Analysis of retractions puts spotlight on academia

Nicola Jones

Nature Medicine 15, 1101 (2009)
doi:10.1038/nm1009-1101

About half of the medical papers retracted over the past few decades were pulled because of misconduct rather than an innocent mistake, according to two new studies. And that fraction is on the increase.
Yet although drug companies are often portrayed by the popular press as the source of all evil in biomedical publishing, just 4% of retractions due to misconduct had declared pharmaceutical sponsorship.>>>

Peer reviewers satisfied with system : TIMES HIGHER EDUCATION

David Schley

But Sense About Science survey finds that two thirds of those polled think it is failing to detect plagiarism.

With the number of learned papers published each year rising to 1.3 million, the peer- review system might be expected to be fraying at the seams.
But an international survey of academics states that two thirds are satisfied with the current system for monitoring the quality of scholarly output, and 90 per cent of those who participate as reviewers remain keen to take part.
The findings were published by the charity Sense About Science at the British Science Festival, held at the University of Surrey, on 8 September.
Tracy Brown, the charity’s managing director, said the issue of whether the system was sustainable was a matter of “public as well as scientific interest”.
But while many of the survey’s findings are reassuring, concerns have been raised.
The vast majority of researchers polled say that peer review should detect plagiarism and fraud, but only about one third think it is doing so.
Similarly, while most respondents say that the system should be able to ensure that papers acknowledge any previous work used, only half think it does so effectively.
Despite these issues, participants caution that expecting reviewers to approach manuscripts with suspicion runs counter to the assumption of honesty and the spirit of collaboration in science.
They add that such a tactic would make the task of peer review unmanageable.
Adrian Mulligan, associate director of research and academic relations at Elsevier, said that the launch later this year of Crosscheck, a pan-publisher plagiarism-detection tool, could resolve some of the problems raised.
Given the principle of openness in science, there is a surprisingly strong desire for anonymity from reviewers, with a double-blind process considered to be most effective.
This consensus has been attributed to a desire to protect junior academics asked to review work by more senior colleagues. According to the survey, editors have warned that completely open reviewing reduced the number of people willing to participate and led to “lame” reviews of little value.
Although more than two thirds of the survey’s respondents state that training would be beneficial, Ms Brown said she was hesitant about the peer-review process being professionalised, as it was difficult to see how any qualification could meet the needs of different disciplines.
Instead, she advocated the nurturing of postdoctoral researchers and postgraduate students by more experienced peers, but noted with disappointment that very few reviews were currently undertaken collaboratively with junior colleagues.
A full report is due to be published in November – following peer review.
For more details, see: www.senseaboutscience.org.uk/index.php/site/project

Further survey findings
A third of respondents say they are happy to review up to five papers a year, with a further third happy to review up to ten.

On average, academics decline two papers each year, principally because they are outside their area of expertise, although workload is another frequently cited reason.

The average time taken to review a paper is six hours. However, there is a great deal of variability: one in every 100 participants in the survey claims to have taken more than 100 hours to review their last paper.

Self-plagiarism: unintentional, harmless, or fraud?

THE LANCET

Volume 374, Issue 9691, 29 August 2009-4 September 2009, Page 664

Editorial

http://dx.doi.org/10.1016/S0140-6736%2809%2961536-1

The intense pressure to publish to advance careers and attract grant money, together with decreasing time available for busy researchers and clinicians, can create a temptation to cut corners and maximise scientific output. Journals are increasingly seeing submissions in which large parts of text have been copied from previously published papers by the same author.

Whereas plagiarism—copying from others—is widely condemned and regarded as intellectual theft, the concept of self-plagiarism is less well defined. Some have argued that it is impossible to steal one's own words. The excuse editors hear when confronting authors about self-plagiarism is that the same thing can only be said in so many words. This might sometimes be legitimate, perhaps for specific parts of a research paper, such as a methods section. However, when large parts of a paper are a word-for-word copy of previously published text, authors' claims that they have inadvertently used exactly the same wording stretch credibility.

There is a clear distinction between self-plagiarism of original research and review material. Republishing large parts of an original research paper is redundant or duplicate publication. Publishing separate parts of the same study with near identical introduction and methods sections in different journals is so-called salami publication. Both practices are unacceptable and will distort the research record. Self-plagiarism in review or opinion papers, one could argue, is less of a crime with no real harm done. It is still an attempt to deceive editors and readers, however, and constitutes intellectual laziness at best.

Deception is the key issue in all forms of self-plagiarism, including in reviews. Few editors will knowingly republish a paper that contains large parts of previously published material. Few readers will happily read the same material several times in different journals. An attempt to deceive amounts to fraud and should not be tolerated by the academic community.