“Kane is polling ahead of Murphy in every media market but Philadelphia and when figures are adjusted for anticipated voter turnout, Kane’s leads the race 42% to 33%,” the pollster wrote. “When accounting for leans among undecided voters, the Kane lead is 8%.”
The at-times negative tenor of the campaign apparently hasn’t had a significant effect on voter opinions. Murphy has a 44 percent net favorable rating among Democrats; Kane’s is 40 percent net favorable – as of April 15 when the poll was taken.
As with any internal poll, take it with a grain of salt. Campaigns typically release their numbers either to boost fundraising or demonstrate viability, or both. In this case, the methodology explained below indicates that the raw survey results may have been favorable to Kane. For example, the pollster was not able to include cell phone users, who tend to be younger and more urban – where Murphy is stronger.
That said, as PoliticsPA noted in its Tuesday report on Kane’s closing ad, which highlights her support from President Bill Clinton:
Their final-week ads show where each campaign thinks it is. Murphy’s most recent spot was a broadside of attacks against Kane on a number of issues. That, plus the fact that Kane appears to be entering the final on a positive note, suggests that both campaigns see her as the frontrunner.
The pollster is Traci Copeland of Zata 3, based in Washington, DC. Here is her explanation for methodology:
We administered the survey using IVR survey technology and recorded responses on 10 questions. We conducted a randomized dial of voter households deemed likely to participate in the upcoming Democratic Primary. The sample size was 1008 respondents. The margin of error is +/- 3.1% at a 95% confidence level.
- Home Phone Ownership. Due to the mode of administration of the survey, the sample is limited to individuals who own home phones.
- Ethnicity. We asked each respondent to identify their ethnicity. The results have been weighted to reflect anticipated turnout.
- Gender. Women are slightly overrepresented in this sample.
- Age. Voters age 55+ are overrepresented in this sample.
- Geography. The survey sample is distributed across the media markets according to recent Democratic primary turnout patterns.
About IVR Surveys. IVR surveys reduce interviewer bias to zero by eliminating the live human interviewer. Every poll respondent hears the same questions read the same way, Independent analyses from publications such as the Wall Street Journal, National Council on Public Polls, American Association of Public Opinion Research and The Pew Research Center all show automated, recorded voice surveys used to record candidate preferences have an accuracy level comparable to live interviewer surveys. In the 2008 primary season, Nate Silver with www.fivethirtyeight.com concluded that two of the three most accurate pollsters were IVR surveys. In 2010 general election, a similar comparison of polls nationwide concluded that IVR surveys were ranked second and fourth most accurate.