Hacker News new | past | comments | ask | show | jobs | submit login

Correlation is not causation. How many times do people need to hear it?



If X & Y are correlated, it absolutely can be because X causes Y. And if X & Y are correlated, it's more likely X causes Y than if they were not correlated.

Parroting "correlation isn't causation" every time someone trots out an argument which isn't water tight is dumb.


When a correlation is observed all around the world, shouting “correlation is not causation” becomes a clearly weak dismissal.

Open flames and house fires are very strongly correlated.


Correct! Correlation is not causation.

But it is a pretty serious hint worth checking to see if a causal link is there.

Outright dismissal of a lead in resolving a problem is NOT helping anyone.


So how should we investigate what _is_ a potential for causation


And only recently has the US government started allowing funds again study the answer to that question:

https://www.nature.com/articles/d41586-019-03882-w




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

Search: