They also took the data they collected from the random telephone sample and cross-referenced with voter registration data so they could find out who caucused in 2004 as well.  In particular they found that Edwards voters are much more likely to be past voters in the Democratic caucus.

Sorry, the sample should be drawn at the very least from registered Ds and Rs–not from random phone numbers.

Preferably from people who have caucused or at least voted in a primary before.

Anyone can tell you on the phone that they will caucus. But will they leave their homes for two hours on a cold January night right after the holidays? I am skeptical.

As Mark Blumenthal over at Pollster.Com says:

Readers should consider that the methodology of this survey, as in August, is different from most of the other Iowa caucus surveys we have seen. According to Professor David Redlawski, who spoke at a Washington press briefing this morning, [...snip...] this most recent survey used essentially the same methodology as their August survey. That is, it used an open-ended vote question, the same screening questions and sampled from a list of telephone numbers drawn from listed telephone directories (i.e. not a registered voter list and not using a random digit dial methodology).

So, when looking at this data it's important to interpret trends – not absolute numbers.  I think the internals to the poll are even more interesting.  Over at the Kevin McCarthy page:

The new University of Iowa “Hawkeye Poll” show Senator Joe Biden making gains since August. Biden was polling at less than 1% support in August and is now pulling close to Richardson with 5.3% in the latest polling done of likely Iowa caucus attendees.

Biden is polling well ahead of Richardson amongst likely caucus goers over 61 years of age – coming in with 9.9% support vs. Richardson's 4.1% amongst this key caucus demographic.

As Mark Blumenthal over at Pollster.Com says:

Readers should consider that the methodology of this survey, as in August, is different from most of the other Iowa caucus surveys we have seen. According to Professor David Redlawski, who spoke at a Washington press briefing this morning, [...snip...] this most recent survey used essentially the same methodology as their August survey. That is, it used an open-ended vote question, the same screening questions and sampled from a list of telephone numbers drawn from listed telephone directories (i.e. not a registered voter list and not using a random digit dial methodology).

So, when looking at this data it’s important to interpret trends – not absolute numbers.  I think the internals to the poll are even more interesting.  Over at the Kevin McCarthy page:

The new University of Iowa “Hawkeye Poll” show Senator Joe Biden making gains since August. Biden was polling at less than 1% support in August and is now pulling close to Richardson with 5.3% in the latest polling done of likely Iowa caucus attendees.

Biden is polling well ahead of Richardson amongst likely caucus goers over 61 years of age – coming in with 9.9% support vs. Richardson’s 4.1% amongst this key caucus demographic.

Sorry, the sample should be drawn at the very least from registered Ds and Rs–not from random phone numbers.

Preferably from people who have caucused or at least voted in a primary before.

Anyone can tell you on the phone that they will caucus. But will they leave their homes for two hours on a cold January night right after the holidays? I am skeptical.

They also took the data they collected from the random telephone sample and cross-referenced with voter registration data so they could find out who caucused in 2004 as well.  In particular they found that Edwards voters are much more likely to be past voters in the Democratic caucus.