There are four things to keep in mind when writing feedback:
(1) Write the kind of comments you expect to get
Most journals suggest that feedback to author/s should be positive and offer concrete advice. However this is not always what happens.
I have seen some truly awful feedback destined for authors. Among the worst… This is the kind of article that gives this methodology a bad name… This is a naïve and simplistic view of… The author has clearly never read beyond… I would fail this if it was a first year essay… This is just awful. There is usually an opportunity to say this kind of thing to an Editor if you must, but it is devastating for author/s to get such off-handed, smart-alec comments.
It’s also not helpful to become teacherly. This is a peer writing, it’s not a student essay. Even if this is obviously written by a doctoral researcher, they are expecting to be dealt with as a colleague, not sent to the kiddy table.
Feedback comments should be appreciatively critical, just like the reading of the article. Generosity of spirit and collegiality in tone is the order of the writing feedback day.
(2) Use a structure for the feedback which allows the author to follow what you are saying.
Remember that the author/s now know that they need to do more, so they are reading with a sinking heart. So you need to be specific as well as kind.
I generally aim for three quarters to a page in length, unless it is an accept without change –this is usually just a paragraph or two saying what I think is great about the paper.
I have a bit of a formula I use for reviews. So here’s what I do – it’s not the only way to write feedback of course, but it’s ONE way to approach the task.
• Write two to four sentences summarising what the paper is about. So something like… This paper addresses… and presents evidence that … . The author/s argue that… This gives the author the chance to see whether you have understood what they wanted to say. If you haven’t got it, they can then consider how they might have produced this misreading.
• If you really enjoyed reading the article, say so now before you start with the concerns.
• Write something about the contribution, as in… The article clearly makes a contribution to/has the potential to add to what we know about/will make a significant addition to … This might be linked to a caveat such as… but needs further work in order to bring this to fruition/realise its potential, needs some revision in order to achieve this.
• Then, if there are suggested revisions, say whether they are major or minor and how many there are, as in .. I have two suggestions for major revisions and one more minor point… or I offer some issues that the author/s needs to consider in the methodological section and a recommendation for some restructuring of the findings…
• Then dispassionately state the changes that you think are necessary, based on your reading of the article. Try to focus on the things that are the most fundamental.
You may just outline the problem (s) and suggest that the author/s needs to find a way of resolving it/them. You might offer one or two suggestions. Or you might have something very definite in mind. Any of these is OK, although just outlining the problem can be a bit scary for the author/s when they come to revise. Whatever, you just need to be explicit about which of these you are doing.
If there is reading that the author/s need to do, give them the references, don’t just say there is literature out there that they ought to know about.
If you are suggesting major revisions, then there probably isn’t much point in outlining twenty five specific things for the author/s to do; it’s the big bits that are the most important for the author/s to grasp. Too much detail and they will be completely confused/overwhelmed/dispirited. And if it’s major revisions you will get another look at the paper, at which time you can pick up any small things that still need resolution.
Finally, succintly list any grammatical, proofing and referencing problems.
• Conclude with some encouragement. This might involve repeating the potential contribution and the importance/value of the author/s continuing to work on the piece.
(3) Be clear
Reviews can be written in a kind of code, just like real estate advertisements. While it is important not to be rude/sarcastic/patronising, it is just as important to be clear. If there is a problem that you can see with an aspect of the paper, say what it is and don’t waffle. Don’t say The references need attention , say The references need to be in the appropriate journal style. Don’t say The methodology section needs to be clearer, say The methodology section needs to include information about the site, sample and types of data generated as well as the methods of analysis. Etc.
The clearer the feedback the more chance the author/s has of deciding whether they agree with you and/or doing exactly what’s needed to get the article to publication.
(4) Don’t tell them to read all your work – unless you really ARE the key figure in the field
This can be really hard. After all, you’ve been sent the article to review on the basis of your expertise and you know how your work might help. BUT refereeing is not about upping your citations, it’s about the author/s writing about their work and they don’t have to cite everything in the field in order to do this. Perhaps the tip is to only require a reference to your work if it’s central to the argument being made.