4

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).

Improve this question
7
  • 8
    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, 2018 at 8:18
  • 2
    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, 2018 at 8:20
  • 2
    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, 2018 at 13:37
  • 2
    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 These may be occasionally outdated as new articles get created, but they're a strong indication that articles listed there don't exist at the time.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, 2018 at 17:18
  • 2
    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, 2018 at 19:27

6 Answers 6

Reset to default
10

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.

Improve this answer
5
  • 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, 2018 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, 2018 at 6:17
  • Also, stating that the WP entry does not exist can be easily verified by the reader.
    – Alexis
    Dec 7, 2018 at 8:15
  • 1
    @Alexis Only until it does. That said, I suppose you can then check the page history.
    – user2768
    Dec 7, 2018 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, 2018 at 21:01
6

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

Improve this answer
1
2

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.

Improve this answer
1

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

Improve this answer
1

I tend to agree with E. Rei's answer, that "to my knowledge" should be sufficient, and is at least somewhat verifiable by consulting Wikipedia's edit history. However, considering your wish for something a bit more solid, two (independent) suggestions:

  1. Expand the scope of "my knowledge" to include "XYZ experts' knowledge." To do this, post a question online somewhere you feel confident that Wikipedia experts exist...here, on Quora, on Wikipedia, etc. Make a simple statement like, "In the the interest of research I am conducting, I'm seeking a definitive answer on whether English Wikipedia contains an article on XYZ topic. I have done a search myself, but just in case my search technique was inadequate, can others with an intimate understanding of Wikipedia confirm this, or direct me to the article?" Then, you can link the discussion as evidence.

  2. Do the search, take a screenshot of the (lack of) results, post it online, and then cite that.

Improve this answer
1

I think you could really just handle it in article text itself. ("I searched" or "a search was done".) I would maybe mention a few details (date of the search, English Wikipedia, what different versions of the name you searched on). I don't think you have to go crazy about the details if it is just a minor point.

You could put details down into a FOOTNOTE ("a" or "b", not a numbered ENDNOTE citation). This gives a little more backup for you, without having to derail the flow of the discussion. You could even consider to just put some [short] description into the endnote citation itself. I have seen people put things like "personal communication" or "Superbowl commercial 2005" or the like into endnotes. Obviously follow formatting strictly when citing normal sources, but if there is some rationale why you need to do something slightly differently, most editors/journals will be fine. The point after all is to communicate.

P.s. Practical advice: Do a Google search of Wiki rather than using the Wiki search engine (it is miserable, too literal). Also if there is more than one name for whatever it is, than search under all the names. Don't want to look stupid if there is a page on it but under a different name.

P.s.s. More practical advice: consider to do a search in Google science or Webofscience or the like to compare how often the funky distribution is returned versus more common ones (normal, etc.) It gives more of a comparison than just the yes/no has a Wiki page or not. You can also look at Google Trends for prevalence on the web (compare to other terms). I believe you can do the same thing even for book text. See below for an example in field I know:

http://2.bp.blogspot.com/-hC819n9EoFM/UfkIFielYMI/AAAAAAAAAlw/wgwD7GDqfqQ/s1600/screenshot_12.jpg

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .