Sunday, October 28, 2012

Gravis Marketing Ohio Poll Results


Results from Ohio Likely Voter Survey

Methodology

Gravis Marketing conducted an automated survey of 730 likely voters in Ohio October 27, 2012. The margin of error for the survey is +/-3.6% and higher for subgroups.  

Results shown by gender and party represent those respondents within those subgroups. All numbers shown in the tables represent percentages rounded to the nearest whole percentage.

The questions were asked in the order of the question numbers which appear in this report.  Results only include respondents who answered that they were registered voters, somewhat likely, likely, or very likely to vote. The statistical methodology comprised weighing various groups for anticipated voting proportions, by using census data and voter turnout models from previous elections.

The poll was conducted on behalf of Gravis Marketing, Inc., Gravis Marketing is a non-partisan marketing and research firm located in Winter Springs, Florida. Contact Doug Kaplan (407) 242-1870 doug@gravismarketing.com

Key Findings

·         President Barack Obama holds a slim one percentage point lead over Governor Romney, 50 to 49 percent. Only 2 percent are undecided. (Note the figures add up to 101 due to rounding to the nearest whole percentage point.)


·         Senator Brown also holds a slim one point lead over his Republican challenger Josh Mandel, 48 to 47 percent. 5 percent of Ohio voters are undecided in this race.

·         Early voters clearly favor the Democratic candidates. President Obama leads Mitt Romney by 28 percentage points among the 20 percent of early voters surveyed, 63 to 35 percent. Sherrod Brown leads Josh Mandel by a similar 26-point margin, 61 to 35 percent.

·         The likely voters who have not yet cast ballots in Ohio preferred the Republican candidates. Mitt Romney leads President Obama by 6 percentage points with these voters (52 to 46 percent), while Josh Mandel leads with these voters by 5 percentage points (50 to 45 percent).

·         This survey shows a 17 percentage point gender gap in the presidential race and a 15 point gap in the Senate race in Ohio. A majority of men favor Governor Romney and Josh Mandel, while a majority of women favor President Obama and Senator Brown.


·          

Ballot, Job Performance & Direction of the Country

Presidential Ballot:
Question 3 & 5 Combined
Early & Likely Voters
Ohio
Voters
Gender
Party
Men

Women
Democrat
Republican
Independent
Barack Obama
50
45
53
83
14
41
Mitt Romney
49
53
44
16
85
53
Undecided
2
2
2
1
0
6

Senate Ballot:
Question 4 & 6 Combined
Early & Likely Voters
Ohio
Voters
Gender
Party
Men

Women
Democrat
Republican
Independent
Sherrod Brown
48
45
51
81
15
40
Josh Mandel
47
52
43
17
79
53
Undecided
5
3
6
2
6
7

Question 7: Do you approve of President Obama’s job performance?
Ohio
Voters
Gender
Party
Men
Women
Democrat
Republican
Independent
Yes
46
42
49
78
16
33
No
48
50
46
16
82
55
Undecided
7
8
5
7
2
11

Question 8: Do you think the United States of America is headed in the right direction or wrong direction?
Ohio
Voters
Gender
Party
Men
Women
Democrat
Republican
Independent
Right Direction
40
37
42
69
12
28
Wrong Direction
50
53
27
18
83
58
Undecided
10
10
11
13
5
13





Early Vs. Likely Vote Balloting

Question 3: EARLY VOTERS ONLY: Which candidate did you already vote for?
Ohio
Voters
Democrat Barack Obama
63
Republican Mitt Romney
35
Rather not say or unsure
1


Question 4: EARLY VOTERS ONLY Did you vote for Democrat Sherrod Brown or Republican Josh Mandel?
Ohio
Voters
Democrat Sherrod Brown
61
Republican Josh Mandel
35
Undecided
4

                                                                                                       
Question 5: DID NOT VOTE YET: If the election for President of the United States were held today who would you vote for?
Ohio
Voters
Democrat Barack Obama
46
Republican Mitt Romney
52
Undecided
2


