[onion2k]

Sometimes correlation isn't causation, but if we just say "correlation =/= causation" every time something happens we'll never discover anything. If two things are correlated it's at least worth investigating further, and quite often turns out to be a causal link.

[AznHisoka]

And its definitely worth examining if there is a rational, simple hypothesis that implies a relationship between the two. Here, that is clearly the case: could what you think about something impact your health? It doesnt seem like such a stretch to believe so (see placebos, etc)

On the other hand, if you found a correlation between butter exports in India and mortality rates, you probably shouldnt dig further into whether there’s a causation relationship. It fails the Occam’s Razor test missrably.

[slx26]

You are right, but in my opinion the page/writer completely fails when they don't explain and clearly state that the possibility of "correlation =/= causation" exists. And it fails because we lose 80% of the people to blind belief towards what the article says, and from the other 20%, 15% are saying "correlation =/= causation", and a 4% like you are saying in response: "sure, but that's something we need to investigate to tell".

I understand that the balance between science and selling the science is hard, but it can be done, and imho they could be doing it better.

[adrianN]

I'm reasonably sure that the set of things that are correlated is much, much larger than the set of things that have a causal relationship. Check for example http://www.tylervigen.com/spurious-correlations

[onion2k]

The number of things that aren't correlated is waaaaaaay bigger. It's still worth investigating when there's a possible link.