File: Larry Pascua carries a rainbow flag at a celebration for the U. S. Supreme Court’s rulings on Prop. 8 and the Defense of Marriage Act in the Castro District in San Francisco, on Wednesday, June 26, 2013. (AP Photo/Mathew Sumner)

Editor’s Note: Since the publication of this post on a study examining how easily people’s minds can be changed concerning same-sex marriage, a co-author has disavowed its findings. Donald P. Green is seeking a retraction of the study from the journal Science, which originally published the research.

For David Fleischer, it was a gut-punch when Proposition 8, the ballot initiative banning gay marriage, passed in California in 2008. He wondered: What could we do to change people’s minds on this subject?

So through his Leadership LAB at the Los Angeles LGBT Center, he launched an experiment. Every other weekend, he sent canvassers to neighborhoods where Prop 8 had passed by a 2-to-1 margin, from the Latino-heavy neighborhoods of East L.A. to black neighborhoods such as Inglewood to the predominantly white Northeast. The goal was to simply talk with the voters at their doorsteps, as they were walking their dogs or sitting in their cars, and engage them on the subject of same-sex marriage.

The experiment appears to have been a success, according to a study being published in the journal Science. Support for same-sex marriage jumped eight percentage points among those voters who had a single conversation with a gay canvasser, according to the study. And not only did that support persist a year later, it also trickled over to other members of the household, the study’s authors found.

Prior to the canvass, the 132 voters who spoke with a gay canvasser resembled Nebraskans in terms of their support for same-sex marriage. After the canvass, they looked more like residents of Massachusetts, the study’s lead author Michael J. LaCour said.

Support for same-sex marriage in this group surged even further in 2013, after the U.S. Supreme Court left in place a lower court’s ruling striking down Prop 8.

LaCour thinks one reason the approach worked is that the canvassers were not trying to collect signatures or garner votes, but simply change minds. “This was a really great way to test out what is the causal effect of talking to a gay person,” said LaCour, a doctoral student at the University of California at Los Angeles.

Another reason could be that “the mental image that came to mind when they thought of a gay person or lesbian was permanently changed by this conversation,” said Donald P. Green, a Columbia University professor of political science and expert on public opinion research who was co-author of the study.

Green said he, like other social scientists, had previously come to believe that changing people’s minds in this way was “very, very difficult.” Typically, he said, voters changed their opinion about hot-button issues in the moment but “when they return to their previous milieu, it snaps back,” he said.

But these results suggest that meaningful, two-way conversations can go a long way toward shifting people’s views on deeply held subjects, he said.

Will this unleash an enormous canvassing campaign come election time?

Green offered a few cautions. On average, the canvassers spoke with voters for 22 minutes — much longer than the average political canvasser. The results were not long-lasting when the canvasser was straight.

Still, Fleischer believes the results offer hope, not only on issues of importance to gays but for on a host of issues. “We’re able to reduce prejudice in a way that lasts,” he said. “That’s promising.”

In fact, LaCour has already started testing this approach with other issues. He just wrapped up a study that found that voters became more supportive of abortion rights when approached by a canvasser who had personally had an abortion, he said. His next experiment will involve undocumented immigrants.

You can watch a video of Green discussing the study here.