The biologist named Ocorrafoo Cobange doesn't really exist. Nor does the Wassee Institute of Medicine where he supposedly works. But when Cobange started submitting a new scientific paper to open-access journals around the world, something strange happened. More than 150 of the 304 journals to which a version of the paper was submitted accepted it for publication.
Here's the kicker. The science in the paper — which purported to reveal a new cancer drug derived from a kind of lichen — was just as fake as Cobange's name and institution. It contained obvious errors and flawed experiments. Very few of the journals noticed any of the flaws during the normal peer-review process, if peer review was conducted at all.
In reality, Cobange was one of many aliases used by journalist John Bohannon whose sting operation, published in the Oct. 4 issue of the journal Science, reveals that many less than scrupulous journals have popped up around the world and may be publishing less-than-acceptable science — and charging researchers for the pleasure of publication.
Open-access publishing is a fairly new thing in academic publishing. Traditionally, peer-reviewed journals were available only to researchers via a pricey subscription model. No, these aren't the $10 annual subscriptions to Wired magazine. Many institutions pay thousands or even millions of dollars a year for online subscriptions to traditional academic journals. If you don't pay to subscribe to the journal, or work for an institution that does, you often can't access the material in these publications.
The open-access model is different. These online publications make their material freely available to anyone who wants to read it. Instead of charging subscription fees, many of these journals charge authors whose papers are accepted for publication.
Bohannon's sting reveals that the open-access movement has created a lower-level tier of less-than-rigorous scientific publications made economically viable by charging publication fees. It also revealed that many of the journals conducted at best cursory peer reviews, if they conducted them at all. Most of these journals were located in India, Pakistan and Turkey. A few were U.S.-based. Some of the journals were from smaller companies, but even journals published by large companies like Elsevier and Wolters Kluwer accepted the paper. (Full disclosure: I used to work for a division of Elsevier.)
None of Bohannon's fake paper was ever actually published — he withdrew it before it got that far — but it was accepted for publication a total of 157 times. Sometimes it seemed as if the journal editor had "rubber-stamped" the paper to accept it. In 16 cases the editor ignored complaints from peer reviews and moved the paper toward publication.
Bohannon ends his report on the sting operation with an important coda: Open-access publishing is an important model that, more often than not, works. But when any journal fails in its peer-review process, it damages all of science.
Related stories on MNN: