7

It is famously known that “visibility” of an article influences its citation count. Is it true that articles cited in Wikipedia page get more academic citations?

5
  • 5
    Generally they are on Wiki because they are well cited for good reasons already... – Jon Custer Nov 13 '18 at 4:33
  • 2
    @Jon Custer It's not always the case, I have seen very recently published articles cited in Wikipedia page. – MBK Nov 13 '18 at 4:38
  • 1
    Anecdotally, someone cited one of my papers on Wikipedia four years ago and in that time it hasn't gotten a single citation. (Not that I particularly expected it to.) – Misha Lavrov Nov 14 '18 at 1:08
  • 1
    @TommiBrander I am interested in academic citations as a result of Wiki visibility. – MBK Nov 14 '18 at 3:12
  • @MBK I edited to clarify. – Tommi Nov 14 '18 at 8:28
7

There appears to be some literature comparing citations on Wikipedia to citations in academic journals, but I have not been able to find any randomised controlled trial (RCT) on the matter. For example, in a comparison of Wikipedia citations to journal metrics, Nielsen (2018) found that "[t]he Wikipedia citation numbers showed high correlation with the [Journal Citation Report's] numbers for the total number of citations to a journal" (p. 4). If go broader than Wikipedia, Lawrence (2011) finds that free online availability is positively correlated with citation rate. Both of these studies use observational data and show correlation between the outcomes, but they do not rule out other causal relations that would lead to correlation (e.g., if papers are more likely to be made available online, or added as citations in Wikipedia, if they become more cited).

If there are any enterprising researchers reading this post, this sounds like a good research question for a randomised controlled trial (RCT). Since Wikipedia can be edited with a free account, it should be possible to set up an RCT where you add a randomised set of presently uncited academic articles to Wikipedia and track their citation progress in academic journals against a control group. First find a set of articles relevant to a subject that are not presently cited on Wikpedia; then randomise them into a control group and treatment group, and add the articles in the treatment group into Wikipedia articles; then watch changes in the citation count in the academic journals over time, and then test for a difference in the groups.

8
  • 7
    I suspect Wikipedia might object to being used as a guinea pig. – JeffE Nov 13 '18 at 12:51
  • 1
    Well, I suppose this is what ethics committees are for. Anyway, I'm not sure they'd necessarily object, given that the research might potentially show the value of their platform to academic researchers. – Ben Nov 13 '18 at 22:55
  • 1
    That doesn't sound like a very good research project. There are a lot of confounds, and it is probably against wiki's ToS. – Trusly Nov 14 '18 at 0:57
  • 1
    Confounders are not generally a problem for an RCT, so long as it is properly constructed. Indeed, that is the primary benefit of doing this as an RCT instead of an observational study. Whether such an experiment would breach ToS/ethics is a separate issue from its merits as a scientific experiment, but is a matter that would need to be considered in an appropriate ethics application. – Ben Nov 14 '18 at 1:00
  • 2
    If all the articles being added are legitimate articles that reasonably should be cited in the Wikipedia articles in question, I don't think most Wikipedians will mind. It might make sense to check with first with someone there, maybe on the Admin Noticeboard before running the experiment, but my guess is the response will not be negative. – JoshuaZ Nov 14 '18 at 3:10
0

Yes.

I once edited an article into Wikipedia. I needed a source for a generic claim (something reasonably well-known to experts in the field, such as "the question of whether P = NP is a major unsolved problem"), and I just cited something I was familiar with. It wasn't even a good source for the claim - the paper didn't aim to prove the claim, it just contained one line that confirmed that the claim is true. But since I was familiar with the paper I didn't bother looking for something else.

A few months later, it turned out that the source, along with others in the Wikipedia article, had been cited by others for the same claim.

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.