Google+ post
Tim HuttonTim Hutton - 2012-01-24 18:23:29+0000 - Updated: 2012-01-24 18:23:29+0000
Originally shared by Mike StayThis is the first I've ever heard of being able to take uncontrolled experimental data (like people smoking--you can't ethically do an experiment where you force someone to smoke) and still extract information about causation. This is really important!

If correlation doesn’t imply causation, then what does? | DDI

Shared with: Public

This post was originally on Google+