Can I write a referee report too fast?

Question

As a math postdoc, I received my first invitation to referee a journal article today, and I accepted the invitation. I've already read a preprint of the submitted article (which has been available on arXiv for a few months and coincides with the submitted version) in great detail some time ago and therefore I already have to say a lot about the article's content and presentation. I was given two months to write a report.

My question, which might be stupid, is now:

Can I write the report "too fast"?

Or, more precisely:

• Does it make a bad impression to the editor if I submit the report too fast (say after a week or two)? Since the editor is an influential and important person in the field, I don't want him to think of me as a bad referee, but I fear that he might think "The report came in so fast that the referee clearly did not put much effort into it. I'll better not ask him next time".

• Even if I've read the article before, is it better to "let it sink" for some more time and not rashly send the report to the editor? At what point should I feel confident that the report is ready?

Update: Thanks everyone for the good advice! I sent the report after around 10 days, and it was really appreciated by the editor!

Answer 1

Fast is great! Just be warned that it means editors will like you and send you more requests so you’ll have to learn to say no. If you also say “no” quickly and suggest alternatives, then you’ll still leave a good impression with the editors.

Answer 2

There's no drawback (from the editor's/author's point of view) to you submitting a quick review, but you should mention to the editor that you've already seen the paper before. It can be suspicious as a case of the author reviewing his/her own paper:

The reviews themselves were not remarkable: mostly favourable, with some suggestions about how to improve the papers. What was unusual was how quickly they were completed — often within 24 hours. The turnaround was a little too fast, and Claudiu Supuran, the journal's editor-in-chief, started to become suspicious.

See the source for more.

Answer 3

Statistical outliers always attract attention.

a) Your speed will be judged against the content and professionalism of your report: Were you speedy or just in a hurry to finish with the damn thing?

b) Even if your speed is not useful to the Editor (because, maybe, there is a second referee, so what will matter for the publication process is the lowest speed, not the fastest), still, if a) above is positively concluded, then your speed will be appreciated and remembered.

c) Always "let sink" anything you write. For how much it depends, but my experience says that if I have a week, I give it a week, so that I am distanced not only from the content but also from syntax and overall structure.

Answer 4

Once your report is detailed and accurate, with substantive and balanced comments, submit it, irrespective of how quick the turnaround is.

Answer 5

A rapid "accept" recommendation with nothing else might seem unusual (but not always, e.g. if the paper is very short, or the work is in your particular area of expertise). But a week is not rapid, if you happen to immediately have spare time.

Regardless of anything else, I advise giving concrete reasons for your recommendation. I suggest including a short summary of the paper, so the authors either know that you understand what's going on, or the authors can identify your misunderstandings.

I've sent off some extremely quick recommendations at times, particularly when it's my area of expertise:

• I think my fastest was in a matter of hours, when I could immediately see that there was simply no hope of it being published in the journal they sent it to. It wasn't even in the same ballpark.

• I've sent off a quick "revise" recommendation when the authors were unaware of a fundamental paper in the area, which would require rewriting a fair chunk of their work.

• Sometimes I send a quick recommendation explaining "this is all I have time for right now", and include as much as I can in a short time. If the editor doesn't like my response, they can enlist someone else.

Answer 6

The waiting time for a review comprises of:

• The reviewer taking care of more urgent/important tasks in their job as a researcher.
• The reviewer doing the actual review.

Since almost everything takes precedence over reviewing, the first part usually takes up most of the time. On the other hand, it is highly variable by nature and it may as well be non existing. For example, if the review request comes at a time where the reviewer is just in a gap between two projects, needs a break from something, or waits for their self-running experiment to finish, it is conceivable that they attend to the review the moment they receive it.

As the first component does not impact the quality of your review, the duration of the review is not useful proxy for its quality. Editors should be aware of this. As an author and reviewer, I have seen sloppy reviews that took excessively long.

Thus the only way a review report can be too fast is if it is shorter than the time needed to diligently perform the actual review. What this is depends on the field and the length of the paper: In my field, an initial review is about half a day’s work, but in fields that require extensive proof-checking, it can be in the order of magnitude of a week. In the latter case, handing in the review after a day would indeed be “too short” in the sense that it will probably arouse unwarranted suspicion. But even then, you can simply state that you were intimately familiar with the preprint, which actually speaks for you being chosen as a reviewer.

While the quality of your review may improve by revisiting it with some temporal distance (as suggested by another answer), I find at least for myself that the quality increase of the review is not worth it. I rather save that time and energy to diligently review the revision (or other papers). On the other hand, for your first few reviews, this is probably a good idea.

Finally, for whatever it’s worth I reviewed a paper within a day several times, and so far I have not received any complaints. Most journals solicited reviews again after this to the extent that I feel that I became a sort of emergency reviewer who is chosen when the originally chosen reviewers failed to deliver and the journal did not want to keep the authors waiting any longer.