Andrew Huberman· PhD
we need 60 people 20 per group right yeah
Loading…
The headline is broadly defensible, but the qualifications matter. Effect sizes vary by population, the strongest claims rest on shorter trials, and credible voices push back on how it's typically framed.
we need 60 people 20 per group right yeah
Every Sunday: the week’s new conflicts and verdict changes — and nothing else.
Native comments, Twitter mentions, and Reddit threads about this claim — surfaced together so the conversation isn't fragmented across platforms.
Bookmarking — the dossier-vs-overview split is the right call. Most of the time I want overview; sometimes I want receipts.
Would love a "what would change this verdict" RSS feed. Sign me up if it exists.
based on this we estimated that an N of 20 n is sample size in each belief condition the final sample would provide ninety percent hour to detect an effect of this magnitude at an alpha of 0.0 0.5 and a two-tailed um test