Comments on: Gallup: Obama Up By 8

By: Big Ern

Big Ern — Wed, 30 Jul 2008 00:45:41 +0000

Those who are sold on Rasmussen numbers should consider one thing…..Rasmussen does NOT poll cell phones. There is a growing number of cell only users in the U.S. Cell users favor Obama about 2-1 over McCain (most cell only users are young….young pretty heavily favor Obama)

By: Donklephant » Blog Archive » Rasmussen: Obama Leads By 1, 2

Donklephant » Blog Archive » Rasmussen: Obama Leads By 1, 2 — Tue, 29 Jul 2008 14:25:54 +0000

[...] that Rasmussen polls “likely” voters, and Gallup polls “registered” voters. And to that point… Remember that a full 25% or more of registered voters NEVER VOTE. Ever. They fill out the [...]

By: Tully

Tully — Mon, 28 Jul 2008 20:55:42 +0000

Gallup also uses a slightly smaller daily sample size, which also adds a touch of variability.

By: Tully

Tully — Mon, 28 Jul 2008 20:53:23 +0000

Nope, that affects the trendline variability a bit but not the daily variability. Gallup is registered voters and Rasmussen is likely voters. And as any old political demographer (and I resemble that remark) can tell you, the portion of registered voters who are not likely voters tend to tell pollsters whatever their Rice Krispies mentioned to them at breakfast. Thus, RV tracking polls always have higher variability than LV tracking polls. Remember that a full 25% or more of registered voters NEVER VOTE. Ever. They fill out the registration form at the Driver's License office and never go to the polls. Some are disinterested, some are illegal immigrants who are afraid to vote, whatever, they don't vote and so their opinion has no bearing at all on the prediction of future elections.

By: Justin Gardner

Justin Gardner — Mon, 28 Jul 2008 20:02:01 +0000

Oh, is it because Gallup is a 3 day poll and Rasmussen is a 5 day?

By: Tully

Tully — Mon, 28 Jul 2008 19:48:14 +0000

Since I follow Gallup and Rasmussen daily, what Iâ€™ve noticed is that Gallup seems to swing quicker to wider margins. In other words, Rasmussenâ€™s numbers seem to be more stable and represent more realistic opinion shifts.

There’s a very good reason for that, and if you look more closely you should find it in very short order. Want hints? ;-)

By: Gold Star for Robot Boy

Gold Star for Robot Boy — Mon, 28 Jul 2008 18:33:46 +0000

Rasmussenâ€™s numbers seem to be more stable and represent more realistic opinion shifts.
Why do you say that? What’s the evidence?
News of Obama’s trip was _everywhere_ last week; shouldn’t there be some reaction seen?