Page 277 - Becoming Metric Wise

269
Research Evaluation

“journals with an impact factor.” Of course, one may consider other
or larger sets obtained by including journals covered by Scopus, or
included in a specially constructed local database (as in the Norwegian
or the Flemish model, discussed further on). Further on, the database(s)
used in the evaluation exercise will be referred to as the designated
database.
• Total number of publications in the designated database, using frac-
tional count (indicator e).
Any form of fractional counting, cf. Chapter 5, Publication and
Citation Analysis, can be used. For example, if an article has 6 authors
and only 2 belong to the research group under investigation, then this
publication receives 2/6 5 1/3 publication credits for the research
group in question (assuming complete-normalized counting).
• The length of a citation window is chosen and for each publication
the number of received citations during the citation window is deter-
mined (indicator f ).
• Global productivity of the group (indicator g).
This indicator is defined as indicator b divided by indicator a. It
can be seen as a maximum productivity.
• Fractionalized designated database-productivity (indicator h).
This is defined as indicator e divided by indicator a.
The values of indicators g and h provide an upper bound and a
lower bound for the productivity of the research group. If the group
publishes mostly articles in journals covered by the designated database
and rarely collaborate with other groups, this upper bound and lower
bound will not differ much. Yet, in most realistic situations there may
be a large difference between these two indicator values.
• Percentage journal articles published in journals covered by the desig-
nated database (indicator i).
Indicator i is defined as indicator d divided by indicator c.
A small value for indicator i may indicate one of three things, or a
combination thereof: either the group published mostly articles of
lower quality which were not accepted in the better journals, or cov-
erage of this group’s field by the designated database is not adequate,
or the group deliberately chooses to support new journals which are
not yet covered by the designated database.
• Percentage of journal contributions with respect to all publications
(indicator j).
Defined as indicator c divided by indicator b.