In this month’s Scientific American, Daniel Willingham describes a friend who clings ferociously to the discredited belief that vaccines cause autism. The puzzle?
“My friend insists that he trusts scientists. In this respect, he is like most Americans. In a 2008 survey by the National Science Foundation, more respondents expressed “a great deal” of confidence in science leaders than in leaders of any other institution except the military. On public policy issues, Americans believe that science leaders are more knowledgeable and impartial than leaders in other sectors of society, such as business or government. Why do people say that they trust scientists in general but part company with them on specific issues?”
Of course, in this specific instance, it could simply be our well-documented willingness to seek information that confirms our beliefs and hold fast to those beliefs in the face of discrediting information. But I think Daniel has a point: People’s expressed trust in scientists doesn’t really mesh with the trust revealed by their behaviour.
So, why do people say they trust scientists but fail to trust scientific findings? I have three thoughts:
- They have low trust in the funders and gatekeepers of scientific research: Government, industry, and the press. They think that the ‘real’ findings are out there, but are being hidden or repressed by scientists’ various nefarious masters.
- They are bad at distinguishing good science from junk science. Both quacks and legitimate scientists hold PhDs. The distinctions that help us make sense of which to trust are either non-obvious to laypeople and/or depend on trust in the other institutions mentioned above.
- It’s hard to know what the scientific consensus is on an issue from press coverage alone. What’s the best-supported position on any given issue? You’d have a hard time gleaning it on just about any topic from press coverage. Press coverage doesn’t meta-analyze — it often provides a he-said/she-said account with equal time allotted to talking heads from two sides of an issue.
Steve Saideman’s suggested answer is perfect in that it offers both snark and parsimony:
4. It requires reading.