|
"A comprehensive, collaborative elections resource."
|
Rasmussen Retorts
|
Parent(s) |
PollingFirm
|
Contributor | particleman |
Last Edited | particleman Sep 09, 2010 09:32pm |
Logged |
0
|
Category | Interview |
Author | Jim Geraghty |
News Date | Friday, September 10, 2010 03:00:00 AM UTC0:0 |
Description | This week began with several big-name polls showcasing dire news for Democratic incumbents. But the results weren’t all that surprising to those who have been following the much-discussed surveys of Scott Rasmussen. Rasmussen took a few moments to talk with National Review Online about polling in the current environment.
JIM GERAGHTY: For much of this cycle, your polls have seen things a little bit better for Republicans than those of a lot of other pollsters. This seems to stem in part from your polling likely voters; many other pollsters survey registered voters and some, simply adults. Do you feel vindicated by the fact that seemingly every pollster has now come out with ever-worsening doomsday scenarios for Democrats?
SCOTT RASMUSSEN: I reject the notion that we have seen things more favorably for Republicans. It’s just that if you were to take a poll of registered voters and compare it to a poll of likely voters, the raw numbers of the registered-voter poll would look better for Democrats. Now that they’re all switching to a likely-voter model, sure, they’re going to end up pretty much in the same place.
JG: Are pollsters wasting people’s time when they give the registered-voter numbers and let people think they’re getting a good sense of how the electorate’s going to look on Election Day?
SR: There are valid reasons to poll different samples. I think what needs to be clear is just what it is you’re doing. We poll all adults for our consumer- and investor-finances polls, because whether you’re a voter or not, you have an impact on economic trends. There are people who would argue that you can’t do a likely-voter poll until after Labor Day, because you don’t really know what the turnout is going to be, so a registered-voter model is preferred in their mind. That’s not a bad argument, except that it implies too much precision. We didn’t know earlier in the year precisely what turnout would be; we still don’t know precisely |
Share |
|
2¢
|
|
Article | Read Full Article |
|
Date |
Category |
Headline |
Article |
Contributor |
|
|