MRuleI

363 points · 10 comments · view on lemmy.world

10 Comments

bjorney@lemmy.ca · 71 pts · 150d (1 reply)

This explanation is a really terrible explanation - it doesn't even explain the point of the study or what multiple comparisons are in statistics...

https://xkcd.com/882/

This is what the salmon study was saying - if you did a whole brain analysis with fmri without an a priori prediction you need to adopt a radically more stringent threshold for significance

The salmon in the MRI was a bowl with a million jellybeans in it

muzzle@lemmy.zip · 71 pts · 149d

The explanation was pretty good without going into the statistical details which is perfectly OK for most people on social media.

Smorty@lemmy.blahaj.zone · 40 pts · 150d (2 replies)

so, as far as i understand it:

  • either dead fish are geniuses
  • or every single MRI result we ever got is bs and thus thousands of papers using them as a main component are invalidated

do i get that right?

if so, mygosh, i gotta get myself a dead fish to see if im faking depression or not.

maybe its an LM type situation: it doesnt matter if its fake "roleplaying" or not, the actions are what matters.

potatopotato@sh.itjust.works · 48 pts · 149d

Not all MRI studies, just a large number of particularly lazy ones. Unfortunately lazy journalists pick up the lazy science and so these studies tend to be really popular in lazy clickbait journalism.

Basically there's still good science, as well as a parallel economy of bullshit from people who have long ago stopped giving a shit.

davidagain@lemmy.world · 29 pts · 150d

do i get that right?

Yes.

i gotta get myself a dead fish to see if im faking depression or not.

No, no, not like that.

WillStealYourUsername@piefed.blahaj.zone · 36 pts · 149d

wow amazing. dead salmon can read human emotions :O

maltasoron@sopuli.xyz · 22 pts · 149d
GalacticGrapefruit@lemmy.world · 16 pts · 149d (1 reply)

Diogenes, holding up a salmon.

NoTagBacks@lemmy.dbzer0.com · 2 pts · 149d

Behold: a salMAN.

cypherpunks@lemmy.ml · 7 pts · 149d