usanow
Veteran
offline
Registered: Jun 2004
Local time: 04:31 PM
Location: Houston
Posts: 403
|
| quote: |
Jim Nasium said this in post #15 :
it's pretty obvious you never had a university course in statistical sampling. the way you phrase it, the polling was done by a grade school kid asking questions to whoever would stop and answer. statistical sampling is done on a highly scientific basis and TAKES INTO ACCOUNT by the margin of error all of the types of things you raise questions about. the results are incorrect about as often as the us olympic hockey team upsets the soviet union team. it DOES happen, but it's highly unusual.
the pre-election polling showed ohio going to kerry. the exit polling showed the election going to kerry by a sizeable margin. something's wrong. believe it! |
It's pretty obvious you never had a university course in statistical sampling. Sayzak asked if they were polled on their way in, or way out. He also mentioned the lines were long at polling stations. I'm not exactly sure where he's going with the comment, but such variables are not factored into the margin of error as you state. Margin of error relates to the population in the subject area and the sample size.
If there is a problem with the sampling, then those problems are resolved via weighting after the data is collected. This adjusts for factors such as gender, race, age, etc. This process in effect normalizes the data so that it fairly represents the target population. Sampling errors are not simply managed by adjusting the margin of error as you stated.
As to your contention that Kerry was leading in Ohio in pre-election polls, that's not correct. Rasmussen and several others showed Bush ahead in Ohio. So if you're going to continue to perpetuate this lie that there is something fishy w/ the results, why don't you at least make your lies somewhat believable by finding supporting data that is grounded in some semblance of reality. Of course, you can't do that because Bush won fair and square.
|
11-19-2004 03:06 PM
