Friday, October 06, 2006

Statistics & Derb   posted by Razib @ 10/06/2006 08:57:00 PM

Chris of Mixing Memory points me to two new statistics weblogs, Statistical Modeling, Causal Inference, and Social Science and Science Science Statistics Blog. Also, Chris tells me he will soon post a review of some papers on visual perception of race, so check his blog in a few days. Meanwhile, Derb has a new article in The New English review. He is in classic Derb form, as he states:
Furthermore, a dreadnought is not a hydrogen bomb. As puny as the MME states may be militarily, culturally, and economically, nukes are, like the Colt 45 in the Old West, equalizers. If, after a careful and judicious weighing of all the evidence, we were to conclude that there was no more than a ten percent chance of a terrorist nuke attack on a U.S. city over the next twenty years-that, in other words, the Sailer/Cochran hypothesis is probably correct, with probability at the 90 percent level-all the doom'n'gloom of Peters, Kurtz, the War Nerd, and me would still be fully justified.

Ten percent is alarming. Frankly, one percent is alarming. My own personal estimation is considerably lower, but what do I know? But, I do have a suspicion that if expectation of the probability of a nuclear attack on a U.S. city within the next 10 years is x, the likelihood that such an attack will occur in the Middle East is considerably higher, perhaps an order of magnitude greater than x.