A Few Things Ill Considered

Is Steve McIntyre an expert statistician?

Posted by coby on September 21, 2012
(99)
More »

I am not afraid of admitting my own areas of ignorance.  The human body of knowledge is enormous and no one can possess all of it, or even be moderately familiar with all of it.  The only shame is in pretending otherwise.

This admission fundamentally shapes my personal approach to the whole Hockeystick/Dendrochronology/Michael Mann brouhaha, which continues to this day despite MBH98 having receded into the rather distant past, in scientific research terms. I have to rely more on networks of trust and take a more removed view of it all and generally park that paper and that famous graph in the “pending full acceptance” bucket.

Well, here comes an opportunity for me to refine my “who to believe” assessments of competing claims of statistical expertise.

Further to the continuing and still amusing “conspiracy to call them conspiracy theorists” episode involving Stephan Lewandowsky, he has a new post responding to some more “auditing” from Steve McIntyre.  His response, after an overview of what exploratory factor analysis (EFA) is, includes this passage:

Applied to the five “climate science” items, the first factor had an eigenvalue of 4.3, representing 86% of the variance. The second factor had an eigenvalue of only .30, representing a mere 6% of the variance. Factors are ordered by their eigenvalues, so all further factors represent even less variance.

Our EFA of the climate items thus provides clear evidence that a single factor is sufficient to represent the largest part of the variance in the five “climate science” items.  Moreover, adding further factors with eigenvalues < 1 is counterproductive because they represent less information than the original individual items. (Remember that all acknowledged standard criteria yield the same conclusions.)

Practically, this means that people’s responses to the five questions regarding climate science were so highly correlated that they reflect, to the largest part, variability on a single dimension, namely the acceptance or rejection of climate science. The remaining variance in individual items is most likely mere measurement error.

How could Mr. McIntyre fail to reproduce our EFA?

Simple: In contravention of normal practice, he forced the analysis to extract two factors. This is obvious in his R command line:

pc=factanal(lew[,1:6],factors=2)

In this and all other EFAs posted on Mr. McIntyre’s blog, the number of factors to be extracted was chosen by fiat and without justification.

Remember, the second factor in our EFA for the climate item had an eigenvalue much below 1, and hence its extraction is nonsensical. (As it is by all other criteria as well.)

But that’s not everything.

When more than one factor is extracted, researchers can rotate factors so that each factor represents a substantial, and approximately equal, part of the variance. In R, the default rotation method, which Mr. McIntyre did not overrule, is to use Varimax rotation, which forces the factors to be uncorrelated. As a result of rotation, the variance is split about evenly among the factors extracted.

This seems a pretty straightforward, black or white, right or wrong issue.  Is it? Is Stephan correct?  Is there any possible constructive response for McIntyre?  Statisticians please respond!

 

 

 

Keywords: , , ,
(99)
More »

Comments

  1. #1 TJR
    September 21, 2012

    The short answer is that an expert statistician would be unlikely to use Factor Analysis at all.

    FA is a method which has been largely developed and used in Psychology rather than Statistics. It does the same sort of thing as Principal Component Analysis but with more assumptions and more computational difficulty.

    Now, it may be that in some particular cases FA might be useful, but I’ve read several Psychology papers where the authors have just banged everything into FA (because everyone else does) when they would have learned far more from a few simple tables and scatterplots.

    One problem with FA is that the solutions are not nested, so that the first factor in the 2-factor solution is not the same as the factor in the 1-factor solution. Hence the objection to using a 2-factor solution when only the first eigenvalue is large seems reasonable.

    However, FA may well not be the best way to analyse these data in the first place. I’ve not looked at the work so can’t comment directly, but past experience suggests it probably won’t be.

  2. #2 AJ
    September 22, 2012

    Steve has already responded to this:

    http://climateaudit.org/2012/09/20/conspiracy-theorist-lewandowsky-tries-to-manufacture-doubt/

    “Lewandowsky’s results are bogus because of his reliance on fake and fraudulent data, not because of replication issues in his factor analysis.”

    Is Steve an “expert” statistician? I’ll leave that question to others, but I think Climate Audit’s influence speaks for itself. The hockeystick has been relegated to the dust heap of bad science. Stieg’s Antarctic temperature reconstruction has been refuted (improved upon?). The Gergis 1000yr Australian temperature reconstruction paper has been “delayed” and now Lew’s paper has been shown to be pure junk (but we already knew that due to poor surveying methods).

    Additionally, Steve has been a champion of better governance in the climate science field. I believe this has had a positive effect, but could still be improved upon.

  3. #3 Wow
    September 22, 2012

    “Is Steve an “expert” statistician? I’ll leave that question to others”

    How about the answer?

    Because it’s “No”.

    “The hockeystick has been relegated to the dust heap of bad science”

    Really? 14 national academies of science seem to disagree.

  4. #4 SanityP
    September 22, 2012

    Instead of bloviating silly comments, why don’t you produce a statistical analysys of your own and present it here for us all to to study? You obviously have much more information about the statistical methods Lewandowsky used than the rest of us combined. Looking forward to it.

  5. #5 Mike Mangan
    United States
    September 22, 2012

    Only the ideologically blinkered still cling to the Hockey Stick and, no, I’m not impressed that bureaucracies stand by it. Gravy train and all that. For great fun watch Richard Muller’s evisceration of the Stick ….

    http://www.youtube.com/watch?v=8BQpciw8suk

    Remember, the overwhelming consensus before Mann was that the MWP and LIA did indeed exist.

  6. #6 Carrick
    September 22, 2012

    Steve McIntyre has made it clear he is working with Roman Mureika on this problem. Roman is an “expert” by any definition.

    Wow:”Really? 14 national academies of science seem to disagree.”

    Regardless of how many appeals to authority you stack on teh head of a pin, it’s fundamentally wrong, his methodology was screwed up (noncentered PCA, failure to replicate via R2, deflation of variance) and most importantly completely at odd with modern, very-non-hockey-stick looking reconstructions including his own favored EIV method from 2008 (the latter agrees well with other modern reconstructions, in spite of more methodological problems).

    Figure.

    [ Can you believe that Lewandowsky used SV to determine the variance to use in his factor analysis? "Wow" is right. :-O ]

  7. #7 AJ
    September 22, 2012

    “How about the answer? Because it’s “No”.”

    What qualifications do you have to pass such an expert judgement?

    “Really? 14 national academies of science seem to disagree.”

    Then 14 national academies must not have looked at the data and methods. If your result can be essentially replicated using random data, that can’t be good. If the selection of you primary components is dependent on a couple of trees from a couple of denro chronologies or the upside down interpretation of contaminated data, then that really can’t be good. Robust my ass.

  8. #8 Wow
    September 22, 2012

    “Regardless of how many appeals to authority you stack on teh head of a pin”

    Do you realise what this is:

    “Steve McIntyre has made it clear he is working with Roman Mureika on this problem. Roman is an “expert” by any definition. ”

    That’s right.

    An appeal to authority.

    Once for McI. Once for Roman.

    So if 14 countries’ physics establishments is ignorable, how much more ignorable are two individuals?

    Wow, you boys certainly have your problems in maths. You think 2 is greater than 14!

  9. #9 Wow
    September 22, 2012

    “What qualifications do you have to pass such an expert judgement?”

    None, hence I’m not an authority and therefore I MUST be right!

  10. #10 Wow
    September 22, 2012

    “his methodology was screwed up (noncentered PCA,”

    Nowhere near as scewwed up as McIntyre’s attempt.

    And, in 2001, a different PCA was applied and the SAME RESULT was seen.

    Oh deary deary me.

