Wednesday, May 20, 2009

Faked journals -- Tip of the iceberg?

Through some investigative reporting, a magazine covering the scientific community has provoked a publisher of scientific journals into admitting it created six "fake" journals between 2000 and 2005 that were essentially secret ads for pharmaceutical companies.

One fake journal reprinted real articles from other sources, virtually all of them favorable to the drug manufacturing giant, Merck, according to magazine The Scientist, which broke the story.

Merck appears to have secretly sponsored the creation of that fake journal, The Australasian Journal of Bone and Joint Medicine, which not only did not have the same controls and editorial oversight as real journals, but also did not in any way disclose its connection to the manufacturer. The superficial impression was that it was a real, independent journal.

The publisher of the fake journals, Elsevier, publishes many other real journals, and is a well-known firm. Many of them might carry ads from manufacturers. But in those legitimate cases, the ads themselves reveal any possible financial entanglements, and the editors of the journal are generally assumed to be independent, relying on peer review rather than corporate input to determine which articles to publish. The six fake journals, on the other hand, appear to have been secretly sponsored by companies that determined their contents, even if some of the articles were real and had been previously published by real journals elsewhere.

Elsevier does not appear to have publicly identified which companies sponsored the other five journals, and the only reason we know about Merck's involvement in the sixth is that The Scientist (the magazine that broke the story) had already identified that relationship in an earlier news story. Elsevier has, however, said that all of the fake journals were published out of its Australian office during a period of five years by people no longer with the company, and the company maintains that this sort of practice does not represent its usual priorities.

Nevertheless, reader comments appearing below the article in The Scientist reveal that the discovery has touched off or renewed some serious concerns: Have we in fact identified all of the fake journals? Many small journals exist for short periods of time. Others held by Elsevier -- or even by other companies -- might be similar to the Australasian Journal of Bone and Joint Medicine. And what of the "legitimate" journals? How neutral are they, really?

Writes one reader, TS Raman:
For all practical purposes, a journal that merely looks like it is a "peer-reviewed" is not different from one that has real peer review, but of a very poor quality. I say this because there are scores of journals published by professional or scientific societies, and in-house journals published by institutions, laboratories, etc., all of which are virtually "captive journals". There is a plethora of such journals in India. They are all nominally peer-reviewed, but the review is often a complete farce. The reviewers may not have expertise in the field of study to which the paper pertains: for instance a person whose [nominal] speciality is nuclear physics, may review a paper on biochemistry.
Both are legitimate concerns, but they are neither new nor unique to the scientific community: Readers have always, and will always, need to be able to discern the difference between suspect and trustworthy material. It can be done.

In one 1SC class I am teaching, a team of students gave a presentation on a journal dealing with climate change. A student in the audience asked whether the journal gave much attention to skeptical arguments (that is, arguments that maybe global warming is a natural phenomenon, not the result of pollution). Even as the student asked his question, it was clear he had reached the same conclusion I had, just from listening to the presentation: No, the journal was strictly in the believer camp. If a scientist or lab had data undermining the consensus on global warming, that would not be a good journal to submit to. The journal was real, in the sense that it had real scientists as editors and authors, and used peer review, but it was pretty easy to tell what its biases were, just from a description of its history, its mission, and a list of articles it had recently published.

This is often the case.

We can all tell when we've flipped the channel to an infomercial, too, without necessarily being able to spell out exactly how we know that's what we're seeing.

