The overwhelming majority of registered New Jersey voters claim that a candidate's weight, gender or sexual orientation doesn't matter to them, but that doesn't seem to be completely accurate.

Newark Mayor Cory Booker (Larry Bussaca, Getty Images)

A new Fairleigh Dickinson University-PublicMind poll shows voter uncertainty about Cory Booker's sexuality is estimated to be costing him six points in his U.S. Senate race against Republican Steve Lonegan.

Over the summer, Booker sparked controversy when he told the Washington Post, "I love seeing on Twitter when someone says I'm gay, and I say, 'So what does it matter if I am? So be it. I hope you are not voting for me because you are making the presumption that I'm straight.'"

The GOP nominee Steve Lonegan, seized on the comments, calling them "kind of weird," and saying that "maybe it helps [Booker] get the gay vote." In public statements made before this election, Booker repeatedly referred to girlfriends, and referred to himself as straight.

Only 5 percent of New Jersey voters think that Booker is gay or bi-sexual, but 74 percent say they aren't sure about the Senate candidate's sexuality. This may seem like a high figure, but almost the same number say they aren't sure about Lonegan's sexuality.

Eighty-four percent insist sexual orientation doesn't matter in how they evaluate a candidate, but the results of the survey tell a different story.

"People today know that they're not supposed to say that a candidate's sexuality matters," said Dan Cassino, the director of experimental research for the poll. "However, the results seen clear that to some voters, it matters a lot."

While only a small number of voters think Booker is gay, these voters are about twice as likely to support Lonegan as Booker. While only about 5 percent of voters think that Booker is gay or bisexual, this group constitutes 8 percent of Lonegan's support and only 2 percent of Booker's. Among those voters who say that Booker is straight, Booker has a huge lead (67 to 17 percent), but among the largest group of voters who say that they don't know whether Booker is gay or straight, the race is fairly close, with the Newark mayor leading by only 10 points (42 to 32 percent). Lonegan's support doesn't change significantly based on perceptions of his sexuality.

Among political moderates, Booker enjoys a huge lead (48 points) among those voters who think he's straight, and a smaller lead (9 points) among those that aren't sure. Among conservatives who think he's heterosexual, Booker is tied with Lonegan; conservatives who think Booker is gay are 82 points more likely to vote for Lonegan.

As a result of the above figures, Cassino said it's possible to calculate how the election results would be different if Booker's sexuality was viewed no differently than Lonegan's. If almost no one thought Booker was homosexual or bi-sexual, and larger numbers thought he was straight, he would be doing a little better among conservatives and liberals, and much better among moderates.

According to Cassino, in the absence of any ambiguity about his sexuality, Booker would be expected to do about six points better in the polls than he currently is.

"Six points may not seem like a lot, but Booker has been trying to make a statement with this race," explained Cassino. "In terms of perception, six points can be the difference between a narrow win and a blowout, and right now, Booker is on the wrong side of that."

Weight And Gender

Just 7 percent of voters say Gov. Chris Christie's weight makes it harder for him to be an effective leader and 90 percent say that it makes no difference at all. Two-in-three voters who say that Christie's weight is a concern are Democrats, and would have been unlikely to vote for Christie anyway.

Seven percent of Democrats say that democratic gubernatorial challenger Sen. Barbara Buono's gender makes it harder for her to be an effective leader, joined by only 3 percent of Republicans. Ninety-two percent of voters say it doesn't make a difference.

"Voters seem to be moving past the idea that we should be evaluating candidates based on in-born characteristics, or appearance," explained Cassino. "In some cases, as with gender, this seems to be working, but with sexuality, there seems to be a ways to go."

The poll of 702 registered voters in New Jersey was conducted by telephone with both landline and cell phones from Sept. 30 through Oct. 5, and has a margin of error of +/-3.7 percentage points.