dinsdag 24 november 2009

Climategate: Violating the Social Contract of Science

The scientific method only works when fellow researchers can implicitly trust the results offered by their colleagues.

November 22, 2009 - by Charlie Martin

On November 19, 2009, climate science was severely shaken by the release of a collection of email messages, together with a collection of data and data processing programs, that were alleged to have been stolen, or hacked, from the University of East Anglia Climatic Research Unit (CRU). (See here and here for previous Pajamas Media coverage.)

So what is this “climate science” of which we speak? Trimmed down to the essentials, what scientists really do comes down to these steps:

1.Look at something happening.
2.Think of a way to explain what’s happening.
3.Make a convincing case, based on evidence and experiment, that this is the best known explanation. Part of this “convincing case” is providing enough information so that a knowledgeable person could, if necessary, perform the same experiments and get the same results. (This should really include some weasel-wording about “within experimental error,” but that’s a technical detail. What’s important is that the knowledgeable third party can get close enough to the same results to satisfy that third part.)
4.Submit that convincing case to other knowledgeable people to review, in order to see if they also find it convincing. This is what is called peer review.
5.Publish that convincing case for the rest of the world, where the results can be seen, commented upon, and challenged.
6.Every so often, others perform the same experiments and confirm or question the results.

Step 4, peer review, is essential to this whole process.

Pajamas Media