Similarly, when one is looking at a Web site or a "news" publication or a white paper that is really advertorial (that is, composed of articles that look real but are basically ads for a product), it seldom takes investigative reporting to smell out the advertising. If you read very often, you start to notice little things like ...
  • The presence of trademark symbols (like the R in a circle, or the letters TM superscripted), which only the original company ever includes. No, Merck probably didn't include these in the faked journal, but you'll see this all the time in press releases and "white papers" that are really marketing gimmicks. Why? It all boils down to Kleenex. I'll explain: When we say we are Xeroxing something, or need a box of Kleenex, or rode a Jet Ski, or got into a Jacuzzi, or drank a Coke, we might not in fact be using devices made by those companies. There are personal watercraft that Jet Ski didn't make. There are photocopiers not made by Xerox. There are hot tubs not made by Jacuzzi. All of those are brand names that have become "generic." Having a name become generic seems like a good thing, but there's a nasty consequence: When you go to court later to try to get someone to stop using your name to sell their product, the court considers whether you have been consistently enforcing your trademark. If you have made it clear that the term can only be used by you, then you can win that case. If you have allowed your term to become generic, then you no longer really "own" the terms, and you might lose. For this reason, legal departments of big companies spend a lot of their time putting trademark symbols (R and TM) next to their names on all of their materials, and sending nasty letters to anyone who uses the terms generically. All of this means that when the marketing department cooks up a really convincing fake document that says XYZ Corporation's new BackShaver is amazing, the legal guys are going to fight to put a little TM next to that word, and when they win that fight, they tell the rest of us that the article was written by the guys who make BackShaver. No one else cares about putting the TM there, and no one else is required to do it.
  • The constant repetition of the company's name or the product name, often starting in the first sentence. Marketing people learn that repetition builds name recognition, and that name recognition builds sales. So they make sure they use those names far more often than a person would in natural conversation or real writing. I bet the Merck journal did do this. If it made any changes to the articles or titles of articles, in fact, my bet would be that they inserted company and product names.
  • The absence of snide, sass, and complexity. Real, professional writers are always trying to sell you on their objectivity; marketing departments are always trying to sell you on their product. These lead to important differences in the copy. For instance, real writers often take little shots at everything and everyone, just to let you know that they're independent. These shots might be as simple as little disclaimers or qualifiers ("The new BackShaver is great at removing back hair, but I really hated it when I had to clean the device") or harmless little jabs at the Marketing Deparment itself ("I wish, though, that they'd given the BackShaver a better name -- who wants to stand in line at a store with a box that says 'BackShaver' on it?") Marketing Departments generally cannot stand these sorts of comments. If it never teases, and never has a complicated opinion involving some sort of negative, it's probably an ad in disguise.
  • Dwelling on details of success that neutral writers care little about. Marketing people know that customers will buy things they think are popular, so they rarely can resist pumping even their faux materials full of user statistics ("90 million copies sold") and testimonials ("'I use it all the time,' says Suzy, a freshman at UCLA"). They also try to build up connections between the current product and previously successful products made by the same company, leading to weird paragraphs in which they say that the 2007 version was great, but the 2009 version is better. A real writer might also talk about popularity, but doesn't necessarily see popularity as a good thing (and often throws in a plug for an underdog -- the software reviewer will talk about Microsoft's popularity, but mention he uses Linux himself, for instance). When describing a new edition or upgrade that improves on the 2007 version, he'll say the obvious: that it fixes problems that the 2007 version had. Marketing people won't ever describe it that way: As far as they're concerned, the old 2007 version was also perfect, just not as super perfect as the 2009 version.
  • Presence of contact information. If it has information for how to contact someone in the company, it's probably an ad, even if it doesn't look like one. (Exception: If it says "Call Bob Smith in the legal department at 111-555-1213 to complain about this critical safety issue!" then it's obviously not an ad.)
A few closing thoughts on this issue:
  1. Don't assume that because Merck and Elsevier's Australian branch did this, that they're "getting away with it" -- or that this means you should be unethical too, in order to get ahead. Generally, the corrupt people you hear about who seem to be getting away with things aren't. You hear about what they did, but you don't see the negative consequences. (See my earlier post on this subject.)
  2. A very enterprising student could do some cool detective work on this sort of thing and then report the results in a paper: Through textual analysis, study a handful of other small, short-run journals published by Elsevier with the word "Australasian" in the title (quite a few exist -- see the reader comments for the article I linked above), and look for signs of shadiness. Have the articles previously appeared elsewhere? (Web of Science will tell you.) Does one company's name or product name keep coming up, even though that company doesn't appear as a sponsor of the journal? Or what about the other five journals? We know their names (see the article), but not who sponsored them. Can you figure out which companies were involved in them by looking at the articles they published? (The enterprising student wouldn't have to cover all of this -- even covering one journal could be interesting.) Why is this a worthwhile project for a student? Among other things, discovering a fake journal while you're an undergraduate would probably lead to real publication for you (or, at least, some news stories about your discovery), and give you something to put on your CV.
  3. Be on the lookout for fakery. The Internet has made this easy, because Web sites are cheap. Faking an actual book or physical journal isn't usually cost-effective, due to printing costs. But Web sites are easy to fake. (In fact, if you plan to do what I describe in #2 above, you might narrow your search to journals that don't have print versions.)

No comments:

Post a Comment