  11. #11 Hank Bowers
    September 22, 2012

    To WoW: No, the mention that Roman M is an expert statistician was not an appeal to authority, it was related to the question whether or not Steve McIntyre is an expert statistician in the title of the article. The title of the article sort of asked whether the “science” in Lewandowsky’s paper is really questioned by experts (presumably if it isn’t then the author would suggest that we should pay no attention to the questioning). The answer to that is: yes, the science in Lewandowsky’s paper is really questioned by experts.

  12. #12 Amused
    September 22, 2012

    Wow,

    You disagree on Mr. McIntyre’s having any expert status, and appeal to an authority of consensus regarding the Hockey-stick.

    Carrick replies saying that Mr. Mureika is an expert as per *your* requirement, and thus in this particular case *you* cannot denigrate the statistics because of a lack of qualification. More importantly, Carrick says that it doesn’t matter anyway because the facts of the analysis trump anything else.

    You proceed to disregard the latter, and accuse Carrick of an appeal to authority and use that to justify your appeal to the authority of a collective of authorities!

    Wow…. indeed.

  13. #13 Wayne2
    September 22, 2012

    Well, Lewandowsky took an online survey of convenience that only included anti-skeptic websites yet claims to investigate skeptics. You can’t make a more basic mistake than that, so anything Lewandowsky says after that point is meaningless.

    Lewandowsky’s ideas about what eigenvalues to keep were outdated back in 1988 in the Psychological Bulletin, while McIntyre’s knowledge seems to be up-to-date. And let’s be honest, McIntyre made his living by his knowledge in the real world, while Lewandowsky has made his living in academia.

  14. #14 Lady in Red
    September 22, 2012

    Thinking — independent thinking — may be hard, but it is important. Even for non-experts.

    Here is Stephan Lewandowsky “on” Steve McIntyre:

    http://www.shapingtomorrowsworld.org/oberauerEFA.html

    Here is Steve McIntyre’s analysis of Lewandowsky’s work:

    http://climateaudit.org/2012/09/20/conspiracy-theorist-lewandowsky-tries-to-manufacture-doubt/

    (Working through the technical comments on both blogs is something of a slog, but important, I think, in tone, style, civility, information sharing, etc. Skip over the stuff that makes no technical sense)

    Here’s more about Stephan Lewandowsky, his feelings about “climate deniers” peer review, and much more. He recorded these short YouTube videos for the public:

    http://www.youtube.com/watch?v=s4GUMMx4sK8

    http://www.youtube.com/watch?v=C8wVfxoPqPA

    Then, stare at the ceiling, chew on a pencil stub and draw your own personal conclusions. Tomorrow, try it again, with some more little bits of insightful information.

    Soon, you will be riding a bicycle, full speed ahead.
    ….Lady in Red

  15. #15 Carrick
    September 22, 2012

    Wow:”That’s right.

    An appeal to authority.”

    Um no. Answering a question asked by the blog author.

    One that you got wrong, genius.

    LOL.

    “Nowhere near as scewwed up as McIntyre’s attempt.”

    “scewwed up”??? Lol. Great timing from the great thinker himself.

    Since the point is obvious wrt to centered PCAs, this just tells me that wow is another noob. All mouth, ad hominem, no knowledge.

    “And, in 2001, a different PCA was applied and the SAME RESULT was seen.”

    Not the same result actually, still a wrong answer because there were other problems (R2 failed to verify). More recent studies (Moberg 2005 and onwards, who your intellectually challenged hero Mann apparently never understood why it was a superior methodology to his own) clearly demonstrate that outside of the calibration period, MBH just selected for and reconstructed red noise.

    Simply because you fix one mistake doesn’t mean you’ve fixed them all (and the list for MBH is pretty long).

    The fact that Mann 2008 doesn’t agree with MBH98, and Mann endorses his EIV as a superior method to CPS should tell you something important.

    Also see comment about mixing SV/PCA and factor analysis and Lewandowsky, who doesn’t know the difference between, even though there’s 1/2 a million of us or so who do.

    Enjoy tripping on your tongue. That’s one thing you’re good out. Peace/out. I don’t waste my time with noobs with nothing to say.

  16. #16 Wow
    September 22, 2012

    “Um no. Answering a question asked by the blog author.”

    Um yes. With an APPEAL TO AUTHORITY.

    Oopsie doopsie.

  17. #17 Wow
    September 22, 2012

    “Since the point is obvious wrt to centered PCAs”

    Since McI selected only 30 measures to correlate to and those were heavily biased to self-correlation, yes, McI’s attempt was laghably incompetent.

    Therefore the answer to the question “Is Steve McIntyre an expert statistician” is “No”.

  18. #18 Wow
    September 22, 2012

    Go on, Steve, feast your eyes on this:

    http://www.cgd.ucar.edu/ccr/ammann/millennium/refs/Wahl_ClimChange2007.pdf

  19. #19 Wow
    September 22, 2012

    “The fact that Mann 2008 doesn’t agree with MBH98″

    Since MBH 2008 still has that hockey stick:

    http://www.ncdc.noaa.gov/paleo/pubs/mann2008/fig3.jpg

    what level of agreement do you count as “doesn’t agree”?

  20. #20 Wow
    September 22, 2012

    And you’re still trying to use R2? You’re no statistician, you’re a dumbass with an Excel spreadsheet and thinking “Pivot Tables” is a cool thing.

  21. #21 Morph
    September 22, 2012

    William Briggs – who is a statistician – already blogged about this paper a few days ago – he is not a fan

    http://wmbriggs.com/blog/?p=6164

    As for Mr McIntyre’s abilities – well, people considering committing to 7 and perhaps 8 or 9 figure investments in mining used to trust his figures, so I’ll go with the money.

    The Hockey Stick has, er, “issues” – it isn’t defended much these days. Mr Mc has blogged about the HS is correct but unless Prof. (“Robust Debate”) Mann is willing to debate openly with Mr Mc then I kind of trust Mr Mc more than him.

    One of them puts his money where his mouth is, one of them puts his mouth in front of the money.

  22. #22 Dan Olner
    September 22, 2012

    Coby, an excellent question and you’re right, it should be absolutely straightforward to get the answer you want (and I’d like to see too). I think we’ve just seen an illustration that online comments as they stand may not be a very good way of doing this. I think it might be better to actually try and find some independent people with credentials and ask them.

    Though there’s also the problem that any level of statistical expertise still leaves plenty of wiggle room for any sort of conclusion-drawing the expert wants. In Shumway’s time series textbook, he takes the opportunity to make some blindingly poor climate-skeptic points – despite, you know, having written a text book that’s clearly technically rich. The climate skeptic points fail some basic logic – which just goes to show, any level of technical expertise is no protection against stupidity.

    So there are two (three?) separate questions: which of the two parties are technically correct (since they can’t both be?) Consulting a small number of technical experts should answer that one quite quickly. Second, are they using the techniques and drawing conclusions in a logically consistent way.

    The third question’s most interesting to me: how does McIntyre choose what to analyse? He seems AFAIK never to have done much in the way of checking over Watts’ work, for example. Is that because it’s absolutely scientifically bomb-proof…?

  23. #23 Ian Forrester
    September 22, 2012

    Carrick is a liar:

    I don’t waste my time with noobs with nothing to say.

    Since you spend lots and lots of time on dishonest denier blogs the above statement is a lie.

  24. #24 Ian Forrester
    September 22, 2012

    Steve McIntyre is both incompetent and dishonest in his use of statistics (note I will not refer to him as a statistician since that would sully all others who are real statisticians). He was completely dishonest in his claim that using Mann’s techniques resulted in “hockey sticks from red noise”. What he, in fact, did was to generate 12,000 runs, throw out all which showed upside down hockey sticks (probably 50%), select 100 of the remainder which had the most hockey stick shape then post the best 12 of these. That is the complete opposite of what any competent and honest statistician would do. It is cherry picking at its very worst. Cherry picking is one of the more egregious forms of scientific misconduct.

