Jonathan Jones and Ruth Dixon have published (see Ruth’s blog here) a comment in Psychological Science on conspiracy theorist Stephan Lewandowsky’s Hoax article, much discussed at CA at the time. Although their statistical points are incontrovertible and clearly expressed, it took considerable persistence – see timeline here. Their first and longer original article was submitted to a different journal, but rejected as being of insufficient interest to readers of that journal. The reviewers were sympathetic but more or less referred them back to Psychological Science, the journal which had published Hoax. Psychological Science has strict word limits on a comment (1000 words) and these word limits are counter-productive when an article is so thoroughly bogus as Lewandowsky et al 2012. Lewandowsky was one of the reviewers for Psychological Science and opposed publication. However, unlike Steig in respect of O’Donnell et al 2011, Lewandowsky was identified to the authors as a reviewer, permitting Jones and Dixon to respond to his review comments knowing of the reviewer’s conflict of interest. Editor Eric Eich accepted the comment, as well as Lewandowsky’s response (response paywalled.) Lewandowsky has a blog reaction here, in which Lewandowsky hypocritically compliments the article as a scientific response in peer reviewed literature, without disclosing that he had opposed its publication.
There are numerous other defects with the Lewandowsky article that are not covered in their comment. One can only do so much with 1000 words and Jonathan and Ruth have unsurprisingly done an excellent job. :
Climate Audit 
56 Comments
Based on 3 anonymous data points – which gave it it’s title – this is how Wray Herbert form the APS -w rote about it on the 19th July 2012 (8 months before actual publication)
http://www.psychologicalscience.org/index.php/news/were-only-human/a-climate-for-conspiracy.html
“The results were unambiguous, and unsettling. First, those who hold a laissez-faire view of unfettered free markets were much more likely to strongly reject climate science. Lewandowsky believes that, because the fundamental importance of fossil fuels (and CO2 emissions) to modern economics, climate science in general (and evidence for global warming in particular) is a threat to free market advocates. Free marketers were also more likely to reject other established scientific findings, even the (undisputed) facts that smoking causes lung cancer and HIV causes AIDS.
Second, conspiracy thinking was clearly linked to climate denial—and to the rejection of scientific propositions in general. This was true even of conspiracy theories unrelated to the environment or climate—the belief that NASA staged the moon landing, for example, or that the CIA killed Martin Luther King. In other words, conspiracy thinking is not simply a convenient way to dismiss a particularly bothersome scientific consensus. Instead, some people seem to have a general personality trait or cognitive style, which leads them to endorse any conspiracy. This paranoid thinking in turn predisposes them to reject completely unrelated scientific facts.
Lewandowsky’s study will be published in a future issue of Psychological Science, a peer-reviewed scientific journal, providing further evidence of a vast and ingenious plot to elevate enlightenment thinking and marginalize the unenlightened.” Wray Herbert APS
The same article was republished by Wray in the Huffington Post! the same day..
http://www.huffingtonpost.com/wray-herbert/climate-change-denial-_b_1686437.html
Thank you, Steve! To be fair, we felt that it just took ‘normal’ (rather than ‘extreme’) persistence, given the well-known difficulty of getting a Comment published in any field. None of the delays were excessive, and the anonymous reviews were fair and constructive. And we were impressed with the way the journal handled Lewandowsky’s review, which we were not asked to respond to.
We agree that there are many other flaws in Lewandowsky’s conspiracy papers, raised here and elsewhere. As you say, the Comment had to be limited to making one clear point.
Steve: I edited the post to reflect this and more accurately reflect the chronology in the timeline attached to your post, which i had not considered in my first cut at a cover post.
Although I am a complete statistical layman, I read the comment and it was so clearly laid out I understood the problem you identified
Indeed I even learned something:
when you ASSUME a linear relationship, you make an *SS out of U and ME!
Unfortunately Ruth is correct that it is exceedingly hard to get comments published. This is a distortion of what science should be. If a wrong paper is not corrected it pollutes the discourse forever.
Bringing such gatekeeping well and truly into the spotlight again. Very well done, JJ and Ruth.
