Saturday, January 26, 2008

Plagiarise, don't let anything evade your eyes

Title taken from a song by Tom Lehrer

A commentary in Nature suggested that duplicate publication is on the increase. Mostly autoplagiarism, apparently, as it seems that the majority of these duplicates share at least one author. A few studies are referenced that suggest a relatively low number of plagiarised articles, but a much higher number of suspected duplicates with the same authors. And it is suggested that those have been published simultaneously, which is, of course, not easy to achieve for alloplagiarism ("simultaneous publication is rarely observed for duplicates that do not share authors").

It also suggested that duplicate publication is bad, particularly in areas like clinical research ("Duplication, particularly of the results of patient trials, can negatively affect the practice of medicine, as it can instill a false sense of confidence regarding the efficacy and safety of new drugs and procedures"). This is no-doubt true, but one wonders if this negative effect is anything other than minor, given the rather widespread publication biases when it comes to clinical trials, such as this one regarding the treatment of depression with selective serotonin reuptake (SSRI) inhibitors: "Thirty-seven studies were assessed by the FDA as positive and, with one exception, every single one of those positive trials got properly written up and published. Meanwhile, 22 studies that had negative or iffy results were simply not published at all, and 11 were written up and published in a way that described them as having a positive outcome." (Ben Goldacre in The Guardian of January 26, 2008). Judging the scientific validity of findings just by counting articles is clearly pretty primitive.

Autoplagiarism is seen as ethically questionable, to say the least. According to the authors of the Nature commentary, Mounir Errami and Harold Garner, "it not only artificially inflates an author's publication record but places an undue burden on journal editors and reviewers, and is expressly forbidden by most journal copyright rules."

This is undoubtedly true as well, but again, placed in context it may be dwarfed by the burden on journal editors and reviewers imposed by the cascading effect of the whole publication process, with its cycle of submission, rejection, submission to another journal, rejection by that other journal, and so forth, until the article is finally published somewhere, meanwhile peer-reviewed at every stage.

What if the motives of autoplagiarising authors are more benign? What if they just want to ensure a wide dissemination of their work and they see multiple publication as a way to achieve that? One might say that publishing in a journal that offers open access would be a better way of doing that, or self-archiving in an open repository (and I would certainly be in favour of publishing with open access). But a quick look at the various open access advocacy email lists shows that cross-posting is rife, even though the archives of such lists are completely open. That complete openness is evidently not being regarded as sufficient by the cross-posting posters to get the attention desired. Multi-publication may in essence be the same phenomenon, or at least driven by the same motives. Is it so much different from having multiple versions of an article, as in one in a journal, another one in a central repository, another one in an institutional repository, et cetera? Sure, those should all refer to the same formally published article, so the authors can't get extra credits for them, so maybe it is very different. But hey, the scientific ego-system is a pretty cut-throat arena, and multiple publication seems amongst the smaller of possible misdemeanors, with a least the positive effect of wider dissemination of research results.

I am not convinced that autoplagiarism is anything other than a minor problem in science. It seems to me that non-publication of negative results is a problem of an order of magnitude greater. It is high time that this bias is addressed, and with the kind of indignation now seemingly accorded to autoplagiarism.

Interesting irony:
A Google search on 'non-publication of negative results in 2007' (search done on 26 January 2008, 16:30 GMT) shows as first result an article in the Journal of the American Medical Informatics Association, with the link: which leads to a screen saying "The article you requested is not currently available online".

Further down in the Google results is a link to an abstract that seems to be from the same article, and it is online, albeit not open. From the abstract: reasons why studies were not published range from "results not of interest for others" (1/3 of all studies), "publication in preparation" (1/3), "no time for publication" (1/5), "limited scientific quality of study" (1/6), "political or legal reasons" (1/7), and "study only conducted for internal use" (1/8)

Jan Velterop

No comments:

Post a Comment