Page 276 - Becoming Metric Wise

268 Becoming Metric-Wise

types. Using this technique a group (or university) may have a high total
score without scoring high on every aspect. We think that the use of this
technique may benefit existing university rankings.

8.5 EVALUATION OF RESEARCH GROUPS

In this section, we provide a short overview of indicators used for scien-
tometric evaluations of research groups. This overview is largely based on
work performed at CWTS (Leiden), see (Moed et al., 1983, 1985a; de
Bruin et al., 1993; Wilsdon et al., 2015). We assume that the evaluation
takes place based on publications during the period [Y 2 T, Y], a period
of a discrete length of T 1 1 years.

8.5.1 Evaluation Indicators

• Average number of FTEs for research (indicator a).
In this expression FTE stands for full-time equivalent. If each col-
laborator worked full-time as a researcher the indicator FTE would be
the number of personnel. Yet, some researchers work part-time or
worked just for a few months in a particular year (end of contract;
maternity leave, etc.). This leads to a fractional contribution to the
research group. Finally, some scientists, although working full-time do
not work full-time as a researcher: they have teaching and/or adminis-
trative duties. Taking all these effects into account leads to a yearly
FTE. These numbers are averaged over the period of study (T 1 1
years).
• The total number of publications during the period of investigation
(indicator b).
To determine this number all scientific publications (any type, any
publication outlet) with at least one member of the group as contribu-
tor are included (published during the period of investigation).
Sometimes working papers are included too, especially if they are
published, after initial screening, in a numbered series of such papers.
• Total number of publications in journals (indicator c).
This number is determined in the same way as indicator b but
restricted to journal articles.
• Total number of publications in journals included in a designated set
of databases (indicator d).
Often one considers only one such database, namely the WoS and
the Journal Citation Reports, and these journals are then said to be