News for hackers. Stuff that matters.

A comment on The opposite of a shibboleth is a frisco.

Re:The opposite of a shibboleth is a frisco by Cpoll on Tuesday January 02, 2024 @05:32PM (#38847985)
> It’s not like scientists are physically incapable of saying “correlation does not imply causation,” in the sense that Ephraimites were presumably physically incapable of pronouncing the word “shibboleth.” It’s just that no real scientists would.

Genuinely curious: What do real scientists say instead at the myriad of pop-science articles that superimpose two graphs and declare that jello cures laryngitis?

[ Parent | on HN ]
Re:The opposite of a shibboleth is a frisco by dekhn on Tuesday January 02, 2024 @08:16PM (#38849494)
I am an ex-scientist and I can confirm that scientists frequently say this. Normally it's what you say after you read a paper from your competitor.

Realistically, almost no modern research demonstrates causation; nearly all work is associational, ideally with a prepdonderance of supporting evidence and no counter-evidence (which merely means it hasn't been falsified. Now falsified- that's a work scientists rarely if ever use).

[ Parent | on HN ]
Re:The opposite of a shibboleth is a frisco by TheCoelacanth on Tuesday January 02, 2024 @11:08PM (#38850612)
It depends a lot on the field. In fields where you can ethically and cost effectively set up controlled experiments, demonstrating causation is pretty common. In fields where you can't, it's rare.
[ Parent | on HN ]
Re:The opposite of a shibboleth is a frisco by dekhn on Tuesday January 02, 2024 @11:19PM (#38850682)
I worked in biophysics with molecules and I don't think we considered even our most carefully run experiments as demonstrating causation. Even in physics nearly everything is presented as a probabilistic claim.
[ Parent | on HN ]
Re:The opposite of a shibboleth is a frisco by RecycledEle on Tuesday January 02, 2024 @05:35PM (#38848012)
If someone dies not say "Correlation does not imply causation" then they are a liberal arts major.

All scientists say "correlation does not imply causation."

[ Parent | on HN ]