SHORTER JOHN LOTT….Tim Lambert has another update on John Lott’s dubious grasp of arithmetic.
Tim’s latest post is another discussion of the weighting controversy, which, of course, is impossibly abstract for most of us. However, it’s also a crucial piece of evidence in the question of whether Lott actually performed the 1997 survey that has since vanished.
So, courtesy of the Calpundit Technical Research Department, here’s the Reader’s Digest version:
Method A (lots of little weights) is obviously bogus, but it does produce the number Lott claims.
Method B (a few big weights) is OK, but there is no way it could possibly produce the number Lott claims.
So did he actually perform the 1997 survey? You decide:
Yes, but he inexplicably decided to use a weighting method that no one in the field would have taken seriously once his data was released.
Yes, but he used a weighting method that can’t produce the number he says he got.
No, he never did the survey.
Only one of these choices strikes me as consistent with the real world. Can you guess which one?