    McIntyre brings to climate science all the worst characteristics of his primary occupation, mining and resource development. That area is full of dishonest scam artists and fraudsters. See for example Bre-X Mining, Cartaway Resources and Timbuktu Gold. McIntyre should have stayed in mining exploration where he would be quite at home with these scoundrels instead of becoming involved in climate science and bringing all those deceitful and dishonest tactics with him.

  25. #25 Wow
    September 22, 2012

    “As for Mr McIntyre’s abilities – well, people considering committing to 7 and perhaps 8 or 9 figure investments in mining used to trust his figures, so I’ll go with the money. ”

    Well, considering the welfare of the fossil fuel industry whose members pay McIntyre, I doubt he’d be welcom if he showed that AGW was real and a problem.

    You’re right: go with the money.

    McIntyre’s been bought off.

  26. #26 Wow
    September 22, 2012

    “The Hockey Stick has, er, “issues” – it isn’t defended much these days.”

    Neither is the spheroidal shape of the earth. I guess that means that the round earth theory has, er, “issues”, right?

    But no, the hockey still has maintained its validity despite a score or more of further more exhaustive trials. What you still hear about is MBH 1998 and how it has “issues” which, as it turned out, were of academic importance. The shape, however, was real.

    ‘sfunny how you never hear about McI’s paper which had so many problems in it the paper was abandoned except in the denialosphere.

  27. #27 Wow
    September 22, 2012

    “He seems AFAIK never to have done much in the way of checking over Watts’ work, for example. Is that because it’s absolutely scientifically bomb-proof…?”

    He never looked at Wegman’s work, either.

    But the answer to your question can be found at tamino’s blog site.

  28. #28 Wow
    September 22, 2012

    “Well, Lewandowsky took an online survey of convenience that only included anti-skeptic websites”

    Well, apart from the web sites like CA and BH and WUWT, but they “couldn’t find” the email asking them for their input.

  29. #29 Wow
    September 22, 2012

    “You disagree on Mr. McIntyre’s having any expert status, and appeal to an authority of consensus regarding the Hockey-stick.”

    I disagree that he has any claim to being an expert statistician in the same way as a priest who buggers little boys has no claim to being a man of god.

    If someone “as an expert statistician” makes the mistakes that McIntyre did in his paper “rebutting” MBH98, then they are not a statistician, let alone an expert one.

  30. #30 Wow
    September 22, 2012

    “the mention that Roman M is an expert statistician was not an appeal to authority,”

    No, it WAS an appeal to authority.

    Unless you’re going to say that there was no point in bringing his name up.

  31. #31 Vince Whirlwind
    September 22, 2012

    When McIntyre posts stuff like this on his blog,
    “Lewandowsky’s results are bogus because of his reliance on fake and fraudulent data”,
    you get a pretty clear picture of where he is coming from.

    *Obviously* McIntyre isn’t a trusted source. He’s just a guy who posts mostly nonsense on a blog to support an opinion is clearly not deride from any facts. Why anybody pays attention to him is a mystery, although his reaction to Lewandowsky has been pure GOld.

  32. #32 AJ
    September 22, 2012

    Sorry coby… I don’t think that esteemed statisticians will be showing up any time soon to answer your question.

    They’ve probably done the analysis and concluded that there is a very high correlation between blog posts on climate and trolls of all varieties. For example, on slashdot they should automatically categorize any climate related post as “troll”. Or maybe “trolling for trolls”.

    Good question though. My thoughts are that Climate Audit’s value is not necessarily with Steve alone, but more so with competency of the collective.

  33. #33 Wow
    September 22, 2012

    “there is a very high correlation between blog posts on climate and trolls of all varieties”

    Yes, you for example, turning up here, trolling away.

    CA has no value whatsoever execept to those who value their personal fortune over the future of their family.

  34. #34 AJ
    September 22, 2012

    Dan Olner… Watt’s doesn’t have an extensive publishing record. Also, I think that any ethical person would recuse themselves from reviewing a pal, otherwise they’d be accused of “pal review”. Maybe the climate science community could learn from his example.

  35. #35 AJ
    September 22, 2012

    coby… re: Trolls

    Wow… exhibit A.

    The rest of us (exempting TRJ)… exhibits Bee thru Zed ;-)

  36. #36 Neil Fisher
    Campbelltown, AU
    September 22, 2012

    “The third questions most interesting to me: how does McIntyre choose what to analyse? He seems AFAIK never to have done much in the way of checking over Watts work, for example. Is that because its absolutely scientifically bomb-proof?”

    This question has been asked many times – I’m surprised that most people who ask it don’t bother to use the search feature on his blog to find the answer.

    In short, what is said on blogs does not impact political policy, nor appear in the reports of UN IPCC where it does have such impact. That’s first. Secondly, it’s his blog, and he is (at least semi-) retired, so he posts about what interests him, as do most bloggers.

    If you bother to check the archives, you will find that his posts on the subject at hand are unusual, in that he does use the f-word. Given the number of times regular readers/posters at his blog appeared to believe he would be justified in using that word, he almost never has, EXCEPT, as in this case, with relation to DATA, NOT PEOPLE; METHODS, NOT MESSENGERS.

  37. #37 Wow
    September 22, 2012

    AJ, you’re a knobhead.

    Exhibit A is DEFINITELY kai.

    You are, likewise only here to propound denial of the facts because you’re

    a) An idiot
    b) Told that it’s all a commie/leftie/eco-hippie conspiracy to steal “your” money. I.e. an idiot useful to others.

  38. #38 Wow
    September 22, 2012

    AJ, Watts is no idol to hold up as an ideal.

    On interview, he slagged off BEST (despite earlier saying that no matter what the answer, he could accept the results BEST came up with. Turned out he lied) saying “It is not peer reviewed”.

    He then quotes answers from papers that WERE NOT PEER REVIEWED.

    And the answers he gave were in contradiction to a paper that WAS peer reviewed THAT HE WROTE.

    That’s right, a peer reviewed paper Watts was second author on said that there was no noticeable difference between badly sited recorders and well sited ones. BUT NOT EVEN HE agrees with the results, preferring to pretend it never happened.

  39. #39 Wow
    September 22, 2012

    No, Neil he gets his moderator who also doubles with a different account to use the f word liberally.

    The mafia don NEVER gets his hands dirty when whetting his beak.

  40. #40 Stewart
    September 23, 2012

    As someone who uses factor analysis/PCA, and has written on the topic, McIntyre’s work was not even a beginner’s error, it was much worse than that. The data shows one factor, there should have been no rotation, and any attempt to say otherwise is ludicrous. (my grounds for expertise? I’ve written 7 papers on factor analysis/PCA in psychological research, cited 250+ times, according to Google Scholar). He may be an expert in some areas, in this area he’s not even a beginner, and any attempt to claim otherwise just shows ignorance.

  41. #41 Dave
    United States
    September 23, 2012

    Wayne2>

    “Lewandowsky took an online survey of convenience that only included anti-skeptic websites yet claims to investigate skeptics. You can’t make a more basic mistake than that”

    Actually, one can – and Lewandowsky did. About the most basic principle of applied psychology is that you don’t investigate things about which you have a strong view, because, intentionally or not, you will skew your results. Any undergrad knows that. Lewandowsky has either committed deliberate fraud, or demonstrated that he is more incompetent than a typical undergrad in his own field – those are the only possible options. I think it’s pretty evident that he is in fact severely lacking in intelligence – does he have velcro on his shoes? I imagine tying laces is beyond him – so it’s unfair to conclude anything other than sheer idiocy has been at work here.

