Mistake not to vote for Hillary as nominee.

[CARLHAYDEN]

Giuliani is not possible as the Republican nominee because he is unacceptable to the National Rifle Association and because he's entirely out of the American mainstream on social issues like abortion and gay marriage.

The majority of Americans are pro-choice.

http://www.cnn.com/ELECTION/2004/pages/results/states/US/P/00/epolls.0.html

55% think abortion should be always or mostly legal.

That poll only asked 14,000 people, not really a fair representation of the 200,000,000+ people in the country.

14,000 people is statistically a massive sample for a poll. If we were only so lucky as to have more polls with that. It makes for a very small margin of error, and is among voters. You couldn't ask for a better poll, really.

First of all, the poll was of 13,660 people, not 14,000.

Second, exit polls are increasingly unreliable as a large percentage of the voters cast their ballots  by other means (just how many voters in Oregon were sampled).

Third, beyond a certain point, the size of samples becomes pretty irrelevant.

In conclusion, yes you can ask for a better poll.  Really!

[Alcon]

Giuliani is not possible as the Republican nominee because he is unacceptable to the National Rifle Association and because he's entirely out of the American mainstream on social issues like abortion and gay marriage.

The majority of Americans are pro-choice.

http://www.cnn.com/ELECTION/2004/pages/results/states/US/P/00/epolls.0.html

55% think abortion should be always or mostly legal.

That poll only asked 14,000 people, not really a fair representation of the 200,000,000+ people in the country.

14,000 people is statistically a massive sample for a poll. If we were only so lucky as to have more polls with that. It makes for a very small margin of error, and is among voters. You couldn't ask for a better poll, really.

First of all, the poll was of 13,660 people, not 14,000.

Second, exit polls are increasingly unreliable as a large percentage of the voters cast their ballots  by other means (just how many voters in Oregon were sampled).

Third, beyond a certain point, the size of samples becomes pretty irrelevant.

In conclusion, yes you can ask for a better poll.  Really!

I was reading someone else's post who said it was 14,000. I am not very knowledgable about mathematics, but is there a significant difference in the reliability of a 13,660-person poll versus a 14,000 one?

This may be true, but were there actually any late exit polls that were significantly wrong? All the state ones I saw were within 1 or 2 percent of the actual result, at the high range. Most nailed it.

[CARLHAYDEN]

First of all, there is NO statistically significant difference between a poll  of 13,660 and a poll of 14,000,  Just thought I would correct one of the many errors stated about that 'poll.'

Second, I have not had time to go over the exit polls for 2004 to compare exit polls with actual results due to a death in the family (its amazing how long it takes to settle an estate).

Third, to cite just one example of defective exit polls, I cite California (1982 Gubenatorial election) in which Bradley won the exit poll and lost the election (the percentage of voters casting ballots other than at the polls has dramatically increased over the past twenty two years).

Fourth, you interpreted the poll results by including sixty per cent of the respondents who favor restrictions (but not prohibition) on abortion as being pro=choice, when they made it clear that they do NOT favor the abortion on demand agenda.

[Alcon]

First of all, there is NO statistically significant difference between a poll  of 13,660 and a poll of 14,000,  Just thought I would correct one of the many errors stated about that 'poll.'

Second, I have not had time to go over the exit polls for 2004 to compare exit polls with actual results due to a death in the family (its amazing how long it takes to settle an estate).

Third, to cite just one example of defective exit polls, I cite California (1982 Gubenatorial election) in which Bradley won the exit poll and lost the election (the percentage of voters casting ballots other than at the polls has dramatically increased over the past twenty two years).

Fourth, you interpreted the poll results by including sixty per cent of the respondents who favor restrictions (but not prohibition) on abortion as being pro=choice, when they made it clear that they do NOT favor the abortion on demand agenda.

I am sorry to hear about the death in your family. That can be very difficult.

Although exit polls can be wrong, they are not wrong that frequently - this year they were fairly spot-on and, in all due respect, they probably have improved since the 1982 Gubernatorial election in California.

In your fourth point, I think you are confusing me with someone else. I have not said anything about the content of the poll (regardless, I am pro-life.)