Consider this comment suitably refined based on Ruth’s earlier one – earlier but not yet visible when I wrote mine!
I just read Dixon’s chronology covering the publication of the article. She clearly states that Lewandowsky recommended rejection to the editor of PS: “We received two reviews from Psychological Science. One was from the previous Reviewer 2, who suggested publication after a few minor changes to improve clarity. The other was by Lewandowsky et al. (this review was not anonymous), who advised rejection.”
At his blog, Lewandowsky writes:
“We are very pleased that, more than two years after the first article in Psychological Science became available, a critical commentary on our work was submitted for peer review.”
Words fail me.
One would say two-faced but with Dr Lew two may not be enough by the end of this.
Heh, watch the pea and compare ‘submitted’ with ‘published’.
===================
Weaverish in your twistedness
I remember my immediate reaction when I first heard of Lewandowsky’s corpus. It was ‘Projection with a capital P and it stands for Paid’. But I denounce myself in the interest of incursing recursion.
==================
‘Curses and recurses’ the good doctor was heard to mutter.
The money quote: “All the data really shows is that people who have no opinion about one fairly technical matter (conspiracy theories) also have no opinion about another fairly technical matter (climate change). Complex models mask this obvious (and trivial) finding.”
They just stick to the dazzle them with BS model, come to the conclusion they want, and the press leads with it. And they think they need to work on communication??
One of the most remarkable bits of disinformation by Lewandowsky is that even by his own measure, conspiracy ideation, even with fake responses, is lower among climate blog participants on both sides of the aisle than in the general public.
Ruth and Jonathan, this paper is distressing. The entire Recursive Fury fiasco centred on Lewandowsky’s trolling skeptics on various blogs after activating his Skepticalscience associates to record comments critical of the LOG12 paper. Much of the questions would have been easily answered if the metadata would have been released. The element of surprise lay entirely in what the data suggested and its source. Lewandowsky was able to paint his critics as conspiracists by withholding this information from them and tabulating their consequent speculations.
The LOG12/13 data is in all likelihood falsified. The authors state they surveyed Skepticalscience when it is clear they did not. They state they analysed comments at Skepticalscience to estimate a mind-boggling 78,000 skeptics were exposed to the survey. Re-analysis shows Skepticalscience hosted comments from all of 36 commenters during the period. Tom Curtis stated skeptical commenters from Junkscience.com must have been exposed to the survey, but the paper shows no data from the website to be included. In short, there was no exposure to online blog commenting skeptics in the paper. The question of the provenance of hundreds of skeptical commenters remains open.
Lewandowsky now states, as he would, that he ‘welcomes’ this paper as it legitimizes the LOG12/13 data.
Shub:
I am not sure I understand the thrust of your comment. Are you suggesting that Jones and Dixon should not have submitted their article without raising the issue of the metadata?
It seems to me thy have successfully eviscerated the Lewandowsky et al’s articles within their own terms.
Lewandowsky’s rebuttal comments are painfully hollow.
bernie, you are right. Lewandowsky and co-authors are yet to establish where they got their blog data from. They won’t release the data. The analysis paper is published with no mention of the fact the very provenance of the data is in question. The title suggests ‘alternative analysis’ is possible, when no analysis of fake/faulty data can lead to meaningful conclusions.
I think the article does a solid job of explaining, with full references, how you vet survey data.
As I see it, one doesn’t need to point out everything wrong with a paper to demonstrate it isn’t usable for it’s original stated purposes. The fact is Lewandowsky made basic analysis errors that resulted in him arriving at erroneous conclusions.
What’s left to say? The paper is dead and it needs a proper burial.
another entry for the timeline –
Public, September 2010 presentation of Moon Hoax results.. naming sceptics..
From the slides – Lewandowsky publicly presented LOG13 ‘results’ at Monarsh University Sept 2010… Email invites/reminders were still going out to sceptics at the time..
Lewandowsky & Gignac (forthcoming)
•Internet survey (N=1100)
•Endorsement of climate conspiracy (“hoax by scientists to get grants”) linked to endorsement of other conspiracies (“NASA faked moon landing”)
•Conspiracy factor without climate item predicts rejection of climate science
In the slides he publicly names Andrew Bolt and Jo Nova, and labels Jo an irrational hyper emotional conspiracy theorist..
