Taking "correlation does not imply causation" back from the internet

by sixes_and_sevens 7y3rd Oct 201270 comments

43


(An idea I had while responding to this quotes thread)

"Correlation does not imply causation" is bandied around inexpertly and inappropriately all over the internet.  Lots of us hate this.

But get this: the phrase, and the most obvious follow-up phrases like "what does imply causation?" are not high-competition search terms.  Up until about an hour ago, the domain name correlationdoesnotimplycausation.com was not taken.  I have just bought it.

There is a correlation-does-not-imply-causation shaped space on the internet, and it's ours for the taking.  I would like to fill this space with a small collection of relevant educational resources explaining what is meant by the term, why it's important, why it's often used inappropriately, and the circumstances under which one may legitimately infer causation.

At the moment the Wikipedia page is trying to do this, but it's not really optimised for the task.  It also doesn't carry the undercurrent of "no, seriously, lots of smart people get this wrong; let's make sure you're not one of them", and I think it should.

The purpose of this post is two-fold:

Firstly, it lets me say "hey dudes, I've just had this idea.  Does anyone have any suggestions (pragmatic/technical, content-related, pointing out why it's a terrible idea, etc.), or alternatively, would anyone like to help?"

Secondly, it raises the question of what other corners of the internet are ripe for the planting of sanity waterline-raising resources.  Are there any other similar concepts that people commonly get wrong, but don't have much of a guiding explanatory web presence to them?  Could we put together a simple web platform for carrying out this task in lots of different places?  The LW readership seems ideally placed to collectively do this sort of work.