25 January 2016

Journal circulation and citations

Stuart Cantrill, writing at The Chemical Connection blog, had this plot showing the citations of one article that was published in multiple venues:

The point is that citations are often used as a measure of the quality or impact of a paper. But since this is the identical article, “article quality” cannot explain the variation in citations. Stuart hypothesized:

(P)erhaps more people read the New England Journal of Medicine than the Medical Journal of Australia and so a wider audience will likely mean a wider potential-citation pool.

I left a comment that this should be testable, because magazines (American ones, anyway) were required to disclose their print circulation annually. Well, one thing led to another (“another” in this case meaning, “a bunch of Google searches for ‘[Journal name] print circulation’”). I couldn’t find the circulation figures for the Croatian Medical Journal. Here are the results:

I’m surprised. For a while, I’ve been thinking about what relationship there is between a scientific journal’s readership and its Impact Factor. It seemed to me that Impact Factor might just be a proxy measurement for readership. So, I genuinely expected the correlation would be similarly tight as the one Stuart had for Impact Factor.

Of course, these circulation figures are mostly based on printed copies distributed. This confuses the issue for many reasons. First, a print issue can be read by many people. Nature estimated about eight people read one copy of their journal. Plus, print is drying up as a medium for magazines and journals. It may well be that website visits or some other measure is a more important measure of a medical journal’s “readership.”

Update, 31 March 2016: Daniel Shanahan has a new paper in PeerJ that shows the same thing as Cantrill’s blog post. When you publish the same article in multiple venues, the Impact Factor predicts how many citations the article gets, which suggests that Impact Factor mostly predicts Impact Factor, not how scientifically interesting the articles within are.

External links

Imperfect impact

Appendix 1: Journal and circulation (click journal name for source of circulation)

New England Journal of Medicine: 120,000
The Lancet: 29,103
Journal of the American Medical Association (JAMA): 292,902
British Medical Journal: 121,762
Canadian Medical Association Journal: 81,083
Medical Journal of Australia: 30,706
Annals of Internal Medicine: 100,014

No comments: