I am reviewing a paper on a topic on which I have published some work previously as an author. While going through this current paper as a reviewer, I felt I could recommend my paper on the similar topic, which addresses a particular point better than in the paper I am reviewing. My question in this regard is whether in such conditions it's a common practice for a reviewer to recommend their own paper?
This is a delicate issue.
- On the one hand, a reviewer is most familiar with their own work, and will certainly know if their work is relevant to the paper that they are reviewing.
- On the other hand, it can feel like a crass "citation for publication" trade.
The way that I typically approach it is to only recommend citation of my own work if I am also finding it appropriate to recommend citation of other related work at the same time.
Put another way: if the hole in their literature review is "me," then I find it likely that I'm wishing they cited me because of personal bias and my own ego. If the hole in their literature review is "me and a whole bunch of other things like me," then I find it likely that it's a legitimate gap and it's OK to recommend my work as part of the set of things to fill said gap.
The metric I use is the same as for everything else I do when refereeing a manuscript: Would I make the suggestion to the authors face as a non-anonymous referee? How would the authors feel if they knew it came from me? Would they think I was trolling for citation, or would they think they missed an important paper for discussion?
Probably, if the manuscript tells a complete story and has a fair picture of the state of the literature without it, I probably would not bring up my own paper in a review. If my paper were seminal in the field, and the authors missed it, then I'd probably bring it up. Oddly enough, papers I tend to recommend during review are not my own, usually.
Is it a common practice for reviewers to recommend their own papers in the review?
While this is likely to vary by discipline, in my field/subfield (electrical engineering/circuits and signal processing), reviewers recommending that one of their papers be cited in the revised manuscript is fairly common.
Now, this does not mean that the authors are obligated to include the reference in their revised manuscript. But, as a reviewer, if you have authored a paper that is highly relevant to the paper under review, I think it's helpful to the field if these links to other relevant works are established.
Coincidentally, I have found out about other related works in closely-related fields (but not exactly the same as my "little corner of the world") by reviewers suggesting that I cite certain papers in my manuscripts (of which, some of the suggested papers are the reviewer's own papers [because, let's face it, sometimes you can tell]). And, of course, I will take a look at these suggested works and make sure they are a good fit for my paper before I comply with the request to cite them.
I find that the existing answers do not address the actual question. The question was not whether a reviewer should ask for his work to be cited but whether he should point out his work regarding the relevance to suggested fixes to the content of the reviewed paper.
The reviewer here appears to me more interested in changes to the text rather than the bibliography.
And that's where I'd actually be drawing my personal line: trying to get your work cited feels distasteful to me. Trying to get your work heeded where doing so would substantially improve the paper's relation to the status quo of research seems appropriate to me.
Now this heeding can be done by taking the results of that paper, removing the appropriate section in the paper itself and obviously putting in a bibliographical reference. It can be done by reworking that passage, possibly just throwing out material that seems no longer motivateable in light of your work even when the act of throwing the material out does not actually warrant a separate documentation and citation.
In contrast, if the work remains unchanged but the bibliography is amended, this amounts to a do-it-yourself science kit for the reader. That makes no sense. The bibliography is not supposed to be a field survey but rather contain references actually supporting the article. Dumping unrelated or contradicting references into it does not improve the article when the article does not actually work out its relation to the reference.
You are really asking several distinct but related questions.
- Is it common to ask people to cite your own work? Yes, it's reasonably common, it has happened to me and to others. On the other hand, I am not aware of any statistics on the practice and it's certainly not something everybody is doing.
- Is it appropriate/ethical to do it? It's delicate, others have already provided excellent answers on this point but do realise that it's extremely difficult to be perfectly honest with yourself in this context. We might be dead certain we are making a reasonable, neutral assessment of our work's relevance to the paper's topic but the psychological literature suggests we humans are usually not very good judges of that.
- How will it be perceived? You did not ask about this but it seems to me that it's likely to look bad. At the very least, follow Jake's and user35124's advice and only mention your own papers if you are doing more than merely asking for a citation. Even if you do it in good faith and you are right about your paper's relevance, only suggesting publications from one author will look odd.
- Should you do it? Anecdotal evidence suggests that some people in my field have been doing it for some time in a rather heavy-handed way. They probably suffered a bit of damage to their reputation with authors and possibly editors but it's definitely not something that's universally frowned upon and could get you in deep trouble like fraud or plagiarism. So it's up to you to assess the ethics and decide if you want to do it or not.
If you feel strongly that your paper is more than peripherally relevant, that you have addressed the issue in a particularly relevant way, why not make the suggestion to the editors rather than directly to the authors and let the editor assess it as a third party in a perhaps more objective way?