Question:
Now, if the 2012 presidential election was being held today and the race was between Barack Obama and Mitt Romney who would you vote for? (OPTIONS ROTATED)

Are you leaning more toward voting for Mitt Romney or Barack Obama?

About this Poll
Number of Interviews: 430 Likely Voters in the Commonwealth of Pennsylvania
Margin of Error: +/-5% at a 95% level of confidence
Fielding Dates – November 1-3, 2012
Method of Interviewing: Live telephone interviewers

Method of Sampling: The sample for this study was drawn randomly from both a list of registered voters in the state of Pennsylvania and from the universe of wireless numbers in the Commonwealth. First, voters were randomly selected from a list of registered voters in Pennsylvania that is provided by the Commonwealth of Pennsylvania. This list has been updated to reflect changes in phone numbers for voters since their initial registration. However since the list updating procedure will not match voters to wireless phone numbers a supplemental sampling procedure is employed. The supplemental sample includes a random selection of wireless numbers for cell phones in Pennsylvania. While the proportion of registered voters in Pennsylvania who exclusively use cell phones is undetermined, it is estimated that approximately a quarter of Americans use only cell phones. We thus supplement the sample drawn from voter files with a sample drawn randomly from wireless phone numbers in Pennsylvania. Respondents from the cell phone sample were asked to identify if they used only cell phones for their personal use, with those using both cell phones and landlines excluded form participation in the survey. The overall sample included 314 responses from the list of registered voters and 116 responses from the cell phone sample.

Likely Voter Screens: To identify likely voters a multiple stage process is utilized. For the sample drawn from the Pennsylvania voter registration files only individuals who voted in at least 1 of the last 8 general or primary elections in Pennsylvania OR who had registered to vote in the state since 2009 were included in the sampling frame. For the cell phone sample respondents were asked if they had voted in 1 of the last 8 elections in the state or had registered to vote since 2009. Among those that met the voting history requirement, two follow up questions were asked to determine voter likelihood. First respondents were asked to confirm their voter registration status (Question 1) with those indicating that they were either not registered or unsure about their registration status being excluded from participation. Finally, individuals were asked to express the likelihood that they will vote (Question 2) in the November election, with only those indicating that they will definitely or most likely vote allowed to participate in the full survey.

Weighting: The data was weighted by gender and region of the state.