02/13/03: 9:50 Cause of the Columbia crash
I'm going to put on my statisticians hat. We statisticians have more than one hat. The one that works most often in the physical world is a Bell curve.
Other hats are more conceptual: the Poisson hat, the Students' hat and the Chi Square hat. The whole point of this blog is derived from the NYTimes report this morning that said a solar phenomenon hit the earth at exactly the same time as the space shuttle Columbia was re-entering the earth's atmosphere.
The phenomenon doesn't appear to have a name: "It is a disturbance, a discontinuity, and it did deliver a punch," according to Dr. Devrie S. Intriligator. (Have you ever seen a name like that?) She describes "the phenomenon as a sort of gigantic wave of electrically charged particles, magnetic fields and radiation that was moving toward Earth at roughly 400 miles a second."
According to the rest of the article, these phenomenon seem to occur several times a year.
As a statistician, I can state unequivocally that this "phenomenon" caused the Columbia disaster. The calculation is simple: something that happens "several times a year," or any number less frequent than weekly, that occurred at the exact same time as an equally rare event, a shuttle re-entering the earth's atmosphere, must be the cause of the disaster. It simply must be the cause of the disaster. There could be other problems and system weaknesses in the shuttle, there always are in "accidents," but the cause was this 400 mile per second storm hitting the earth. Near statistical certainty.
Of course the NY Times found someone to make sure that no real human thinks like a statistician: "Spiros Antiochos, an astrophysicist at the Naval Research Laboratory, said there was a 'very low probability' that a storm in space played a role in the Columbia's demise."
Thank you, Spiro ... you read your statistical table upside down.