Subterranean War: Some Reasonable Questions and Answers

By Stephan Lewandowsky
Professor, School of Experimental Psychology and Cabot Institute, University of Bristol
Posted on 10 November 2013
Filed under Climate denial, Culture
and Michael Mann

Further authors: Gerard Hastings and Linda Bauld, University of Stirling

The “Subterranean War” article that we published earlier this month caused considerable interest. Notably, we received much further confirmation of the common pattern underlying attacks on scientists from colleagues in multiple disciplines, including in particular medical research. One article in the New England Journal of Medicine from 1997, entitled “the messenger under attack—intimidation of researchers by special interest groups” reads like a prequel to our own article.

There have also been critical voices on the internet, not all of which were constructive. We take up some of the more incisive questions that have been raised by various commentators, in particular by Warren Pearce.

Warren raises three points that we take up in slightly different order:

2) 3rd party re-analysis of data is surely a staple of science. Of course, those doing so may have particular motivations (as in the Philip Morris example), but one would have a hard time preventing this taking place. Recent history shows the perils for scientific credibility of not making data available.

We agree and like most scientists, we make all our data are routinely available upon publication of an article. We agree that this is a healthy staple of science. (There are some important discipline-specific exceptions involving confidentiality of participants which are important to understand but need not concern us here.) We are however concerned with the way in which this basic scientific principle can be abused. It is helpful to underscore those abusive techniques here:

  • There have been many instances of re-“analyses” of epidemiological data (or other data with regulatory import) by industry bodies or their affiliates, in which inconvenient results were “sanitized” by elimination of data or other statistical statistical sleight of hand. This is well documented and is an abuse of the transparency of the scientific process. It may not be possible to prevent this from occurring, but it is possible to draw the public’s attention to those strategies so it can make an informed choice about how much credence to lend to such activities.
  • Similarly, if requests for data persist after all results of any potential scientific value have been made available, those requests are difficult to reconcile with good-faith attempts to contribute to new knowledge. Such requests are more likely to be harassment than attempts to aid in scientific discovery. Recent decisions by the UK Information Commissioner support this perspective by rejecting requests for prepublication data quite decidedly.
  • Finally, if requests for data have been met by scientists, but they are nonetheless accused of “hiding data,” this is a fairly clear fingerprint of denial. (We wonder what “recent history” Warren is alluding to; this might well be mythical problem rather than an actual one.)

3) The piece vividly depicts some troubles and tribulations of science (and indeed, life) in the modern world. However, it might benefit from a stronger counterpoint than the final paragraph’s nod to the “public’s right to access to information”. The activities of climate sceptics may well represent an “insertion into the scientific process”, and I do not offer a blanket defence of their multifarious criticisms and approaches. In particular, where bullying is identified it should not be tolerated anywhere in modern society. However, the arrival of online fora has demonstrated that the public are not always a passive group waiting for the latest scientific knowledge to be visited upon them. On occasion they can be somewhat unruly and, if sufficiently motivated, they may wish to “insert themselves” in any way they can with the limited tools available to them; especially as members of the public do not enjoy the same access to journals as academics. This may be an inconvenient truth, but it is also a fact of modern life. With better systems for dealing with this, we can hopefully focus more on transparent and robust methods of managing conflicts – both legitimate and otherwise – between science and society, rather than seeking to devise new laws to protect the former from the latter.”

We agree that the public need not (indeed, should not) be a “passive recipient” of knowledge. There is nothing wrong with vigorous public debate in blogs or elsewhere. Both of us contribute to public debate on an on-going basis, and we regret that our time commitments are insufficient to engage even further and in more detail. There are however clear boundaries between vigorous (perhaps even polemical) debate and the fingerprints of denial. To give but a few examples,

  • Posting email addresses of scientists or executives of universities on the internet with the explicit or tacit encouragement to launch complaints, on the basis of the flimsiest of accusations, is not a means of public discussion. It is difficult to consider this to be anything but harassment. It is also a waste of the tax-payers’ money because someone has to respond to whatever correspondence to a university ensues.
  • Refusal to follow proper paths by which complaints and concerns can be redressed—e.g., by refusing to make a formal approach to a university but continuing nuisance email correspondence—is not a matter of public debate. It is harassment, pure and simple.
  • Refusal to take note of the outcome of complaints, by continuing to air concerns that have already been adjudicated, is not a meaningful contribution to debate but, likely, a further tool of harassment.
  • Refusal to submit one’s criticisms of academic work to peer-review, while at the same time seeking to suppress research by bullying of editors is not public debate. It is harassment, and it constitutes an intolerable and unethical interference with due scientific process. More than anything else, this issue of seeking to suppress academic work must be tackled in light of the scholarly evidence that climate scientists are unduly risk averse. If there is one thing the public must be protected from, it is scientists who have been bullied into downplaying the true risk societies are facing, be it from tobacco, HIV, or climate change.

1) How does one differentiate between ‘vexatious’ or ‘trivial’ requests for data and those which are merited? The authors give the example of timestamps for blogposts as trivial, but one could imagine occasions when such information might be quite important. There appears to be an appeal to lawmakers to act in the final paragraph. Is this really the best way to proceed? An ethics committee containing a rich mix of personnel drawn from different sections and strata of society (i.e., not just academics) might provide better, context-specific judgements.

