4 ms·
Except that's unblinded so placebo would probably dominate over actual effects. You'd have to make a placebo pill as well and not know which one you were taking
by shalmanese 11y ago
Except that's unblinded so placebo would probably dominate over actual effects. You'd have to make a placebo pill as well and not know which one you were taking until after you gather all your results.
- gwern 11y agoThat is what I meant, yes. Please forgive me for not including a full description of my self-blinding method in a quick comment. > until after you gather all your results. Not all of them; you only need to be blinded for each unit. Once the data is collected for that unit, then you can unblind yourself (and this is the easiest way to go).