Don’t you want to read this? I’ll make it easy for Brad Plumer to cite: It’s about his great post

Brad Plumer shared some interesting research on the reach of research. One bit, at the end of his post, is about work I’ve posted on before. Here’s what’s new to me from Plumer:

new study in the journal Scientometrics finds that scientific papers whose titles were phrased as questions were more likely to be downloaded. […] Papers with colons in their titles, by contrast, receive fewer downloads and citations.

Question-mark papers tend to be downloaded more, but they’re cited less. And studies with longer titles are downloaded less but cited more. Perhaps long, laborious titles help cull casual thrill-seekers — only the truly interested and dedicated readers click. Also, papers with “highly amusing” titles get noticed more and downloaded more, but are cited far less — perhaps there’s a disappointment factor at work, or perhaps scientists who write less important papers feel more of a need to spruce things up with a sexy title.

Maybe, but there’s more. You’ve got to put yourself in the mind of someone trying to support his work by citing others. Bloggers do this too. Only bloggers get to write their own anchor text. A blogger can write, “This is true,” and link that statement to something. However, the title of that something could be “Is this true?” Good enough for blogging!

A scholar has less freedom. A scholar has to cite the paper and include it in the list of references with the actual title. If the scholar wants to claim “This is true” it’s way better if the title of the paper expresses that. Longer titles with colons probably make very specific statements that are useful for supporting an argument. The perfect tile might be “This is true: Here’s why.”

Of course, one need not cite a paper for its title. It’s perfectly fine to cite a paper in support of an argument because some clause of an end note has the supporting material. Good enough! But if it is right there in the title, what could be better?


On a related note, Goldacre also points to a 1991 study finding that medical research results that get mentioned in the New York Times were far more likely to receive citations by other academics. Is that just because the papers themselves were inherently more important? Not necessarily. The study’s authors managed to conduct a natural experiment: During a strike by Times printers, the journalists were still putting together an “edition of record” every day; it just wasn’t getting distributed to the public. The medical research written about during that three-week span didn’t receive an extra citation boost. The New York Times really can make a scientific study more influential. Academics seem to be swayed by more than just the content of a paper.

Well, academics need to both know about the paper and be swayed by the content. Citation in a circulating NYT issue helps, doesn’t it? There are a lot of papers academics don’t read because, well, there are a lot of papers. And now we have to read blogs?!?!?! 🙂

Hidden information below


Email Address*