re: #157 SixDegrees

Peer review has it’s limitations, but doesn’t seem to have broken in this case. First, when a paper is submitted for publication it is reviewed for basic adherence to established scientific principles, based on the procedures presented in the paper. The actual data is assumed to have been collected or obtained as stated, and no investigation is done on this point; it’s taken as a given. Once accepted, the paper is published for a wide audience of peers within the field. It’s at this stage that someone may decide to attempt to replicate the research, although that’s fairly rare unless what’s reported falls well outside expectations.

Still, it happens. What Walter is currently doing is peer reviewing the data from within his own area of expertise. So the peer review process appears to be working just fine.

The problem at this point seems to be the inability to accurately reconstruct the original data that was used to derive the filtered data that CRU obtained it’s published results from. This points to a problem with reproducibility - a hallmark of scientific process. It isn’t clear at this point whether the original raw dataset can be reproduced by independent researchers, or even by CRU itself. This is a real problem, and it apparently will require CRU’s help to resolve.

But I don’t think there’s anything wrong with the peer review process itself.

The only thing I can say to you on this, click on my name, go to my website and find my email and email me and we can take this issue off LGF. I think you would be very surprised on some things I could show you. There may be an issue with how CRU has handled peer-review and I can let you make your own decision on that.

If you have some throw-a-way email address you use, that would work, like gmail or something.

That’s up to you.