Bogus Polls Alert
Earlier today, I read a Detroit News poll that seemed to be an outliar so I checked further. Here’s what caused me to question it:
Granholm’s job approval ratings have shifted dramatically since June. In a June 13 poll, only 40 percent approved of her job as governor. In the latest survey, 51 percent give her passing grades for the job she’s doing. The governor’s personal favorability rating was at 58 percent in this poll, 6 percentage points higher than in June.
I’m skeptical of any poll that shows an 11 point jump in an incumbent’s JA rating over time. I’m even more skeptical when that jump happens in a two month. If it’s possible, I’m more skeptical now. It’s also worth noting that Granholm’s ratings were in the disaster range because of Michigan’s freefalling economy. Nothing’s happened to change that reality.
Here’s why I’m so skeptical: I wanted to check the poll’s internals so I emailed Mark Hornbeck, who wrote part of the article, asking where I could find the internals. Here’s his response:
We don’t share any polling data other than what is published…if you have questions about methodology, I’d encourage you to contact EPIC/MRA in Lansing.
One of the things that I’ve always demanded before believing that a poll is credible is that they publish the breakdowns. While they aren’t exactly hiding the information, they aren’t being totally open either.
The good news about the Michigan governor’s race is that Dick DeVos is in good position to beat failed Jennifer Granholm. This coincides with Debbie Stabenow’s decline in the Senate race. Stabenow was leading the Republican by double digits but now she leads Mike Bouchard by only 5 points.
EPIC/MRA doesn’t have a great reputation for accuracy. I’ve read other bloggers who’ve said that it’s usually off by 5-10 points, always in the Democrat’s favor.
Kinda sounds like the Minnesota Poll, doesn’t it?
Technorati: Granholm, DeVos, Election 2006, EPIC/MRA, Minnesota Poll
Cross-post at LetFreedomRingBlog
August 16th, 2006 at 3:20 pm
[...] Cross-posted at California Conservative Categories: Midterm Elections, Economy, Blogs, Media Bias | [...]
August 17th, 2006 at 10:53 am
However, the EPIC/MRA poll isn’t the only one showing Granholm with the lead. Rassmusen has her up by 1% and the last Zogby poll had her up by about 5%. So this race is basically a tie/within the margin of error.
As for “the freefalling economy” there has been some news beneficial to Granholm coming out of MI on that front from what I have read. I saw a poll (I believe it was Detroit News) showing that a majority of Michiganders blame Bush for the economy and I’m sure that is going to be repeated again and again until November. Plus, companies like Google, eprize, Whirpool and others creating new jobs defintely help Granholm’s cause.
Plus Granholm and the MDP just started running their ads, which likely had something to do with the change in polling numbers. DeVos can defintely outspend Granholm on that front and he will have right up to the election. I’m not sold on DeVos’ ability to pull this one out, he’s going to have to catch a few breaks.
August 17th, 2006 at 12:23 pm
If a poll is not conducted with a *real* SRS of 2000 of the likes Gallup does for presidential elections, which are extremely expensive (and rare to weekly polls) , then I highly-doubt there significance - other than the propaganda purposes of the said- institution. Gallup does do 500, 750s, 1000 for week to week and small candidates - but I think these are garbage-margin polls anyway.
August 17th, 2006 at 12:25 pm
“We don’t share any polling data ” HAHAHAHAHAHAHAHAH - Neither did Hitler.