Hacker News Reader

is a small proof of concept study not a study anymore? gastric biopsies aren’t exactly easy to obtain at scale.

They are studies, and frankly, without something like this, doing things like the appropriate power calculations and risk assessments for larger studies would be hard to do.

I would consider things like these the equivalent of a test to see if a study is viable, not a study.

Sort of like, okay, if I have a hundred water balloons and I dropped fifty of those water balloons off the top of my house growing up, and fifty off the top of the Empire State building, would the latter always be that much worse? Just a few off the first and I can guess the outcome, and let's call that a control group, but the second, who knows... sure you might get lucky and knock one or two people out (or get a couple of good people showing results or seeming connections you want in a study) but you probably need more than a few to prove it isn't just a fluke. ;) Okay, messing around with the water balloons. Not with the study thing.

Reply on news.ycombinator.com