June 13, 2020

People are too slow to update beliefs in the face of stronger opposing evidence

Bayesian updating

Anecdotes and other less rigorous forms of information can and should inform our beliefs and priors. But if we are presented with opposing evidence in a stronger form (e.g. Randomized-Controlled-Trials or well designed studies), we should be keen to update.

I tend to get irritated when people cling to beliefs that have been discredited. I used to think I was annoyed because people were forming beliefs on faulty sources of evidence. Now I realize the issue was not that they formed beliefs on less rigorous evidence, but that they seemed hostile to updating the belief in light of new, stronger evidence.

I think this is one thing I do relatively well: rather than feeling defensive when presented with new evidence, I generally feel excited at the chance to develop more accurate beliefs. I’m trying to keep an eye out for any beliefs I feel protective over, as they are probably ones I would find it harder to discard in the light of new evidence.

Updated Sep, 13 2020


#epistemics #skepticism