Question 6: DID NOT VOTE YET: If the Ohio election for senate were held today, would you vote for Sherrod Brown or Josh Mandel?
Ohio
Voters
Democrat Sherrod Brown
45
Republican Josh Mandel
50
Undecided
5



                                                         




Screeners & Demographics

Question 1:  Are you registered to vote?
            Ohio Voters
Yes
100
No
0


Question 2:  How likely are you to vote in this year’s presidential election?
            Ohio Voters
Very unlikely
0
Unlikely
0
Somewhat unlikely
0
Somewhat likely
1
Likely
12
Very likely
68                               
Already voted
20
Question 7: What is your political party affiliation?         [FIRST TWO CHOICES ROTATED]
            Ohio Voters
Democrat
40
Republican
32
Independent or in another party
28

Question 8:  What race do you identify yourself as?
Ohio Voters
White/Caucasian
76
African-American
12
Hispanic
6
Asian
1
Other
5

Question 9: Which of the following best represents your religious affiliation?
Ohio Voters
Catholic
23
Protestant/Other Christian
55
Jewish
4
Muslim
1
Other
17

Question 10:  What is your age?
Ohio Voters
18-29
18
30-49
28
50-64
30
65+
24

Question 11: What is your gender?
Ohio Voters
Male
47
Female
53
                                                                                                                

17 comments:

  1. Democrats + 8% Again? Doug, you need to poll something other than Cleveland crack houses.

    ReplyDelete
    Replies
    1. Honestly, this oversampling gets rather silly. Apparently that's the party line, though. Have to make it look close.

      Delete
    2. ...ever maybe consider that democrats outnumber Republicans in Ohio? Or...perhaps more accurately: That Republican-leaning voters are much more likely to self-identify as "independent"?

      Which also neatly explains Romney's edge with "independents". Look at the national graphs for party self identification sometime: those who indicated republican were down to 29 or so percent in January of this year. Democratic numbers suffered as well, but not as much. Romney's lead among independents and the Democratic Oversampling Boogeyman you guys see in every single poll by everyone ever....are one in the same.
      And...even in Right leaning (some would argue 'biased') polls like Rasmussen and Civitas, Romney has, at his _very_ best, only polled as having a 2 point lead. That was in _one_ Rasmussen Poll,which had a 4 point Margin of Error. All the others have shown a tie in Ohio _at best_...and even the 'tie' polls are the exception rather than the norm.
      All that said...you're crying over a one point Obama lead...why?

      Delete
  2. We at The Elephant in the Room agree with RWV's assessment. There is no party affiliation information that claims democrats outnumber republicans 40% to 32% in the state. Why is Gravis continuing this unrealistic sampling?

    If you're interested in seeing unskewed polls that use Gallup's party affiliation data as the standard, check out our unskewed poll database here: http://loudmouthelephant.blogspot.com/2012/09/lme-presidential-poll-tracker-database.html

    ReplyDelete
    Replies
    1. then
      Gravis isn't the only one, but the polling this year has been especially screwy in Ohio, and not just because of the warped Party ID breakdown.
      http://right-winggenius.blogspot.com/2012/10/oh-oh.html

      Delete
    2. ....this would be because in Ohio there _is_ no real party affiliation information, save for voter participation in the primaries.
      A little over 2.3 million voted in the Ohio 2008 democratic primary.

      About 1.2 million voted in the Ohio 2012 Republican primary.

      I would also add this: John Kerry lost Ohio in 2004 by about 100,000 votes. He also carried 85 or so percent of the African American vote. Obama is slated to carry 95 to 97 percent of the African American vote in 2012. This more than makes up the 100,000 vote difference.

      Relax. Bend over. And try to enjoy it. : )

      Delete
  3. Tea bags. More people identify themselves as DEMS. Repukes now call themselves independent. When will you idiots understand this.

