In times gone by, Nature used to be a top quality science journal. But more recently it has descended into a mouthpiece for opinions, pseudo-science and political activism. As an example, see the recent anonymous editorial Harassment victims deserve better. It claims just below the headline that “Sexual harassment is rife in science” and “But one thing we do know is that sexual harassment is a serious problem in science.” These assertions are untrue, and the editorial provides no evidence in support of them apart from a couple of cases that have recently come to light. In fact it contradicts itself by mixing phrases such as “Nobody knows” and “We don’t know the answers to those questions” in among the statements of apparent certainty. Why is Nature magazine employing someone who is incapable of logical thought to write its editorials? Furthermore, the article is irresponsible and counter-productive, since inaccurate scaremongering of this type is likely to put women off careers in science if they read and believe this sort of nonsense.
However, the main theme of this post is the Nature comment article Research integrity: Don’t let transparency damage science by Stephan Lewandowsky and Dorothy Bishop. Lewandowsky of course needs no introduction to readers of this blog, being notorious for breaking a fundamental principle of his research field, respect for subjects. So the very idea of him writing an article for Nature on research integrity is an absurd irony, on a par with self-confessed fraudster Peter Gleick being on an ethics panel. His co-author Dorothy Bishop recently achieved notoriety as one of a small handful of academics who led the prejudiced witch-hunt against Sir Tim Hunt, based only on a single misleading tweet from Connie St Louis; she wrote “@profgeraintrees Could we ask that he not be on any appointments or promotions committees, given his views.”
The ridiculous idea of the article is that transparency, in other words free availability of research data, can somehow “damage” science. The data might fall into the hands of bad people, who might misuse it. As pointed out by Brad Keyes, in a comment that the Nature moderators deleted, this is reminiscent of Phil Jones’s notorious response to a request for data,
“We have 25 or so years invested in the work. Why should I make the data available to you, when your aim is to try and find something wrong with it.”
[A brief reminder of what happened in that case: Jones and his university came up with a sequence of bogus excuses for not releasing the data, including claims of confidentiality agreements, then the notion that the data could only be released to academics. This led to two academics making FOI requests, met with further refusals and appeals, ending with a ruling from the information commissioner in 2011 in favour of the requesters.]
The main idea of L&B does not stand up to the most basic scrutiny. They provide lists of vague, subjective and unworkable criteria for distinguishing between “use” and “abuse” of data, and a list of “Red flags” (ironically, Lewandowsky’s own work fits with at least five of these flags) . Speculation about the motivations of data requesters makes no sense, as it cannot be done in any rigorous objective way. Who would decide who was worthy and who wasn’t? Stephan Lewandowsky himself perhaps? Quis custodiet? These issues were thought through by the people who devised the FOIA, with the result that a key principle of the act being that the reason for the request is of no relevance and not required.
The response to the Nature piece was interesting, in that as well as the expected objections from climate sceptics, another group of individuals reacted just as angrily. These were suffers from, or researchers in, a debilitating and poorly understood illness known as ME or CFS (chronic fatigue syndrome). L&B said that some CFS researchers had been subjected to harassment, referring in particular to a large-scale trial known as PACE. In the comments, and on twitter, many people objected to this, saying that the PACE researchers had used dubious research methods (such as changing criteria during the experiment) and had refused to release data for independent scrutiny. See the comments under the Nature article by Keith Gerachty, Peter Kemp, Simon Wright and others, and the comments at PLOSONE on a PACE paper. There is also an interesting discussion forum, Demonization of Patients continues: Nature Blog. Of course those of us not familiar with this story shouldn’t really take sides, but it’s hard not to, given the apparent close similarities of researchers with a pre-set agenda (CO2 is the climate control knob / climate sceptics are conspiracy theorists / CFS is largely psychological and can be successfully treated with CBT) using dubious methods to confirm their beliefs, making exaggerated misleading press releases, refusing to produce data and complaining about being harassed.
A comment from Ken Rice inadvertently highlighted Lewandowsky’s inconsistent claims about his own data transparency. Rice provided a link to this Lewandowsky data set, which says in large letters “Access to this dataset is not open“, but in this article, linked from the Nature piece, Lewandowsky says that he always releases his data. Following the Phil Jones approach, his data is available “to credentialed scholars only”. Presumably this is just an administrative mix-up by Bristol University that will soon be corrected.
The final irony was that yesterday, Thursday, around lunchtime, Nature moderators started deleting comments that were critical of L&B’s flawed and inflammatory piece about openness and transparency (comments were then closed). When Richard Tol noticed the comment deletion and mentioned it on twitter, I saved the comments that I had in an open browser window, see this docx file. Tol’s deleted comment read:
“Research integrity and transparency are great. It starts at home. It would be good if Professor Lewandowsky would come clean about his research on climate change, and if he would tell his student John Cook to do the same. Harassment would be much reduced if researchers would tell the truth, the whole truth, and nothing but the truth about their data, how it was gathered, and how it was processed and analyzed.”
Barry Woods also had a comment deleted which you can read here; although critical of Lewandowsky, it’s entirely factual.
At the time of writing, my comment that ends with “I find it remarkable that Nature should publish such nonsense” is still visible.
I will add in links to other blogs here.
Getting the idea of transparency all wrong from Nicole Janz, a Cambridge academic in social sciences.
James Coyne, Professor of health psychology at Groningen, has a blog post here that criticises the Science Media Centre as well as L&B.
Violating the norms and ethos of science: Judith Curry picks out phrases from L&B that break the principles of science set out by Merton.
Tom Fuller, Transparency in Science Over-rated: Stephan Lewandowsky writes “I have frequently described Lewandowsky as a charlatan at this blog and I see no reason to change my views”.
Trial By Error, Continued: A Few Words About “Harassment” by David Tuller, a well written piece with useful background on PACE.
El nuevo escándalo de Lewandovsky salpica a Nature from the PlazaMoyua blog.
FOI Coyne ridiculous at Bishop Hill.
PACE trial and other clinical data sharing by Leonid Schneider.
More On The Epidemic Of Orthodoxy Enforcement In The World Of “Science” by Francis Menton.
The Committee on Public Ethics is holding a webinar on these issues on February 12th. They claim that the questions were “explored systematically and thoughtfully” by L&B. They invite comments.
Some of Brad’s deleted comments are posted at Canman’s blog.
30 March: Warren Pearce, Sarah Hartley and Brigitte Nerlich have written a response at the Making Science Public blog. They also had a very short comment published in Nature. They say that the L&B article is oversimplified, inflames tensions and makes unsubstantiated claims.