1146 19 St.

, NW, Suite 200 Washington, DC 20036 (202) 463-7300

th

Interview dates: January 30-February 3, 2012 Base: 2475 Americans

Ipsos Poll conducted for Reuters, February 2012 Quarterback Primary Poll NOTE: all results shown are percentages unless otherwise labeled. These are findings from an Ipsos poll conducted for Thomson Reuters from January 30 February 3 , 2012. For the survey, a sample of 2475 Americans was interviewed online. The precision of the Reuters/Ipsos online polls is measured using a credibility interval. In this case, the poll has a credibility interval of plus or minus 2.3 percentage points for all respondents. For more information about credibility intervals, please see the appendix. The data were weighted to the U.S. current population data by gender, age, education, ethnicity and a political values scale. Statistical margins of error are not applicable to online polls. All sample surveys and polls may be subject to other sources of error, including, but not limited to coverage error and measurement error. Figures marked by an asterisk (*) indicate a percentage value of greater than zero but less than one half of a per cent. Where figures do not sum to 100, this is due to the effects of rounding. QUARTERBACK PRIMARY POLL Q. If the age restriction did not exist and only quarterbacks in this years NFL playoffs were eligible to run for President of the United States, who would you vote for?(Select one)
PARTY ID & REGION
th rd

Tim Tebow Eli Manning Tom Brady Drew Brees Aaron Rodgers Ben Roethlisberger Joe Flacco Alex Smith Matt Ryan Matt Stafford T.J. Yates Andy Dalton

Total 26 18 17 15 7 4 3 3 2 2 2 1

Democrat 20 21 17 17 7 5 4 4 1 2 2 1

Republican 39 13 14 15 8 4 1 2 2 1 2 1

Independent 16 20 22 13 8 3 6 4 4 3 1 1

Northeast 18 34 23 9 5 5 1 1 * 1 2 1

Midwest 23 15 16 14 18 4 2 2 1 3 1 1

South 29 12 15 21 3 4 6 3 3 2 2 1

West 29 15 16 13 6 4 1 6 4 2 4 1

DEMOGRAPHICS

Total Tim Tebow Eli Manning Tom Brady Drew Brees Aaron Rodgers Ben Roethlisberger Joe Flacco Alex Smith Matt Ryan Matt Stafford T.J. Yates Andy Dalton 26 18 17 15 7 4 3 3 2 2 2 1

Male 23 15 18 18 8 5 4 3 2 2 2 1

Female 28 20 15 13 6 4 2 3 3 2 2 1

White 28 18 16 15 9 5 2 2 2 1 1 1

African American 18 15 20 21 3 6 4 5 1 4 3 1

Hispanic 24 15 19 12 3 2 8 7 3 2 6 1

18-24 25 15 14 13 9 5 4 5 4 2 2 3

25-34 28 15 21 13 5 5 6 4 1 1 1 1

35-44 24 15 17 20 7 5 1 1 2 2 6 1

45-54 21 26 17 15 5 4 2 1 5 2 1 *

55+ 28 16 16 15 9 4 2 4 1 2 1 1

How to Calculate Bayesian Credibility Intervals The calculation of credibility intervals assumes that Y has a binomial distribution conditioned on the parameter \, i.e., Y|~Bin(n,), where n is the size of our sample. In this setting, Y counts the . This model is often called the likelihood 1 function, and it is a standard concept in both the Bayesian and the Classical framework. The Bayesian statistics combines both the prior distribution and the likelihood function to create a posterior distribution. The posterior distribution represents our opinion about which are the plausible values for adjusted after observing the sample data. In reality, the posterior distribution is ones knowledge base updated using the latest survey information. For the prior and likelihood functions specified here, the posterior distribution is also a beta distribution ((/y)~(y+a,n-y+b)), but with updated hyper-parameters. Our credibility interval for is based on this posterior distribution. As mentioned above, these intervals represent our belief about which are the most plausible values for given our updated knowledge base. There are different ways to calculate these intervals based on . Since we want only one measure of precision for all variables in the survey, analogous to what is done within the Classical framework, we will compute the largest possible credibility interval for any observed sample. The worst case occurs when we assume that a=1 and b=1 and . Using a simple approximation of the posterior by the normal distribution, the 95% credibility interval is given by, approximately:

.. For the QB poll published on February 3 , 2012, the Bayesian Credibility Interval was adjusted using standard weighting design 2 effect 1+L=1.3 to account for complex weighting Analysis Domain American public (online sample) Sample size 2475 Credibility intervals 2.3%
rd

Bayesian Data Analysis, Second Edition, Andrew Gelman, John B. Carlin, Hal S. Stern, Donald B. Rubin, Chapman & Hall/CRC | ISBN: 158488388X | 2003 Kish, L. (1992). Weighting for unequal Pi . Journal of Official, Statistics, 8, 2, 183200.