I am presently writing an article about a certain class of discrete probability distributions, and I want to mention that the distribution is relatively obscure compared to other common distributions. As an indicator of its relative obscurity, I would like to mention the fact that this class of distributions does not have its own Wikipedia page, whereas other common distributions (e.g., the binomial, Poisson, hypergeometric, etc.) do have their own pages.

My question: What is the proper way to cite this evidence? Presumably I will be citing the fact that I have performed a search of Wikipedia at a certain time, and that I found pages for various common discrete distributions, but no page for the one that is the subject of my paper.

What I am NOT asking: I am well aware of the reasons that academic papers do not usually cite Wikipedia as a source. In this instance I only intend to cite it as an indicator of the fact that a topic is sufficiently obscure that no-one has created a page for it. As such, I am not seeking answers on the general objections to citations to Wikipedia (which I am already familiar with).

  • 4
    Have you considered creating the WP page, then referencing it saying "until this work, (...) did not have its own WP page" and then the edit history provides the timestamp to prove it – Alexis Dec 7 at 8:18
  • 1
    academic papers do not usually cite Wikipedia as a source [citation needed]. By the way it’s fine to cite Wikipedia when there is a good reason to do so, and sometimes there is (I cited a Wikipedia article in one of my papers). – Dan Romik Dec 7 at 8:20
  • 1
    Would it not be more interesting to know, how well-known this particular thing is in the professional community and not under general public? Something like the number of Google Scholar hits for your particular distribution and, say, Poisson, might be more helpful. – Oleg Lobachev Dec 7 at 13:37
  • 1
    Alexis' suggestion is probably the most reliable way to back such a statement, but you could also link to a permalink of one of the pages that list missing articles. These may be occasionally outdated as new articles get created, but they're a strong indication that articles listed there don't exist at the time. After a brief search, the most appropriate such page seems to be en.wikipedia.org/wiki/Wikipedia:Requested_articles/…. – waldyrious Dec 7 at 17:18
  • 1
    Instead of citing inexistence of article in Wikipedia (or in addition to it) I'd suggest comparing the number of questions about that distribution in stats.stackexchange versus the number of questions about other distributions. Similar comparisons have often been done in StackOverflow as a measure of popularity of several programming languages. – Pere Dec 8 at 19:27

I don't think one needs a citation for the fact that no citation exists. If you're identifying a gap in the regular literature, you can say something like 'to the authors' knowledge, no study of this kind has been published' or words to that effect and you don't need a citation to confirm that you are identifying a gap in the literature.

I find it highly unlikely that you would be criticised for wording along the lines of 'at the time of writing (possibly give date), no article exists on this topic at Wikipedia' and then maybe give a general citation for Wikipedia as a whole. As Wikipedia time stamps every edit to every article, this is something that could very easily be verified at any time in the future. I'm not sure having a Wikipedia page or not is the best measure of obscurity, but that's a different question entirely.

  • That is a good point about the time-stamp. Still, just to be safe, I'd like to cite a search. It is not just "to my knowledge" - I actually want to do a search on Wikipedia, and then cite the fact that the search did not yield a page for the topic of interest. – Ben Dec 7 at 6:06
  • 3
    Yeah, the 'to my knowledge' wording was in reference to referring to a gap in the wider literature, not the specific Wikipedia article. I'm not sure a citation is really what you're after here though, you can detail a search strategy without a citation, as that is something that YOU have done. Citations are for citing existing work, which is not what you are doing here. I would specify the name of the page you searched for and the time you searched for it. If you still really want a citation, then cite this as you would any other webpage: en.wikipedia.org/wiki/Mystery_distribution – E. Rei Dec 7 at 6:17
  • Also, stating that the WP entry does not exist can be easily verified by the reader. – Alexis Dec 7 at 8:15
  • @Alexis Only until it does. That said, I suppose you can then check the page history. – user2768 Dec 7 at 13:26
  • 1
    I agree this is not useful but if you are really set on it you can make an archive.org or similar snapshot of the page not existing, or of the search results. – Tgr Dec 8 at 21:01

You could also use a completely different, arguably more objective, metric for 'obscurity', such as published articles on your topic per year. This can easily be done in many indexing databases. E.g. in Scopus (biological or life sciences) the 'obscurity' of snow algae research could be demonstrated by comparing the number of documents per year to brown algae research.

Search string 'snow algae': snow algae documents

Search string 'brown algae': brown algae documents

What is the proper way to cite this evidence? Presumably I will be citing the fact that I have performed a search of Wikipedia at a certain time, and that I found pages for various common discrete distributions, but no page for the one that is the subject of my paper.

You can prove that a particular page didn't exist, for instance, by accessing a particular page (e.g., obscure_distributions) and recording raw result data (which will be authenticated). See my answer regarding a similar question for further details. That said, this is probably overkill, and you should probably be doing something completely different anyhow.

Cui bono?

As it was mentioned in the comments, the question originally looks for a way to document something is lesser known to general public. Would it not be more interesting to know, how well-known this particular thing is in the professional community? A measure for this is a number of citations, for example.

So, stating something like the number of Google Scholar hits for your particular distribution and, say, Poisson, might be more helpful.

There are quite some papers on the meta- level, so it would be a somewhat accepted aspect of a research paper.

Also, obligatory xkcd:

too much meta xkcd comic

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

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