Comments on: Replication Policy

By: Kennedy, Editor of Science, on PBS « Climate Audit

Kennedy, Editor of Science, on PBS « Climate Audit — Sat, 04 Dec 2010 18:07:46 +0000

[...] under "Disclosure and Due Diligence". I draw your attention particularly to the post Replication Policy. That post did not draw as much response as some other posts. As a purported rebuttal, Dano asked: [...]

By: fFreddy

fFreddy — Wed, 31 May 2006 16:35:58 +0000

Steve, the link to the McCullough and Vinod paper is not working.

By: Ed Snack

Ed Snack — Sat, 10 Sep 2005 03:22:11 +0000

It would be useful, before making such requests Danzero, if you were to state your own position upon this matter. Do you yourself believe that replicability in a scientific paper is necessary or merely useful ?

Of course you could also put on your “useful shoes” and check those articles yourself and advise us on the results.

By: TCO

TCO — Fri, 09 Sep 2005 19:50:52 +0000

This type of behavior is a black eye for these scientists. Style point: I think you will get further by framing the argument as not meeting the practices of science (include social sciences) vice comparison to business (many academics are vaguelly anti-capitalistic). Also, I think you need to cite the long-standing policies for replicatability and preseervation and sharing of data (cf E Bright Wilson) vice characterizing this as some new thing. IOW, the issue is to hold people to common standards that have long been part of the ethic.

By: àƒ'?anàƒËœ

àƒ'?anàƒËœ — Fri, 09 Sep 2005 19:28:25 +0000

Steve,

How many of the 14 Reports in the latest ed of Science have sufficient archived material to permit replication? And the latest ed of Nature has 20 methodable articles/letters – how many of these have sufficient archived material to permit replication?

Thanks!

Best,

By: John Hekman

John Hekman — Fri, 09 Sep 2005 18:21:51 +0000

Really interesting stuff. I think that since some of these climate authors have not provided some of the necessary code and data, and since you have used the available data that is the same data they have claimed they used, then you have already shown that their conclusions are unsubstantiaed. Since you have made your data and code available, others can dispute your findings, but MBH98 etc. are unreliable until and unless they can demonstrate that their use of the data was valid and reliably produces their results.