mark olson
2006-10-20 12:01:35

The WSJ had an article on this ridiculous study. The biggest but not only weakness in his “study” was the use of far too few clusters in his sample; perhaps 1/5 as many as should have been used.