Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Randomized controlled trials are great, they are incredibly certain to be right. But experience has a whole other quality, it is much less reliable but a whole lot more comprehensive. An RCT can only measure effects on what can be used to select a large group of people. It cannot measure the effects of a weird anomaly that is recorded nowhere. Meanwhile experience can extrapolate from a very few examples.

So it is not at all obvious that we should favor RCTs over experience. Because they have very different things they can measure.



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: