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

Reading the articles left me with the impression not that they were less effective, but that they couldn't prove efficacy. So they could win, or lose, but it was close.

It is also worth noting that one of the reasons why this happens in stage 3 despite having passed through a test in stage 2 that was meant to whittle this out is that improved expectations increase the strength of the placebo effect. Which makes the real effect harder to detect.



> not that they were less effective, but that they couldn't prove efficacy

Could you explain the difference?


To prove efficacy, you have to have a statistically significant win.

If you're ahead by a small margin, it isn't evidence that you're less effective. But you did not prove efficacy.




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

Search: