Thursday, August 11, 2011

Scientific Misconduct in India: An Analysis of Retractions in PubMed


That is the title of a study I have done, and the preliminary results appear in a paper embedded below I have posted online [links appear at the end of the post]. The document is labeled Version 1.0, since study is yet to be wrapped up. Still, there's enough in there for it to be opened up for comments / suggestions.

I presented this work at the Workshop on Academic Ethics organized by Rahul Siddharthan, Gautam Menon and N.S. Siddharthan about a month ago.

Quick summary: PubMed database lists ~103,000 papers published by Indian authors during the previous decade (2001-2010); 70 69 of these papers have been retracted, and 45 of the retractions are due to some form of misconduct. Plagiarism is overwhelmingly the primary mode of misconduct: all but one of the 45 misconduct-related retractions were due to plagiarism.

If that doesn't sound bad enough, consider this: At 44 per 100,000 papers, India's misconduct rate is far higher than that of countries such as the UK, the USA, Germany and Japan.

There's some silver lining, though: Retraction of papers from Indian authors show a steep fall since 2007 -- either because Indian researchers know better now, or because plagiarized papers are ever less likely to make it to print in the first place due to increasingly widespread use of plagiarism detecting software by journals.

Here's the paper. Okay, the embed didn't quite work out well; here's the html version; if you prefer a pdf, get it from here.

Comments welcome.

* * *

Update (13 August 2011): The html version will keep getting updated with minor corrections which will be duly noted in the footnotes. There has already been one correction: the number of retractions of Indian papers for the decade 2001-10 is 69 -- the original version had it at 70.

6 Comments:

  1. Rainbow Scientist said...

    Since bio-medical research is still not well developed in India, this study seems to be incomplete. Most of the research in Indian universities and institutes is in basic sciences...

  2. Abi said...

    @Rainbow Scientist: Yes, I do say (in the last paragraph) that a limitation of this study is that it's based on just one source: PubMed.

    But the proportion of biomed research in India is not as low as your comment seems to imply. PubMed lists over 13000 papers from India for the year 2007 -- and this number is nearly half of India's total number of papers for that year, estimated to be over 27000.

    In terms of whether PubMed papers are representative of India's research output, I don't think anyone has evidence to show that India's biomed scientists are significantly more honest (or dishonest!) than its physical and mathematical scientists ...

  3. ktwop said...

    Congratulations. A long overdue analysis of research publications from India and hopefully will eventually lead to to the establishment of something like the ORI in the US.
    I have a theory that the chain of responsibility - from institution practices to researcher to supervisor to colleagues to reviewers to Journals to Editors - must have a corresponding chain of accountability for the responsibility to be taken. The solution may lie in a concept of liability (based on tort or good faith) applying even to scientific work and publications.

  4. Rahul Siddharthan said...

    Thanks! It is interesting and pretty much the same, I think, as what you talked about at the meeting. Regarding Rainbow Scientist's comment -- pubmed lists all journals in basic biological sciences of any significance whatever (if it's not in pubmed, it's an extremely obscure journal or a dubious fly-by-night journal), and also lists papers from a lot of physics/chemistry/computer sciennce journals (eg, many of Chiranjeevi's papers are chemistry rather than biology). I'd guess that these days at least half the serious output in basic sciences from India (including, certainly, all of the basic biological sciences research) would be listed in pubmed.

  5. VRS said...

    1) Elsevier, Springer, Wiley etc journals often tell younger, less established scientists that there is not much publishing space so they have to be rejected or are put through the peer review process only to be asked by reviewers who clearly don’t have the time / inclination / ability to do proper reviews of the manuscript and propose experiments that are out of the authors’ scope or ask idiotic questions that demonstrate the lack of interest and knowledge of the reviewers.

    Even if a certain volume of work was being published in a given year, editors often tell authors submitting manuscripts with very similar volume of work (in the scope of the journals) that they have decided (suddenly) to stop accepting manuscripts with such content. This kind of whimsical behaviour makes younger scientists go towards the dubious or fly-by-night journals, some of which are quite predatory and demand money to publish the manuscripts though no such charges are mentioned in the beginning.

    2) On the flip side, these journals are happy to publish some very shady work from well established investigators. Plagiarism may be a major symptom of scientific misconduct but data falsification, manipulation, redundant publications, gift and ghost authorship are also quite abundant and often go unobserved and unchecked. Many papers have figures of gels where the proteins have been stained with Coomassie Blue due or silver after trying to express and purify them.