1.(b.) Conspiracy Theorists and Science
•Conspiratorial thinking usually not isolated
–If you believe in one (FBI assassinated MLK), you are more likely to also believe in others (e.g., AIDS created by U.S. government; Goertzel, 1994)
–single factor for 14 conspiracies (34.6% of variance; Swami, 2009)
–predicted by alienation and cynicism and disengagement from mainstream politics
•Joanne Nova called 9/11 a “building accident”
Following slide
Two Classes of Contrarians
•Free-market ideologues (Bolt)
–pragmatic (?) and not overtly irrational
–driven by opposition to any form of regulation
–provide fodder for the second class …
•Conspiracy theorists (Jo Nova)
–outside mainstream politics and society
–hyper-emotional and often irrational
ref the ‘building accident’ and 9/11 – Jo wrote a blog post wrote MOCKING Lewandowsky’s, May 2010 article.
Lewandowsky is no doubt just quoting Watching the Deniers – Michael Marriott (co-author Recursive Fury) – utter misrepresentation of Jo..
Watching the Deniers –
Jo Nova: claims she is no conspiracy theorist while stating 9/11 was a “building accident”
https://watchingthedeniers.wordpress.com/2010/05/07/jo-nova-claims-she-is-no-conspiracy-theorist-by-stating-911-was-a-building-accident/
Jo’s actual article said, which was mocking Lewandowsky conspiracy theorists May 10 article –
http://joannenova.com.au/2010/05/name-calling-fairy-dust-conspiracy-theorist/
“Is the planet warming from man-made CO2? Lewandowsky “knows” it is. Why? Because the 9/11 truthers are conspiracy theorists (and conspiracies are always wrong). O’ look, a few people ask odd questions about an accident in a building years ago, and sometimes those people are also the species Homo Sapiens Climata Scepticus (!). So it follows (if you are insane) that because some people still doubt the official story of an unrelated past event, man-made global warming will contribute 3.7W/m2 in the year 2079, and we’ll all become souffles in the global Sahara.
I’m not making this stuff up. I’ve tallied up the obvious errors from both articles. His power to confuse himself with red herrings is … “impressive”. – Jo Nova – May 2010
slides: (Cook is included as author)
Click to access msi-seminar_10-09-23_lewandowsky_presentation.pdf
Talk:
http://monash.edu/sustainability-institute/seminars-workshops/talks.html#2010
Professor Stephan Lewandowsky: The psychology of climate change communication and ‘skepticism’ – Held on 23 September 2010
What motivates the seemingly growing number of climate ‘skeptics’? How is climate science best communicated in the face of contrarian voices that are small in number but highly organised and vocal? Do we need to communicate the science at all, or is there a better way forward? Australian Professorial Fellow and Winthrop Professor from the University of Western Australia – Professor Stephan Lewandowsky – explores the insights and recommendations provided by psychology and cognitive science.
Lew reply online here http://www.climateaudit.info/pdf/irreproducible/lewandowsky%20Psychological%20Science-2015-Lewandowsky.pdf
Steve, Ruth, and Jonathan, any thoughts on their rebuttal?
Thanks.
Re: Chuck L (Mar 27 13:55) Chuck, we are not impressed by Lewandowsky’s points, and we plan to write a new post shortly.
We’ve now posted some responses to Lewandowsky’s reply:
https://mygardenpond.wordpress.com/2015/03/27/our-responses-to-lewandowskys-reply/
Excellent analysis at your blog, Ruth. Especially the skewed plot showing almost no skeptics sampled and the clear effects of heteroskedasticity. The fact that the authors will never admit you are right does not say much for their statistical chops. Doesn’t Lew teach a psych stats class? Pity the fools who take it.
So if more skeptics had responded, is it likely that they would have gotten worse correlation?
All four posts are great, but this is my favorite:
https://mygardenpond.wordpress.com/skew/
Figure III-2 is positively diabolical.
