Hacker News new | ask | show | jobs
by pbhjpbhj 4823 days ago
By "assume causation" I meant "assume there is a causative link", which for cancer appears to be true to a degree. So if I specify and say "is it sound to assume a causative link [without specifying the cause and effect direction or that link] where a strong direct correlation is observed" (or similar wording) would you go for that?

There are many examples of course to prompt this clarification - fatness causes over-eating, bruises cause people to get hit, et cetera. There are also likely many examples where the direction of the causation is not clear - poverty and [minor] theft say.

Can you [or anyone] give an example where a causative relationship as a first hypothesis is ludicrous without questioning direction of the relationship?