synthesize a large body of research, using quantitative analysis, with network logic underpinning much of it.

What types of bibliographic reviews are there?

  • Five major methods:
    • Citation: looks for most cited works in the field, using the list of references. This penalizes newer works, but shows key articles.
    • Co-Citation: looks at what works are cited together. Unit of analysis is the cited articles, NOT the downloaded articles. This is the most validated way to show central works. This shows knowledge base clusters, not research fronts.
    • Bibliographic coupling: this looks at what articles share a bibliography, the more they share, the more they are on the same topic. This maps research fronts, but does not say anything about importance.
    • Co-author; looks at who writes together, and shows the research social structure of the field.
    • Co-word: looks at the content of the bibliography and key words to find what words often come together, and thus maps out the topical structure of the field.

–> See Zupic and Cater 2015 for great article on what it is, and how to perform a bibliographic study.

How does one get started:

  • First step is to download a complete set of articles on a given topic with boundary conditions, WITH reference list for each article. Scopus is best, but use Web of Science when this is available. https://apps.webofknowledge.com
    • Log in and save searches, they start new sessions every few hours. Very annoying to lose searches!
  • This set needs to be cleaned, as in:
    • include all spellings, forms and synonyms for a term. (Eg. Expatriate, expatriates, sojourn, etc.). limited to topics / field, possibly journals / sources.
    • Words need to by reduced so same word, but different spelling are grouped
    • Different spellings of names are equated (eg: J.S. Black; John S. Black, J. Stewart Black and John Stewart Black, all load to same person)
    • Decided on downloading on time periods, or group this in software if possible.
    • Can also be categorized by importance (as in, key to defining field, uses constructs, mentions)

Software tools

  • SciMAT: Difficult to get started with, but the most complete software so far. Some issues exporting. Read userguide.
    • See also the website: http://sci2s.ugr.es/scimat/description.html
  • VOSViewr: Easy to start with, gives great visuals; can export to pajek. Limited analysis
    • Read more, and find the guide and manual at: http://www.vosviewer.com/getting-started
  • Pajek: to further analyze network data.
  • BibExcel: not yet tested, but one of the first.

 

Youtube video (in portugese)

 

Output:

Below is an an example (based on raw, uncleaned data) of a co-author network; based on Web of Science data; for researchers who have published on expatriates. Search term: “TS=Expat*”

Share This