So the mistake is not a major bug, the contributions of the paper remain the same, but still, I feel it was a silly statistical error. Can I correct it, even if the reviewers missed catching it?
Should I correct a mistake for the camera-ready version even though the reviewers missed that mistake?
That depends on the exact nature of the mistake. When it's "not a major bug" what is it, exactly? In my view, every word below depends on that exposition. I see "bugs", here, as obvious typographical errors. Do you see this instance as something else?– Robbie GoodwinMay 10, 2022 at 19:14
Yes, you should if you can. Presumably the reviewers are not the last people who will be reading the text.
27The sooner corrections are made the better.– BuffyApr 24, 2022 at 21:26
While the answer by markvs already tells you the right course of action, I think the following addendum is important:
The phrase "even if the reviewers missed catching it" indicates that you might have an inaccurate impression of the peer review process and should maybe change your perspective on it:
As an important part of scientific quality ensurance, peer review serves several purposes. One purpose is to decrease the probability that a published paper contains errors. It is not a guarantee that a published paper does not contain any errors.
It is not among the purposes of peer review to partially relieve the authors of their responsibility for the correctness of their results. If your readers find a mistake in your paper, they will blame you for it, not the reviewers.
Given these two points, it is clear that you should take all necessary measures to make your paper as correct as you possibly can, also if the reviewers overlooked a mistake.
Important note. If you make non-trivial changes after the completion of the peer review process, I strongly recommend to inform the handling editor of these changes - the editor then has the opportunity to approve (or, sometimes maybe, disapprove) the changes; and if they think that the changes are also in need of peer review, they then also have the opportunity to send the revised file to the reviewers once again. (EDIT: As pointed out by Danica in a comment, in some fields and for some publication venues this last paragraph does not apply.)
21This! You should be clear that if there is a mistake in the paper, that is your mistake. It's your paper, your mistakes, everything in the paper is yours. Everything. Everyone else just "helps" you and may set a minimum quality standard to allow you publication, but it's up to you - and attributed to you - to publish great or bad stuff. One point to maybe add: make sure to record the change in a changelog, mention it somewhere.– Mayou36Apr 25, 2022 at 12:20
10Or perhaps the impression was that it might not be OK to change a reviewed paper, because that would mean publishing something that has not fully been reviewed (pedantic, but technically true). If so, the question is asking whether changing it is "allowed" despite this. (yes, and encouraged even)– tjallingApr 25, 2022 at 15:21
1@tjalling: Yes, that might also be an intended meaning of the question. The last paragraph that I now added hopefully clarifies for the OP what they can do to ensure that the changes are allowed. Apr 25, 2022 at 18:28
3For conference publications, at least in CS, it is usually the case that there is not a way to communicate with the "handling editor" after acceptance (generally called an "area chair" or "meta-reviewer" in my area), and there is absolutely no expectation that one do so. In the case of extremely-major changes, you should perhaps communicate with the program chairs. For relatively small changes (even on the scale of, say, swapping out a nontrivial section of a proof), this is not needed, and you should just do your very best to ensure that your correction is itself correct.– DanicaApr 26, 2022 at 1:14
2The last paragraph is important: Don't dismiss the possibility of your "fix" being erroneous itself. Maybe your fix makes the paper worse, so you should definitely inform and make sure it is approved.– FalcoApr 26, 2022 at 15:27
It is your responsibility to ensure that the text is correct.
It is not the responsibility of the reviewers. The reviewers only act as gatekeepers to offer an expert opinion and state that the paper looks to be correct and interesting. They will not check every detail (necessarily) and will likely not see small mistakes and errors.
The Question isn't really clear but if, from another perspective, you're Asking whether the fact that there are reviewers absolves you of responsibility for accuracy then no; of course it never could.
When I worked in publishing everyone, at every level, knew that if anyone - specifically including the telephone sanitisers - questioned anything, the correct procedure was to amend the wording of that thing.
Of course right there and then, I could give the sanitiser an explanation, and so what?
What explanation could I give to 50-100,000 readers, after publication?
1Sorry but this does not answer the question from an academic perspective, and it clearly misrepresents the role of the reviewers.– Massimo Ortolano ♦Apr 27, 2022 at 18:41
@Massimo. Sorry but if that's your view then you have no idea either what publishing involves, or what the role of reviewers is. If you don't follow that, look above mine at Tom's Answer. Apr 27, 2022 at 19:04
2I think your first paragraph technically offers an answer to the question, even if it adds nothing to Tom's answer (so, I will decline the "not an answer" flags). That said: I found the remaining three paragraphs very hard to follow. Moreover: I think there may be some confusion about the distinction between academic reviewers and other types of publishing (e.g., for undergraduate textbooks) -- proofreaders in the latter may indeed spend months polishing every word, but reviewers for the former do not tend to do much polishing at all.– cag51 ♦Apr 27, 2022 at 22:19
@cag51 The Question still isn't clear and if you looked at it in other ways, could you still see no difference between conceptual problems and simple factual errors? Don't you see that the mention of "camera-ready" moves the Question outside the range of review, or any responsibility of reviewers, and despite your experience, should show it it to be about publishing in general, not in any way specific to academia? Try it, why not? Apr 30, 2022 at 21:54
2Sorry, I don't understand. This is the academia site, so all questions (and answers) are assumed to be about academia; otherwise, they would be off-topic (for questions) or non-responsive (for answers). Further, if one feels that the question is not clear, we should seek clarification in the comments rather than posting answers (in any case, this question mentions "papers" not "books"). As for the substantive issues: I'm not sure that we're understanding one another, but that's okay, we don't need to agree. Cheers!– cag51 ♦May 1, 2022 at 3:07