Page 217 - Becoming Metric Wise

208 Becoming Metric-Wise

Table 7.2 Publication-citation (p-c) matrix of a fictitious scientist
Publication (date) # citations

A (2010) 8
B (2013) 5
C (2011) 5
D (2011) 4
E (2008) 4
F (2014) 1
G (2010) 0

The scientist represented by Table 7.2 has h-index 4. Indeed, the first
four publications have each at least 4 citations and four is the largest natu-
ral number for which this statement is true. Alternatively we may say that
the first four publications have each at least 4 citations and the fifth
ranked publication does not have 4 1 1 5 5 citations. Using the conven-
tion to rank publications with an equal number of citations in antichro-
nological order the h-core consists of articles A, B, C, and D. The
extended h-core is {A,B,C,D,E}. When the context is clear, a scientist
such as the one represented in Table 7.2 will be represented more suc-
cinctly by the array [8,5,5,4,4,1,0].
We note that citations are collected from some database, and hence a
scientist’s h-index depends on the used database or subdatabase. In prac-
tice a scientist can have a Web of Science (WoS) h-index (or a Social
Science Citation Index h-index, when restricting the WoS to the part
covering the social sciences), a Scopus h-index, a Google Scholar (GS)
h-index and maybe also an h-index according to a local or other database.
What does the h-index measure? What type of measure is it? It is not
an average, percentile or other basic statistical measure. It is another way
of combining citation and publication counts. Almost overnight it
became very popular, leading to its inclusion in the WoS, Scopus, and
Publish or Perish, among others.
Before Hirsch most scientometricians agreed that scientometric indicators
were not suitable for the evaluation of single scientists. In the introductory
chapter of the Handbook of Quantitative Science and Technology Research (Moed
et al., 2004) van Raan (2004a) wrote “the process of citation is a complex one,
and it certainly does not provide an ‘ideal’ monitor on scientific performance.
This is particularly the case at a statistically low aggregation level, e.g., the indi-
vidual researcher.” However, van Raan continues by stating that the application
of citation analysis to the work of a group of researchers over a long period of