On Fri, 05 Jan 2001 20:38:07 GMT, [EMAIL PROTECTED] (J. Williams)
wrote:
< snip, SS about "data quality" >
> Another corollary issue is the media calling Florida for a candidate
> based on common exit poll data prior to the closing of the polling
> stations due to differing time zones within the state. Needless to
> say, some citizens from the western part of the state may not have
> bothered voting once the electronic media has consecrated a "winner."
< snip, stuff about several things >
?
What is your corollary issue? I don't see that you name one ... I can
suggest "a minor question" or two that exists, but they are not
important statistical questions.
What fraction of the vote takes place in the last hour?
What fraction of voters arrive in the last 10 minutes?
If a "winner" is announced 10 minutes before the polls close,
how many people who are on-queue, or headed there,
will (a) hear about the result, and (b) decide not to vote.
And (c) Does that new decision depend on which candidate
the voter favored, and whether that one was favored?
Here is a statistical question, or maybe it would just be helpful
to have statistics to confirm my guess. I think that a very large
majority of the undervotes (and probably, the overvotes) were
cast by first-time voters. I know that I read comments about
how many first-time voters there were, in various places -
- Were these reports totally subjective ones, from election day?
- or does someone ever have access to such records?
- and publish some information like this?
--
Rich Ulrich, [EMAIL PROTECTED]
http://www.pitt.edu/~wpilib/index.html
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
