Why I said no to revising my submission to BMJ Rapid Response
Dear Professor Coyne,
Thank you for your rapid response (copied below).
We would like to post if you would you resubmit concentrating on the science. The talk of spinning data suggests that the authors have a hidden agenda, which we cannot post without proof and is anyway difficult to prove and tends to add more heat than light to a debate..
Thank you very much for your help.
Sharon, thank you for consideration of my Rapid Response. After giving your condition serious thought, I have decided not to resubmit.
Posting a Rapid Response from BMJ website that has advantages, beyond the obvious one of getting considered for publication in the print edition. Anyone going to the BMJ website and looking at an article will see that a response has been made and can choose to read it. BMJ has been an important innovator in this respect. It should serve as a model for the many journals lagging far behind in post publication peer review.
Sharon, as you are well aware, I have taken advantage of the BMJ Rapid Response option numerous times. On some of occasions, you have suggested that I had to tone something down in order to be posted. I have always done and, in hindsight, sometimes with beneficial effects.
Rapid Responses used to be the only game in town for someone wanting to respond to a BMJ article. A commentator had to comply with requests to edit comments or risk the post disappearing into oblivion.
But then blogs came along as an alternative offering a more unfettered opportunity to comment. Blogs have usually carried the disadvantages of not having pre-post peer-reviewed, not having the same access afforded a Rapid Response attached to an article, or not being archived or indexed. But things are changing and some of these disadvantages are beginning to be eliminated.
Then PubMed Commons emerged, which allowed posting of comments that would be available to anyone accessing a BMJ article or any other of 23 million articles through PubMed. It has less restriction and grants more discretion to commentators. It also allows commentators to post links to blog posts, where points could be elaborated.
PubMed Commons is a challenging alternative to letters to the editor and also takes away control of comments from editors and journals. It competes with BMJ Rapid Responses, but I would not be surprised if it stimulated other journals to adopt the model of BMJ to better accommodate post publication peer review.
It may seem the high ground to insist that Rapid Responses stick to the science, but it is naïve and misleading. BMJ, like Nature and Science, does not simply publish just data, and certainly not all of the data that their projects generate. The data that are presented are selective, framed, and interpreted. Ironically, the author of the BMJ article claiming the link between fast food outlets and obesity has stated elsewhere that you can never take authors’ role out of the data they choose to present and interpret in their articles, and that authors have a strong hand in creating or knocking down “obesogenic realities.”
As for the issue of “spin,” the concept can be operationalized and scientifically studied. Numerous examples of spin can be cited. For instance, it is been shown that much exaggeration in media coverage can be traced to spin in the abstracts of scientific articles. Over a series of blog posts, I showed that was the case with a recent Lancet study of CBT for schizophrenia (1, 2, 3). And before that there was that churnaled study of a breakthough blood test for postpartum depression.
It is the interest of science and the public to point out spin in abstracts and the body of articles, and, in the case of this particular BMJ article, to point to discrepancies between what is said in the abstract and what is said in the results section. Too often, readers and the media rely only on abstracts and few people bother to make comparisons between abstracts and what was actually said in the text, which could substantially contradict or at least qualify what is said in the abstracts.
My intended Rapid Response raised methodological and statistical points about key aspects of the article on fast food outlets and obesity, but was also intended to alert readers to the spin in the abstract of the BMJ article and invited readers to decide for themselves.
But I did more than complain about the author, I complained about policies of BMJ that serve to encourage, reward, and maybe even insist on spin. I have in mind the requirement that cross-sectional observational studies claim clinical or public health importance, particularly in the absence of a requirement that they preregister hypotheses, outcomes to be examined, candidate covariates and analytic plan.
So this time, at least, I will sacrifice the advantages that go with posting a Rapid Response at the BMJ website, but I would keep writing a series of blog posts about the article. When the post at PLOS Mind the Brain has been uploaded, I will then post a PubMed Commons comment. Anyone accessing the BMJ article through PubMed will have the opportunity to view and respond to my comment and learn about it in my blog.
PubMed Commons is truly revolutionary. It builds on the earlier act of PubMed making MEDLINE abstracts universally accessible, and not just limited to academics having the gateway of University libraries. Previously, journals, maybe even the BMJ, fought to use copyright of abstracts to keep this from happening. I recall cheering the dramatic moment on American national TV when Vice President Al Gore publicly made use of PubMed to learn more about his wife’s illness.
BMJ Rapid Responses has also been revolutionary in its conception. But now BMJ must response to the challenge posed by PubMed Commons or risk being left behind. I am sure that the journal is up to the task, and I again thank you for offering the opportunity to revise my proposed BMJ Rapid Response.