3 ms·
Regression to the mean is a huge problem with anecdotes. Some people get better. It may be due to some treatment, or they may just get better. This is why we ha
by csours 7y ago
Regression to the mean is a huge problem with anecdotes. Some people get better. It may be due to some treatment, or they may just get better. This is why we have structured studies.
- jschwartzi 7y agoYeah, I guarantee you that even with a 10% fatality rate 90% of people will get better spontaneously on their own. So any prospective treatment would have to do better than the recovery rate without treatment. This is really a situation where the phrase "the plural of anecdote is not statistics."