This is the first of the posts with stuff that didn't go into the working paper on misconduct in India.
Richard Grant Steen's paper, Retractions in the scientiļ¬c literature: do authors deliberately commit research fraud?, and discussions of some of its controversial findings were the ones that motivated me to take a detailed look at retractions of papers by Indian authors. Here's one of Steen's findings:
Retracted papers are more likely to appear in journals with a high impact factor (IF), are more likely to involve certain ‘repeat offender’ authors and are more likely to involve authors from the USA. [Bold emphasis added by me]
Just the other day, Retraction Watch linked to a study which looked at retractions by a bunch of journals spanning a wide range of impact factors, and found "a surprisingly robust correlation between the journal retraction index and its impact factor."
All that is about retractons. But Steen also found that "journal impact factor was higher among papers retracted for fraud than among those retracted for error." Interestingly, his definition of 'fraud' includes fabrication, falsification and data plagiarism; plagiarism -- just copying text from one's own papers or from otehrs' papers -- constitutes an 'error' in his classification.
Since almost all the misconduct cases from India are due to plagiarism, they too tend to be published in (and later, retracted from) lower profile journals. Indeed, the list of retracted papers by Indian authors doesn't appear to be populated with high profile journals. This should not really be surprising, because we expect plagiarizers to try to "fly under the radar" by sending their papers to lower tier journals where, presumably, plagiarism detectors were not much in use (at least until 2007-08 or so).
On the other hand, not only do the fraudsters from the West (or, more generally, the rich countries) tend to commit serious fraud (fabrication and falsification), they also appear to be ambitious: they work in fields where the stakes and payoffs (by way of high profile publications) are higher.
[The lead case (which is from Japan) discussed in this WSJ story by Gautam Naik is illustrative, The stakes just can't get any higher than playing with the lives of hundreds of thousands of people! The Retraction Watch blog is filled with cases of this kind.]
3 Comments:
Dr. Abi,
For several papers, academic malpractice is identified at the review-stage by anonymous peers and/or editors. How do you account for such cases? Is there any data on this?
If we were to study trends in what I will term as 'aggregate honesty of a system', should we be measuring (number of papers rejected for malpractice + number of retractions)? This data may be harder to get. In many such cases, there is no public castigation of the authors who attempted to submit a plagiarized papers. The fact that a paper isn't published doesn't change the maliciousness of such authors.
Is the drop in malpractice partly due to increased diligence of reviewers and journal editors? And the widespread availability software to check plagiarism?
Hi Abi,
Check out your 15 ns of fame here here.
Regards, Anant
@karatalaamalaka: You are right about the intent to deceive of the unsuccessful plagiarizers / fraudsters.
But we will never get to know about those cases that are caught at the peer review stage. The journal editors do, though. When they find egregious cases -- for example, a repeat offender -- they banish that person from being able to get anything published. However, many of these (adverse) decisions are just internal to the journal, and not published on its website (I have heard of just one case where the journal devoted a page to a 'notice of banishment' -- just imagine how egregious that author's conduct must have been! Sadly, I haven't saved that link ...)
@Anant: 15ns of fame it is! Thanks for that link.
Post a Comment