    For what it’s worth, you don’t need to be an expert of any kind to judge the Lew paper. Google up a GCSE science mark scheme for coursework – Lewandowsky’s paper wouldn’t even get a decent grade at high-school level.

  42. #42 Ian Forrester
    September 23, 2012

    Seems Dave is incapable of understanding what the thread is about. It is about scumball McIntyre who is so dishonest that I expect that even AGW deniers like you don’t like to discuss his antics.

    However, you just come on this thread and start spreading ad hominem comments about someone who you obviously know nothing about and in an area which you also don’t seem to understand. Typical denier troll behaviour.

  43. #43 Neil Fisher
    September 23, 2012

    WOW:

    Perhaps. And perhaps you are a sock-puppet for our host here – how would I know?

    This is beside the point anyway – I have no desire to enter into a slanging match about who is smarter etc. How about you just stick to the data and methods and defend them against SMs claims? Show the data, show the methods in all their full detail and show WHY he’s wrong instead of attacking the messenger!

    Honestly, if you feel SM (or anyone else for that matter) is striking below the belt, why lower yourself to that standard? If he’s so obviously wrong, why doesn’t anyone appear to be prepared to show exactly why and where he’s wrong instead of attacking the man himself? That’s what normally happens in science, isn’t it? It would be completely devastating to his case, wouldn’t it?

    That’s all a very serious set of questions, BTW. I remain stunned that such a simple and effective technique never seems to be employed. IMO, that is why such people get traction in this debate – to any interested outsider (such as myself), the whole thing reeks of a clique that simply wants to ignore the annoying outsider and hope he goes away. It comes across as childish and clique-ish crap rather than science, and frankly it looks really, really bad for the consensus side that they act this way, and it comes across as really, really concerning (to me) that no-one who is on the consensus side seems prepared to stand up and say “you know what, the guy has a point – there IS a minor mistake here, we should correct it”. Not even ONCE in more than a decade of following this debate have I seen that, even when outsiders with no axe to grind found in SMs favour. How can anyone who claims to hold science in high esteem defend such behaviour?

  44. #44 Sou
    September 23, 2012

    @ Neil Fisher – you have been looking in the wrong places.

    It’s telling that you are not more concerned about the appalling behaviour of science deniers than you are about the reluctance of scientists, whose reputations are daily being attacked by same, not thanking them for their efforts to reject science.

    On topic, sort of – McIntyre might or might not know general stats but he is not at all familiar with EFA. In any case, as his multiple posts demonstrate, he doesn’t give a damn about statistics in this context. Nor about cognitive science. He and others of his kind are using the paper as a distraction from the record melt in the Arctic – and an excuse to trash climate science and foment false rage (particularly among the extreme right wing ideologues and conspiracy theorists who reject science).

    The latest suggestion on CA (including by Watts, who recently demonstrated he can’t do arithmetic let alone advanced stats) is to harass the DVC-Research at UWA and someone else suggested harassing the ARC. (I don’t think their onslaught will get much response other than:- Lewandowsky et al might be interested in this additional swathe of evidence if they are planning follow up research.)

  45. #45 Pete
    September 23, 2012

    Not that I am an expert but there are some pretty solid looking publish papers being linked at CA that support SMs position. I haven’t seen the same from the other side yet.

  46. #46 Turboblocke
    September 23, 2012

    Wayne2>

    “Lewandowsky took an online survey of convenience that only included anti-skeptic websites yet claims to investigate skeptics. You can’t make a more basic mistake than that”

    If true, (which it’s not as he did contact “skeptic” websites) what makes you think that the anti science mob don’t look at nor comment on the mainstream science/reality sites?

    Actually the sample might be biased in that from what I’ve seen, the anti-science commentators on the mainstream science/reality sites seem more rational than those who don’t leave the confines of the “skeptic” websites. Mind you that does set the bar pretty low.

  47. #47 Wow
    September 23, 2012

    “This is beside the point anyway – I have no desire to enter into a slanging match about who is smarter etc”

    Except that is PRECISELY what you came here to do.

    Except now that you’ve found you’re a jackass and quite a bit dumber than people who try to educate themselves, you’re pretending otherwise.

    McIntyre is sold body and soul to the fossil fuel industry. And someone who will knowingly lie in their sphere of expertiese (if indeed we take for the sake of argument McI is an expert) has thrown away their expertise and no longer can be classed as such.

    Doctors who go round killing patients are no longer called doctors and are stripped of their certification if they bring their profession into disrepute.

    McIntyre has done that.

  48. #48 Wow
    September 23, 2012

    ” Show the data, show the methods in all their full detail and show WHY he’s wrong instead of attacking the messenger! ”

    Since this paper is pretending to show how MBH98 are wrong, how about you see whether M&M are right?

    Given there are papers out that destroy the laughable attempt at M&M to recreate a hockey stick (e.g by selecting out of hundreds of random walk simulations those half dozen that look a bit like hockey sticks, but not very much), why don’t you prove M&M right?

    Or is this all about someone else do the work that has been done scores of times so that they can’t do anything else?

  50. #50 Wow
    September 23, 2012

    But if all you’re going to say is “M&M did that”, then you need to prove them right because this paper:

    http://www.cgd.ucar.edu/ccr/ammann/millennium/refs/Wahl_ClimChange2007.pdf

    gets the same answer with different methodologies.

  51. #51 coby
    September 23, 2012

    Neil Fisher:
    ” If he’s so obviously wrong, why doesn’t anyone appear to be prepared to show exactly why and where he’s wrong instead of attacking the man himself? ”

    I believe that would be this bit here:

    “In contravention of normal practice, he forced the analysis to extract two factors. This is obvious in his R command line:

    pc=factanal(lew[,1:6],factors=2)

    In this and all other EFAs posted on Mr. McIntyre’s blog, the number of factors to be extracted was chosen by fiat and without justification.

    Remember, the second factor in our EFA for the climate item had an eigenvalue much below 1, and hence its extraction is nonsensical. (As it is by all other criteria as well.)”

    Do you have a defense?

  52. #52 Sou
    September 23, 2012

    A person who goes by the handle of A Scott did a mangled copy of the Lewandowsky survey on WUWT. I spotted on CA that McIntyre looks like being the one to analyse it.

    He said he has to work out how Lewandowsky et al crunched their numbers first. Going by the lack of progress so far, it will be a very long time before the A Scott survey sees the light of day – if it ever does.

    I also notice that McIntyre seems unable to cope unless he’s told exactly how to do the sums. He’s pointing out that A Scott’s survey had a different Likert scale (and from the sound of it a bad one – not just adding a neutral response, which is arguably not the way to go for this study, but reportedly having the ‘neutral’ response double as a ‘don’t know’ response.) McI is complaining that it makes it difficult to compare the two.

    While it’s not an identical survey instrument, and the mangled scale will work to dilute relationships, I can’t see why a competent person couldn’t work out associations between ‘skepticism’, right wing ideology, conspiracy ideation, and the other constructs that were in the original research. (If the results are similar it adds support to the original work. If not it doesn’t . ) This discomfort with a differently designed instrument, to my mind adds support the suggestion that McI is not adept at stats for sociological and psych studies.