This is a difficult question and like many other things in public life, it requires judgment. There are however valuable sources of constraint that are beginning to emerge:

  • The literature on querulous complainants has yielded a fairly good set of markers that administrators can use to differentiate between true grievances and vexatious agenda-driven complaints. It is important to recognize that every vexatious request ties up time that could otherwise be put towards resolving a true grievance—in that sense, vexatious complaints and requests are no different from prank calls to fire or police emergency lines.
  • There is a growing tendency, at least in the UK, to recognize the problematic implications of FOI legislation in the age of electronic communication, where private conversations among scientists are now considered to be “public documents” because the technology—but not the context or intent of the parties involved—has removed the right to privacy that citizens are entitled to in democratic societies.
  • In a recent judgment, the UK information commissioner seems to have recognized that scientists are entitled to a private space of debate that is not subject to FOI. We cite paragraph 34 of that judgment: “All too often such [FOI] requests are likely to be motivated by a desire not to have information but a desire to divert and improperly undermine the research and publication process–in football terminology–playing the man and not the ball. This is especially true where information is being sought as part of a campaign–it is not sought in an open-minded search for the truth–rather to impose the views and values of the requester on the researcher. This is a subversion of Academic Freedom under the guise of FOIA and the Commissioner, under his Article 13 duty must be robust in protecting the freedom of academics from time-wasting diversions through the use of FOIA.” (Emphasis added.)

Are we calling for lawmakers to act? We consider this to be an open question. As we noted in “Subterranean War”, daylight is the best disinfectant. The daylight should enable a public conversation about ways in which inconvenient scientists who are conducting research in the public interest can be protected from harassment and vexatious complaints, while they continue to be accountable to ethical and professional bodies as they already are.

We must not forget that science denial can kill. It killed thousands in South Africa because vital anti-retroviral drugs were withheld from AIDS sufferers by a government that considered Western medicine to be racist. It killed tens if not hundreds of thousands when firm medical knowledge about the effects of tobacco on human health was questioned by organized denial. And, yes, to the extent that climate denial delays mitigative action, it too will come with a cost that is measured in human lives as well as money.

This week’s typhoon that is now estimated to have killed 10,000 people in the Philippines might have occurred in the absence of climate change, although global warming likely put it on steroids. Nonetheless, right on cue, some individuals have already denied the strength of the storm by claiming that the typhoon was “another overhyped storm that didn’t match early reports.”

The issue of science denial and the attacks on scientists it provokes is thus too important to ignore or to put into the “too-hard” basket. The scholarly literature on denial—some of which is reviewed here— provides some initial criteria by which it can be differentiated from legitimate scientific or public debate. It is in the public’s interest to become conversant with that distinction so it does not confuse the noise generated by attention-seeking or agenda-driven individuals with genuine scientific debate.

Bookmark and Share

40 Comments


Comments 1 to 40:

  1. I find Warren Pearce's complaints to be specious. Making false and abusive assertions under the pretense of "asking questions" is a political ploy--- not a scientific one, nor a ploy that is designed to address legitimate concerns. One sees this behavior not only in deniers of the observed facts about human-caused climate change, but also deniers of the evidence for evolution, a spherical Earth, the heliocentric model of the Solar System, believers in "free energy," and the like.

    There is absolutely no legitimate reason why scientists should pay attention to "questions" (i.e., assertions spewed out by denialists under the pretense of "just asking questions") that are designed to be abusive, asinine, acidic, and harassing.
  2. Stephan/Mike,

    I am very pleased to see this rebuttal but concerned by Barry Wood's continuing insistence (i.e. on 'Science Made Public') that UWA have refused to investigate his complaints about alleged misdemeanors relating to timestamps on emails...?

    I am also pleased to see your reference to the events in the Philippines this week. This is because, in researching a response to an anthropogenic climate disruption (ACD) 'sceptic' on my blog, I have discovered this recent report (PDF) published by the Taiwanese Environmental Protetction Authority. This appears to indicate that the number of typhoons hitting Taiwan has more than tripled since the 1970s. This implies that more than 70% of the typhoons now hitting Taiwan could be the consequence the multi-decadal industry-funded campaign to deny that burning fossil fuels is the primary driver of ACD.
    Moderator Response: UWA has advised me that no one has contacted them with a request to investigate anything in the last 6 months. I do not respond to allegations on the internet when the people who make them fail to take their own opinions sufficiently seriously to either submit them for peer review or refer them to the appropriate institution for examination.
  3. Brandon Shollenberger at 05:38 AM on 12 November, 2013
    There's an interesting dichotomy in this post. It says:

    We agree and like most scientists, we make all our data are [sic] routinely available upon publication of an article.


    Yet it has paragraph after paragraph of unsubstantiated accusations. An uninformed person reading this post could not possibly verify most of what it says. Similarly, anyone who might disagree cannot respond meaningfully as they can't show they're actually responding to what the authors are referring to.

    The effect of this dichotomy is to render this post meaningless save as propaganda. Even if everything said in this post were true, there'd be no reason to believe it.
  4. It is good of you to reply to Warren Pearce’s criticisms, which for some reason didn’t actually appear under your article, though they were apparently intended by the author to be comments on your article.
    Will you be replying to any of the criticisms and requests which did in fact appear under your article? For example, the assertion made by myself and Barry Woods that your claim to have posted your “Moon Hoax” survey at SkepticalScience is false?
    You say:
    “...we make all our data routinely available upon publication of an article.”
    Will you therefore publish the data identifying the source blogs for your 1300+ respondents to LOG13 (“Moon Hoax”)? This would clear up definitively the question of whether you or Barry and I are telling the truth.
  5. I am afraid that Stephan Lewamdowsky (and his LOG12 co-authors) really needs to prove that the survey for LOG12 posted on the Skeptical Science website...

    If I repeat this request, it is because the authors have NOT addressed the issue...

    As on 2 occasions now a Skeptical Science moderator no less (named Tom Curtis) has categorically stated that the survey was not held there...

    Tom Curtis' blog:
    http://bybrisbanewaters.blogspot.co.uk/2013/09/skeptical-science-and-lewandowsky-survey.html


    If Stephan recalls last year Stephan told me he had this url , but had lost it. (whicj I thought a bit sloppy at the time) but now we know it never existed.

    Which is an issue, because a number of claims for LOG12 about the audience for this survey, depends absolutely on Skeptical Science's involvement.

