[EM] advocacy by means of exit polls

[RLSuter at aol.com]

It makes no sense in this case to define "wrong" as anything other
than a wrong count of how people actually voted. Therefore, the
official results of an election are just as capable of being wrong
as exit poll results.

As for your claim that the margin of discrepancy is "closely
comparable between exit polling and phone polling," that is
such an improbable sounding claim that you are going to have
to provide some pretty convincing documentation to convince
me of it. Exit polling is widely understood as much more accurate
than phone polling. It is so reliable that in many countries where
paper ballots are used, news organizations use exit polls when
reporting results of elections and except in very close elections
the vote counts that often come much later rarely contradict the
exit poll results.

But if you have good evidence to support your claim that phone
polls are as accurate as exit polls, I'm certainly willing to take a
look at it.

-Ralph Suter

<< > At 03:28 PM 8/27/2006, Michael Poole wrote:
 >>There are well-known cases where exit polls get the margins wrong
 >>(e.g. Phillipines 2004) or even the results wrong (e.g. USA 2004).
 >
 > I'd hardly call USA 2004 a "well-known case" of this. As Mr. Suter
 > pointed out, serious controversy still exists regarding that
 > situation. In the absence of some kind of verification, we really
 > can't know for sure which occurred: error due to faulty polling, error
 > due to actual deviation between voter actions and voter poll answers,
 > or error due to vote fraud or other tabulation problems.
 
 Since you and Mr. Suter clearly share some confusion over what "wrong"
 meant, let me clarify it to mean any divergence from the official poll
 results.
 
 Whether the difference is intentional or accidental, or which counting
 of public preference is less representative, is absolutely irrelevant
 when it comes to determining what different results other methods
 might reach: some discrepancy exists between the collected data and
 the official results, and (ceteris paribus) this "margin of
 discrepancy" is closely comparable between exit polling and phone
 polling.
 
 By which I meant to point out that, to get results as useful as you
 are likely to ever get, you do not necessarily need to convince exit
 pollsters to do things your way.  Phone pollsters could work as well.
 
 Michael Poole >>