[Mpls] Minnesota Poll

[Steve Brandt]

Several views have been expressed on the Minnesota Poll printed yesterday on the Rybak-McLaughlin contest.

One argument made is that the poll supports the editorial endorsement of the newspaper.  This is what a statistician would call an association, but not causation.  If polls are slanted to support the newspaper's editorials, then can someone please explain this poll result? 


Date: 05/01/05
Section: NEWS
Page: 01A
Edition: METRO
Byline: Rochelle Olson; Staff Writer
Graphic: CHART;ILLUSTRATION
Length: 63.64
Column: Minnesota Poll
Subject: poll;minnesota;new stadium;baseball;finance;building;taxation;demographic
Slug: poll0501

   Most Minnesotans still oppose the use of public money for a new 
Minnesota Twins stadium, even though support is somewhat stronger 
for a new plan offered last week by Hennepin County and the team, a 
new Minnesota Poll has found.

Another correspondent argues that the election-eve poll in 2001 showing a 20 percentage point gap for Rybak caused a 30-point victory by depressing turnout.  That's one explanation.  But another possibility considers that 12 percent of the likely voters surveyed in the election eve poll were undecided.  It's possible that that most of them broke toward Rybak.

Still another person says that the polls are not scientific; perhaps she could post further to explain for us with which law of probability she takes issue.

I would agree with Brauer, who faults the newspaper for cutting back substantially on election-year polling in both Minneapolis and St. Paul.  In the last two city elections, there was an extensive early-summer poll that essentially took the temperature of voters on issues like crime, whether the parks and schools were better or worse, and so on.  That was lost to budget cuts.  But I think that the increasing suburban focus of the newspaper also is to blame.

Steve Brandt
Star Tribune/Kingfield