    ReplyDelete
    Replies
    1. More typical hate speech from the party of openness, tolerance, and love. Apparently, you're not good at reading, either. The issue isn't if more people identify themselves as republicans or democrats; the issue is how large of a difference there is. This poll, for example, claims there are 25% more democrats in Ohio than there are republicans. The precise amount is ambiguous, but there is no indication, as I said, that the difference is this large. So yes, you can continue to sit there and sling insults and talk of "repukes," or you can open your eyes to reality and data.

      - www.loudmouthelephant.com

      Delete
    2. lil Black, what a foul mouth hater you must be...typical democrat I guess.

      Delete
  4. Democrats will out vote Republicans by 8%? What a joke. I live in Ohio and party ID is 40-40-20. Romney wins Ohio by a MINIMUM of 5%

    ReplyDelete
  5. UPDATE: Want to see this poll unskewed, which would eliminate the unrealistic +8 democrat oversampling? Visit our unskewed poll database at The Elephant: http://loudmouthelephant.blogspot.com/2012/09/lme-presidential-poll-tracker-database.html

    ReplyDelete
  6. decreasing Dem advantage by just one point to (39-32-29) and holding all other findings equal results in 0.4896 for Romney and 0.4889 for Obama. Factor in undecideds at the same percentage as independents (a fair enough assumption since 86% of undecideds in this survey are independents) you will get a result of Romney 0.501 to Obama 0.498

    If you drop the Dem advantage to +5 (37-32-31) and hold all other findings equal you get 0.497 Romney and 0.481 Obama

    Notice in those cases, I did not increase GOP turnout by a single point.

    Someone mentioned a (40-40-20) split in Ohio. Plugging that turnout in and holding all of Gravis's other findings equal you get
    Romney 0.512 and Obama 0.472. If the undecideds break in the same percentage as all the other independents you get Romney 0.521 to Obama 0.4789 a 4.2% victory for Romney

    The race is definitely tight and it will come down to turnout, but unless the Dem's turnout actually reaches a +8 advantage, I don't seem them winning Ohio. If you drop the Dem advantage just one point, the results favor Romney.

    I don't see Romney winning by as much as 5% either, but right now I would have to give the advantage to Romney in spite of what the polls say.



    ReplyDelete
  7. Gravis (or Doug Kaplan) claims his data is collected by random dialing and then application of a likely voter screen on those agreeing to be interviewed. It is fact the Ohio voter registration is NOT 40/32/28 (D/R/I) a he presents so it seems that if his calling routine is truly random such a skewed response distribution should be unlikely? Are more republicans refusing to take the poll? Also a slight female to male skew (53/47)?

    ReplyDelete
  8. Traditional voter turn out is equal number of Republicans versus democrats. Usually republicans vote at a 1 - 2% higher. So 8% more democrats is not likely. Even in 2008 Democrats had a 2.5% advantage. So that tells me the race is really +4% or so for Romney.

    ReplyDelete
    Replies
    1. Um....no. Not....at...all. Ever. Nationally registered Democrats outnumber registered Republicans to the tune of a few million. This will vary from state to state, of course. You have states like Utah, where....Republicans eat us for lunch without fail every time. Then you have New York, where we don't even bother trying to hunt you guys down as a food source because your numbers are so scarce its not worth the effort.

      In Ohio, voter registration is only official via the requirement to register in order to vote in the Republican and Democratic primaries. 2.3 million voted in the Ohio Dem Primaries in '08. 1.2 million voted in the Ohio GOP primaries. At least a couple thousand or so of those registered Republican in order to vote Santorum to mess with you and because we had nothing better to do. Do the math from there.

      Delete
  9. This comment has been removed by the author.

    ReplyDelete
  10. In some of the Ohio polls where they breakdown by region, I have seen where the Northeast is highly over-polled, and the Southwest is highly under-polled. The Columbus Dispatch polled 515 people from the Northeast (Obama 54%, Romney 44%), and just 151 from the Southwest (Romney 61%, Obama 38%). I don't know the exact population for each region, but there is no way that the Northeast will out vote the Southwest 515 to 151.

    ReplyDelete