Reading a study without getting fooled

Cheat sheet I wish I had: check sample size, whether there was a control/placebo, whether it is in humans, and what "significant" actually means (it is about probability, not importance). Abstracts oversell — read the methods.