Hacker News new | ask | show | jobs
by jeffxtreme 1319 days ago
As in general with associative studies, it is hard to make a distinct conclusion on causation.

Do we know if alcohol decreases gray and white matter volumes? Could the causation run in the reverse direction? Bi-directionally? Or could there be some common cause (e.g., proneness to "risky" behavior) that leads to both?

https://en.wikipedia.org/wiki/Correlation_does_not_imply_cau...
2 comments
ryandvm 1319 days ago
Every study that comes through here somebody invariably trots out the old "correlation does not equal causation" argument. You're not wrong, but you know, sometimes it is causation.

Rather than disregard every study, I just sort of internally give it a 50% weight and move on. Maybe it's bogus, maybe it's not, but surely some of them are correctly identifying causation.

Life is too short to quibble about this crap.

link
yjftsjthsd-h 1319 days ago
> Every study that comes through here somebody invariably trots out the old "correlation does not equal causation" argument.

Well, it's still true, and still applicable.

> You're not wrong, but you know, sometimes it is causation.

No, I don't know, and neither do you. That's... kind of the problem.

link
TrueSlacker0 1319 days ago
"Life is too short to quibble about this crap. "

Yup, so have a drink when you want, or don't. We will all die eventually.

link
Mathnerd314 1319 days ago
... Or will we? Anti-aging research is an area.
link
DefineOutside 1319 days ago
even with every disease cured and without aging, you still will die by some random event unless you hide away in a bunker - in which case the bunker will eventually collapse and you will still die
link
superhuzza 1319 days ago
>Life is too short to quibble about this crap.

Then why bother doing science at all?

There are ways to determine the correlation vs causation problem, but yes it takes more effort - possibly redesigning studies, follow up studies etc.

But that's just the price of knowledge. And knowledge is worth quibbling over.

link
arnaudsm 1319 days ago
Indeed, we'd need an interventional study. Forcing a random study group to drink for decades seems unethical and expensive, but I wish we had the results.
link