A crash course in scientometrics

Have you ever heard about scientometrics? While chances are that you have never heard about this term, you have probably heard about its applications. 

Citation metrics like the h-index and impact factor are used by researchers and academic institutions worldwide to determine career advancement, funding, and reputability. 

In this blog, I will briefly introduce you to the field of scientometrics. 

What is Scientometrics? 

Scientometrics is a field that studies scholarly publications. It does this in two major ways: quantitatively and qualitatively. The quantitative aspect of scientometrics involves aggregating and analyzing data. In general, this data will describe authorship, citations, and research themes. The authorship data will help identify the most prolific researchers and institutions within a field. This type of information is readily available on PubMed, Google Scholar, Web of Science, Scopus, and ResearchGate.  On the other hand, citation metrics serve as proxies of popularity. The interactions between authors, articles, and themes can be visualized using special maps called social network maps. These maps are made of circles called nodes and these nodes may represent authors, academic institutions, countries, articles, or keywords. Nodes sharing a relationship will be linked by a line called a link. The link lengths are inversely proportional to the proximity between the nodes. As such, the more two nodes have in common, the shorter their link. Groups of nodes that share a lot in common will aggregate due to their short links. Nodes within these groups will share a color together and will be called a cluster. Cluster analysis helps identify the major groups within a field but also, it helps identify isolated nodes. Nodes with multiple links will be larger in size and generally more influential.

The data needed for scientometric studies can be obtained from major databases like PubMed and Web of Science. While PubMed is free, Web of Science is not! So, you will need access via an institution. Also, you will have to build a comprehensive search strategy to identify relevant articles. Once you have that data, you can use one of a multitude of scientometric software including the Bibliometrix package in R, VosViewer, and CiteSpace.

For more on scientometric studies, check out some of my publications on Google Scholar.