The Lewandowsky view is Drilling into noise. The McIntyre response: Lewandowsky’s Fake Correlation
My favourite Lewandowsky line is: “We cannot get into the details here…”
McIntyre can and does in gory depth. He posts the equations, the code, the tables, everything. He graphs the residuals, and shows the “severe non-normality” of them. He tests the correlation and finds that the two most obvious fake responses heavily affect the results:
“Lewandowsky is absolutely off-base in his assertion that the examination of outliers is inappropriate statistical analysis. In fact, exactly the opposite is the case: proper statistical analysis REQUIRES the examination of outliers.”
“One can readily see that the two super-scammers (889, 963) contribute essentially 100% (over 100%) actually of the negative correlation between CauseHIV and CYMoon in this calculation.”
Lewandowsky says: “no one who has toyed with our data has thus far exhibited any knowledge of the crucial notion of a latent construct or latent variable.”
McIntyre replies: “Principal components, a frequent topic at this blog, are a form of latent variable analysis.”
As a former graduate of UWA, this is embarrassing. Does UWA not teach and use rigorous statistical methods? Is there no one who can help him?
Plus, when will that “in press” paper be published?
Lewandowsky’s paper was in press as of July 27th, when the Guardian announced its results. But it doesn’t seem to have been published in the September edition of Psychological Science. Nor is it mentioned in the “early releases”. Stan points out most of the September stories were first published in late July. It may mean nothing (a delay of a month), or it may mean the paper is being rewritten, or possibly presages a silent “withdrawal”? Certainly skiphil found a comment by Lewandowsky that suggests the moonlanding paper was being “extended” and was not quite the complete and settled science it was presented as being at The Guardian by Adam Corner, and The Telegraph too. h/t to Stan, Barry, Wayne and Skilhil in comments at CA.
64. Stephan Lewandowsky at 22:04 PM on 14 September, 2012
Questions continue to be raised for further information relating to this paper. My response is threefold:
1. I see little merit in treading over ground that is already clearly stated in the paper (e.g., the elimination of duplicate IP numbers).
2. Several questions concern material that is presently subject to an FOI request. I will let that process run to completion rather than pre-empt it.
3. The supplementary online material for the article is being extended to contain additional information (e.g., the outlier analysis from the preceding post). The online supplement will be released when the typesetting of the article is complete.
Time permitting, I may also write another post or two on topics relating to this paper that are of general interest.
This Friday it will be eight weeks since The Guardian article. In this modern era where anyone can self-publish a book in a day on their home computer, it does seem odd that Psychological Science needs nearly 2 months to typeset an article.
Note point 2 also: No Stephan, no one cares if you “preempt the FOI” — there is no penalty for releasing information that is public property. As a public servant and a scientist(?) the emails, the data and the methods belong to all Australians. Sure, redact the private details, but no one should have to FOI those answers in the first place. That you use the FOI as an excuse to delay providing the answers you owe the public sends a message about your dedication to the honest process of discovery and your conscientious duty as a man who is supposed to serve the public. If you had a clear conscience, and were proud of your work, you’d be only too happy to help people understand your careful responsible impartial dependable work, right?
Lewandowsky’s forgotten warning about computer models
I searched the entire Psychological Science site and found only a few older references to Lewandowsky.
Notable was this one from 1993 where Stephan warns us about models.
The Rewards and Hazards of Computer Simulations
- Stephan Lewandowsky
Psychological Science, July 1993; vol. 4, 4: pp. 236-243.
Does Stephan Lewandowsky realize he hangs his entire career not on fundamental properties of the greenhouse theory, but on not-so-pragmatic-choices made by modelers about feedbacks?
He certainly must believe his own PR — that the deniers are dog-stupid flawed brains who deny the evidence, otherwise he would not have published material which was such a lay-down-misere-gift for for his targets to shred. Could it be that he’s realizing that he’s playing statistical games against full-time statisticians instead of Arts grads with “training”? I don’t think so. It takes a certain acumen to know that.
REFERENCE:
Lewandowsky, S., Oberauer, K., & Gignac, C. E. (in press). NASA faked the moon landing—therefore (climate) science is a hoax: An anatomy of the motivated rejection of science.. Psychological Science.
————————————–
My posts on this topic:
PART I Lewandowsky – Shows “skeptics” are nutters by asking alarmists to fill out survey
PART II 10 conspiracy theorists makes a moon landing paper for Stephan Lewandowsky (Part II) PLUS all 40 questions
PART III here Lewandowsky hopes we meant “Conspiracy” but we mean “Incompetence”
PART IV Steve McIntyre finds Lewandowsky’s paper is a “landmark of junk science”
PART V Lewandowsky does “science” by taunts and attempted parody instead of answering questions
PART VI Lewandowsky gets $1.7m of taxpayer funds to denigrate people who disagree with him
Part VII Lewandowksy, Oberauer, Gignac – Is the paper bad enough to make history?