Saturday, December 19, 2009

It's Not Science to Slant Data

In the latest Science Saturday at bloggingheads, George Johnson tries to diminish the relevance of the climategate emails by saying that "everyone does it". As if, that's just locker room science talk.



There is no excuse for trying to hide the medieval warming period statistically, or diminish the credibility of critics. It isn't simple talk, and more than a white guy can say they use the N-word casually but that means nothing: everyone knows that someone who talks that way isn't credible on racial issues.

Talking about coalitions, and hiding conspicuous trends, is pure slanting data to one's predisposition. If someone shaded beta-return data from the 1970s because they were not relevant to the CAPM ,that would be conscious fraud. Luckily in economics we are all looking at the same data, so no one would even try to do that, but if they did, it would be a big, black stain. There is no good excuse for what they said in those emails.

No comments:

Post a Comment