    An academic (a maths professor at Nottingham university, coincidentally from the same university as Dr Warren Pearce) who raised this concern on a sceptical blog, was named in the Recursive Fury paper, under conspiracy ideation, for just making the observation that the methodology had not been followed as described in the paper.
  6. REF:

    "Moderator Response: UWA has advised me that no one has contacted them with a request to investigate anything in the last 6 months. I do not respond to allegations on the internet when the people who make them fail to take their own opinions sufficiently seriously to either submit them for peer review or refer them to the appropriate institution for examination."

    I would like to submit a comment, I asked both Psychological Science and the authors for the raw (anonymous) kwik survey data, including survey domain refering urls (ie which blog hosted the link to the survey) and proof of a claim that Skeptical Science was surveyed in the methodology of LOG12.

    and this has NOT been supplied..

    making it shall we say difficult to submit a comment..


    Chief Editor Erich Eich has invited me to submit a comment and I still wish to, I just require the data that was requested..

    When I originally asked for it over a year ago. Professor Lewandowsky told me he had it, but had lost it, and that John Cook (skeptical Science founder) may have deleted it.... (which seemed sloppy at the time)

    and now we know thanks to Tom Curtis who is a moderator of Skeptical Science, (the people that helped set up this very blog, and a co-author of Professor Lewandowsky) the information that was requested never existed...
    yet the authors published anyway..

    This is the claim made in LOG12, which is made false by Tom Curtis and all the evidence from the wayback machine archive:


    LOG12 - Prevalence of "skeptics" among blog visitors

    All of the blogs that carried the link to the survey broadly endorsed the scientic consensus on climate change (see Table S1). As evidenced by the comment streams, however, their readership was broad and encompassed a wide range of view on climate change. To illustrate, a content analysis of 1067 comments from unique visitors to
    www. skepticalscience.com, conducted by the proprietor of the blog, revealed that around 20% (N = 222) held clearly "skeptical" views, with the remainder (N = 845)
    endorsing the scienti c consensus. At the time the research was conducted (September 2010), www.skepticalscience.com received 390,000 monthly visits. Extrapolating from
    the content analysis of the comments, this translates into up to 78,000 visits from "skeptics" at the time when the survey was open (although it cannot be ascertained how
    many of the visitors actually saw the link.)

    For comparison, a survey of the U.S. public in June 2010 pegged the proportion of "skeptics" in the population at 18% (Leiserowitz, Maibach, Roser-Renouf, & Smith, 2011).
    Comparable surveys in other countries (e.g., Australia; Leviston & Walker, 2010) yielded similar estimates for the same time period. The proportion of "skeptics" who comment at www.skepticalscience.com is thus roughly commensurate with their proportion in the population at large."

    -----------------

    This becomes a substantial issue, because a) the survey was never shown at Skeptical Science, making claims for the potential size of audience, and diversity (20% sceptical, false.. and b) the content analysis was somehow applied to represent all the other 7 blogs...

    I wish to comment, I have been invited to be the Chief Editor of Psychological Science, I have also reported a factual error in the paper..

    yet the data has not been forthcoming, nor has the matter of the factual LOG12 methodological error been resolved. (it is clearly non-partisan, as Tom Curtis, who writes at the very same blog, Skeptical Science, has stated publically , twice) that the survey was NOT held at Skeptical Science.

    Please supply me with the requested data, and I will submit a comment to the journal for review.
  7. Here is my original request for the data described above, over 14 months ago. pls note I also alert Stephan Lewandowsky to the possibilty of their being no survey held at Skeptical Science:


    Sent: Tuesday, July 31, 2012 5:00 AM
    To: Stephan Lewandowsky
    Subject: Links to surverys - Skeptical Science - Guardian Article about you recent paper.

    Hi Stephan

    sorry to approach you one more time.

    I cannot find the link to Skeptical Science survey, this is probably the most high profile blog with the most media/public recognition (i.e. won awards) of the ‘pro-science’ vs. the “Skeptical” blogs

    (I’m guessing Climate Audit, WUWT, Bishop Hill & maybe The Air Vent (ie Condon) and Jo Nova ;-) )

    I’ve found six of the links to the opinion surveys, and the range of comments on the blogs are quite interesting as well, did you consider this feedback in the research?

    but, I would expect that Skeptical Science would have the most comments and opinions and probably the largest readership.

    Can you send me the link to the Skeptical Science blog article/comments?

    And was the survey able to capture the referring blog, as this might also give indicators of relative popularity of the blog,

    does the survey break down by referring blog and are these figures available?

    Best Regards

    Barry

    rather than lots of questions, if you have the supporting data, etc in an easily accessible package (without too much trouble for yourself) could you send that as well.

    If not quickly to hand, that’s fine please don’t waste any time, as I’m mainly just curious on the couple of point above.

    there were the links I found:

    http://scienceblogs.com/illconsidered/2010/08/counting-your-attitudes/
    http://profmandia.wordpress.com/2010/08/29/opinion-survey-regarding-climate-change/
    http://scienceblogs.com/deltoid/2010/08/29/survey-on-attitudes-towards-cl/
    http://hot-topic.co.nz/questionnaire/
    http://tamino.wordpress.com/2010/08/28/survey-says/
    http://bbickmore.wordpress.com/2010/08/28/take-a-survey/

    I’m missing this blog survey link as well.
    http://www.trunity.net/uuuno/blogs/

    -end email

    At the time of my writing (July 2012) the paper was 7 months away from being published, the supplementary data paper, with the claims of a diverse audience based on a content analysis of Skeptical Science was not known nor available.

    Thus the claims made became of more interest, when the paper was finally published in March 2013:

    to quote Skeptical Science moderator Tom Curtis (extract):

    "2) Lewandowsky’s new addition to his paper is silly beyond belief. While the issue of whether the SkS notification was by blog or by tweet was of no consequence when it was simply notification of the publicizing of the survey, once Lewandowsky drew conclusions from the population commenting on SkS it becomes a substantial issue."


