by M. Koivula

Here are guidelines on how to review a manuscript for a journal. Some of you are already involved in the process, and the rest will be as soon as you get your first paper printed. This document contains a checklist on reviewing, a sad example statement, and guidelines on writing two letters: comments to the authors, and comments to the editors of the journal.

Note: you will not be paid for your statements, and if you write high-quality statements you will certainly be requested to review again.
Some idealistic points

    * A referee helps the journal and the authors to improve their work
    * Take the reviewing as a learning opportunity (new methods & ways of thinking)
    * Avoid giving "opinion" comments on methods unless these truly will produce new perspectives to the ms. Opinions are welcome, however, on Introduction and Discussion sections.
    * Spend as much time as you need with the review!

REVIEWING CHECKLIST

    * Title adequately describes the content
    * Necessary contact information, authors' full names, apropriate key words (no or only 1-2 title words repeated) and a short version of the title are provided
    * Abstract presents why the study was done, how it was done, what were the main results, and what they mean in terms of the general context
    * Introduction provides adequate background for understanding the context, and ends with study questions and/or hypotheses (that are linked to the context)
    * Material and methods describe the design, sampling and other methods so that you could duplicate the study, and no open questions remain e.g. on sample independence, reasoning for the techniques used, etc.
    * Results section presents what was found in relation to the study questions, no more or less, and no discussion or speculation is included
    * Analyses are correctly done, no undescribed analyses are presented, necessary analyses are all done, and they are not redundant (compare study questions)
    * Discussion handles all the results, statements are justified by data, and relevant literature is cited
    * References seem to be error free - many (but not all) reviewers cross-check references and citations
    * Figures and tables are all necessary and of high quality, and their captions adequately explains their content (e.g., no unexplained column names)
    * Terms are used in a consistent way, and are well explained
    * Exotic methods are adequately explained and reasons for their use – instead of using standard methods – given
    * If you do not know "enough" on something in the ms, e.g. on species ecology or a fancy method used, tell it clearly to the editors (what a shame)

EXAMPLE "Comments to the authors"

A paper on beetles of highway central reservation strips by Koivula, Kotze and Salokannel. The below letter is partly imaginary, as I could not find the original one, so a couple of points are missing from the criticism list. The bad thing is that the authors are usually routinely requested to respond to these sorts of comments. However, in this case the journal editor wrote (translation mine):

    "After studying the comments of the referee, I conclude that they are not justified. Therefore I ask you to do only the small correction the other referee suggested, and return the final text, tables and figures a.s.a.p. to the journal for page proof process."

Here is the referee statement:

    This paper deals with beetles on grassy strips between highway lanes.
    I cannot recommend the publication of the ms because:
    (1) The graphs and tables are not particularly good.
    (2) The text must be shortened by at least 50%.
    (3) Sampling period was short: several years of sampling are needed to provide reliable evidence for effects of roads.
    (4) I do not like pooling species into some "ecological" groups.
    (5) Pitfall trapping leads to unreliable catches.

LETTER 1: COMMENTS TO THE AUTHORS

Write a review that you would like to receive if you had written the ms. This does not mean "This is a wonderful and probably the most significant scientific contribution I have ever seen. I wholeheartedly recommend it to be urgently published as it is. I just wonder why the authors did not submit this to Nature!" Rather, be critical but constructive. (Do not include recommendations on publishing the ms - that decision is on the editors.)

    * Start with a very brief overview on the ms content (main results) and sum up some nice stuff that the ms includes. These can include different technical issues, such as that the ms is well written and the text is logically proceeding, graphs and tables are of high quality, etc. (Be honest, of course.)
    * Do not just list criticisms, but provide reasons on why these comments should be taken into account when further preparing the ms
    * Provide a clear list of issues that you think must be corrected or changed, and give fact-based reasons why
    * If shortening is needed, give suggestions on where, and what to leave out
    * List other comments (no typo pin-pointing unless you have nothing else to say)
    * Preferably be neutral, i.e., do not let personal issues affect your reviewing. This may of course be difficult if you do not personally like the authors, or if you are currently preparing a manuscript on your own similar results
    * Try to take into account cultural differences if possible.
    * In most journals you can sign your statement!

LETTER 2: COMMENTS TO THE EDITORS

Write a short cover letter that includes the following issues:

    * Your brief overview of the paper: topic, main results, and their significance
    * Your opinion on whether the ms fits to the journal scope, and for whom the ms may be of interest (if you can think of any)
    * Your recommendation on accepting (usually after no (rare), minor, moderate or major review) or rejecting the ms. If you think the ms is not publishable either in the journal in question or generally, give the editors clear and justified reasons why you think so
    * List aspects that need to be changed before publishing, and give reasons why
    * If you are unsure of some part of the ms (e.g., an unknown method), suggest requesting another opinion on that aspect

SUMMARY

Reviewing is a time-consuming piece of cake and should be done properly, although you run the risk of being requested ad nauseam.