Re: interpreting poll margin of error?

From: George Kahrimanis (anakreon_at_hol.gr)
Date: 10/23/04 

Date: 23 Oct 2004 05:21:44 -0700

I read with great interest the responses by Rich Ulrich and
Stan Brown. They both make good sense to me.

If the meaning of "3% margin of error" is that the number of
valid and pertinent responses is such that a fair coin would
tally between 47% and 53%, with confidence 95% ("two sigma"),
then we are talking about some number a little below 9000.
Not too high.

With 9000 replies, a candidate whose true portion is 3% will be
polled as between 2.6% and 3.4% (roughly) 95% of the time.
In that sense, the late polls worked well. (I still suspect
that some early polls were manipulated to *create* trends.)

I was amazed at the way all TV stations presented every
tiny change, as an indication of a trend! Obviously anyone
who would point out the flimsiness of the evidence -- in other
words, this is not big news, or even news at all -- was not
allowed to be heard on air.

Wrt the tacit 95% confidence level, I admit that due to my
background I had supposed a 68% CL. Moreover, I had unconsciously
presumed that the behavour of voters is too capricious to merit
a higher CL. I mean, even if you asked everyone, you would be far
from sure how the vote would turn out. Besides random waffling
and general trends, you must also know to some extent the playing
hand of each opponent and of each potential ally, and the skeletons
in everyone's closet, and you must be aware that the interviewees
may be lying or misleading in certain ways (as Rich mentions).

Well, it seems that I was too "pessimistic"! Unless what I have
described was a fluke, voters *can* be predictable, at least in
Greece. I only wish that such predictability would be derived
from rationality. Now, if voters are predictable but not rational,
oops... this is a subject for another discussion group.

~George Kahrimanis