RealClimate features a social science study by Jack Zhou of Duke University that goes beyond “evidence doesn’t matter” to “evidence backfires”.
It makes the usual mistakes that these sorts of study make, reducing persuasion to a trivial and contrived exposure in a lab setting, a “treatment”.
It’s certainly true that such experiments can detect real effects. But is the effect really what you are looking for? Changing an opinion is a gradual, not an abrupt, process.
Marketers understand this. Here is renowned marketing consultant Seth Godin in a short entry called “The Flip Is Elusive”
the flip isn’t something that happens at the first glance or encounter with new evidence.
This doesn’t mean the evidence doesn’t matter.
It means that we’re bad at admitting we were wrong.
If you want to change people’s minds, you need more than evidence. You need persistence. And empathy. And mostly, you need the resources to keep showing up, peeling off one person after another, surrounding a cultural problem with a cultural solution.
Perhaps “scicomm” people need to get in touch with people actually in the business of changing people’s minds.
In the meantime, if someone has a study of how evidence influences opinion that doesn’t address persistence, never mind empathy, I would like them to go away and leave us alone. We have work to do.