On Sat, 18 Nov 2000 22:49:41 -0500, Bob Wheeler <[EMAIL PROTECTED]>
wrote:
> Undervotes are represented by vote cards which
> have no punches for a presidential candidate -- at
> least none that were detected by tabulating
> machines. It is reported that there are some 28000
> of these in the three contentious Florida
> counties.
>
> Because these counties seem to have favored Gore
> more than Bush (by 69%,54% and 64%), the Democrats
> believe that manual inspection of these cards will
> result in votes for Gore. The Republicans fear
> this. Is there any statistical justification for
> these views?
- Here is what I posted on the 12th, where I suggested your V was
small for some others, but about .5 for Palm Beach County....
"With 10,000 no-punches where only half that many no-votes should be
expected (in Palm Beach County), they re-counted a 1% sample and came
up with 47 additional votes -- about half of the 100 or so that were
possible, and consistent with the number of no-votes that typically
are seen. There was no report of how many no-punches had existed.
Gore gained, as he was expected to, because Gore carried the county by
almost 2-to-1. "
That's why I thought the re-count was a slam-dunk for Gore, and
(therefore) why the sec't'y of state had to oppose it to salvage a
Bush victory.
>
> There are two possibilities: (A) a vote was
> attempted; (B) no vote was attempted. Let us
> assume that for (A) the probability of a Gore vote
> is 0.62, and 0.50 for (B), then if V is the
> proportion of the 28000 corresponding to (A), one
> has the mean vote for Gore being
> 28000(0.62V+0.50(1-V)), and similarly one can
> obtain a SD. One finds:
> V Delta
> 0 0
> 0.1 336
> 0.2 692
> 0.3 1008
> where Delta is the incremental number of votes for
> Gore. Since the SD in all cases is about 84 votes,
> it may be seen that V must be in the neighborhood
> of 0.3 before Gore will pick up enough votes to
> overcome the 960 vote lead of Bush.
>
> Is this likely? Early results from the counts seem
> to suggest it is not.
--
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/
=================================================================
