Page 281 - Becoming Metric Wise

273
Research Evaluation

This approach eliminates the AoR versus RoA problem and is nowa-
days preferred by many colleagues.
We already mentioned that the approach presented here is based on
the evaluation procedures used at CWTS (Leiden, the Netherlands).
Another, excellent approach is that explained in the Bibliometric Handbook
for Karolinska Institutet (Rehn et al., 2014).

8.7 COUNTRY STUDIES
Most indicators that can be applied to universities can also be applied to
countries. However, some are (even) less suited for countries than for
universities, the h-index being a typical example. This indicator depends
too much on the number of publications, so that lists according to
h-index are very similar to lists according to number of publications. For
that reason Molinari & Molinari (2008) proposed a size-independent
0.4
h-index, defined as h/T , where T denotes the number of publications.
Yet, based on Mahbuba & Rousseau (2010) we suggest that this indicator
overcompensates and hence favors countries with smaller numbers of
publications. In this context we note that in Chapter 9 (The Informetric
Laws) we will show that in a so-called Lotkaian framework with expo-
1/α
nent alpha, h/T 5 1.
When performing a scientometric study of a country the final scien-
tific output is what counts. Yet, results must be placed in context: a study
of a developing country or one of the G7-countries (Canada, France,
Germany, Italy, Japan, UK, USA) is a totally different matter. Hence such
a study may start by the description of the country’s educational system
and its basic economic indicators such as GDP, GDP per capita, percent-
age of GDP spend on R&D, GERD, compounded annual growth rate,
etc. Gantman (2012) for instance studied the influence of economic, lin-
guistic and political factors on the scientific production of countries.
Among these the size of the economy had the largest positive influence.
International think tanks have constructed indicators, not scientometric
ones, allowing comparisons between countries such as the Human
Development Index and the Global Innovation Index.
Combining economic, demographic and scientometric indicators may
lead to indicators such as “Research papers per GDP per capita”
(Mashelkar, 2015).