Monday, June 30, 2008

Women's study statistics

Previous BS installment here.
Statistics are like a bikini. What they present is suggestive, but what they conceal is vital--Aaron Levenstein
Megan McArdle writes:
But many of the worst statistics come out of women's study and feminist advocacy. There are the appallingly shaky statistics on the number of rapes based on badly designed surveys manipulated with statistical methods so crude that Bayes must be spinning in his grave fast enough to power a high-speed monorail between New York and LA. The confident assertions that abortions have not increased significantly since legalization, when the pre-Roe figure is obviously unknowable, and the data we do have--on pregnancies--points firmly in the other direction. The various numbers on domestic violence that are thrown around with abandon even though a moment's thought is enough to dismiss them as ridiculous--the infamous Super Bowl claims being only the worst of the breed. And, of course, the silly assertion that we know how many women are helped with guns vs. hurt by them, when the data needed to decide such a claim are unavailable, and the coding problems enough to make Jesus weep.

No comments:

Post a Comment