One of the more fascinating manifestations of the split between the elite press and the newer media is in how they treat the robotic (or automated) polls of Scott Rasmussen . Mr. Rasmussen uses a variety of screens and can poll so often and so cheaply because he does not use humans to ask the questions. A computer-generated questioner interacts with the voter, who presses the corresponding button a touch-tone phone.
The press -- and the pollsters who consult for them and most campaign pollsters -- dismiss Rasmussen out of hand. For the latter group, there may be an element of competitive jealousy in the mix, but there are also valid reasons to be skeptical about screening and sample bias in automated surveys and contact and cooperation rates with voters. Democrats in particular tend to be skeptical because Rasmussen is a conservative.
For conservative activists, they are the equivalent of crack, and much, much less deadly.
Occasionally, the campaigns whose pollsters denigrate Mr. Rasmussen's polls often drop their skepticism and cite them when those polls bounce in a positive direction.
Yesterday, Fred Thompson's press shop issued a release with the headline: "Thompson Tied for First in National Poll; Giuliani in Free Fall."
While Fred Thompson travels the nation with his message of consistent conservatism and touts his border security and immigration reform proposal, a recent national tracking poll shows that Thompson has pulled into a virtual tie with Rudy Giuliani. This poll further shows that Rudy has lost almost a third of his support in the last 12 days. (Margin of error is +/- 4 percent)
The source for these numbers is a Rasmussen survey.
This press release was clearly not intended for those of us who work for established publications.
It was aimed at the newer generation of opinion leaders.
But if and when anyone from Thompson's team ever dumps on the Rasmussen robots, I will be the first to let you know.
Fun fact: Scott Rasmussen is a co-founder of ESPN.
| 
The problem with Rasmussen isn't that he uses automated calling. SurveyUSA does the same, and no one has a problem.
The problem with Rasmussen is that he is, as you mention, a conservative, and that this seems to affect his methodology. He consistently gets a right-ward bias in his numbers, compared to all the other pollsters.
In short, he's got an ideological agenda, and his numbers are consistently an outlier in that ideological direction. Questions of intellectual honesty are thus raised.
All that said, his daily tracking numbers are useful, since even if his methodology skews right-ward, it is consistent from day to day.
Posted by Petey | October 26, 2007 11:20 AM