[throwaway48540]

Empirically researched data would be better - if people went and read the study in detail and saw that the questions asked don't lead to the conclusion that they claim, such as this case. They read a headline or check a sheet of numbers and think they understand the world.

Since people are apparently unable to do that, talking to people is the next best thing.

[moi2388]

If people are unable to get the results from a study, but get confused by a headline, what makes you think they’ll be able to draw correct statistical conclusions from talking to a couple of people with inherent bias and small sample size?

[throwaway48540]

People are social creatures, and if they see first hand that the world doesn't match their expectations, it might get them to think about it in more detail.