6/20-22/08; 580 LV, 4.2%
Obama 48, McCain 46
Why not do larger sample sizes? 500 seems low in a state like Ohio. Why not sample at least 1000 like qinnipiac?
Posted on June 27, 2008 3:46 PM
You all know that I am an Obama supporter, but the previous SurveyUsa poll that had Obama up by 9% was just plain crazy, 52% democrats in Ohio? In my dreams!! Just 28% of Republicans? I wish!!!
Now, this poll sounds more "normal" with 47% democrats and 35% republicans. I think that after Newsweek and Bloomber/LATimes pollsters are paying attention as to what % of democrats and republicans they include. This numbers are closer to the reality. Remeber that Rasmussen had McCain up by 1% and Qunnipiac Obama up by 6%, so we can assume that this SurveyUsa is more in the middle, which makes some sense to me.
Posted on June 27, 2008 4:01 PM
@carl29: You may be right. What strikes me as a little odd about this poll is the gigantic gender gap. A 47 point gender gap is just not believable. The gender gap in Quinnipiac was more than half that.
The past few SUSA polls have been showing this gigantic gender gap. Is there something screwy lately with the sampling method? I understand that a gender gap may exist, but almost 50 points seems HUGE. Unbelievably huge.
Posted on June 27, 2008 4:40 PM
Tim Rasmussen explained why Democrat/Republican proportions are different in the polls generally as follows:
One group uses the approach that the random sample results in the right proportion, i.e. partisan ID is fluid.
The other group (Rasmussen included) believes that partisan affiliation are more stable, thus these polls are adjusted to reflect such. Rasmussen Reports measures changes in partisan identification on a monthly basis using 15,000 phone interviews.
Posted on June 27, 2008 4:42 PM
Make that SCOTT Rasmussen.
Posted on June 27, 2008 4:44 PM
Here we go again with the wild variations in SUSA's party ID numbers. I really don't have much of a feel for what it's like "on the ground" in Ohio but the variations in party ID between this poll and their May poll - as carl29 noted - are problematic. Did the GOP, in month, see an increase of 8% in Ohio voters, while the Dems lost 5%? Did Ohio voters suddenly go from being 48-27 moderate v. conservative to 40-32? Either one of these is wrong or the answer is somewhere in the middle. The party/ideology ID shift in favor of the GOP/conservatives absolutely explains the change from a 9% Obama lead to a 1% Obama lead.
If I had to take a shot in the dark, I'd say Obama leads in Ohio currently by at least 3% but no more than 5%. It's almost as if SUSA is correcting their overstatement of Dem/moderate ID in May with and overstatement of GOP/conservative ID here in June.
Make that a 2% Obama lead...I said 1% in my last post.
RE: the gender gap. In SUSA's May Ohio poll, the gender breakdown as a percentage of those polled was about the same as this one. Back then, Obama LED McCain with male voters, 45-42, and led with women voters, 50-37. This month, McCain leads with men voters, 58-35 (????), while Obama leads with women voters, 59-35. I can understand Obama's lead with women increasing - in May, we were still at the height of the Dem primary season - the increase is Hillary voters coming home. But did male voters, in the short span of a month, swing away from Obama (10%) and to McCain (16%) that drastically?
Is this, again, all due to the big differences in party/ideology ID between their May and June polls? (ie - men tend to be more conservative and more Republican)
Posted on June 27, 2008 4:53 PM
SUSA has shown itself to be more dependable as per 538's polling of their outcomes for the primaries. To continue to take this poll as a give all tell all is a bit short sighted given that many polls are starting to balance out between a range. SUSA has their moments of calling it high and low but their mean has been very dependable and its pretty expected that Ohio be close anyhow.
Posted on June 27, 2008 5:00 PM
SurveyUSA doesn't not weight by party ID. Rasmussen does. Weighting by party ID will the impact of any change from the norm. The reason why party ID weighting is particularly problematic in this cycle is because Democrats are seeing a sustained surge in those that recognize themselves as Democrats, the size of which hasn't been seen for decades.
So when you see a SurveyUSA poll that says 47% Dem to 35% Rep, that means that this is what it was after they corrected their sample standard demographics only. When you see such numbers from Rasmussen, that means that this is what they roughly targeted, and it can have a very large effect on the outcome of your results.
Although one 500 person sample doesn't make or break a race, I do see trouble in this poll for Obama in that it is showing almost all of the undecideds are white and of them, 3/5ths are independents. That doesn't leave much ground for widening the gap should this be more static.
Obama can still lose Ohio though and win easily. McCain has to win one of Colorado or New Mexico. This is why this election won't be all about just Ohio.
Posted on June 27, 2008 5:26 PM
Nate had a very interesting comment:
"...it's the SurveyUSA result in Ohio that I want to focus on. Obama leads by 2 here, but had been ahead by 9 in SurveyUSA's may poll of the state. That previous poll had shown a heavily Democratic sample -- 52 percent Democrat, 28 percent Republican, 18 percent independent -- and had triggered a lot of discussion about whether pollsters should be weighting their results by party ID. SurveyUSA does not do so -- although if it had applied the May distribution of party IDs to this poll, it would have shown Obama ahead by 10-11 points rather than by 2. Conversely, if SurveyUSA had applied the June party ID distribution to its May poll, that poll would have shown a dead heat rather than Obama ahead by 9."
Posted on June 27, 2008 5:42 PM
After a few debates I bet Ohio will swing for one candidate.
Posted on June 29, 2008 8:56 PM
Comments: (you may use HTML tags for style)
Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.
Please email us to report offensive comments.
See our comment policy here. Note that we require commenters to share their email address via Typekey. We will never share your email address with anyone without your explicit permission.
MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR