Inaccurate Polling Data

[Wickwire]

I just read an article, published yesterday, Nov. 13, which was about Huckabee's sudden climb in the Iowa polls. There was no mention of Ron Paul. At the end of the article was the following statement:

"CBS News and The New York Times conducted telephone interviews with 1273 likely caucus-goers in Iowa November 2-11, 2007 and 719 likely primary voters in New Hampshire November 9-12, 2007."

I was wondering... how many of you actually have a land line? My only phone has been a cell phone for quite some time.

How many of you have ever been polled? I'm 27, and have never been polled in one of these surveys. I also believe that Ron Paul's message has redefined the typical "likely voter" this election cycle, given the fact that many of his supporters have never been involved in politics or supported a campaign before.

Given the outdated method of polling, I believe that they are missing a significant portion of the potential voter demographic, a large portion of which will be voting for Ron Paul. I would not be surprised if Ron Paul wins in Iowa and New Hampshire, further baffling the so-called "experts".

 

[noxagol]

I love how they claim that a sample size that is about .1%-1% of the numbers that will actually show up can be used to speak for the rest of the group.

 

[american.swan]

I love how they claim that a sample size that is about .1%-1% of the numbers that will actually show up can be used to speak for the rest of the group.

If you ever study statistics, sample sizes can be rather small. Like someone else said, 1500-2000 is enough. But the problem that we need to harp on isn't the 1500-2000 people, it's where they got the 1500-2000 people from. If you limit your demographic to X then Y isn't covered.

 

[Midnight77]

I'm 30, have a land line, and I have never been polled in my life, nor have my parents.

 

[american.swan]

Lets say we decide to do our own random poll.

We decide to use some common statistics and poll 2200 Iowans with land line phones.

First off 1000 of them probably don't care about politics.
600 won't vote in the primary.
That leaves us with 600 people. But only 300 of them are Republicans.
So now we have a potential sample size of 300 likely republican voters.