I would add to this a bit actually. Telling people that correlation doesn't equal causation is all well and good, but helping them spot the difference is even better.SciMal said:1) I really, really, really wish that people understood science better. How it's performed, when it's good, when it's bad, and what you can conclude from the findings. Very few things piss me off more than going onto a fluff article on CNN or something about "Eggplant may cure cancer!" and seeing a bunch of "It totally worked for ME..." in the comments. The plural of personal experience is not data! Also, correlation is not causation!
Anytime you see an article talking about how A cures B, or may reduce C you need to stop and ask two questions about the original study:
A) Is it an observational study or a clinical study?
B) Did they control their variables?
If the answer to the first is observational, and/or the answer to the second is no, then the study doesn't prove a damn thing and no conclusions can be drawn. Such a study may be used to form a hypothesis that you can later test in more rigorous circumstances, but that's it. Unless it was a clinical study and the variables were controlled so they know some other factor than the one they were testing didn't significantly impact the results then the study proves nothing, and should not be treated as proof of anything by anyone at any time.
And if anyone is wondering what the answers to those questions are for a lot of health, medical, nutritional, etc. studies are it's that most of them are observational and didn't control their variables, yet people (usually those who zealously agree with their conclusions) will happily toss them out there as proof of whatever the topic was, regardless of how weak they are as evidence.