Page 282 - Becoming Metric Wise

274 Becoming Metric-Wise

8.8 SOME REMARKS ON EVALUATIONS OF INDIVIDUALS
8.8.1 General Remarks
Most, or probably all, bibliometric indicators should not be used on the
individual level (cf. Leiden Manifesto, point 7). With Phelan (1999) and
many other colleagues we stress that when making inferences about indi-
vidual cases, considerable caution should be used. The two main reasons
for this are that, most of the time, numbers are too small, and more
importantly numbers of citations are highly skewed so that averages or
even medians become virtually meaningless. Nevertheless, Wildgaard
et al. (2014) have even collected and discussed 108 author-level indica-
tors. Indeed, individual researchers must be evaluated and this for multi-
ple reasons. Here is an incomplete list of such reasons:
• Recruitment
• Promotion
• Tenure
• Funding allocation
• Yearly (or, more generally, periodical) review answering questions
such as: Is this doctoral student on the right track to get their degree
during the available time span? Shall we extend this person’s contract
for another period?
• In centers of excellence, being among the top 5% may not be good
enough so one wants to find answers to questions such as: does one
expect this person to move up to the top 1% in the field?
• In mission-oriented institutes one wants to make sure that researchers
perform investigations relevant to the mission of the institute.
All these questions leading to the stated aims must be answered by
review committees.

8.8.2 Evaluation and Author Roles
One aspect of scientists’ publication careers is the role they play in pub-
lished research, more concretely: are they usually first author or usually
the last one? In the experimental sciences the first and last positions in the
byline of a publication are usually considered to be the most important
ones. The scientist occupying the first position is then doing the bulk of
the experimental work, while the colleague in the last position supervises
the research. In a study on cooperation between scientists Wardil and
Hauert (2015) refer to authors who contribute mostly as last authors as