Friday, September 28, 2012

McIntyre v Lewandowsky — Can we call in a statistician at UWA to help Lew?

By Jo Nova,
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?  Read More Here….

No comments: