MN: Survey USA: Dayton & Emmer basically tied, Horner surging

[Tender Branson]

New Poll: Minnesota Governor by Survey USA on 2010-09-15

Summary: D: 38%, R: 36%, I: 18%, U: 4%

Poll Source URL: Full Poll Details

[Dgov]

Another SUSA Poll showing the Republican winning big with the Young (46-30 vs 36-38 Statewide).  Either them or PPP is wrong in that regard, but I'm leaning towards SUSA given that this is yet another poll showing the Republicans running even or close with the Black vote.

Also, Emmer's winning in the Twin Cities but Losing in Southern MN, which also seems very odd.

[Associate Justice PiT]

     Given that this is Minnesota we are talking about, the breakdown of the black vote in polling is far too small to be read into statistically. SUSA is developing something of a pattern of showing Republican-friendly results this cycle, though. My guess is Dayton +6 at this point, though I'm not at all sure of that.

[Dgov]

     Given that this is Minnesota we are talking about, the breakdown of the black vote in polling is far too small to be read into statistically. SUSA is developing something of a pattern of showing Republican-friendly results this cycle, though. My guess is Dayton +6 at this point, though I'm not at all sure of that.

I know that it's far too small a sample to be statistically significant, but we're talking about a Demographic that went like 95% for the Democrats last cycle, going almost 50/50 in this one.  That kind of shift (Especially in multiple, consecutive polls) cannot adequately be explained soley by statistical error.

[Associate Justice PiT]

     Given that this is Minnesota we are talking about, the breakdown of the black vote in polling is far too small to be read into statistically. SUSA is developing something of a pattern of showing Republican-friendly results this cycle, though. My guess is Dayton +6 at this point, though I'm not at all sure of that.

I know that it's far too small a sample to be statistically significant, but we're talking about a Demographic that went like 95% for the Democrats last cycle, going almost 50/50 in this one.  That kind of shift (Especially in multiple, consecutive polls) cannot adequately be explained soley by statistical error.

     Point taken, it shouldn't deviate by more than 10% from the true spread (if the true spread were 50-50, the permissible for a 95% confidence value would be in the realm of 22% in either direction). That is, if I remember my statistics well, which I probably don't.

[fezzyfestoon]

Damn, and I thought this one was pretty much wrapped up.

[JohnnyLongtorso]

SUSA historically leans Republican in Minnesota. Something about their methodology, it might be because they don't catch voters who register on election day.

[Small Business Owner of Any Repute]

This one will likely be decided by the size of the GOP wave.

[Hash]

Another SUSA Poll showing the Republican winning big with the Young (46-30 vs 36-38 Statewide).  Either them or PPP is wrong in that regard, but I'm leaning towards SUSA given that this is yet another poll showing the Republicans running even or close with the Black vote.

Also, Emmer's winning in the Twin Cities but Losing in Southern MN, which also seems very odd.

STOP LOOKING AT CROSSTABS

[Small Business Owner of Any Repute]

Yes, we need a forum wide education on this.

Crosstabs that involve the whole sample (i.e., what proportion of the sample is R, D, and I) are valid and worth looking at for curiosity sake and poll validation. Crosstabs that involve part of the sample (i.e., what proportion of the sample's Republicans are voting for x) are notably less accurate because they carry a higher margin of error, but may still carry some valuable information. Crosstabs that involve a tiny (<[/u ]10%) part of the sample (groups like "aged 18-29" and "black voters") have a gigantic margin of error and are not worth looking at for serious statistical purposes.

[I spent the winter writing songs about getting better]

SUSA historically leans Republican in Minnesota. Something about their methodology, it might be because they don't catch voters who register on election day.

Basically. Their last poll in 2008 only had Obama up by 3.

[Torie]

SUSA historically leans Republican in Minnesota. Something about their methodology, it might be because they don't catch voters who register on election day.

Basically. Their last poll in 2008 only had Obama up by 3.

You are just worried that all of your Minnesota gerrymandering fantasies might come to naught.

[TheDeadFlagBlues]

When push comes to shove, a good chunk of Horner's supporters will vote for Dayton. No one wants a Governor Emmer.

[bgwah]

     Given that this is Minnesota we are talking about, the breakdown of the black vote in polling is far too small to be read into statistically. SUSA is developing something of a pattern of showing Republican-friendly results this cycle, though. My guess is Dayton +6 at this point, though I'm not at all sure of that.

I know that it's far too small a sample to be statistically significant, but we're talking about a Demographic that went like 95% for the Democrats last cycle, going almost 50/50 in this one.  That kind of shift (Especially in multiple, consecutive polls) cannot adequately be explained soley by statistical error.

>95% Democrat? Oh really? I will play along with this silly little game... Let us compare SUSA's last MN poll for President, where blacks voted:

9 Obama (60%)
6 McCain (40%)

SUSA's last MN Senate poll showed blacks voting:

7 Franken (47%)
5 Coleman (33%)
1 Barkley (7%)
1 Undecided (7%)

Some shift!