Stereotypes tend to be so deeply engrained that it can take months of intense intervention to truly reduce bias. But a new study suggests that a 10-minute conversation can also make a difference, if it's done the right way.
The study tested "deep canvassing," an approach to door-to-door campaigning that involves a carefully designed two-way conversation instead of a canned message.
"In a lot of standard door-to-door canvassing, you as a canvasser have a script, you knock on a door, you read through the script, and you go on to your next house," says Josh Kalla, a PhD student in political science at the University of California, Berkeley, who worked on the study with Stanford University professor David Broockman. "It's very much a one-way interaction, where basically you the canvasser deliver a monologue. Here, a lot of the canvass involves the voter themselves doing the work."
In the study, researchers surveyed voters in Miami-Dade County about their attitudes towards transgender people. Then, canvassers later showed up at their doors, and launched into a conversation. Voters were asked to remember a time when they had been judged for being different, and then asked to think about how that might help them understand transgender people's experiences. (The control group heard a message about recycling instead.)
"It really builds upon this theory of self-persuasion, where the canvasser is not just reading talking points to a voter, but to get the voter to see the world through a little different lens, and draw their own conclusions, and hopefully persuade themselves," says Kalla.
The intervention shifted attitudes—and the effect lasted. Three months later, the voters who had the conversation about transgender people were less phobic than before, and more supportive of a new local anti-discrimination law than those who hadn't had the visit.
In a previous study, different researchers tested whether a "deep canvassing" intervention could reduce homophobia and increase support for same-sex marriage. That study was later retracted because data was fabricated. But the new study proves that the approach can work.
Unlike the older study, which had suggested that the intervention might only work when it came from gay and lesbian canvassers, the new study found that both transgender and non-transgender people were equally effective. "Which theoretically is interesting, but also practically is very interesting," says Kalla. "If you're trying to create a social movement, and trying to change minds, it's very important for the allies to be able to be effective allies."
That doesn't mean the technique will always work—another intervention that attempted to change attitudes about abortion didn't seem to have an effect. (The researchers are redesigning the intervention to see if changing the message might make a difference.) But there's already evidence that deep canvassing can increase voter turnout. And it's possible that this approach could be used in presidential campaigns.
"We're definitely having conversations with people who want to see if it can be applied," Kalla says.