"J. Williams" wrote:
> Much of what passes for "scientific" is back of the envelope statistics
> IMHO. Oh well, it sells papers and gives talking heads something to
> mull over.
Especially after the election, when they have to explain why the polls
didn't give the right answer. (Before, why my poll gives a different
answer than yours is easy: mine is right).
I recall having heard a pollster say on the radio that over half of people
contacted for polls about political candidates refuse to answer. This
means that polls depend on the (polite and helpful) answerers being the
same as the (ornery and cantankerous) non-answerers.
Doing my bit to keep the pundits guessing.
Jon Miller
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