    In science methodology matters, errors in methodology should be corrected, otherwise what you are doing is not science.
  8. link to Tom Curtis' quote:
    http://climateaudit.org/2013/03/28/lewandowsky-doubles-down/#comment-408051
  9. Brandon Shollenberger at 08:59 AM on 13 November, 2013
    The moderator response @2 confuses me. It says:

    I do not respond to allegations on the internet when the people who make them fail to take their own opinions sufficiently seriously to either submit them for peer review or refer them to the appropriate institution for examination.


    First, the moderator's response doesn't list an identity. It's peculiar to make a response using pronouns like "I" and "me" without providing an antecedent.

    Second, this paper, and the paper it references, offers almost no evidence for any allegations made. In effect, it's a bunch of hand-waving. That seems to indicate the moderator in question believes publishing opinions in peer-reviewed literature is important even if peer review couldn't possibly check those opinions.

    It's difficult to see the value in submitting claims to peer review if peer review won't bother checking them. Unsupported allegations do not become more valid simply because they've been rubber stamped by peer review.

    Third, there are many legitimate ways to handle concerns. The fact people may not resort to two (ineffective) methods this moderator mentions in no way indicates anything about how seriously they take their concerns. This is especially striking as the authors of this post have both used other methods in pursuing their own concerns.
  10. Ref:

    "UWA has advised me that no one has contacted them with a request to investigate anything in the last 6 months."


    What is still unresolved is the fact that the paper was published with a major methodological error - described above - despite apparently (according to Tom Curtis) the authors knowing about it.

    As this was reported to UWA in March April.. that would make it.. over 7 months ago ! ;-) ;-) !!
  11. With reference to:
    "Martin Lack at 04:09 AM on 12 November, 2013
    Stephan/Mike,

    I am very pleased to see this rebuttal but concerned by Barry Wood's continuing insistence (i.e. on 'Science Made Public') that UWA have refused to investigate his complaints about alleged misdemeanors relating to timestamps on emails...?"

    -------------------
    Martin.
    My concern is whether the survey for LOG12 actually appeared at Skeptical Science.. Tom Curtis a Skeptical Science moderator, says categorically not.

    I have not written to Prof Lewandowsky, or his co-authors, or any journal, or any university, about any timestamps on any emails.

    So please withdraw that claim above, about my continuing insistence about timestamps on emails. I do hope the moderator (whoever they are), in their reply to you will verify that as well.

    This is simply not true, please get your fact right Martin, maybe you are thinking about somebody else.
  12. geoffchambers at 07:23 AM on 5 December, 2013
    Professors Lewandowsky and Mann, in this rather long article, say:
    “There is nothing wrong with vigorous public debate in blogs or elsewhere. Both of us contribute to public debate on an on-going basis, and we regret that our time commitments are insufficient to engage even further and in more detail.”

    Surely they can find the time to answer the two extremely simple questions in the above comments posted by Barry Woods and myself over three weeks ago? Does Professor Lewandowsky have the URLs of the blogs linking to the survey that was the subject of LOG12? Was the survey linked at SkepticalScience?

    We’ve been waiting a year now for two yes-or-no answers.
  13. Barry Woods is incorrect on some details. I was not a SkS moderator at the time of the survey, and am no longer a moderator at SkS. His repeated emphasis claim that I am an SkS moderator is false, and my status as a former SkS moderator is irrelevant in that I have, and hand, no access to relevant information not in the public domain.

    Further, while I have emailed both John Cook and Stephan Lewandowsky about the findings of my analysis, I have not received a reply from Stephan Lewandowsky to that email, and have not way to no it was actually read. Further, the reply I received from John Cook clearly indicated his continuing belief that he had posted the the survey on SkS itself, rather than just as a tweet. That belief is mistaken, but in my opinion, sincerely so. I have no reason to believe he has actually gone through the wayback machine evidence to reach his own conclusion on that point. The upshot is, Barry Woods has no basis to claim that Lewandowsky or Cook know that the survey was not posted at SkS, as distinct from being merely informed as such.

    Finally, I regard the non-posting at SkS a trivial matter having no substantial effect on the outcome of the survey. It is of no more consequence than the posting at the "skeptical" blog Junk Science, also unmentioned by, and probably unknown to, Lewandowsky (search for "hanich" on the page to find the article). Barry Wood's obsessive pursuit of the issue is, IMO, an example of the type of harassment of scientists this article refers to.
  14. Whether the survey for lOG12 appeared on SkS would have made what difference to the results pray tell?
  15. test
  16. Hi Tom..

    Yes, the survey was held in 2010, I know, you know, everyone knows that, it is irrelavant whether you were a moderator ornot in 2010)

    I have written that when you investigated to see if the survey was held (as you have written about on your blog), you had moderator access at the time of your investigation, this was in September 2012.

    Thus, my claim appear NOT to be false, please do not misrepresent me.

    By your actions at the time (in September 2012), as you described in the email to Climate Audit, you were able to look at deleted comments, etc. You had, at the time access to the Skeptical Science website that no member of the public has..

    Also by your comment on September the 5th 2012, around the time you investigated for yourself you said this:

    "First, apparently some people are treating my comments as representing an official view of Skeptical Science. They are not. I have only spoken for myself; and will never speak as an official spokes person of Skeptical Science except in my limited role as a moderator. Nothing I say or write outside of moderation boxes on SkS represents any view other than my own, although other SkS authors may agree with me to a greater or lesser extent." - Tom Curtis, Skeptical Science comment, Sept 5th 2012..

