August 12, 2003
From the desk of Jane Galt:You know how I've said that data from surveys is vastly less trustworthy than data from revealed preference? This is why.
Posted by Jane Galt at August 12, 2003 12:59 PM | TrackBack | Technorati inbound linksEven the "lie detector" test probably still shows bias--surely there are a signficicant number of women who (a) don't believe lie detectors work, or (b) sensed that they were being conned. In any event, to the extent that this receives wide publoicity, the technique can never be used again.
The problem of survey data is complicated by the facts that (1) people are probably particularly likely to lie about sex and (2) in sex research, for obvious reasons, surveys are usually all you get. I would suspect, to give an obvious example, that surveys would show that there is sufficient demand to support maybe 10 or 15 full-time prostitutes in New York City, whereas the police blotters alone would contradict that.
There are some objective measures of sexual activity. Rates of STDs or DNA testing of children's paternity. (I recall hearing a study that found 10% of childrens fathers were not who they thought they were.)
Jane,
I'm sure you would look the fool if you tried to debate that point with a talmudic scolar!
I'm sure you would look the fool if you tried to debate that point with a talmudic scolar!
Suddently these statements start appearing in the talkback posts, leading me to wonder -- Did I miss something? Like the auditions announcement for a village idiot? Or what the...?
And "scholar" is spelled with an "h." Sure, typos happen, but...
Nick the Dick,
I would trounce all Talmudic scholars, as I know the One True Way!
Captain Catholic
Comments are Closed.