Page 111 - Becoming Metric Wise

102 Becoming Metric-Wise

5.1.6 Criticism on Citation Analysis
MacRoberts and MacRoberts (1989, 2010) noted that many influences
on the content of articles stay unmentioned. As an example they quote
the field of biogeography. Observations in this field often end up in floras.
When these floras are used in analytical studies they are rarely cited and
the observations on which they are based are certainly not. The point
made by MacRoberts and MacRoberts is that being uncited or rarely
cited is not the same as being unused or of little interest.

5.1.7 Citation Networks
Articles are connected to other ones by citations. These citation links lead to
a directed and possibly weighted network of citations (for more on networks,
see Chapter 10: Networks). Hence, one can derive citation networks of arti-
cles, journals, authors, universities, countries and many more. Scientific
authors may also be connected by coauthorship, cocitation and bibliographic
networks (see further for a definition of these notions). The term network
theory is often used when one has applications in mind, while the term
graph theory is more used in a purely mathematical context. Yet, this distinc-
tion is not very strict and these words may be used interchangeably. Needless
to say network theory plays a basic role in citation analysis. Some aspects of
network theory will be studied in Chapter 10, Networks. Price (1965),
Xhignesse and Osgood (1967), and Garner (1967) were among the first to
study the citation relation as a graph.

5.1.8 Applications of Publication and Citation Analysis

We mention three main application areas of publication and citation anal-
ysis (Hu et al., 2011).
• Information retrieval: returning to the origin of the citation network,
or finding related or subsequent articles. In information retrieval this
method is called citation pearl growing or cycling.
• Research evaluation: contributing to the evaluation of scientists,
research groups, institutes and universities, and groups of countries.
Note the word “contributing” as research evaluation can never be
based on publication and citation indicators only.
• Study of science: research related to the structure of publication and
citation networks. This can be done by tracing the origin of an idea
or the relations between subfields. Garfield already applied citation
analysis when tracing the history of the breaking of the genetic code,