    Thus when you said 'in my limited role as a moderator' I think quite fairly believed you to have moderator access to the website in September 2012, and were able to check as a moderator, at the time you wrote to Stephan Lewandowsky, another Co-author, and Cook in 2012,


    If you were not moderating at that time, please conform this? (Sept 2012), if not, what rights did you have to allow you this level of access?

    Also. I have written directly to Stephan Lewandowsky and his coauthors, weeks ago, to formally inform them of this error, as I have also written to the journal and UWA.

    thus the authors are WELL aware of the error....

    Additionally in July 2012 I wrote to Stephan Lewandowsky, and he told me he had, had the link to the SkS survey, but had lost it!!

    I find this comment below fascinating.

    "The upshot is, Barry Woods has no basis to claim that Lewandowsky or Cook know that the survey was not posted at SkS, as distinct from being merely informed as such."

    That seems to suggest that it might have been and we have to prove a negative! If he cannot prove that the survey happened (all the evidence suggests it did not), the paper should be withdrawn until it is corrected.

    very very sloppy, a moments webarchive/citation was all that was required..

    If Prof Lewandowsky were to provide the raw kwik survey data (as has been requested) the referring domain from each participant in will clearly show whether anybody at all was refereed to take part in the survey, from Skeptical Science website.

    please Professor Lewandowsky provide the raw kwik survey as requested. (hey, who knows, it might prove me wrong!!)


    The paper makes a claim, that he survey was viewed by a diverse and broad audience, with 20% sceptical, similar to the wider public, this is based SOLEY on Skeptical Science being involved, and a 'content analysis' and traffic figures by John Cook. Thus it is non-trivial..

    As I have suggested to he authors a new content analysis will be required for each of the other 7 blogs, to show what % sceptical, and how diverse the audience was, and traffic figures for the other 7 blogs need to obtained.


    Tom says that it would have no effect on the outcome, however, would it have been published at all, based on we did a survey on 7 blogs, with a low % of sceptical readers, dominated by 2 blogs that hate sceptics (ie relative alexa ranking) in a high profile journal such as Psychological Science.

    I asked Prof Lewandowsky (July 2012) to show whether a key part of his data/methodology existed, a key part of methodology..

    ...he said it lost it..

    ...then critics show it seems to have never existed..
  17. Hi Tom

    When the LOG12 paper was published in March 2013, some months after your initial investigation) you said this in a comment at Climate Audit:

    Tom Curtis comment-Climate Audit-Lewandowsky Doubles Down

    "2) Lewandowsky’s new addition to his paper is silly beyond belief. While the issue of whether the SkS notification was by blog or by tweet was of no consequence when it was simply notification of the publicizing of the survey, once Lewandowsky drew conclusions from the population commenting on SkS it becomes a substantial issue. Specifically, there is no a priori reason to assume the readership of the blog and followers of the tweets are the same. In fact, followers of the tweets are likely to be even more heavily weighted in favour of acceptors of the IPCC position than is the readership of the blog. That is, the very point Lewandowsky tries to make is the substantive difference between notification by blog or by tweet.
    (I note that in my comments to Geoff Chambers I was unaware of the changes made by Lewandowsky to his paper.)"
    -Tom Curtis, 29th March 2013

    -------------------

    "once Lewandowsky drew conclusions from the population commenting on SkS it becomes a substantial issue." Tom Curtis

    This is the exact basis of my concerns, once the paper was published we both realized how the paper depends on Skeptical Science's involvement for conclusions based on audience populations. I formed my concerns based on my reading of the published paper and partly based on your own work/investigations and comments!

    What exactly made you change your mind from being a 'substantial' issue, to as now you say it is 'trivial matter'...
  18. Barry Woods @16, I was a moderator at the time of my original investigation, but while that did give me access to deleted comments, deleted comments were only accessible to moderators for about a week after deletion. It had no bearing, therefore, on my ability to discover whether or not the survey had been posted at SkS. As I have stated, everything I was able to discover was available on the public record. Indeed, the only thing I actually added to McIntyre's evidence was a more exhaustive search over the relevant period.

    Your description of me on this site as a moderator of SkS implies that I had privileged access to relevant information. That is simply not true. Therefore that description, while true at the time of the investigation only misleads without qualifications you have not made.

    @17, I stand by what I said at Climate Audit, and there is no contradiction. The additional argument by Lewandowky is irrelevant given the failure to post the survey at SkS. It is also, however, irrelevant to the actual results of the survey. There are sufficient "skeptics" reading the various blogs, and I presume following SkS on twitter to account for the small number of "skeptics" who took the survey. The suggestions by McIntyre that more than a very small number (about 4 from memory) of responses were scammed have always been based on ludicrous evidence.

    Indeed, if you analyse the results of the survey carefully, you will see it strongly supports a hypothesis that "skeptics" are less prone than "believers" to express certainty that conspiracy theories are false. You could parse that as simply showing that "skeptics" are less dogmatic than "believers" if you wished. I consider the biggest flaw in the paper is that Lewandowky did not identify this alternative hypothesis, nor that the data did not support the claim that conspiracy theorists are more prone to be "skeptics" (his actual hypothesis) over the alternative.

    Further, although the failure to post the survey on the SkS website probably resulted in fewer "skeptic" responses, the posting of the survey on Junk Science probably resulted in more "skeptics" taking the survey, as did the fact that at least one (and possibly more) skeptics took the survey at the request of "skeptic" blog owners who never-the-less did not post the survey on their website.

