In an Elon University Poll released today, Barack Obama trails John Mcain by 5.3 percentage points in North Carolina, just beyond a margin of error of 4.9 percentage points. However, this data is based on a poll question that asked respondents if they would vote for the "Democratic party" or "Republican party" in the upcoming Presidential race. In the Elon news release, parroted by WRAL, the real news--involving actual people--is that North Carolinians have a far less favorable opinion of Barack Obama than of John McCain: 37 to 54. By comparison, the latest New York Times/CBS Poll shows respondents across the country as having a roughly equal opinion of the candidates: 45 percent held a favorable opinion of Obama; 44 percent for McCain.
The Elon poll and the Times/CBS poll differ in several significant ways: First, on the "favorable" question, the Elon poll provided a sliding scale of 1-10, and counted 6-10 as "favorable" and 1-4 "unfavorable." The Times/CBS poll simply asked respondents if their opinion of the candidates was favorable, not favorable, undecided, or if they had no opinion. Also, the Times/CBS poll asked respondents the "favorable" question first, before asking who they would vote for. Meanwhile, the Elon poll posed the "favorable" question after first asking respondents who they would vote for in the Presidential race, in addition to several questions about the North Carolina candidates for governor and U.S. senator--including, "Do you know people that will not vote for a Senate candidate that is too old?" (16.8 percent said yes, by the way.)
It's impossible to know the precise effect that voters' "favorable" or "unfavorable" opinions of candidates will play on the election, but in the Times/CBS poll, which gauged an equal public opinion of Obama and McCain, Obama wins, 48 to 43, within the poll's margin of error.