Page 56 - Becoming Metric Wise

46 Becoming Metric-Wise

well-written reviews to other reviewers (of the same submission)
as part of a teaching activity as for scientists too, reviewing is a learning
process.
Following Varn (2014), we note that a reviewer is not expected to
replicate experiments, redo statistical analysis or go through all calcula-
tions in detail. The primary objective is to say whether the results are
important enough to publish, more specifically, to be published in the
journal to which it is submitted. The purpose of peer review is not to
establish that the results in the submitted manuscript are surely correct.
Instead, it does seek to determine whether the paper has obvious mistakes
or omissions, is original, and is relevant to the journal. Peer review fur-
ther checks whether the conclusions made by the authors are supported
by the evidence presented, and whether the work is presented in a way
experts in the field would understand. As for the issue of fraud, peer
review is not the place to catch it, unless it was rather sloppily done.
Indeed, who can say that an observation wasn’t made?

3.1.6 Editorial Decision Schemes
We already mentioned that reviewers offer advice to the EIC who, in the
end, has to accept or reject the manuscript. Now we take a closer look at
how an EIC may proceed once he or she has received the suggestions of
the reviewers. If one of the reviewers found a serious error, then the
submission is rejected. Assume now that this is not the case and that the
submission has been sent to two reviewers.
If the advice is YY (Y stands for yes: accept; N stand for no: reject)
then the decision is accepted, and if it is NN the decision is rejected.
Decision schemes differ in the case of NY. In the clear-cut system
used e.g., in the journal “Angewandte Chemie” (see Bornmann &
Daniel, 2009b), this leads to a rejection. Otherwise, the EIC plays the
role of a third reviewer and, hence, follows his or her own advice, or
sends it to a third reviewer, who, possibly without knowing it, makes the
decision, as the EIC follows a majority vote.
One may wonder if three reviewers are better than two. Schultz (2010)
investigated 500 manuscripts submitted to Monthly Weather Review. In this
journal, Schultz, being the EIC, chooses an editor, who then invites
reviewers. This editor is free to choose of the number of reviewers. Schultz
found that manuscripts were usually reviewed by two (61.2% of the time)
or three reviewers (31.0% of the time). Occasionally a manuscript got zero,