    So, to summarize, while the failure to post on SkS destroys the premise of Lewandowsky's additional argument, that argument was inconsequential in any event, and the failure to post is irrelevant to the main claims and flaws of the paper.
  19. Tom

    As you know the concern is about this:

    LOG12 - Prevalence of "skeptics" among blog visitors

    All of the blogs that carried the link to the survey broadly endorsed the scientic consensus on climate change (see Table S1). As evidenced by the comment streams, however, their readership was broad and encompassed a wide range of view on climate change. To illustrate, a content analysis of 1067 comments from unique visitors to
    www. skepticalscience.com, conducted by the proprietor of the blog, revealed that around 20% (N = 222) held clearly "skeptical" views, with the remainder (N = 845)
    endorsing the scienti c consensus. At the time the research was conducted (September 2010), www.skepticalscience.com received 390,000 monthly visits. Extrapolating from
    the content analysis of the comments, this translates into up to 78,000 visits from "skeptics" at the time when the survey was open (although it cannot be ascertained how
    many of the visitors actually saw the link.)

    For comparison, a survey of the U.S. public in June 2010 pegged the proportion of "skeptics" in the population at 18% (Leiserowitz, Maibach, Roser-Renouf, & Smith, 2011).
    Comparable surveys in other countries (e.g., Australia; Leviston & Walker, 2010) yielded similar estimates for the same time period. The proportion of "skeptics" who comment at www.skepticalscience.com is thus roughly commensurate with their proportion in the population at large." - LOG12


    ----------------------------

    What is the % of sceptical readers that actually took part in the survey from the other 7 blogs?

    Where are content analysis for the 7 blogs that actually held the survey? (SkS' is irrelevant to these anyway!)

    What is the traffic of these blogs?

    Did Deltoid/Tamino dominate the responses (ie checking Alexa ranking, show, the other blogs to have much lower traffic than these 2 much higher profile blogs (Lew/Cook used Alexa for Recursive Fury.

    The issue is basic scientific principle. Lewandowsky and his co-authors cannot show that Sks held the survey,a key part of his methodology in fact the evidience suggest it did not happen.

    A key claim of a broad and divers survey audience participating, in line with the wider public, is based on Skeptical Science's involvement,which seems clear did not happen

    I have asked for the raw kwik survey data, the referring domain per participant, should quickly demonstrate, or not whether SkS held the survey

    Therefore the paper,and the whole prevalence of sceptics section, and claims about the audience, should be corrected accordingly.
  20. Am I understanding Tom's position correctly?

    We have here a piece of work which purports to be peer-reviewed science. Various people discover that part of the methodology explained in the paper was never implemented.

    The authors repeatedly deny that this is the case and resort to lying about it to people who point out the evidence.

    Tom, as an external third party, understands all this but thinks that, in his opinion, using the proper methodology wouldn't have made much difference to the results - so the paper should stand.

    In what parallel universe would this deceitful rubbish be accepted as any kind of "science"?

    I only hope somebody with a scrap of integrity at the University of Bristol or the Royal Society is reading this.
  21. Foxgoose, as you well know I have argued quite forcefully on a number of occasions that the papers results do not stand. Even taking the data at face value, it shows only that:
    1) Extreme "skeptics" (ie, dragonslayers) and AGW proponents tend to have stronger opinions on conspiracy theories (ie, are more likely to choose 1 or 4 rather than 2 or 3) than do "moderate skeptics";
    2) "Moderate skeptics" and and AGW proponents are approximately equally likely to reject conspiracy theories; and
    3) Extreme "skeptics" are more likely to accept conspiracy theories than other groups.
    That does not support Lewandowsky's conclusions (nor contradict them). These results obtain even if you exclude the four almost certainly gamed results. On top of that, there are several methodological flaws which IMO render the results dubious in any event. The paper is ripe for a peer reviewed response completely rebutting it.

    In fact, considering how bad I think the paper is, I have been laughing the whole time at the "skeptic" response to the paper which has repeatedly concentrated on trivialities, and not launched substantial criticisms except occasionally, were they have missed the point. It has been frankly incompetent as skepticism (but I guess that is no surprise).

    So, even if the failure to post the survey at SkS were a substantial matter, it would be irrelevant because the results of the paper do not stand in any event.

    Having said that, the failure to post at SkS is irrelevant except as rebutal to Lewandowky's subsidiary argument in any event. The proportion of "skeptic" and "non-skeptic" responses is easily determined from the data itself, once the four almost certainly scammed responses are excluded. There have been no substantial arguments that call the apparent ratio into question. Even Barry Woods obsessive attempts to make an issue out of this depend essentially on ignoring the fact that a link to the survey was posted at a "skeptic" blog, albeit not one that was requested to post it.

    Finally, I know of no evidence that either Lewandowsky or Cook have lied about the survey. They have both obstinately, IMO, failed to revise their views in the light of contrary evidence with regard to the posting of the survey at SkS, but if that is all it took to be dishonest, every AGW "skeptic" would be classed as irredeemably dishonest.
  22. Barry Woods, see my response to Foxgoose. In particular you should note that harping on about the failure to post at SkS while not acknowledging the post at Junk Science shows your efforts to be biased. If you acknowledge the latter, of course, your argument that there must have been less than 20% "skeptic" respondents evaporates.

    It is fair enough to notify Lewandowsky, and/or his editors of what is undoubtedly an error in the paper. At that stage, it is pointless second guessing the editors response. You would be better of writing a comment focusing on the analysis of the data as obtained, but mentioning the problems with how the data was obtained, ie, both the SkS and the Junk Science issues in a single short paragraph. If you did it properly, I suspect it would be published.
  23. The editors/journals /APS response was get lost. (to paraphrase) Prof Erich Eich evenrefused to ask Lew et al to supply me the data i had requested from them(and had been ignored)


    Why am I biased..

    I knew that the link had been shown at Junkscience, because that is how the sceptics found out that Hanich sent out the email request to sceptics.

    BUT, and it is a BIG BUT

    As far as I'm aware if there were any Junkscience respondences (unlikely given the 'warning about the survey', at Junkscience?)they were NOT used in the paper...!

    because IF JunkScience responses were used in LOG12, and Prof Lewandowsky and his co-author knows this, but did not include this in the methodology, then the errors get worse!

