31

I am asking this question simply out of curiosity. I have reviewed around 4-5 papers of double-blind conferences. One pattern that I have found in common is that when authors use techniques/results from their previously published works, they do not properly cite that particular section/theorem since they are the same authors and they do not take this task very seriously.

Due to this, it becomes easy to figure out the identities of the authors.

Therefore, I want to know the other ways in which a reviewer can figure out the identity of the authors? Then, perhaps, we as authors can avoid making these mistakes when submitting our papers.

9
  • 6
    When the paper builds on existing work, it is very hard to stay anonymous. One example I tried is to remove some of the relevant references and leave a remark like "removed for peer-review". These references are then only sent to the editor and are reinstated after the manuscript is accepted. I have no clue if this worked or not since I never managed to figure out the identity of the reviewers, though.
    – Aolon
    Jul 22 at 5:44
  • 7
    Authors may do that intentionally, especially if the paper has a big name author. Jul 22 at 8:57
  • 28
    Opposite thing. We were so good at hiding our identity in one of our papers that it backfired. We cited our old papers as if they were 3rd party ones. Reviewer claimed that it was just a rehash of previous work of Captain Emacs (it wasn't, there were substantially new results). It would have been funny if it weren't that sad. Had they realised (or we allowed to indicate) it was a continuation, they may have found it easier to identify the innovation. Jul 22 at 12:05
  • 2
    It's not just possible to figure out the authors. Often, it's just as easy for the authors to figure out who the reviewers are, simply by noticing "where they are coming from" in evaluating the paper.
    – alephzero
    Jul 22 at 17:21
  • 1
    I once saw a sole author chemistry paper with 53 references: the CRC Handbook plus 52 papers by the sole author. Hard to hide that one.
    – Ed V
    Jul 24 at 15:26
46

I work in a small experimental field where the groups around the world know each other very well . There's no double-blind peer-review, but things that make papers from my colleagues easily recognizable are the following (I think I'd recognize with high probability the authors of virtually any work in my field):

  1. The presence of certain specific pieces of equipment. Certain pieces of equipment are owned by only 2-3 research groups around the world, and finding it in the description of an experiment clearly delimits the possibilities.
  2. The description of a long-term experiment Certain experiments have been developed and described in several papers along ten-twenty years and are quite unique of certain groups. A paper on the topic clearly identifies the group.
  3. The manufacturing of certain specific devices. There are devices for which only one or two groups possess the manufacturing capabilities.
  4. The style of symbols in schematic diagrams. Each group typically develops along the years certain personalised symbols.
  5. The choice of symbols for certain quantities. For instance the usage of U to denote a voltage is typical in Germany, Czech Republic and some other countries in Eastern Europe. If there are just a few groups in that area, this can be a revealing detail.
  6. The presence of certain idiosyncratic expressions.
2
  • 19
    Maybe a little stronger on point 6. Someone who writes a lot can often be recognized by their style.
    – Buffy
    Jul 22 at 13:48
  • 6
    @Buffy Yes, I wanted to expand but didn't have time. Will do.
    – Massimo Ortolano
    Jul 22 at 14:14
22

"This paper is based on the dissertation of Noche (2013)."

1
  • 2
    Even if it is changed to "This paper is based on the dissertation of Author (2013)," if the dissertation is listed in the bibliography, then the title and the university it is from can be used to search for the author's name. Jul 22 at 5:57
14

In some cases the subject matter makes it almost impossible to anonymise the work. For instance, a couple of years back I was asked to review a paper about a proposed improvement to the methods used for estimating the national accounts of a certain country.

That work is done by a specific national agency (the only agency that has access to all the relevant data) so it was obvious where the authors worked. Looking up recent publications connected to that agency would probably have allowed me to make a good guess as to who in particular was working on this topic.

Unfortunately I'm not aware of a good way to avoid this particular issue.

9

Because of the rules of my PhD program, I have to state sentences akin to the following in the acknowledgements: „This paper is part of pbaer‘s dissertation project at the University of Academiaville.“

I dont think one could make it any easier for a reviewer to identify me as the author during a double-blind review.

