An article appearing the Proceeding of the National Academy, well, youi know the rest has added some more fuel to the slow burn about the number of retractions of articles in the scientific literature. Some analyses previously published seemed to support the suggestion that retracted articles were being pulled because the authors had made mistakes and had somehow become aware of this fact and were now asking the journal editors to yank the article.  But the PNAS article offers some pretty strong evidence that such a comforting interpretation is, to quote Eliza Doolittle, ‘not bloody likely’. The authors made careful surveys of the reasons stated for retraction by  the editors, or in a statement written by the retracting authors. Some of those are so bland and anodyne (‘the data are unreliable’) that it’s not possible to figure out what went wrong. But the authors dug deeper and found from secondary sources such as newspaper accounts that research integrity problems were factors in many retractions.  The blog Retraction Watch covers the story quite extensively here:

http://retractionwatch.wordpress.com/2012/10/01/majority-of-retractions-are-due-to-misconduct-study-confirms-opaque-notices-distort-the-scientific-record/

Tuesday is the day the New York Times science section appears, and there is a story on the PNAS article also:

http://www.nytimes.com/2012/10/02/science/study-finds-fraud-is-widespread-in-retracted-scientific-papers.html?_r=1&hpw

There has been a large increase in the number of scientific papers retracted by authors or journal editors. And, typically, the journal publishes a notice which explains the reasons for the retraction, soft of anyway. Some observers complain that, sure, there are notices, but very often they don’t tell you much.   A recent study in the Journal of Medical Ethics followed up on publications involved in cases of research misconduct and found that in some cases the article describing the research was not retracted and so lives on. In others, the not terrible helpful  word RETRACTED was added. In others yet, a vague, circumlocutary statement was issued, talking about ‘mistakes’, or ‘errors’, without any admission of hanky, much less of panky. The guys at Retraction Watch have been crusading for greater ‘transparency’ in the area of retraction statements. Instead of reading that ‘mistakes were made’, they want to know what kind of mistakes, and who made them. The good news is that journals seem to be getting tougher. When the retraction craze started, there was a certain diffidence, but as things advanced, editors began to resent more openly that they, the reviewers, and the readership had all been conned, and the more they thought about it, the less they liked it. Some retraction statements have become more detailed. Image manipulation still seems to lead the way as major cause for retraction. And a cadre of quasi-professional image sleuths has grown up, quite skilled in the detection of image re-use, repositioning, clipping, inserting and other tricks. It’s really a vulnerability for the would-be faker: your efforts are there, just waiting to be uncovered. Problems with informed consent, plagiarism or other forms of ‘creative recycling’ are frequent causes. And then there is the not inconsiderable number of data fakers.   Takes all kinds.

http://retractionwatch.wordpress.com/2012/09/05/surprise-surprise-study-says-retraction-notices-often-arent-honest-about-misconduct/#more-9559