Eliminate skeptics from your sample completely! *Then* you’ll be able to show just how conspiracist-minded they are!
BWAAAHAAHAAAHAAHAAA!
For chutzpah, it’s hard to beat this statement from Lewandowsky et al.: “Alternative models should reflect alternative theoretically motivated hypotheses, any mention of which is conspicuously lacking in Dixon and Jones’s Commentary.”
Well, gee, if you’re limited to 1000 words (250 more than the average newspaper op ed), good luck so much as citing an alternative theory, let alone developing hypotheses from it.
@JunnkPsychology
We did in fact include something along the lines of our concluding statement “All this really shows is that people who are undecided about one fairly technical matter (conspiracy theories) also have no firm opinion about another (climate change)” in the first version of the comment which we submitted, but one of the referees told us to take it out, so out it went.
In reply to MikeN, “So if more skeptics had responded, is it likely that they would have gotten worse correlation?”
Yes, exactly! In the ‘panel’ survey more skeptics (and undecided) did respond than in the ‘blogs’ survey, and they got a much worse correlation.
Congratulations to Ruth Dixon and Jonathan Jones for getting a serious critique past the gatekeepers.
Not only are Psychological Science’s length limits a real problem—they’re moving online, as all journals are, yet a full report of original empirical research is supposed to be no more than 5000 words. The decision to allow Lewandowsky to serve as a reviewer on a critique of his work is so terrible, it’s hard to know where to start in on it.
JunkPsychology
It was probably a courtesy review. It looks like the editor was not sufficiently impressed with Lewandowsky’s response so as to nix the article. Certainly giving Lewandowsky the opportunity to respond is in order.
Re: JunkPsychology (Mar 27 13:17), Provided the review was not anonymous, we don’t see a problem with that. The editor accepted the view of the other reviewer over that of Lewandowsky.
I don’t think an author in Lewandowsky’s position is entitled to a courtesy review, even if he is also required to identify himself.
The author can respond to the commentary after it’s accepted for publication. Whether he agrees with the decision to accept the commentary is irrelevant.
Any thoughts about your comment being embargoed (a 3 month delay from acceptance till publication online), when the same journal had imposed no embargo on the article you were commenting on: 8 months till actual publication, while the journal’s parent society actually ordered up prepublication publicity?
Lewandowsky’s Hundred Day March into absurdity and oblivion.
==========
Re: JunkPsychology (Mar 29 14:46), the embargo was certainly unusual, as the stated policy of Psychological Science is not to have embargoes.
The justification in this case was that the comment and response should be released as a package. While this makes some sense to me it remains odd that the embargo was uniquely imposed in this case, rather than this being a general rule for comments. However as the embargo was imposed as a condition of the paper being accepted for publication we were not in a strong position to complain about it.
This story looks similar to McI and McK trying to get a comment published in Nature from my reading of The Hockey Stick Illusion.
I also want to congratulate Ruth Dixon and Jonathan Jones on their excellent work, especially considering the hurdles many of us know too well.
I’d make one point – the “skew” is even more important than noted.
Lost in the fog of time is the fact the HOAX (LOG13-blogs) paper was intended to ascertain and review the beliefs of climate SKEPTICS who posted on online climate blogs. See the running head of the paper “MOTIVATED REJECTION OF SCIENCE” and extensive authors comments in the paper.
Ignoring the data collection issues, and focusing only on the data, the responses were dramatically skewed with the “anti-skeptic” position representing the vast majority of the respondents. Which only reinforces the criticisms about the data collection methods of the authors.
Most rational people would NOT go to virulently anti-skeptic sites to obtain data on SKEPTIC beliefs.
The original N=1377 was whittled down to N=1145 in the authors initial parsing. Looking at the numbers of respondents that answered “disagree” or “strongly disagree” with the five CLIM questions, we find an average total for each CLIM question ranging from N=125 to N=245, with an average across all CLIM questions of N=177.
An average of just N=177 out of N=1145 total responses can be considered as responses from someone actually skeptical of global warming.
The HOAX authors hypothesis and findings claim a link between skeptical climate beliefs and conspiracy ideation. Yet only 15% of their responses could be considered to represent data from the target “skeptic” demographic.
