Excellent critique of the Ontario checklist study I previously wrote about. It gets into a little more technical detail about the role of sample size than I did and makes the excellent point that post-hoc power calculations on negative studies will always show they were underpowered.
At the intersection of medicine and technology