    Though they claim to have “purified” the protein, a look at the pictures will reveal how Photoshop has been used to change the brightness and contrast to make certain bands stand out and the others fade into the glare. Figures are incorrect, the experimental details are incorrect and clearly impractical and the same experiment is used to publish multiple papers that are quite similar. The mistakes in one paper are repeated in the others. Still all the reviewers of all these papers missed these mistakes and have OK’d the publication of these papers.

    That the same thing was recycled is never caught by anyone, much less the fact that it was incorrect in the first place when it was published. Even when it is pointed out to these journals, many case studies reveal that the journals are unwilling to confess that their peer review systems are flawed and prefer to push the matter under the carpet rather than deal with the problems. A Committee on Publication Ethics (COPE) is supposed to be an authority on, well, publication ethics and many publishers such as Elsevier, Wiley, Springer etc are all members of COPE.

    However, it would be difficult to regulate individual or even organizational ethics externally. Ultimately, people need to be ethical just so that they can expect it from others and live peaceful lives. Problems such as gift and ghost authorship are more difficult to prove. Ghost authorship can be proved if the person who feels that they ought to be an author has proof that they have worked on the manuscript and its data. But often the people who are dropped are students or research associates or junior faculty who are too afraid to question their seniors.

    Gift authorships are usually given to those who are powerful or well known in some area or who have been wronged by the principal investigator in some way or are being set up to make someone else take a fall and are being compensated in this way. The fact that an investigator is publishing in an area that they were not trained in should indicate that there is some gift authorship issue going on but if journals don’t act against research falsification and open errors like manipulated images, how and why would they act against gift authorship?

    What is insidious is that many of these senior scientists in India are helped by the policy makers and funding agencies who close ranks with their peers and help these autocrats perpetuate their unethical conduct and punish the younger scientists if they protest about the theft of their work.

  6. VRS said...

    1) Elsevier, Springer, Wiley etc journals often tell younger, less established scientists that there is not much publishing space so they have to be rejected or are put through the peer review process only to be asked by reviewers who clearly don’t have the time / inclination / ability to do proper reviews of the manuscript and propose experiments that are out of the authors’ scope or ask idiotic questions that demonstrate the lack of interest and knowledge of the reviewers.

    Even if a certain volume of work was being published in a given year, editors often tell authors submitting manuscripts with very similar volume of work (in the scope of the journals) that they have decided (suddenly) to stop accepting manuscripts with such content. This kind of whimsical behaviour makes younger scientists go towards the dubious or fly-by-night journals, some of which are quite predatory and demand money to publish the manuscripts though no such charges are mentioned in the beginning.

    2) On the flip side, these journals are happy to publish some very shady work from well established investigators. Plagiarism may be a major symptom of scientific misconduct but data falsification, manipulation, redundant publications, gift and ghost authorship are also quite abundant and often go unobserved and unchecked. Many papers have figures of gels where the proteins have been stained with Coomassie Blue due or silver after trying to express and purify them.

    Though they claim to have “purified” the protein, a look at the pictures will reveal how Photoshop has been used to change the brightness and contrast to make certain bands stand out and the others fade into the glare. Figures are incorrect, the experimental details are incorrect and clearly impractical and the same experiment is used to publish multiple papers that are quite similar. The mistakes in one paper are repeated in the others. Still all the reviewers of all these papers missed these mistakes and have OK’d the publication of these papers.

    That the same thing was recycled is never caught by anyone, much less the fact that it was incorrect in the first place when it was published. Even when it is pointed out to these journals, many case studies reveal that the journals are unwilling to confess that their peer review systems are flawed and prefer to push the matter under the carpet rather than deal with the problems. A Committee on Publication Ethics (COPE) is supposed to be an authority on, well, publication ethics and many publishers such as Elsevier, Wiley, Springer etc are all members of COPE.

    However, it would be difficult to regulate individual or even organizational ethics externally. Ultimately, people need to be ethical just so that they can expect it from others and live peaceful lives. Problems such as gift and ghost authorship are more difficult to prove. Ghost authorship can be proved if the person who feels that they ought to be an author has proof that they have worked on the manuscript and its data. But often the people who are dropped are students or research associates or junior faculty who are too afraid to question their seniors.

    Gift authorships are usually given to those who are powerful or well known in some area or who have been wronged by the principal investigator in some way or are being set up to make someone else take a fall and are being compensated in this way. The fact that an investigator is publishing in an area that they were not trained in should indicate that there is some gift authorship issue going on but if journals don’t act against research falsification and open errors like manipulated images, how and why would they act against gift authorship?

    What is insidious is that many of these senior scientists in India are helped by the policy makers and funding agencies who close ranks with their peers and help these autocrats perpetuate their unethical conduct and punish the younger scientists if they protest about the theft of their work.