Current St. Norbert College Poll: Good News for Democrats?
A poll released this past Wednesday by Wisconsin Public Radio seems to have good news for liberals.
It shows, for example, that Gov. Jim Doyle has a comfortable lead over Mark Green, with Doyle being favored by 51% of likely voters and Green by 38%.
Both the Death Penalty referendum and the ban on Gay Marriage are favored in the poll, but by smaller margins than one might expect.
The former appears to be favored by 50% of likely voters (with 45% opposed) and the latter favored by 51% of likely voters (with 44% opposed).
But the poll might have a Democratic bias. It included 44% Democrats and 33% Republicans. That’s an 11 percent Democratic lead.
Yet other highly reputable polls show Wisconsin more balanced between Democrats and Republicans. The 2004 Annenberg survey shows Democrats having a 5% advantage in party indentification in Wisconsin.
And a 2006 Pew Research Center analysis showed Republicans having a 1% advantage in Wisconsin.
Only one factor complicates the issue. The St. Norbert poll asked about party identification only among people who had previously been identified as “likely voters.” So it’s impossible to know whether they had too many Democrats in their sample, or whether their sample was accurate and more Democrats made the cut by saying they would “definitely” or “probably” vote. If that latter supposition is true, Democrats are more mobilized to vote this year.
Given that their results are more pro-Doyle and less pro-Marriage Amendment and pro-Death Penalty than other polls, we tend to believe that a sampling fluke has simply produced a poll with a liberal, Democratic-leaning bias.
This happens every so often, and happens to quite reputable polls.
[Update]
Wagner on the Web notes that the St. Norbert poll has a history of producing results biased in a Democratic/liberal direction. Perhaps they need to carefully evaluate their sampling procedures.
It shows, for example, that Gov. Jim Doyle has a comfortable lead over Mark Green, with Doyle being favored by 51% of likely voters and Green by 38%.
Both the Death Penalty referendum and the ban on Gay Marriage are favored in the poll, but by smaller margins than one might expect.
The former appears to be favored by 50% of likely voters (with 45% opposed) and the latter favored by 51% of likely voters (with 44% opposed).
But the poll might have a Democratic bias. It included 44% Democrats and 33% Republicans. That’s an 11 percent Democratic lead.
Yet other highly reputable polls show Wisconsin more balanced between Democrats and Republicans. The 2004 Annenberg survey shows Democrats having a 5% advantage in party indentification in Wisconsin.
And a 2006 Pew Research Center analysis showed Republicans having a 1% advantage in Wisconsin.
Only one factor complicates the issue. The St. Norbert poll asked about party identification only among people who had previously been identified as “likely voters.” So it’s impossible to know whether they had too many Democrats in their sample, or whether their sample was accurate and more Democrats made the cut by saying they would “definitely” or “probably” vote. If that latter supposition is true, Democrats are more mobilized to vote this year.
Given that their results are more pro-Doyle and less pro-Marriage Amendment and pro-Death Penalty than other polls, we tend to believe that a sampling fluke has simply produced a poll with a liberal, Democratic-leaning bias.
This happens every so often, and happens to quite reputable polls.
[Update]
Wagner on the Web notes that the St. Norbert poll has a history of producing results biased in a Democratic/liberal direction. Perhaps they need to carefully evaluate their sampling procedures.
posted by John McAdams at 2:03 PM
0 Comments:
Post a Comment
<< Home