  53. #53 Stewart
    September 23, 2012

    A couple of points, based on the recent discussion.
    1. There are some AGW deniers who read science-based websites (see above). Failing other options, some would participate in such a survey (I remember reading the complaints by some on the blogs).
    2. Self-proclaimed statistcal experts may have pockets of expertise, and willingness to attempt statistics, but no-one can be expected ot acquire techniques flawlessly without supervision. SM is no exception, his use of PCA betrays lack of any practical knowledge, and hence, when he tries ot do something new, he is forced to beg for help from the same people he accuses of incompetencfe or dishonetsy (note that dishonesty is the only mortal sin in science). I know PCA well, but if I had to use SEM, I also know the statistician I would be collaborating with for a month or two, until I got up to speed. McIntyre is in southern Ontario, correct? Why hasn’t he contacted statisticians at McMaster, Guelph, Toronto, etc., wwho can help him understand the techniques he wishes to use and critique, so he does nto embarrass himself?

  54. #54 Sou
    September 24, 2012

    The thing that struck me most about McI’s approach was he was going about everything back to front. He seemed to look inside R for a formula that would ‘fit’ the answer that Lewandowski et al provided, rather than work out what EFA is designed to do, why someone would use it as part of an analysis – and then figuring out how to apply it in the context of this particular study.

    (From the little I’ve read, R may not be the best program for EFA – although I understand there are modules that can be added.)

    Until/unless he gets a better conceptual understanding of EFA and SEM then I don’t think he’ll get very far.

    He might get further if he took a step back and thought about the research itself – and maybe even took a different approach altogether to the analysis if he thinks it’s warranted. I’m not sure he is capable of that sort of holistic thinking.

  55. #55 Wow
    September 24, 2012

    Sou, like with his search for a pink-noise graph that looks like a hockey stick, he starts with wanting an answer and keeps looking until he finds something that produces the answer he wants, then stops.

  56. #56 Sou
    September 24, 2012

    Wow, the main question is, of course, why he is so obsessed with this study. It’s far from being the first to obtain these sort of findings (eg Kahan et al and others). It’s not climate science per se. Seems far removed from his area of interest (unless you include slagging off any scientist who accepts AGW).

    Another question is why is he assuming the data is ‘faked’, and even if a few of the responses were ‘scammed’, he or at least some others, seems to be of the view that means the study itself is flawed – which doesn’t make sense. (He offers no substantive reason for either assertion other than he thinks so.)

    Surveys have been conducted since the year dot (well, maybe I exaggerate a tad). They form quite a big part of psych research in general in one form or another. Does he dismiss all surveys? (Maybe it’s a prelude to dismissing the results of A Scott’s survey of WUWT deniers – lol.)

  57. #57 Wow
    September 24, 2012

    “Seems far removed from his area of interest (unless you include slagging off any scientist who accepts AGW).”

    You just answered your question, sou…

  58. #58 Sou
    September 24, 2012

    Maybe that’s a only secondary reason, Wow. It could be that he wants to raise his flagging profile and attract people back to his blog again. This research seemed to get more traction in the MSM and on denier blogs than other similar research. (He even lowered himself to the extent of writing a piece for WUWT on the subject, which probably boosted Watts’ ego.)

    McI’s been losing relevance after the past couple of years of weather disasters in the USA and the horribly dramatic Arctic summer. He hasn’t done anything new for ages. (I imagine most of his potential audience is US based.)

    I doubt it will give him much more than a short term bounce.

  59. #59 Jesse Jones
    Florida
    September 24, 2012

    I would like to extend to you personal invitation to check my feature length documentary on the “cold facts” of Global Warming entitled “The Boy Who Cried Warming,” available in full at http://www.theboywhocriedwarming.com. The virtual premier has been enjoyed by over 12,000 viewers due to a grassroots campaign effort of handing out flyers and emailing people just like you! We are independent filmmakers without corporate sponsorship, every view counts to us, and we would truly appreciate if you would take a look and (if you enjoy the film) encourage others to check it out. The list of websites mentioning our film growing, and we would be honored if you would join the growing list distinguished sites below:

    “The Boy Who Cried Warming” has enjoyed recommendations from:

    Watts Up With That?
    Examiner.com
    Digging in the Clay
    Bishop Hill
    Junk Science
    Climate Depot
    No Trick Zone
    Before it’s News
    Climate Change Dispatch
    Climate Ponderings
    Jammie Wearing Fools
    Oh What Now
    SCEF.org.uk
    Tom Nelson

    And the list keeps on growing… PLEASE feel free to Google the name to check out the comments, and as always, enjoy the show!

    Jesse Jones
    Producer/Writer “The Boy Who Cried Warming”

  60. #60 Wow
    September 24, 2012

    No need. The reason for people flocking to his blog is to engender support for the fossil fuel industry’s PR.

    The way things are now, the only ones left are the converted who will never shift until they can find someone else to blame for their destitution.

    His need now (or, rather, the need of his employers) is to maintain doubt for the use of the US senate bought and paid for members.

    If they can keep the USA doing fuck all about this, they

    a) ruin Obama
    b) stop everyone else from doing anything substantive about it

    the latter being because there are similar greedy fuckwits who will go “Oh, we can’t hamstring ourselves in this global market, so we can’t go until the USA goes first…”.

    All these idiots want is to keep the scam going until they can retire, leaving the mess for the next sucker to clean up.

    Rather like the rapine of companies in hostile takovers, really. Or, indeed, the stock market itself.

  61. #61 kai
    September 24, 2012

    wow and fellow co2 hysterics and climate church fellows

    fossil fuel industry and similar bla bla bla from corrupted postnormal climate science is insanely strange. dagw is the biggest public manipulation effort by leftist and green activists around the planet to attract people’s awareness in order to satisfy ego narcisstic neurotic profiling dreams of desperate individuals from the hysteric climate church

  62. #62 kai
    September 24, 2012

    wow, you are extremely poorly informed, unfortunately.

    you should know that there is no anthropogenic warming. what you will never understand, because of missing formation and knowledge in the subject, is that the climate always changes with or without you and your bad co2.

    your level of scientific understanding is sooooooo low that i have to go to the toilet now for six hours of continous vomiting

  63. #63 Wow
    September 24, 2012

    Jessie, it was hilarious.

    Though I don’t think that was your intent.

    Not since Lord Mockton last got up in front of the fat men in the Heartland Institute lectures was such a pile of self-serving garbace made of not merely half-truths but outright rubbish has been noted in public before.

    The last time something like this was seen outside the denialosphere was a youtube clip of a little boy with chocolate all round his mouth, being asked by his mother “did you eat the chocolate cake” to be replied with angelic dissemblance “No”. Not once, but several times.

  64. #64 mandas
    September 24, 2012

    Lots of discussion here in a very short period of time. Unfortunately, I think both Coby and all the other posters are missing the point.

    It doesn’t matter whether or not Lewandowsky or McIntyre are right with their own interpretation of the statistics. Firstly, because whether or not deniers are also more likely to believe in conspiracy theories does not change the science one iota. The science is strong, and humans are changing the climate because of emissions of GHG. Of that there is no doubt.

    The whole point of this is that Lewandowsky produced a paper of very minor importance and the deniersphere went into paroxysms of indignation about it. It is so funny to watch – watts, mcintyre, nova and all the usual crowd are frothing at the mouth with rage and calling Lewandowsky every name they can think of. Lewandowsky statistical analysis may be flawed (or it may not be) because of biased samples etc, but he now has a huge sample of comments to draw on. And they are as funny as hell!

    The deniers have made Lewandowsky’s case for him with their comments and indignation. After years of calling scientists ‘liars’ and ‘scammers’, and calling AGW a hoax, the deniers are getting a taste of their own medicine, and they don’t like it. The irony and humour in this whole situation is wonderful to witness.