Nearly 85% of the data is from those who can be considered strongly “anti-skeptic” regarding global warming. And all but irrelevant to the HOAX hypothesis and findings.
There are several examples of what can be uncharitably called pseudoscience in the service of policy. Lewandowsky’s papers are among them,as are Anderegg, Prall et al, Cook’s 97% solution and Oreskes seminal one page paper unable to find climate skeptics in the literature.
Each of these works can be criticized along multiple lines, from sampling error and confirmation bias to inability to use proper statistical tools in analysis.
The question from a policy standpoint is which line of challenge should be prioritized. In this case, commenters at Lewandowsky’s blog (including myself) highlighted the gross errors in methodology that should have been embarrassing enough to prompt a correction, if nothing else.
These challenges were not effective.
Much as our host’s challenges of paleoclimate papers were more effective because he went after the hard math aspects of flawed papers, I suspect Dixon and Jones will gain more traction due both to going after the math and having the persistence to get their reanalysis published in the forum where the original paper was published, even if it wasn’t their first choice of venue.
There’s a lot to think about in how this plays out. It would suggest to me that Steve might re-examine the utility of his criticism of Mr. Weaver, as justified as it seems to be, on the grounds that it won’t move molehills instead of mountains.
And for people such as myself, Ben Pile, Jose Duarte, Barry Woods and others, we should realize that criticism of methods is not as effective as pointing out errors in math.
Tom, do you REALLY think that math criticism is going to make a difference? It will be ignored as usual. Very sad.
Lew et al are propagandists and advocates who rely upon a veneer of science. What they are up to needs to be called out at the scientific as well as the political level.
It isn’t either or.
More comments like this one being published help the science heal itself from this abuse (and no doubt then move on to the next one). And in this modern age the blog is another means to help get rigour (and disinformation) promulgated.
Re: thomaswfuller2 (Mar 27 23:50), “These challenges were not effective.”
Personally, I think they were effective, not least in alerting Eric Eich to significant problems with the paper. At the critical moment when I appealed the initial rejection of our Comment (a request we considered had almost no chance of success), the previous criticisms may (of course I have no evidence) have influenced his decision to allow us to resubmit.
Also, perhaps our Comment will mean people will take the blog criticisms more seriously henceforward… (one can always hope!)
A crocus peeps through the snowbank.
=========
+1
Eric Eich has a history with this paper.
A part of a long email chain which led to the removal of my name is below:
Bold added here. There were a lot of emails back and forth but the eventual reply from Dr. Eich was this:
Of course, the Air Vent was the only blog source in the study.
Thanks to both of the authors of the comment, for steadfastly facing the hostile review and publication process on an article which shouldn’t have existed. Hopefully Dr. Eich’s political magazine has a tiny bit higher hurdle for publication in the future. Your work is appreciated by many more than you realize.
The editorial correspondence that Jeff Id quoted here is consistent with some other exchanges that I’ve seen.
This was not the usual relationship between an editor who has accepted an article and the author of the article the editor has accepted.
Eric Eich was behaving like a sponsor or a protector of Stephan Lewandowsky.
As Jonathan’s pointed out on Twitter Chris Mooney covered the controversy in the Washington Post yesterday. Not terribly well but he seems to accept the game’s up for the attempted correlation. Without apology, needless to day, and he can’t help himself at end:
The D-cognate still assumed to apply to us as well as the anti-vaxxers. What a climbdown looks like at this stage of the game I guess.
From one point of view there are only four things wrong with the original LOG13-blogs paper. Unfortunately those four things are the design of the experiment, the implementation of the data collection, the analysis of the data, and the reporting of the results. As a consequence of this interlinked network of ineptitude it is very difficult to disentangle all the errors from each other.
The LGO13-panel paper, by comparison, is much better. The design is relatively standard: no worse than many papers in the field. The implementation is still very poor (see for example the discussion at our post on satisficing), but it’s not so bad as to render the data completely useless. The analysis is still incorrect, but this time it is possible to tease out how and why it is incorrect, rather than just noting that it’s all a horrible mess. The reporting is still poor, but that doesn’t matter for a reanalysis.
