Tuesday, December 01, 2009

Data Deniers update

Well we now have word that Hadley CRU "lost" the original data and only has the "value added (i.e. quality controlled and homogenized) data".

Notice again the use of words which are themselves lies: "value added". No value is added to raw data by homogenizing it. While discarding outliers may be necessary to the construction of reasonable statistical models, value is not thereby added to the data. Sometimes outliers are hints at unguessed at phenomena, things that should be measured to be included in a better statistical model. The other homogenization process that comes to mind, averaging, simply loses information. Again it is useful, but it does not add value to the data from which the averages were constructed.

Given what has been revealed about the code used in the CRU's analyses, complete with "fudge factors", one wonders what "quality control" means in this case, one wonders and suspects nothing good.

And, from down under, we learn that not only New Zealand but Australia is afflicted with the tendency for its data deniers (a.k.a "climate researchers") to publish climbing temperature graphs that have little resemblance to the relatively flat raw data on which they are "based".


Post a Comment

Subscribe to Post Comments [Atom]

Links to this post:

Create a Link

<< Home