    My recommendation – stop worrying about whether or not the paper and the analysis is any good. Grab a bucket of popcorn and sit back and watch the fun.

  65. #65 mandas
    September 24, 2012

    And thank you Stewart, for the most rational comments on the questions posed by Coby – even though I have my own take on the issue.

  66. #66 Sou
    September 24, 2012

    @ Mandas, it’s true that whatever people believe the science doesn’t change. What people believe does affect what is/can be done about the situation. (Outliers and cranks aren’t important – it’s the middle group that makes a difference.)

    A better understanding of why people reject facts will hopefully provide some insights into how to communicate science and options for policy response, and enable democracies to come up with workable strategies to deal with the awful situation we’ve got ourselves into

  67. #67 mandas
    September 24, 2012

    Sou

    I take your point about gaining insights so we can better develop policy responses. However, I think that discussions to that end often start out with the naive assumption that it is possible – if we improve our method of communication – that we will end up convincing all but the ‘outliers and cranks’. Unfortunately, I don’t believe that is possible.

    To explain, there are plenty of examples around that demonstrate that people will reject even the most convincing arguments because of culture or ideology, or because they just don’t like the answer. Things like UFOs, the moon landings, 9/11 truthers, Kennedy assassination et al are easy to point to and suggest that it is only cranks that believe this sort of conspiracy nonsense. But there are many, many otherwise logical and rational people out there who totally reject evidence because it does not fit with their preconceived worldview, and the vast majority of them will never change their minds. At one end of the spectrum of evidence rejecters are young Earth creationists and evolution deniers, but there is also a large percentage of the population that still believes in god despite the complete lack of evidence for such a view. How do you communicate science and options for a policy response to such people when they are not interested in facts and evidence – and never will be?

    So I am going to stick to my originally stated view. I don’t think Lewandowsky is interested in contributing anything of significance to the issue of communicating science. I just think he is having a bit of fun with deniers. He gave the hornet’s nest a very gently poke, and got a very typical, frothy-mouthed, hypocritical response from people who have no sense of humour about their ideology

    And the evidence for that is on his site, here:

    http://www.shapingtomorrowsworld.org/lewandowskyVersionGate.html

    And some of the comments as follows:

    “But most of all I’d like to thank our skeptic friends for providing the material in undoubtedly the most enjoyable thread ever in the climate blog wars. To quote the psychiatrist from Fawlty Towers, “there’s enough material in here for an entire conference”

    “Oh God, the poor souls just can’t help themselves. Do they have any idea how completely ridiculous and desperate they look? This thread suggests no.”

  68. #68 Sou
    September 24, 2012

    We’ll have to differ on that score, mandas. There is ample evidence that an understanding of motivation/behaviour can help design strategies that are more effective and more properly, change behaviour. (This is at the heart of the fear of the ultra-right wing ideologues IMO. They are scared of and opposed to ‘social engineering’ – including what others might think of as being a responsible member of society.)

    Examples include encouraging people to quit smoking, drink alcohol in moderation, use public transport, recycle waste, stop littering, keep fit, use sun protection etc etc.

    By the way – there’s nothing to say one cannot get some enjoyment from seeing one’s research confirmed in multiple ways :) Prof Lewandowsky’s writings and research show that his professional interest is in cognitive science plus he cares strongly about the adverse impacts of global warming. This study fits both. I don’t believe it was a flippant piece of research, nor do I think it was a seminal study (and I doubt he does either).

  69. #69 Vince Whirlwind
    September 24, 2012

    I don’t know why you lump in the Kennedy assassination in with UFOs, and in turn UFOs in with 9/11 and moon-landing hoax nonsense.

    What’s wrong with UFOs? The possibility of undocumented aircraft flying around isn’t monstrously low, and the possibility that they are extra-terrestrial in origin might be even smaller, but I think it’s reasonable to keep an open mind about the idea.

    As for Kennedy….the motives are pretty clear, and even the guy they fingered for it (although he conveniently never made it to a lawyer let alone a trial) was a CIA employee. Shades of Thomas a Beckett there, at the very least…

  70. #70 Wow
    September 24, 2012

    I don’t think deniers are “getting a taste of their own medicine”. For decades the only recourse for AGW deniers is conspiracy theory. They’ve learned only that now, for a generation they’ve relied upon it.

    At most, their learned reaction is being taken advantage of and documented for future generations and the resume of psychologists.

    And it isn’t “the middle” that are important. Light cavalry attack a larger forces’ flanks, removing the numbers bit by bit until the core is alone and irrelevant as a force.

    And here the outliers are being shown how DUMB their compatriots are. The most extreme on the rational side will leave as the evidence of the nutbar nature of the core becomes obvious.

    Their departure will leave a gap that will be mostly filled by the less self-aware fringe, now the “rational fringe” of the remaining group.

    After a while, more hardcore people will leave, either realising they want to be on “the winning side” or just because they’re in the core because all the people they know were there or nearabouts, and followed them in. Well, now they’re following some out.

  71. #71 Sou
    September 24, 2012

    Yes, after I wrote that about ‘the middle’ I had further thoughts. For example, it’s the outliers (at both ends of the spectrum) that influence what I believe is referred to as the Overton window. Extremists can pull the middle – and the majority – in one direction or the other.

    In the USA it’s the cranks to the right of the spectrum who appear to have influenced lawmakers – pulling both Republicans and Democrats towards extremist right wing views.

    Shining a spotlight on the craziness of this extremist nonsense, the irrational basis for rejecting science (I don’t accept facts because I don’t want to pay tax), plus the stark evidence of the damage the shifting climate can wreak (Arctic, weather disasters around the world etc) will hopefully wake enough people up to the importance and urgency of the situation we’re in.

  72. #72 Wow
    September 24, 2012

    Part of the problem is you’re thinking of a straight line from “us” to “them”.

    This is why the US political system is fucked up.

    Think of a flat plane of spacetime with black holes and massive objects sitting on it.

    You can be extremely IN on point and that may constitute astronomical density.

    But you can’t go round drawing lines on there and saying “between these two points is “the middle”".

    The absolute simplest you should go is to imagine a triangle with you on one vertex and “the two extremes” on the other two vertices.

    From each one’s point of view, THEY are the moderate.

    Imaginging a line and two extremes and a middle where, for some unexplained reason, “the truth” should be is not only unhelpful but actually damaging.

  73. #73 Sou
    September 24, 2012

    Wow – we’re probably talking past each other. The downside of short comments on a blog :(

    You don’t need to convince me of what you are saying. I am most certainly not of the view that ‘truth’ must be a compromise middle position on some straight line spectrum..

    My comments were in the context of the ability or otherwise of a government to develop and implement sensible policy, given the tendency of elected officials to pander to what they perceive of as popular opinion (or the opinion of their donors, which is not necessarily the same thing) in order to get donations or to be re-elected.

  74. #74 Wow
    September 24, 2012

    Well, it’s a common phrasing of the problem and is, moreover, the underlying process analogised in the Overton Window.

    And quite unhelpful unless you have a specific implementation in mind to paint two ends wrong and an intermediate “middle ground” as best.

    You can’t compromise on the value of Pi without recognising the effect of that error.

  75. #75 Sou
    September 24, 2012

    We’ve moved way off the topic here, Wow, but I’ll just make this observation.

    I’m thinking you are talking science fact (ie “truth” or the closest we can get to it given the information available at the time) while I am talking about policy response.

    Policy development and more especially implementation is almost always about compromise, in a democracy at least, but also in many other forms of government (eg the party-based system in China). (The Overton window as I understand it is a convenient way of describing a cluster of public opinions in terms of acceptable policy responses to a range of issues. Eg when I was in the USA in the 1970s, wearing a seatbelt was considered by many people I met to be an infringement of civil liberties. I don’t know if attitudes have changed since – or if any states have since passed legislation making seat belts compulsory. IMO there is no ‘right or wrong’ or ‘truth’ in any particular policy response in that instance. The facts are there but what you do with the facts can include a range of options.)