So the original point of our comment was to see what we could say about the analysis of the data from LGO13-panel. Somewhat to our surprise we found that, once we knew what to look for, the same analysis also worked for LOG13-blogs, albeit not so clearly because of the appalling skew in that dataset. We don’t say much about other issues, not because we don’t believe they are important, but simply because it’s best in a comment to pick one important issue, where the argument can be made very clearly, and then run with it.
Prof Jones:
Your opening paragraph is one of the best and funniest eviscerations of a research article that I have seen. Mark Steyn or the late Christopher Hitchens or even Oscar Wilde could not have done it better. Well done.
Indeed, ‘interlinked network of ineptitude’ could be very useful elsewhere.
With such an interlinked network of ineptitude, how do we explain the decisions by two or three reviewers and an action editor at Psych Science?
How many worse-quality articles have appeared in that journal since its inception?
Reblogged this on I Didn't Ask To Be a Blog.
Very late in this thread, but I wanted to point Steve to a peice by a UWA faculty member on Lomborg’s new Consensus Conference.
Dave Pannell is an agricultural economist you may have heard of (he builds decision tool models) and has always seemed admirably open-minded. Here is his piece on Lomborg:
http://www.pannelldiscussions.net
Steve: David Pannell commented several times at CA some years ago. I think that he also emailed me in 2005 in connection with our early articles.
David Pannell had some very sensible comments about Climategate e.g. here. Pannell’s take was that climate science extended well beyond the Team, but that the Team were out-of-proportion influential. He also noted the extraordinary stubbornness of Mann and associates in refusing to concede even ordinary errors, noting with some amusement that Mann et al 2009 perpetuated the use of the contaminated sediments even after the Kaufman corrigendum.
Pannell would surely be even more amused that Mann and coauthors continue to use the contaminated sediments many years later for Rahmstorf’s reconstruction of Atlantic ocean currents.
Thanks for the link although reading it now six years after it was written is rather disheartening. Dave Pannell had some very good things to say about the state of climate science, but – notwithstanding his point that the field was more than just the Team – nothing seems to have changed. As you have pointed out: the proxy data is still bad (contaminated etc.); it is still(!) used in the wrong orientation; the statistical methods are used inappropriately; and there is still no acknowledgement of any criticism as valid.
The fact that the UWA administration should have – essentially – covered up for Lewandowsky four or five years after Climategate just shows how little these words by Dave (others) actually got through to anyone.
Kimbremetrics finds great value in the Lewandowsky/Lonborg switcheroo, and wonders about the players to be determined later.
===============
7 Trackbacks
[…] https://climateaudit.org/2015/03/27/jones-and-dixon-refute-conspiracy-theorist-lewandowsky/ […]
[…] https://climateaudit.org/2015/03/27/jones-and-dixon-refute-conspiracy-theorist-lewandowsky/ […]
[…] https://climateaudit.org/2015/03/27/jones-and-dixon-refute-conspiracy-theorist-lewandowsky/ […]
[…] First off is a doozy of a new paper by Stephan Lewandowsky, Naomi Oreskes and colleagues that complains that skeptical viewpoints are disproportionately influencing the science of climate change. Recall that Lewandowsky and Oreskes are quixotic climate change denialslayers—conspiracy theorists of somewhat ill-repute. […]
[…] First off is a doozy of a new paper by Stephan Lewandowsky, Naomi Oreskes and colleagues that complains that skeptical viewpoints are disproportionately influencing the science of climate change. Recall that Lewandowsky and Oreskes are quixotic climate change denialslayers—conspiracy theorists of somewhat ill–repute. […]
[…] First off is a doozy of a new paper by Stephan Lewandowsky, Naomi Oreskes and colleagues that complains that skeptical viewpoints are disproportionately influencing the science of climate change. Recall that Lewandowsky and Oreskes are quixotic climate change denialslayers—conspiracy theorists of somewhat ill–repute. […]
[…] https://climateaudit.org/2015/03/27/jones-and-dixon-refute-conspiracy-theorist-lewandowsky/#comment-7… […]