Associated Press
Republican presidential candidate, former Massachusetts Gov. Mitt Romney, left, greets fellow candidate, former Utah Gov. Jon Huntsman, Jr., prior to the start the Fourth of July parade in Amherst, N.H., Monday, July 4, 2011.

With two Mormons in the current race for the Republican nomination, the poll question, "Would you vote for a Mormon for president?" keeps appearing in the news, showing that a large percentage of Americans say "no." Does that mean that neither Mitt Romney nor Jon Huntsman Jr. has a chance to be president? Let's talk about polls.

Polls consist of questions. They can show a candidate with far more support than he really has if the person drawing up the questions steers them towards the candidate's strengths while downplaying his weaknesses. The way questions are worded influences the answers.

Next, they are only a snapshot in time. Conditions change, and people's opinions change with them. Using one poll to predict how far a campaign will go is like looking at a photo of a plane taking off and predicting how high it will fly.

Finally, polls are a sampling of the larger group. To make sure that the sample is truly representative of the group, pollsters "weight" the results to make them more accurate. For example, if the group is 54 percent female but 54 percent of the poll respondents are males, the pollster will remove enough male responses to bring the gender ratio of the final sample in line with the group ratio. Weighting strategies differ between polls, which is why two polls asking the same questions on the same day can report different results.

With that in mind, let us turn to the recent incident in Texas where a pastor labeled Mormonism a non-Christian cult. Clearly, he would have told a pollster that he would not vote for a Mormon. However, when asked about his feelings for Romney specifically, he said that Romney "is a wonderful man, a highly moral man." Nonetheless, he still didn't want to vote for him and said, "I hope I don't have to."

I think that means the pastor would "have to" vote for Romney against Obama. If the initial poll question had been, "Are there any circumstances under which you would consider voting for a Mormon?" instead of "Would you vote for a Mormon?" his reply might well have been different.

Current polls suggest that resistance to Romney's Mormon faith is not as great as it was when the question was asked four years ago. He is now doing surprisingly well in Iowa, where the Mormon question hurt him last time. He is ahead by double digits in New Hampshire, which he also lost four years ago, and he is running stronger in Florida and South Carolina than he did the last time around.

That looks good, but history shows that warnings about the uncertainties of polls still apply. In 2008, primary polls in New Hampshire said Obama would defeat Clinton; in 2010, they said Mike Castle would win in Delaware and Lisa Murkowski would win in Alaska. Utah's last primary polls had Bridgewater ahead of Lee. Primary polls are notoriously subject to error because it is very difficult to draw an accurate sample of what is often a mercurial group. Polls are most useful when dealing with a straightforward question posed to a stable group, as in the final days of a presidential election.

That's why the Romney campaign touts the poll that goes above the primary voters and shows him as Obama's strongest challenger when their names are placed before the entire voting population. If that advantage makes Romney the nominee, the Texas pastor — and many others like him — won't like it, but they will "have to" go with Romney no matter what they told a pollster about not voting for a Mormon.

Robert Bennett, former U.S. senator from Utah, is a part-time teacher, researcher and lecturer at the University of Utah's Hinckley Institute of Politics.