On Nov 24, 2004, at 7:06 PM, Dan Minette wrote:
Strippers at a local "high class"
strip club, had put together a charity fund, and then voted to have it
go
to groups of sexually abused children. Teri talked to them, and they
all
had been sexually abused as children.
Do you find this surprising? I mean, the group self-selected. Of course
they were all abuse victims. They set up a charity to *help* abuse
victims. But it doesn't follow that all strippers are abuse victims, or
even more likely to be abuse victims. Your sample size is *far* too
small and is *not* random.
That's not the scenario. The strippers at club XXX decided to donate a portion of their tip money to charity.
Yep, I see I missed that part of it. So here's the next question: Did these women know each other before they went to work at that club? Did they recommend it to each other as a place to work?
Suppose a group of CPAs put together a charity to assist victims of mugging. Would you be even remotely surprised to learn that most CPAs who participated in or gave to that group were themselves victims of mugging? Probably you wouldn't. But you would be very off center to suggest that this locally high correlation of CPAs to mugging victimhood meant that a "preponderance" or even "overwhelming" numbers of CPAs were victims of mugging.
But, if it was the CPAs at company XYZ who did that, then one would have to
ask why CPAs who are mugged all work at company XYZ.
Right, which is why I asked the question above. IOW, if those CPAs were all classmates or lived in the same neighborhood, it still wouldn't necessarily be remarkable that they were all mugged.
And of course you know you've still got a problem with sample size.
-- Warren Ockrassa, Publisher/Editor, nightwares Books http://books.nightwares.com/ Current work in progress "The Seven-Year Mirror" http://www.nightwares.com/books/ockrassa/Flat_Out.pdf
_______________________________________________ http://www.mccmedia.com/mailman/listinfo/brin-l
