Things looking up for Pawlenty; down for Kennedy

Wall Street Journal has the latest polls from their provider — Zogby.

In the Senate race, Amy Klobuchar has 49.2% to Mark Kennedy’s 41.8% — a slight increase for Klobuchar, a slight decrease for Kennedy.

In the governor’s race, Mike Hatch has been caught by Tim Pawlenty.

Anyway, Pawlenty is at 43.5% vs Hatch at 42.6%. Pawlenty is 6% better than Becky Lourey, and 6.5% (roughly) ahead of Steve Kelley.

Add Peter Hutchinson into the mix, and I’d bet Pawlenty’s lead would double in each of those races. The Journal, by the way, says all of the polls show “Tim” Hutchinson polling 6-8%. Great. They can’t get the guy’s name right.

(Update) – I notice a couple of political sites, in an effort to dismiss the Zogby methodology, posted one paragraph of Zogby’s methodology, making it look like it’s little more than an online survey. Here, is the actual methodology, unedited .

These polls were conducted by Zogby International. Online polls were conducted by the company’s Zogby Interactive unit. Phone polls were conducted by Zogby International.

Online polls were conducted by Zogby Interactive, a unit of Zogby International of Utica, N.Y. Zogby has assembled a database of individuals who have registered to take part in online polls through solicitations on the company’s Web site as well as other Web sites that span the political spectrum. Individuals who registered were asked to provide personal information such as home state, age and political party to Zogby, which in turn examined that data and contacted individuals by telephone to confirm that it was valid.

Zogby International telephoned about 2% of respondents who completed the interactive survey to validate their personal data. To solicit participation, Zogby sent emails to individuals who had asked to join its online-polling database, inviting them to complete an interactive poll. Many individuals who have participated in Zogby’s telephone surveys also have submitted e-mail addresses so they may take part in online polls.

The Interactive polls were supplemented by 20 to 50 telephone calls in 19 states (AR, CA, CO, FL, GA, IL, MD, MI, MO, NV, NM, NJ, NY, OH, PA, TN, TX, VA, WI) to ensure proper representation of all demographic groups.

Margins of error for each candidate vary by state and range between 3.2 and 4.3 percentage points. Margins for specific states are available on the state panels.

Zogby International President John Zogby says 15% of the company’s U.S. database of online-poll participants are “regulars,” who take part in half of the interactive polls the company conducts; the balance of the names of respondents in the database change frequently. Likely voters in each of the 25 states followed instructions sent by an e-mail that led them to the survey located on Zogby’s secure servers. Those polled were asked unique questions pertaining to the races in their state.

As is usual in polling, weightings are applied to ensure that the selection of participants accurately reflects characteristics of the voting population, including region, party, age, race, religion and gender.

For the overall party breakdowns shown in the governor and senate panels: Races that aren’t being polled are assumed to stay in their current party’s possession.

Regarding 2006 races, the actual matchups that voters will see on ballots haven’t been decided. Primary elections that will narrow the field will be held beginning in early March. Until the fields are narrowed for 2006, the Zogby polls match up multiple candidates in each state and identify some of the strongest candidates from each party. Voter sentiment is gauged on declared candidates along with others whom political observers have identified as potential contenders.

This graphic highlights a sampling of matchups for each race. In some cases, Zogby has polled additional matchups that aren’t included in the graphic. Full polling results are available from Zogby.

There’s a lot more information in there than just one paragraph and the way I read that is that obtaining names of people via online is part of a vetting process that identifies them and their characteristics are then worked into whatever science pollsters use to determine whether there’s a proper sampling.

How is that any different than me picking up a telephone and calling someone and asking them for the same information and then deciding whether their characteristics fit the science?

Now, I don’t dispute that John Zogby doesn’t appear to like Republicans very much, all the more interesting that the Wall Street Journal, not exactly the Onion, chose his firm.

But Zogby’s personality aside, I think if you look at his numbers in the waning days of the 2004 campaign, you’ll find them relatively accurate.

Some of his detractors point to polls he produced in January and May of that year and compare them to the November results and say, “see, he’s wrong.” That, for the record, is what I call the Bill Cooper method. He led the fight against the Star Tribune’s polling system by comparing polls to November results. That’s not even bad science. That’s just bad logic.

I understand the concerns against Zogby, but if you actually read more than one paragraph — well, assuming someone actually tells you there’s more than one paragraph — it makes sense. And I also trust the Wall Street Journal not to pin its polling with a hack.

Could I be wrong? Of course. But seeing some polling that showing a different trend would be the place to start. Anybody?

We (MPR) have been looking at polling firms to see if there’s something we’d want to do locally, and I think Rasmussen is pretty interesting — mostly because they’re polling Minnesota. But in checking them out, a lot of folks put up the red flag because they use automatic dialing. I’m not sure what that means, but AP doesn’t touch Rasmussen stuff.

  • http://www.kennedyvmachine.com Gary M. Miller

    You would be well-served by going with Rasmussen — if accuracy is the desired outcome. They were the most accurate in the last presidential election. Zogby was wrong in 7 of 9 Senate elections in the last mid-terms.

    But as I said, that’s only if accuracy is the aim.

    * * * *

    (Bob notes: I think this is the problem I have evaluating polls, however. I think we need to look at the last poll before an election to determine ‘right’ vs. ‘wrong.’ And track the various polls over a period of time. I do think we do a disservice in creating the obvious impression that polls predict anything. I like Rasmussen, myself. I guess the automatic dialers are the thing.)

    …by the way…

    Nice shot in mischaracterizing what I posted about Zogby on your site (hey, turn the comments on! What are you afraid of? (g). Truth is just collaateral damage sometimes, I guess. Especially if it’s not the point in the first place.

  • http://www.kennedyvmachine.com Gary M. Miller

    Truth doesn’t serve our evil purposes.