90% confident?

Apparently, the most recent summary (not the full report) by the IPCC's leaders states that there is a 90% confidence level that human activity is causing global warming. Such is the evidence that climatologist--er, columnist--Ellen Goodman is apparently willing to compare those who doubt the word of climatologists with Holocaust deniers. So perhaps it behooves us to explore what that 90% figure means.

For starters, Goodman is incorrect that 90% is about as good as science ever gets. If you doubt this, throw an apple up in the air, and I guarantee you that you will be more than 90% sure that Newtonian mechanics works. Turn on the radio, and you are more than 90% sure that Maxwell's electromagnetics is operative--not to mention semiconductor physics and some quantum physics to boot. In the testable areas of science, confidence levels exceeding 99% are routine.

But what does that 90% mean? Traditionally, a confidence level is the statistical likelihood that your test sample differs from a control. At this very point, the global warming hypothesis runs into a big problem; the whole earth is the test sample, and there is no control. Oops.

So we get to contrive a control, right? Well, the climatologists run into a problem there, too; none of their models have been validated against the historical record. So they can't exactly contrive a good control then, either.

So what is this 90% confidence level? Well, it's what Six Sigma trained statisticians like myself gently call a "rectal data extraction," pulling numbers out of one's nether regions. In other words, an elegantly crafted lie.

And, for what it's worth, I have ten times more CF bulbs than Ellen Goodman does, and I don't drive to the hardware store solely for the purpose of buying just one of them, either. In that way, my V8 powered pickup is being used far more efficiently than Ellen Goodman's Prius.

Gosh, does that mean that Ellen Goodman is denying global warming by her actions? I guess she'll enjoy getting to know her new friends.