Trump is a good president” Or when an article says 45% of people saw no improvement; 10% of people reported side effects. I think just training people to ask what’s missing from the claim - what does side effect mean? What were the sample size and demographics?