    Did Prof Lewandowsky use Junkscience responses?

    If he(or UWA) provides the raw Kwik Survey data (which should have responses including the refering domain) then we can resolve this potential extra can of worms, you have just brought to my attention.
  24. The survey was also shown I believe at Planet 3.0 and at UWA internally, I wasn't concerned about those locations for the same reasons as I was not concerned about JunkScience (ie had assumed that this non described data had not been used in LOG12).

    But, if it was it shown at any other places.. ?

    Were responses from these locations, and any others used in the LOG12 paper.. despite the LOG12 methodology saying nothing about these other locations(I had always assumed, that other location data was not included)

    I guess neither Tom or I have any idea, we know of at least one fault in this area of methodology already, missing SkS survey. The raw kwik survey data would help to resolve this.
  25. Barry, you are biased because you simply dismiss the possibility that data was obtained from the Junk Science link. In this case, that is all "AFAIAA" means. You have no relevant data on the issue. The only relevant data on the public domain is that the number of responses reported at about the time of the Junk Science post was about 300 (from memory) less than that finally reported in the paper, which may be because the number reported in Melbourne was a round figure, or that there were additional responses including, potentially, from Junk Science.
  26. Tom

    I am biased? because I thought that when the paper said it obtained data from 8 named blogs. I assumed it did not take data from another unnamed blog..

    Is that what you are saying...

    a very odd sort of scientific method applied in this paper!

    This of course distracts from the point. The methodolgy was not as described, and specific claims made that rely on the content analysis of Skeptical Science are not supported by any evidence.
  27. Tom

    The issue here is not the quality of the data, methods and conclusions of the paper.

    It is whether the authors have been and are being honest.

    You, I, Barry, Geoff Chambers and quite a few others know with absolute certainty that a link to the survey was never posted at the SkS website.

    Lewandowsky and Cook continue to assert that it was and refuse to correct the paper.

    They have now had clear evidence, including yours, put before them that the link was never posted - but they continue with the pretence that it was.

    You try to minimise their conduct by saying - " They have both obstinately, IMO, failed to revise their views in the light of contrary evidence".

    This is IMHO not an honest statement of the facts. Because Lewandowsky & Cook themselves KNOW whether or not the survey was posted - they shouldn't need "contrary evidence" to persuade them to tell the truth.

    If they KNOW they didn't post it and continue to pretend that they did - they're telling lies.

    It's that simple.
  28. Barry, may I remind you that you said:
    "Therefore the paper,and the whole prevalence of sceptics section, and claims about the audience, should be corrected accordingly."

    You said that only about the failure to post at SkS, even though you knew about the posting on a "skeptic" blog. I agree a corrigendum about the SkS survey, and the Junk Science posting would be nice, but concentrating on one and ignoring the other shows bias. In this case the bias feeds your unsubstantiated and almost certainly untrue claim that the number of "skeptics" taking the survey was much lower than is apparent from the data itself.

    Foxgoose, Lewandowsky and Cook do not know (with or without capitals) what happened. They rely on their memory of what happened, and the memory is fairly obviously confused. Beyond that, they have no more information than is now in the public domain, either from the wayback machine, or from the foi emails. IMO Lewandowsky ought to examine that evidence, and ought to issue a corrigendum; but it is not dishonest of him to not do so, though nor does it show strong scientific integrity to ignore the issue. If he declines to do that, the judgement of the editor as to whether the issue is serious enough to require him to do so is final - and he has made that judgement.

    My problem with the points you try to make on this issue is that they are overblown. You are trying to make a mountain out of a slag heap. At the same time, you completely ignore full blown dishonesty from well known "skeptics".

    This will be my last post on this thread. Both Barry and Foxgoose have well deserved reputations of never letting anything rest, and never acknowledging the valid points made by the people with whom the are disputing. Having clarified my views, there is, therefore no point in continued discussion with them. Readers should bear in mind, however, that if either Barry or Foxgoose (particularly Foxgoose) purport to represent my opinion on something, they will almost certainly be misrepresenting it for tactical purposes. If you don't read it in my words, in context, don't believe them.
  29. Tom

    So your explanation is that the two lead authors of an academic paper had a simultaneous memory loss - which prevents them from remembering that they they didn't carry out the procedure they said they had.

    Do you think an alien abduction may have figured in this?

    It sounds marginally more likely.
  30. Well yes, to paraphrase Leó Szilárd's answer when Fermi asked "where are the Martians", they are here, Foxgoose and Barry Woods are doing passable imitations, and to extend the metaphor with something that Eli saw recently from someone who met one of the "Martians" meetings them is a strange and disturbing experience.

    Have you guys ever actually written or refereed a scientific paper? Tom C has it right, this is a distinction without a difference. He also has it right that everyone is free to interpret the results as they wish.
  31. Man pretending to be bunny rabbit makes predictable sideways hop - to avoid confronting the issue.

    Also leaves a few little bunny name droppings and invokes the divine status of academe.

    Personally, I haven't written or refereed "a scientific paper" since I've never been an academic. I have however spent a career running technology businesses based on the work of academics. Oftentimes, I've risked my entire livelihood on judging the quality of such work and also had to run the risk of ignominy or even jail if I was ever found to exaggerate the value of it in a share sale prospectus.

    The academics involved were always keen to share the proceeds of any success arising from their efforts - but curiously reluctant to attach their names to anything which might result in adverse consequences, if their work turned out to be deficient.

    Since academics are generally a group who risk no adverse consequences from the accuracy or otherwise of their ruminations - it puzzles me that they should be allied to be arbiters of risk & reward on a planetary scale.
  32. allied = allowed
  33. "Foxgoose, Lewandowsky and Cook do not know (with or without capitals) what happened. They rely on their memory of what happened, and the memory is fairly obviously confused" Tom Curtis


    I see we are worlds apart...
    if that is the standard set for science these days...
  34. I think it's post-modern scientific ethics.

