So is this true? About the polls . . .

[boondoggle]

This is my second time today making a thread that's centric to a Youtube video, but I promise it's the last one. :)

I'm on a youtube spree today and I ran into this, but I don't know if it's the truth or not. Check it:
http://uk.youtube.com/watch?v=lDIiQcCyuxU&feature=related

[RoyalShock]

My understanding is that when a poll references "likely Republican pirmary voters", it is assumed that means people who voted in the 2004 primary.

Have the pollsters come right out and said it? I don't know. Maybe there are others here who follow the polls more closely who can answer that.

[AtomiC]

Yup that's exactly why he's low in the "scientific polls."

[MsDoodahs]

Yeah, if pollsters are using participation in the 2004 primaries to determine "likely" primary voters this time, then yeah, their initial "sample" is screwed from the outset as the videomaker notes. Combine that with the steady and ongoing migration from land line to cell phone only...

I think Zogby has realized what's coming, that's why he's gotten on the record saying Ron Paul will surprise. Zogby is trying to play CYA, but I think they're all going to miss with their so called "scientific" numbers pretty badly. Maybe enough to kill the polling business.

:)

[N13]

If you read the assumptions behind the polls, you will see that this is true.

[MsDoodahs]

My understanding is that when a poll references "likely Republican pirmary voters", it is assumed that means people who voted in the 2004 primary.

Have the pollsters come right out and said it? I don't know. Maybe there are others here who follow the polls more closely who can answer that.

Some have been quite up front with that info, yes. I think most of them include how they came up with their definition of "likely GOP primary voter" some place in the guts of their poll data.

[RoyalShock]

Some have been quite up front with that info, yes. I think most of them include how they came up with their definition of "likely GOP primary voter" some place in the guts of their poll data.

I suspected as much, but hadn't seen it for myself. That's good information to know.

[fuzzybekool]

Yes, it is TRUE 100%

[gravesdav]

Don't make excuses for low poll numbers. He has low name recognition. The poll I work for (Marist Institute of Public Opinion) doesn't just call people who voted for Bush. We randomly call numbers in the area we're polling then ask people their chances of voting in the primary. Then they only use the very likely and likely responses in the data. What polls can't measure is the strong support and voter turnout that Paul will get.

[shasshas]

Don't make excuses for low poll numbers. He has low name recognition. The poll I work for (Marist Institute of Public Opinion) doesn't just call people who voted for Bush. We randomly call numbers in the area we're polling then ask people their chances of voting in the primary. Then they only use the very likely and likely responses in the data. What polls can't measure is the strong support and voter turnout that Paul will get.

gravesdav -- do you mean to say that all scientific poolsters do not restrict their sample to just the previous list of republican voters? then the poll, apart from cell phonee bias, would be a representative sample ?

[boondoggle]

Don't make excuses for low poll numbers. He has low name recognition. The poll I work for (Marist Institute of Public Opinion) doesn't just call people who voted for Bush. We randomly call numbers in the area we're polling then ask people their chances of voting in the primary. Then they only use the very likely and likely responses in the data. What polls can't measure is the strong support and voter turnout that Paul will get.

How's he doing then??