Even if you manage to identify an effective argument to support
your position, there are things you can do to improve your chances of
persuasion. First, the delivery method can matter. There’s evidence
that people are much more likely to complete a survey if asked in
person rather than by e-mail,[69] for example. Other experiments
have come to similar conclusions, finding that people can be more
convincing face-to-face than by phone, post or online.[70]
The timing of messages can also make a difference. According to
Briony Swire-Thompson, a psychologist at Northeastern University,
researchers are increasingly thinking about how ideas wane. ‘It’s this
concept that once you change someone’s mind, it doesn’t stick
permanently.’ In 2017, she conducted a study asking people whether
they believed certain myths, like carrots improving your eyesight or
liars moving their eyes in a certain direction.[71] The study found that
they could often correct false beliefs, but the effect didn’t necessarily
last. ‘If you get a correction, you might reduce your belief initially, but
as time goes on you’re going to re-believe in the initial
misconception,’ Swire-Thompson said. It seems repetition matters:
new beliefs survived longer if people were reminded of the truth
several times, rather than just given one correction.[72]
Thinking about the moral position of others. Having face-to-face
interactions. Finding ways to encourage long-term change. All of
these things can help improve persuasion. And it happens that they
are also part of the deep canvassing approach advocated by the Los
Angeles LGBT Center. Which brings us back to that dubious LaCour
and Green paper. Although the study was retracted in 2015, the story
didn’t end there. The following year, David Brookman and Joshua
Kalla – those two Berkeley researchers who’d found the problems in
the original paper – published a new study.[73] This one focused on
transgender rights. And this time they’d definitely collected the data.
Comparing deep canvassing with results from a control group,
they’d found that a ten-minute conversation about transgender rights
could noticeably reduce prejudice. It didn’t matter whether the
canvasser was transgender; the change in voters’ opinion persisted
regardless. The change in belief also seemed to be resistant to
attacks. After a few weeks, the researchers showed people anti-