hedgehog
(1000+ posts)
Send PM |
Profile |
Ignore
|
Thu Sep-27-07 10:11 AM
Original message |
How many people who are being polled now will actually turn out for |
|
the caucuses or primaries?
|
DemocratSinceBirth
(1000+ posts)
Send PM |
Profile |
Ignore
|
Thu Sep-27-07 10:13 AM
Response to Original message |
1. The Final Polls Should Roughly Match The Primary Results |
|
Caucuses are much harder to call...
|
CTyankee
(1000+ posts)
Send PM |
Profile |
Ignore
|
Thu Sep-27-07 10:16 AM
Response to Original message |
2. If the poll asks people who self identify as "likely voters" the primaries |
|
votes should pretty much be the same later. I am on a monthly poll online and it always asks me to identify whether I am a likely voter.
|
hedgehog
(1000+ posts)
Send PM |
Profile |
Ignore
|
Thu Sep-27-07 10:30 AM
Response to Reply #2 |
4. Here's what I'm getting at. |
|
Someone says they 'll be a likely voter - after all , who wants to admit they have no idea an election is coming!
Now, the person is asked who they are voting for. How many people respond Clinton because that's the only name they recognize? How many of them will actually vote?
|
CTyankee
(1000+ posts)
Send PM |
Profile |
Ignore
|
Thu Sep-27-07 11:37 AM
Response to Reply #4 |
5. Well, a poll can't read minds. However, I think as you get closer to the election |
|
you get further away from people who only know the candidate from previous "name recognition." They ask "How likely is it thatyou will be voting in the next election" and give you multiple choice and I always choose "very likely."
Your margin of error is smaller the larger the poll and vice versa. So you'll get a more accurate reading with a poll of 1,000 very likely voters than a poll of 500. That's how the veracity of the poll works. It's a matter of mathematical probability, which is why those exit poll numbers in Ohio in the 04 election were questioned by mathematicians. Their MOEs were ridiculously high compared with the actual vote, so something had to be wrong with the "data," i.e. the actual votes cast.
I am not a mathematician, but I believe in polls, if they are scientifically sound.
|
hedgehog
(1000+ posts)
Send PM |
Profile |
Ignore
|
Thu Sep-27-07 12:39 PM
Response to Reply #5 |
6. There is some science and some art in poll taking. |
|
Also known as garbage in, garbage out.
Depending on who is asked and how the questions are phrased, the poll could be accurate or inaccurate.
|
CTyankee
(1000+ posts)
Send PM |
Profile |
Ignore
|
Thu Sep-27-07 03:12 PM
Response to Reply #6 |
7. Exactly, but it is in the best interests of candidates to get as close to the truth |
|
as possible for their own purposes, to gauge where they are. I'm talking about polls that take a large segment of the voters who self identify as "very likely" to vote and then asks them to choose which candidate they will vote for. That's pretty straightforward.
|
hedgehog
(1000+ posts)
Send PM |
Profile |
Ignore
|
Thu Sep-27-07 04:05 PM
Response to Reply #7 |
8. But if the candidate is using poll sto convince people that he or |
|
she is the winner and therefore they should donate money........
|
CTyankee
(1000+ posts)
Send PM |
Profile |
Ignore
|
Thu Sep-27-07 04:10 PM
Response to Reply #8 |
9. If their internal polling is saying something different than what they tell people |
|
then they are lying. I am talking about polls done by organizations outside of their own campaigns. If you doubt the poll's veracity you should try to see the methodology and how the questions are asked.
|
Tellurian
(1000+ posts)
Send PM |
Profile |
Ignore
|
Thu Sep-27-07 10:20 AM
Response to Original message |
3. Hillary in the lead in IOWA... |
|
Obama buried himself last night for ALL time! September 27, 2007
Strategic Vision Iowa poll:
Clinton 24%, Obama 21%, Edwards 22%, Richardson 13%
|
DU
AdBot (1000+ posts) |
Fri Apr 26th 2024, 02:10 AM
Response to Original message |