    There is more than one way to skin a cat, as they say.

    However, before good policy can be agreed upon, decision-makers need to agree/accept the ‘facts’ of the situation.

    Apologies, Coby.

    McI has published another post, but hasn’t yet shown any sign he’s learnt about statistics appropriate to cognitive science.

  76. #76 Wow
    September 24, 2012

    It doesn’t have to be scientific truth.

    If I say I want to kill you, you say you don’t want me to kill you, we compromise on me cutting your legs off.

    Compromise?

    The political scene (as typified by the illustrations of the Overton Window) relies far too much on some assertion of only one dimension, left to right, and the extremes ALWAYS being wrong and, though it doesn’t specifically say so, the median point being “the truth”.

    All that means is that the nutters then go extreme, asking for the moon, stars and all the creatures therein, knowing that the median point will be moved to “compromise” on the “true middle ground”.

    It’s how the overton window moves.

    And it is enabled by this meme that there is a single dimension and extremes are at either end.

    Its an infinite plane, with each mass being a position to take. And all the people in orbit around them, occasionally moving off to orbit another nearby idea to go back again when disturbed from that orbit.

    There is no extreme end.

    There’s an orbit never moving away from a point, and that’s the only extreme.

  77. #77 Sou
    September 24, 2012

    Yeah, well that wasn’t my understanding of either the meaning or context of the term, Wow. But you could well be correct. It’s an american term AFAIK – a different culture to my own (and to some extent, speaks a different language :D ) so I’m not the expert here.

    In any case, I hope my post clarified what I was talking about, regardless of the definition of the Overton window.

  78. #78 Dan Olner
    September 25, 2012

    Seconding mandas, Stewart’s is the only reply so far that fits the original ask. Coby, I wonder if it might be worth re-doing the thread, this time specifically asking for people competent at the specific stats and moderating to only allow them to post? (So that would exclude my post above as well!)

    I think that’s probably the only way a thread like this can function: laying out moderation policy to start with and sticking to it. Nowt wrong with having a specific goal for a thread, and this one has only seen the one reply from Stewart that fits.

  79. #79 Wow
    September 25, 2012

    Coby would be far far better off asking specifically statisticians.

    E.g. see if Lewandowski has some spare time or check Tamino’s blog, post there to Tamino.

    But this post, if it really IS about questioning actual bona fide statisticians, has built in fail.

  80. #80 Wow
    September 25, 2012

    PS Coby already has a response from a real statistician, included in the thread topic.

    Hence, since the claim of error has not been refuted, the mechanism may be opaque, but the reasoning for requiring a high binding count (i.e. if you’re going to pick a synthetic value, it should correlate with many other genuine values from which that synthetic one is derived) is pretty clear and incontrovertible and therefore the criticism of McI’s work stands.

    About all that remains is whether McI is an expert statistician.

    Since, as I’ve said before, someone who will prostitute their expertise to pretend an incorrect result has given up their claim to expertise, this answers that question.

  81. #81 Sou
    September 26, 2012

    Seems to me the more McI opens his mouth (so to speak) the more he puts his foot in it. If EFA is a ‘popular technique’, how come there are so few experts blogging on these threads? If ‘EFA assumes normality’ means it can only work with normal distributions (as McI seems to be saying), why can it be used with non-normal data (which is not uncommon from what I read)? Why does McI keep calling ‘fraud’, which is a very strong word even if any responses were an attempt at ‘gaming’ – of which there’s no evidence.

    Maybe Stewart will pop in again and give his thoughts.

    (One of the authors on LOG12 specialises in multivariate design, factor analysis, SEM and psychometrics and has worked in the private sector – validating proprietary instruments and developing psychometric inventories – as well as academia. I’d hazard a guess he knows a bit more about these techniques than McI.)

  82. #82 Dan Olner
    September 26, 2012

    I thought PCA type analyses did ask for normal data. I’m sure many people use it anyway, but that’s the case with a mass of statistics. Non-parametric stats is a pain in the ass as I recall so many people seem to go lalalala when it comes to checking for normality. Doesn’t make it right, but that might be why you see so much of it Sou.

  83. #83 Wow
    September 26, 2012

    The paper talked about here was EFA where you try to find a pseudo factor (made up of a combination of other factors) that can be used to replace those other factors to reduce the number of degrees of freedom that you have to run a component analysis (or any correlation analysis) on, making the job a darn sight easier.

    And though you’re probably right in saying “that might be why you see so much of this”, the point is that you shouldn’t be seeing it from somone claiming (or being hailed as) an expert.

    Proctologists don’t go round claiming expertise in cardiology, though they may have done some work on that in their medical degree.

  84. #84 Stewart
    September 26, 2012

    EFA/PCA doesn’t require normal data, that’s uneccessary because the statistical tests aren’t particularly widely-used, and results are pretty robust (see Buja & Eyuboglu, ‘Remarks on Parallel Analysis’, 1992).
    No statistician is an expert in all fields – and an expert knows where their expertise ends. Bu tthe mistakes McI made are not even novice mistakes, as could be made by someoine unfamiliar with the field picking up a textbook and fumbling. A scientist or statistician would start by wondering where they went wrong, and then work through it, or contact the author with their questions. A real expert knows where to turn for help. Uri Simonsohn has identified actual fraud in science, and puts contacting a statistical advisor as a major step.

  85. #85 Sou
    September 26, 2012

    Thanks, Stewart.

    One good thing – it’s re-awakened a dormant interest in stats and thinking about how I might improve some of my work – and develop some other ideas I’ve been mulling over. Might be time for me to go back to school :D

  86. #86 Neil Fisher
    October 13, 2012

    “Do you have a defense?”

    Do I? No, but SM does – that the method was insufficiently and or incorrectly described to allow him to KNOW what was done, so he had to GUESS.

    From what he says, the number of factors used was determined by using what the authors now claim is “normal practice” in the discipline. OK, fair enough – yet for them to pillory SM instead of pointing out the REAL reason for the choice is somewhat odd to me. If you would be so kind as to point out where the author has made such a point, I will happily concede – if you cannot, will you do the same? I would hope so.

    To be frank, the math and stats are over my head (at least in the detail – I will flatter myself to say that I understand the concepts if not the minutia), however SM lays out what he did so that the interested (and competent) can follow and duplicate his work. It is a pity that the original author cannot make the same claim, isn’t it?

  87. #87 kai
    October 13, 2012

    wow, why are you everywhere so stupid: “Proctologists don’t go round claiming expertise in cardiology, though they may have done some work on that in their medical degree”

    you should said:
    “climate scientists go round claiming expertise in atmospheric physics, though they may have done some work on that in their former jobs as lawyers, judges, teachers, partisan green politicians, opera singers, blue collars, investment bankers, housewifes, etc, etc.”

    wow, you should really work on your intellectual capabilities and missing knowledge

  88. #88 Vince Whirlwind
    October 13, 2012

    Neil, why on earth would anybody want to “duplicate” Steve McIntyre’s work?

    Why on earth would “the original author” be expected to facilitate anybody wanting to “duplicate his work”?

    I think you’re a bit confused about the difference between ignorant amateur-hour blog-driving and academic work.

  89. #89 Wow
    October 13, 2012

    “SM does – that the method was insufficiently and or incorrectly described to allow him to KNOW what was done”

    That, however, seems due to his incompetence at statistics.

