Information Technology Reference
In-Depth Information
more than outline the basic idea. Such authors perhaps want to establish that an idea
is theirs, without going to the trouble of demonstrating its correctness, or are simply
tired of the work and hope referees will supply details they haven't bothered to obtain
themselves. Such papers don't deserve a thorough evaluation. However, don't be too
quick to judge a paper as being in this category.
Referees should make an effort to search for errors that don't affect the quality
of the work but should be corrected before going into print. These include spelling
and syntax, written expression, errors in the bibliography, whether all concepts and
terms have been defined or explained, errors in any formulas or mathematics, and
inconsistency in just about anything fromvariable names to table layout to formatting
of the bibliography. Some of these kinds of errors may be picked up in the typesetting
process, if the paper is to appear in a journal, but many of themwon't be. In particular,
only a referee is likely to find errors in mathematics.
Such errors can become more serious defects that might make the paper unaccept-
able. A few typographic errors in themathematics are to be expected, for example, but
if the subscripts aremixed up or the notation keeps changing case then it is quite likely
that the author has not checked the results with sufficient care; it may well be reason-
able to reject the paper and expect the author to review it before submitting again.
Similar arguments apply to the presentation: to a certain extent poorly written
papers should be accepted (however reluctantly), but real incompetence in the pre-
sentation is grounds for rejection, because a paper is of no value if it cannot be read.
But note that the converse does not apply: excellence in presentation does not justify
acceptance. Occasionally a referee receives a paper that is well written and shows
real care in the development of the results, but which does no more than reproduce
existing work. Such papers must, regrettably, be rejected.
A difficult issue for some papers is whether to recommend outright rejection or to
recommend resubmission after major changes. The latter means that, with no more
than a reasonable amount of additional work, the paper could be of acceptable stan-
dard. This recommendation should not be used as a form of “soft reject”, to spare
the author's feelings or some such, while asking for changes that are in practice
impossible; eventual acceptance, perhaps after several more rounds of reviewing, is
the usual final result of such a recommendation. If getting the work to an acceptable
standard will involve substantial additional research and writing, rejection is appro-
priate. This verdict can be softened in other ways, such as suggesting that the paper
be resubmitted once the problems have been addressed.
As a consequence of the peer review system, active researchers should expect to
referee about two to three times as many papers as they submit (or somewhat less if
their papers are usually co-authored) and only decline to referee a paper with good
reason. For many papers, there may be no potential referee who is truly expert in the
area, so be prepared to referee even when you are not confident in your judgement of
the paper. Always state your limitations as a referee—that you are unfamiliar with the
literature in the area, for example, or were not able to check that certain proofs were
correct. That is, you need to admit your ignorance. Ultimately, a referee should not
recommend acceptance if the paper is not of adequate standard in some respect—the
onus is on the referee to fully evaluate the paper.
Search WWH ::




Custom Search