Then they were exposed to one of two messages, one positive (he won the Gandhi Peace Prize) and one negative (he flip-flops on the issues.) Some students were told the messages came from one of their friends, and the others were told they came from a reputable radio program.
"I shouldn't expect my friend to know more about someone like Kucinich than myself," so the friend would not be very reliable as a source, Johar says. The radio report, she adds, from a "reliable program" would be a better source.
There was a "significant difference" between the people who had a strong opinion of Kucinich, and those who didn't, she says. The students who were ambivalent about the congressman were easily swayed by the report, regardless of its source. But students with a strong opinion were likely to be swayed only if the report came from the radio broadcast, not a friend.
Thus, she adds, ambivalent voters are likely to be less discriminating about the source of a report and more likely to be led astray by a disreputable source. That finding was reinforced by two additional experiments on ambivalence, including one involving consumer attitudes toward a shower gel.
Of course, there are many factors influencing an election. And there are many ways that voters can be swayed.
"If you hear something often enough, you tend to believe it," even if it's not plausible, Johar says. "If you attack Barack Obama 10 times, and I hear it 10 times, it might be from disreputable sources, but just the fact that it's being repeated so many times gives it some weight. After awhile, you quit asking where you heard it, and just kind of absorb it."
And that's regardless of whether you have a weak or strong opinion of the Illinois senator, she adds.
All of that could lead to a lot of hostility toward many of the candidates, but that's not necessarily all bad. According to a study at Ohio State University in Columbus, a little hatred would be good for voter turnout. The research was based on a nationwide survey covering the presidential elections from 1972 to 1988.
The bottom line: People who liked both candidates were less likely to vote. People who really disliked one of the candidates were much more likely to vote.
So what does all this tell us about the coming election? It's probably going to get nasty, but a lot of people will show up at the polls.