HHH/MPR: Obama easily defeats all GOPers in Minnesota

[Mr. Morden]

I think the Pawlenty buzz started when it came out he was on the short list for McCain VP.  Evidently, that convinced him he was up for the top job.

Meh, I think there was some Pawlenty buzz even before that, as far back as when he was reelected in 2006.  He was definitely mentioned on here in 2006 as a longshot possibility to run in 2008.

Presidential buzz seems to bubble up for nearly every prominent politician who fits a certain profile, and it's just a matter of whether the guy swats it down or nurtures it.

[Alcon]

I thought Pawlenty was popular in Minnesota?

The sample size is 750 out of a state with a population of over 5 million, so there is a lot of room for error.

Ten thousand, ten million, at n=750, it doesn't really matter much.

[Ebowed]

Unless you think it's unusually hard to make up the last couple of points in Minnesota (and I see no compelling reason why this is the case, unless you believe in personifying states as decision makers)

It is difficult, because Minnesota is polarized and the Democrats retain a generic advantage in national elections (a less liberal parallel to Washington or Oregon).

It is easier for Republicans to win at the state level because the Independence party draws from voters of all views who are dissatisfied with the major parties.

[Nichlemn]

Unless you think it's unusually hard to make up the last couple of points in Minnesota (and I see no compelling reason why this is the case, unless you believe in personifying states as decision makers)

It is difficult, because Minnesota is polarized and the Democrats retain a generic advantage in national elections (a less liberal parallel to Washington or Oregon).

It is easier for Republicans to win at the state level because the Independence party draws from voters of all views who are dissatisfied with the major parties.

For a state with such a modest Democratic lean (and D+2 is modest), Minnesota would need to be exceptionally polarised to not be winnable by the Republicans in years where they won the popular vote by a few points. Gore and Kerry only won by a few points, so you're saying that almost all of their voters are hardcore Democrats. I find this hard to believe.   

[pbrower2a]

I thought Pawlenty was popular in Minnesota?

The sample size is 750 out of a state with a population of over 5 million, so there is a lot of room for error.

Ten thousand, ten million, at n=750, it doesn't really matter much.

750 isn't enough for Texas -- or California if it is close.

[Franzl]

I thought Pawlenty was popular in Minnesota?

The sample size is 750 out of a state with a population of over 5 million, so there is a lot of room for error.

Ten thousand, ten million, at n=750, it doesn't really matter much.

750 isn't enough for Texas -- or California if it is close.

Someone doesn't understand statistics...

[I spent the winter writing songs about getting better]

I thought Pawlenty was popular in Minnesota?

He's not. He never particularly was and his ratings have gone even further south with him only concerned about winning the nomination instead of governing the state.

And all you folks need to take a polling class.

[The Mikado]

I thought Pawlenty was popular in Minnesota?

The sample size is 750 out of a state with a population of over 5 million, so there is a lot of room for error.

Ten thousand, ten million, at n=750, it doesn't really matter much.

750 isn't enough for Texas -- or California if it is close.

Someone failed stat.

Off the top of my head, 750 is somewhere between 3-4% Margin of Error.  If his approval was 50%, that'd be a range from about 46.5-53.5.  Granted, there'd be a pretty big difference between the two ends there.

EDIT: I checked it: 3.65% MOE.

[Gustaf]

I thought Pawlenty was popular in Minnesota?

The sample size is 750 out of a state with a population of over 5 million, so there is a lot of room for error.

Ten thousand, ten million, at n=750, it doesn't really matter much.

750 isn't enough for Texas -- or California if it is close.

Someone failed stat.

Off the top of my head, 750 is somewhere between 3-4% Margin of Error.  If his approval was 50%, that'd be a range from about 46.5-53.5.  Granted, there'd be a pretty big difference between the two ends there.

EDIT: I checked it: 3.65% MOE.

That's a 95% confidence interval. It's unlikely to end up close to those tails anyway.

[pbrower2a]

I thought Pawlenty was popular in Minnesota?

The sample size is 750 out of a state with a population of over 5 million, so there is a lot of room for error.

Ten thousand, ten million, at n=750, it doesn't really matter much.

750 isn't enough for Texas -- or California if it is close.

Someone doesn't understand statistics...

It's Texas -- not statistics as applied to polling in most states. Texas is simply so diverse, with so many different communities and some very different cities (nobody would ever confuse Amarillo with Austin) and three  large ethnic groups that vote very differently. It's not statistics that is the problem in polling Texas; it is the difficulty of getting a representative sample.

[Franzl]

I thought Pawlenty was popular in Minnesota?

The sample size is 750 out of a state with a population of over 5 million, so there is a lot of room for error.

Ten thousand, ten million, at n=750, it doesn't really matter much.

750 isn't enough for Texas -- or California if it is close.

Someone doesn't understand statistics...

It's Texas -- not statistics as applied to polling in most states. Texas is simply so diverse, with so many different communities and some very different cities (nobody would ever confuse Amarillo with Austin) and three  large ethnic groups that vote very differently. It's not statistics that is the problem in polling Texas; it is the difficulty of getting a representative sample.

But increasing the sample size, as you suggested, would not make that sample more representative.

[Alcon]

I thought Pawlenty was popular in Minnesota?

The sample size is 750 out of a state with a population of over 5 million, so there is a lot of room for error.

Ten thousand, ten million, at n=750, it doesn't really matter much.

750 isn't enough for Texas -- or California if it is close.

Someone doesn't understand statistics...

It's Texas -- not statistics as applied to polling in most states. Texas is simply so diverse, with so many different communities and some very different cities (nobody would ever confuse Amarillo with Austin) and three  large ethnic groups that vote very differently. It's not statistics that is the problem in polling Texas; it is the difficulty of getting a representative sample.

But increasing the sample size, as you suggested, would not make that sample more representative.

Not to mention that random sampling does not fail because of "geographical diversity" or whatever.  pbrower you make no sense.