    If you believe enough - it's true.
  35. It is not true, as Tom Curtis claims, that “Foxgoose, Lewandowsky and Cook do not know ... what happened. They rely on their memory of what happened, and the memory is fairly obviously confused.”
    We rely on emails that are in the public domain, from Lewandowsky to Barry Woods, and from Cook to me, both claiming that the survey was posted at SkS, and on the evidence of the Wayback machine,that shows that it wasn’t.
    There are also the emails between Lewandowsky and Cook prior to the posting of the survey in late August, (obtained via a FOI request by Simon Turnill) in which Lewandowsky asks Cook to post the survey, and Cook replies that he’ll do so when he posts his own survey - something he never did.
    They don’t have to rely on their confused memories. All they have to do is to consult their own emails to each other.
    Of course the question could be easily cleared up by another yes/no answer.
    Professor Lewandowsky, is your memory as to where and whether you posted your survey confused? You’ve found the time to go twice round the world giving interviews and writing articles ithe press on three continents publicising the results of your research. You can hardly claim that you don’t have the time to answer this simple question.
  36. Tom Curtis:
    You say (comment 13)
    “..the reply I received from John Cook clearly indicated his continuing belief that he had posted the survey on SkS itself, rather than just as a tweet. That belief is mistaken, but in my opinion, sincerely so. I have no reason to believe he has actually gone through the wayback machine evidence to reach his own conclusion on that point. The upshot is, Barry Woods has no basis to claim that Lewandowsky or Cook know that the survey was not posted at SkS, as distinct from being merely informed as such.”
    Well, I told him that the Wayback machine indicated that he was in error back in September 2012. If he hasn’t “actually gone through the wayback machine evidence to reach his own conclusion”, as you believe, that can only mean that he doesn’t want to know the truth. So his “mistaken belief” can hardly be described as “sincere”.

    You say (comment 25)
    “The only relevant data on the public domain is that the number of responses reported at about the time of the Junk Science post was about 300 (from memory) less than that finally reported in the paper, which may be because the number reported in Melbourne was a round figure, or that there were additional responses including, potentially, from Junk Science.”
    Could you enlarge on this “relevant data on the public domain”?

    The Junkscience article is dated September 30, 2010 http://web.archive.org/web/20110710004658/http://junksciencearchive.com/sep10.html
    But Lewandowsky had already reported results at Monash University on September 23rd based on 1100+ respondents.

    It is quite possible, as you seem to be suggesting, that about 300 responses came from Junkscience. Why not? But in that case, what is one to make of Lewandowsky’s presentation at Monash, September 23rd, which, it was claimed, ws based on 1100+ responses?

    You say (comment 28) “You are trying to make a mountain out of a slag heap”.
    Well, there I can agree with you. Though I would employ a more vulgar term.
    The Lewandowsky “Moon Hoax” paper has multiple problems. No-one knows who answered the questionnaire, or why, or where they came from. Lewandowsky’s account in the paper of the source of his respondents is faulty on at least three counts: he claims to have responses from SkepticalScience which he clearly hasn’t, and he ignores the possibility of responses from Cook’s personal tweet and Junkscience.

    No-one knows who responded to the Moon Hoax paper three years ago because Lewandowsky won’t tell us. He could do so here and now, but he won’t.
  37. Well Goosey, indeed the motto of Rabett Run is droppings along the bunny trail, But certainly you and Woods are honking a lot about not very much. As TC said, no one is questioning your right to an opinion on the value of the study, and no one is required to take it very seriously.
  38. geoff chambers points out that no one knows the identity of those who responded. For anything that gets past a human subjects committee, no one could be able to know the identity of the respondents.
  39. Tom Curtis #25 says:
    “... the number of responses reported at about the time of the Junk Science post was about 300 (from memory) less than that finally reported in the paper, which may be because the number reported in Melbourne was a round figure, or that there were additional responses including, potentially, from Junk Science.”

    So the timeline we have now is as follows:
    28/29 Aug 2010: Seven “pro-science blogs post invitation to participate in survey. Plus Cook’s tweet to his thousand followers.
    28 Aug - 5 Sep: About 80 on-topic comments made on six blogs
    6 Sep: Hanich tries unsuccessfully to get the survey posted at sceptical blogs
    23 Sep: Hanich tries again
    23 Sep: Lewandowsky announces preliminary results based on about 1100 responses at Monash University
    30 Sep: Junkscience posts a link to the survey, at which time (according to Curtis) there are about 800 responses.
    Note that, according to Lewandowsky, there were about 1300 responses in total, but 200 were thrown out for nonsensical ages or for coming from the same address.
    It doesn’t make sense, does it? It’s quite possible that Junkscience article elicited a lot of responses. It was reported neutrally in the context of a science news roundup. But in that case, what was Lewandowsky doing reporting preliminary results in Monash based on a sample of 1100?

    Eli Rabett:
    Thanks for the correction. Rather than the “identity of respondents” I should have said “identity of the blogs which elicited respondents”. A nit well picked.
  40. Correction: The Junkscience link was on 24 September.

    Of course, I realise no-one gives a tinker’s whether the number of responses was 1300 or 1100 or 800 when Lewandowsky announced what were to become (two years later) the definitive results of a scientific paper which was featured in many of the world’s most prestigious journals, and is quoted by government ministers.

    I used to work in market research. An interviewer who suddenly produced 200 questionnaires without addresses would be sacked. A junior executive who couldn’t explain the source of his interviews would be sacked. A company director who wouldn’t answer a simple yes-or-no question about the sources of his information would be sacked.

    Professor Lewandowsky has received a five-figure sum and a medal from the Royal Society for his contribution to science.
Comments Policy

Post a Comment

You need to be logged in to post a comment. Login via the left margin or register a new account.