4
  • 15
    OTOH, that kind of thing can be removed by an editor before passing the paper along for review. Jul 22 at 7:27
  • 18
    And some venues explicitly mention that acknowledgments must be added only when submitting the final camera-ready version.
    – GoodDeeds
    Jul 22 at 10:57
  • 2
    Shouldn't such a statement be removed from the manuscript prior to review? It can be added back in later if the manuscript is accepted.
    – Tom
    Jul 22 at 20:52
  • 2
    @GoodDeeds I have not seen any mentions akin to what you describe in my field yet. As far as I am concerned, the journals I submitted to asked to send in the paper with the acknowledgements already inserted. I should take a closer look at this issue.
    – pbaer
    Jul 22 at 21:15
9

Probably the biggest and hardest to mask is writing style. A given author will structure their writing in certain ways, and use certain turns of phrase and grammatical quirks; someone who's read a number of that author's works can identify the author just from those cues.

As an example: my writing tends to make greater use of colons, semicolons, and qualifiers than average, and there are probably some other indications that I'm not aware of. Here on StackExchange, there are other users I can identify just from patterns of grammatical errors, or the use of large quote blocks, or how the introductory sentence of an answer is structured.

5

They do so by citing themselves unnecessarily. They do so by inappropriately ranking a reference to their own previous work ahead of more relevant references.

4

I actually just refereed an “anonymous” submission. After a simple copy-paste of the title in GoogleScholar, the search immediately produced the arXiv version with full author details.

8
  • Haha. Here authors want to reveal themselves intentionally.
    – IY2
    Jul 23 at 5:40
  • 7
    You actively trying to find out who the authors are is hardly the authors' fault.
    – Maeher
    Jul 23 at 7:32
  • 8
    Conferences like NeurIPS explicitly prohibit searching for working papers: "do not attempt to find out the identities of the authors for any of your assigned submissions (e.g., by searching on arXiv). This would constitute an active violation of the double-blind reviewing policy."
    – Xi'an
    Jul 23 at 11:45
  • 1
    @Meaher I have in the past asked a trusted colleague to do exactly this search in my place, but only because I strongly suspected who the authors were and had a conflict of interest with them (my colleague would only confirm or deny my suspicion here). I feared accidentally unblinding the authors halfway the reviewing process, when googling some more info on the papers topic. The authors were who I suspected and I denied to review. They unblinded themselves by use of a certain symbol to indicate a part of the paper was moved to the appendix.
    – user53923
    Jul 23 at 11:53
  • 1
    @ZeroTheHero No, they are not. Conferences like NeurIPS and ICML are very selective and it's normal for a work to be resubmitted in different years without any major changes. My paper spent two years on arXiv before being accepted. Jul 24 at 8:55
2

In my area of research double blind reviews are used to minimize unconscious bias. The earlier a reviewer knows the authors, the more she may be biased.

The goal of anonymity should thus only be to remove the obvious hints. Hide author names. Instead of "our previous paper [5]" write "the paper [5]", etc. not making major changes. If you think the authors could improve in this regard, you could mention it as 'minor comment' similar to how you would mention typos.

In my area anonymity is a fickle thing. In particular, for papers the following make it night impossible:

1) Experts: Commonly there are only few people who are likely to publish on a given topic and if I am asked to review a paper I am an expert and probably know them.

2) Follow up work: They build strongly on previous work and seem to know all the ins and outs. Moreover, structure, writing style, idiosyncracies of this particular work are identical. Guess what.

3) Preprints: There is a good chance that you google the title and find the paper on arXiv.

4) Writing style: The most devious one because it pervades even reviews, which should be anonymous. Do they write British or American English? Do they make certain grammatical mistakes typical for non-natives of a certain background? Do they overuse CERTAIN words?

If you combine these, there is really no chance to anonymize papers in many areas. In fact, 1) and 4) suffice for an expert to make pretty good guesses about stuff that definitely should be anonymous, like reviews.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.