Is there any research that verifies the results of Rasmussen polls?

[randr]

I have always questioned the accuracy of any polls associated with Rasmussen and now they claim that 33% of Colorado voters align themselves with the teabaggers.
http://www.huffingtonpost.com/2010/04/23/rasmussen-33-of-colorado_n_550095.html
This is so out of the park incredulous. Even in the county I live in, overwhelmingly Republican, only a few dozen baggers showed up for the latest party.
Somehow we need to address Rasmussen when they fire off such bullshit.

[NoNothing]

1. Rasmussen is actually pretty accurate for elections

However their polling always includes only "likely voters" which tends to oversample conservatives.

[FrenchieCat]

6. they are only accurate, a day or so before an election......

The rest of the time, they are consistantly typified as the outlier.; them and Zogby.

[Jackpine Radical]

2. You raise an interesting question.

The ultimate test of any political pollster is in the ability to predict elections--or is that actually to predict how the voting machines are rigged?

For other polls, that don't have external validation at the ballot box, the only real test is whether they agree with other polls. Technically speaking, consistency among polls is really a measure of reliability and only indirectly of validity. In cases such as this, where there are no other polls asking the question, the pollster could simply make up the results and nobody would be the wiser. Everyone would assume the results are at least roughly accurate because the pollster has been able to predict elections. But, especially if the pollster has a motive to bias the results (e.g. by asking questions that "pull" for a given response, selecting a nonrandom sample, or, for that matter, just lying), there is no way to be sure he hasn't done exactly that.

[tritsofme]

3. You can look at the results and judge for yourself

Final Rasmussen 2008 Poll Results – Presidential Election

Final Rasmussen 2008 Poll Results – Senate Elections

[randr]

4. Interesting lean to the right

[NaturalHigh]

5. I took a graduate course in statistics a few years back...

and I still don't really understand polling.

Lies, damned lies, and statistics...