  90. #90 Marco
    October 13, 2012

    Neil Fisher, it is quite common that methodology in a scientific paper is insufficiently clear for those who are not experts in the field. I would almost say it is standard. I remember my first steps into the field in which I am now considered an expert: frustration, errors, and much more, simply because many steps are considered trivial by the experts.

    Rule number 1 when failing to replicate a study that is outside your area of expertise is to assume you did something wrong. McIntyre’s rule appears to be that if he cannot replicate something, the OTHERS did something wrong.

    Also, if you write multiple posts in which you accuse a group of scientists of scientific misconduct (both explicit and implicit, as McIntyre did), don’t expect that same group to be very willing to help you.

    It was quite funny to see the Auditor being audited…and found wanting. Not for the first time, it should be noted. I remember when deepclimate showed that McIntyre’s code mined for the 1% of most extreme hockeystick shapes. Something not described in the method section of his paper. Oddly (well, actually not), none of his regulars called him out on that omission when it was exposed.

  91. #91 Neil Fisher
    November 3, 2012

    “Neil Fisher, it is quite common that methodology in a scientific paper is insufficiently clear for those who are not experts in the field. I would almost say it is standard.”

    Sure – and when you started, did the experts tell you you were stupid, ignorant, etc etc, or did they point out the error of your ways, show you where to learn etc etc? The latter, right? And if you DID find a mistake, did they complain bitterly that you were still an idiot, or did they say something like “Hey! You’re tight – let’s see what that means…”? Did they fail to correct the mistake even when it didn’t “matter” to the conclusions, or did they “fix” the error? Yeah – exactly, they fixed it, even when it “didn’t matter”, right? Contrast to Phil Jones’ actions – reprehensible in comparison, aren’t they? Who is anti-science here?

    I have followed this as an interested observer for over a decade and I have to say that the more bullying, obscuration, failure the correct and failure to explain I have seen – and frankly, the more lies – the worse it looks for the establishment/consensus side. Doesn’t mean they are wrong, but it sure is suggestive that they cannot be trusted to be objective (NOTE: NOT “cannot be trusted”, but “cannot be trusted to be objective” – an all too common failure, myself no doubt included).

  92. #92 Marco
    November 3, 2012

    Neil Fisher, when I started my science career the experts told me to first get up to speed with the basic literature, ask questions to those close to me, and then try and add my knowledge. Unlike Steve McIntyre I did not go in with the rule “I am right, others are wrong, and this is the way it must be”. I have seen very few cases where the experts went in and “fixed” any errors, especially when those supposed errors have little impact and when the field has moved on years ago (as in: no longer using that procedure).

    I’m not sure what actions of Phil Jones you are referring to, but if it is the non-sharing of data, I think someone needs to stay away from Climateaudit for a while and work in a lab that has to deal with proprietary information. I have, and under absolutely no circumstance will you ever get access to that data from me. You can go to the actual data owners, and if they decline, well, tough luck for you! It’s interesting you don’t see the FoI requests sent out through Climateaudit are not bullying (why don’t they go to the actual data owners? Oh, that’s right, no interest in doing actual work).

    Oh, and has Steve McIntyre already sent a correction to GRL for not disclosing in his methodology that his method to show the supposed hockeystick from red noise included an explicit selection of some examples of the 1% of most extreme shapes? And that even then the hockeystick shape was still *much* smaller than the actual signal observed in MBH? No? Gee, and somehow you trust this guy.

  93. #93 Richard Simons
    November 3, 2012

    when you started, did the experts tell you you were stupid, ignorant, etc etc, or did they point out the error of your ways, show you where to learn etc etc? The latter, right?

    I don’t imagine that Marco’s first reaction was to loudly proclaim that all the experts were wrong and either incompetent or frauds.

  94. #94 Mal Adapted
    November 4, 2012

    Neil Fisher:

    I have followed this as an interested observer for over a decade and I have to say that the more bullying, obscuration, failure the correct and failure to explain I have seen – and frankly, the more lies – the worse it looks for the establishment/consensus side. Doesn’t mean they are wrong, but it sure is suggestive that they cannot be trusted to be objective (NOTE: NOT “cannot be trusted”, but “cannot be trusted to be objective” – an all too common failure, myself no doubt included).

    This is one of the aspects of professional scientific culture that’s hardest for the lay public to understand: why are so many scientists jerks?The best answer I’ve seen is here. Summary:

    “Science doesn’t work despite scientists being asses. Science works, to at least some extent, because scientists are asses. “

    I trained to the doctoral level for a career in science, before I found an easier way to make a living. One of the reasons I dropped out was that I couldn’t take the heat, but I didn’t reject the science because of that. My classmates who stayed are still jerks competitive personalities, but they’re now leading the field by virtue of the originality, power and fruitfulness of their contributions, which have held up to the unsparing scrutiny of their equally competitive peers. It’s how science is done.

  95. #95 Mike Haseler
    Lenzie
    December 5, 2012

    Mal Adapted. The question is not whether there is human nature in science. It is whether there are robust systems to prevent that human nature distorting science and leading to the level of certainty required of research to carry the title “science”.

    Your argument is that because people are human, it’s perfectly OK for referees to favour one side, for policemen to accept bribes for climate researchers to run propaganda attack sites aimed at pushing their own politics.

    None of these are acceptable. They might all be a result of human nature, and you can’t blame us for being human. But you can blame the systems and processes in science that have failed to maintain the standards that are required for the public to have any trust.

    Let’s put it this way. Would you be happy for Lewandowski to be tried before a jury of ordinary human beings … who just happen to be selected from those posting at WUWT?

    They are only human? They are only doing what humans do? What on earth could possibly be wrong with having a totally biased jury that is bound to hang draw and quarter this jerk?

    So, why do you think it acceptable to have a jury of climate nutters pretending to be scientists deciding what the world should do on the climate?

    And to be quite honest, given your answer is it any surprise the public do not trust scientists who as you say are a bunch of jerks?

  96. #96 citizenschallenge
    Southwest USA
    March 30, 2013

    Morph says: “As for Mr McIntyre’s abilities – well, people considering committing to 7 and perhaps 8 or 9 figure investments in mining used to trust his figures, so I’ll go with the money.”

    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    I’m confusing – convincing investor to sink there money into various mines. . . . . . . has what to do with understanding the complexities of our global heat distribution engine, {you know climate}???

    It does have a lot to say about his ability to play with and present numbers and also the finer arts of sweet talking investors. . .

    And that has what to do with understanding the complexities of our climate???

    Cheers
    http://whatsupwiththatwatts.blogspot.com

  97. #97 citizenschallenge
    Southwest USA
    March 30, 2013

    Morph says: “As for Mr McIntyre’s abilities – well, people considering committing to 7 and perhaps 8 or 9 figure investments in mining used to trust his figures, so I’ll go with the money.”

    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    I’m confusing – convincing investors to sink their money into various mines. . . . . . . has what to do with understanding the complexities of our global heat distribution engine, {you know climate}???

    It does have a lot to say about his ability to play with and present numbers and also the finer arts of sweet talking investors. . .

    And that has what to do with understanding the complexities of our climate???

    Cheers
    http://whatsupwiththatwatts.blogspot.com

  98. #98 freddy
    March 31, 2013

    @citizenschallenge, I fiddled through your junk site and found only stupid “anti-denier” religious fervour from a further warm climate hypcrite?

    Are you willing to disclose your political climate agenda? Be honest and tell us here that you are a green activist who hates people like Whatts, McIntyre, Lindzen, Singer, etc. etc.

  99. #99 Wow
    March 31, 2013

    Yes, the definition of a site is whether it agrees with you, right, kai?