I said below that Geoff Nunberg “fudged his numbers.” That wasn’t nice – I should have said that Boyd was unable to replicate the results of Nunberg’s survey. Nunberg engages in spin, but that’s not the same as outright lying (except about NOW, where he seems to be involved in wishful thinking.) Civility is very important.
I tend to give people more credit for knowing what they’re doing than others might, I’ve learned. When Richard Peterson examined the data Lenore Weitzman collected on post-divorce incomes and found that they didn’t support her argument that women suffer a 73% decline in standard of living whle men increased theirs by 42% (in reality, Weitzman’s figures showed that including tax effects, both had the same decline, about 10%), I said she lied and he said she erred. Take your pick.
Nunberg’s study is suspect on several methodoligical grounds, especially the small sample size and the short distance between noun and adjective he requires. It would seem that he had a point to prove, and chose a set of variables that appeared to prove it; that’s spin.